TOPS-MODE model of multiplexing neuroprotective effects of drugs and experimental-theoretic study of new 1,3-rasagiline derivatives potentially useful in neurodegenerative diseases

Article abstract of DOI:10.1016/j.bmc.2013.01.035

The interest on computational techniques for the discovery of neuroprotective drugs has increased due to recent fail of important clinical trials. In fact, there is a huge amount of data accumulated in public databases like CHEMBL with respect to structurally heterogeneous series of drugs, multiple assays, drug targets, and model organisms. However, there are no reports of multi-target or multiplexing Quantitative Structure-Property Relationships (mt-QSAR/mx-QSAR) models of these multiplexing assay outcomes reported in CHEMBL for neurotoxicity/neuroprotective effects of drugs. Accordingly, in this paper we develop the first mx-QSAR model for multiplexing assays of neurotoxicity/neuroprotective effects of drugs. We used the method TOPS-MODE to calculate the structural parameters of drugs. The best model found correctly classified 4393 out of 4915 total cases in both training and validation. This is representative of overall train and validation Accuracy, Sensitivity, and Specificity values near to 90%, 98%, and 80%, respectively. This dataset includes multiplexing assay endpoints of 2217 compounds. Every one compound was assayed in at least one out of 338 assays, which involved 148 molecular or cellular targets and 35 standard type measures in 11 model organisms (including human). The second aim of this work is the exemplification of the use of the new mx-QSAR model with a practical case of study. To this end, we obtained again by organic synthesis and reported, by the first time, experimental assays of the new 1,3-rasagiline derivatives 3 different tests: assay (1) in absence of neurotoxic agents, (2) in the presence of glutamate, and (3) in the presence of H₂O₂. The higher neuroprotective effects found for each one of these assays were for the stereoisomers of compound 7: compound 7b with protection = 23.4% in assay (1) and protection = 15.2% in assay (2); and for compound 7a with protection = 46.2% in assay (3). Interestingly, almost all compounds show protection values >10% in assay (3) but not in the other 2 assays. After that, we used the mx-QSAR model to predict the more probable response of the new compounds in 559 unique pharmacological tests not carried out experimentally. The results obtained are very significant because they complement the pharmacological studies of these promising rasagiline derivatives. This work paves the way for further developments in the multi-target/multiplexing screening of large libraries of compounds potentially useful in the treatment of neurodegenerative diseases.

Full text of DOI:10.1016/j.bmc.2013.01.035

Bioorganic & Medicinal Chemistry 21 (2013) 1870–1879
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
TOPS-MODE model of multiplexing neuroprotective effects of drugs
and experimental-theoretic study of new 1,3-rasagiline derivatives
potentially useful in neurodegenerative diseases
b,
c
c
c
Feng Luan a,b, , M. Natália D. S. Cordeiro , Nerea Alonso , Xerardo García-Mera , Olga Caamaño ,
⇑
⇑
Francisco J. Romero-Duran c, Matilde Yañez d, Humberto González-Díaz e,f
a Department of Applied Chemistry, Yantai University, Yantai 264005, People’s Republic of China
b REQUIMTE, Department of Chemistry and Biochemistry, University of Porto, 4169-007 Porto, Portugal
c Department of Organic Chemistry, Faculty of Pharmacy, University of Santiago de Compostela (USC), 15782 Santiago de Compostela, Spain
d Department of Pharmacology, Faculty of Pharmacy, USC, 15782 Santiago de Compostela, Spain
e Departament of Organic Chemistry II, Faculty of Science and Technology, University of the Basque Country (UPV/EHU), 48940 Leioa, Bizkaia, Spain
f IKERBASQUE, Basque Foundation for Science, 48011 Bilbao, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
The interest on computational techniques for the discovery of neuroprotective drugs has increased due to
recent fail of important clinical trials. In fact, there is a huge amount of data accumulated in public dat-
abases like CHEMBL with respect to structurally heterogeneous series of drugs, multiple assays, drug tar-
gets, and model organisms. However, there are no reports of multi-target or multiplexing Quantitative
Structure–Property Relationships (mt-QSAR/mx-QSAR) models of these multiplexing assay outcomes
reported in CHEMBL for neurotoxicity/neuroprotective effects of drugs. Accordingly, in this paper we
develop the first mx-QSAR model for multiplexing assays of neurotoxicity/neuroprotective effects of
drugs. We used the method TOPS-MODE to calculate the structural parameters of drugs. The best model
found correctly classified 4393 out of 4915 total cases in both training and validation. This is represen-
tative of overall train and validation Accuracy, Sensitivity, and Specificity values near to 90%, 98%, and
80%, respectively. This dataset includes multiplexing assay endpoints of 2217 compounds. Every one
compound was assayed in at least one out of 338 assays, which involved 148 molecular or cellular targets
and 35 standard type measures in 11 model organisms (including human). The second aim of this work is
the exemplification of the use of the new mx-QSAR model with a practical case of study. To this end, we
obtained again by organic synthesis and reported, by the first time, experimental assays of the new 1,3-
rasagiline derivatives 3 different tests: assay (1) in absence of neurotoxic agents, (2) in the presence of
glutamate, and (3) in the presence of H2O2. The higher neuroprotective effects found for each one of these
assays were for the stereoisomers of compound 7: compound 7b with protection = 23.4% in assay (1) and
protection = 15.2% in assay (2); and for compound 7a with protection = 46.2% in assay (3). Interestingly,
almost all compounds show protection values >10% in assay (3) but not in the other 2 assays. After that,
we used the mx-QSAR model to predict the more probable response of the new compounds in 559 unique
pharmacological tests not carried out experimentally. The results obtained are very significant because
they complement the pharmacological studies of these promising rasagiline derivatives. This work paves
the way for further developments in the multi-target/multiplexing screening of large libraries of com-
pounds potentially useful in the treatment of neurodegenerative diseases.
Received 1 October 2012
Revised 13 January 2013
Accepted 17 January 2013
Available online 27 January 2013
Keywords:
Neuroprotective effects
Rasagiline derivatives
CHEMBL
Multi-target QSAR
Multiplexing assays
Moving averages
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
the common signals and substrates involved in the different pro-
cess in the neurovascular unit may reveal useful paradigms in this
Neuroprotective strategy has evolved in the last two decades
from targeting a signal pathway in neurons to protect all neurovas-
cular components.1 According to Xing et al.2 an understanding of
area. Nevertheless, regardless of the growing knowledge of the
physiological mechanisms of the ischemic penumbra, no effective
neuroprotective therapy has been found so far.1 Indeed, recent
failed clinical trials have reduced fervour for the discovery of neu-
roprotective drugs.3 In any case, there is a huge amount of informa-
tion in public datasets about structurally heterogeneous drugs as
well as multiple assays, biochemical pathways, and drug targets,
⇑
Corresponding authors.
E-mail addresses: fluan@sina.com (F. Luan), ncordeir@fc.up.pt (M. Natália D.S.
Cordeiro).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.01.035

F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
1871
which should be taken into consideration towards the develop-
ment of drugs with neuroprotective effects. In fact, Xing et al.4 have
summarized many pathophysiologic cascades involved in ischemic
stroke. For instance, Capettini et al.,5 have updated very recently
the evidence on different cannabinoid-triggered avenues to reduce
inflammation and neuronal injury in acute ischemic stroke. In a re-
cent review, Dudas and Semeniken,6 discussed that the mechanism
by which the glycosaminoglycans exhibit neuroprotective proper-
ties may involve their impact on amyloidogenesis and in the apop-
deviation of the value of one molecular descriptor of one com-
pound from the average hDji of these values for compounds that
give a positive result in the same assay conditions mj. Next, we up-
load this preprocessed data to one Statistics or Machine Learning
software to seek the model. In the particular case of a linear model,
the mx-QSAR equation based on moving averages has the follow-
ing general form:
m
m
X
X
SiðmjÞ ¼ a0 þ
¼ a0 þ
bi ꢁ Di þ
bi ꢁ Di þ
ci ꢁ
DDi
totic pathway. Another example is Cerebrolysin,
a
peptide
i
i
m
m
preparation that mimics the pleiotropic effects of neurotrophic fac-
tors, which are indispensable for survival, differentiation, and pro-
tection against damage under pathologic conditions of developing
neurons. Allegri and Guekht,7 reviewed very recently several clin-
ical trials investigating the therapeutic efficacy of Cerebrolysin in
Alzheimer’s Disease (AD) and Vascular Dementia (VD). In a last
example, Yañez et al.,8 demonstrated that Resveratrol has neuro-
protective effects in patients suffering Amyotrophic Lateral Sclero-
sis (ALS).
X
X
ci ꢁ ðDi ꢀ hDjiÞ
ð1Þ
i
i
where, Si(mj) is a numerical score of the biological activity of the ith
compound measured under the jth assay defined by the set of con-
ditions mj. In the case when mj refer only to different targets, we are
in the presence of a mt-QSAR model based on moving averages.16 In
a more general picture, we are in the presence of a mx-QSAR model
when mj refers to different multiplexing assay conditions, for exam-
ple, targets, assays, cellular lines, organisms, organs, etc. See also
the excellent works published after Speck-Planche et al.17–22 on dif-
ferent ways to seek similar mt-QSAR/mx-QSAR models. We illus-
trate the general workflow of the mx-QSAR procedure in Figure 1.
In principle, we can use any one of the existing software for cal-
culation molecular descriptors to seek mt-QSAR or mx-QSAR mod-
els. Some of the more used nowadays are: DRAGON,23,24 MOE,25
TOMOCOMD,26,27 CODESSA28–30 and MARCH-INSIDE.31–33 Spe-
cially, the method TOPS-MODE implemented by Estrada et al. in
the computer program MODESLAB, is one of the more widely used
techniques for both QSAR and mt-QSAR studies. Very recently,
On the other hand, the large number of experimental results re-
ported by different groups worldwide has led to the accumulation
of huge amounts of information in large databases. This deter-
mines, in turn, the necessity of new algorithms to perform data
mining of these databases. CHEMBL is more probably the largest
public database containing Binding (B), Functional (F), and ADMET
(A) information for a large number of drug-like bioactive com-
pounds. Access is available through a web-based interface, data
downloads and web services at: https://www.ebi.ac.uk/chembldb.9
Currently, CHEMBL contains >5.4 million assay outcomes for >1
million of compounds and 5200 protein targets. Nonetheless, the
database describes only some selected tests for almost all drugs
with respect to the huge number of assays reported. Consequently,
multi-target/multiplexing techniques useful to measure or predict
neurotoxicity/neuroprotective profiles of drugs are highly desired
in order to increase the safety and efficacy of compounds poten-
tially active against neurodegenerative diseases.10 In fact, Mok
and Brenk11 postulated the data mining of CHEMBL using different
computational tools as a very interesting source of new knowledge
for drug discovery. Regrettably, almost current Quantitative Struc-
ture–Activity Relationships (QSAR) techniques are able to predict
new outcomes only for one specific assay. In our opinion, we can
evade this problem developing new Multi-target/Multiplexing
QSAR models (mt-QSAR/mx-QSAR). These methods are especially
powerful when we need to process very large collections of com-
pounds assayed against multiple molecular or cellular targets in
different assay conditions (mj) as is the case of CHEMBL.12,13 This
step may be of the major relevance for the future of QSAR.
Notably, mt-QSAR models are able to predict the results of the
assay of different drugs for multiple targets. However, mt-QSAR
models are unable to predict different results for a given series of
targets when we change the set of specific assay conditions for
each target. Fortunately, the new class of mx-QSAR models applies
not only to different targets but also to different multiplexing assay
conditions (mj) for all targets. Specifically, we have reported the
first mx-QSAR model for multiplexing assays of anti-Alzheimer,
anti-parasitic, anti-fungi, and anti-bacterial activity of GSK-3
inhibitors in vitro, in vivo, and in different cellular lines.14 In a first
step, we need to calculate the molecular descriptors using Di of a
given ith compound using one or more software for generation of
molecular descriptors. In a second step, we expand the raw dataset
of molecular descriptors adding new variables
D
Dij = Di ꢀ hDi(mj)i.
These deviation-like parameters
D
Dij are inspired in the idea of
moving averages used in time series analysis.15 In any case, hDi(mj)i
is the average of the Di of compounds active in an assay carry out
under the set of conditions mj and does not quantify an interval of
time like in time series. These additional terms express the
Figure 1. Workflow of the mx-QSAR study.

1872
F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
Tenorio-Borroto et al.34 upgraded the method TOPS-MODE to carry
out mx-QSAR studies using the MAD methodology. For instance,
the mx-QSAR model developed by our group recently correctly
classifies 8258 out of 9000 (Accuracy = 91.76%) multiplexing assay
endpoints of 7903 drugs (including both train and test series). Each
endpoint correspond to one out of 1418 assays, 36 molecular and
cellular targets, 46 standard type measures, in two possible organ-
isms (human and mouse).34 However, there is no report of mx-
QSAR analysis of multiplexing assay outcomes reported in CHEMBL
for neurotoxicity/neuroprotective effects of drugs, until the best of
our knowledge. Accordingly, in this paper we develop the first mx-
QSAR model for multiplexing assays of neurotoxicity/neuroprotec-
tive effects of drugs. We used the method TOPS-MODE to calculate
the structural parameters of drugs. The second aim of this work is
the exemplification of the use of the new mx-QSAR model with a
practical case of study. To this end, we obtained again by organic
synthesis and report by the first time, the biological assay of new
2,3-rasgiline derivatives in three different tests for neuroprotective
effect. Rasagiline is a promising drug for treatment of Parkison’s
disease with an interesting inhibition effect selective for MAO-
B.35,36 After that, we used the mx-QSAR model to predict the most
probable results for this compound in a large number of assays not
carried out experimentally in this work.
Table 1
Overall results of the mx-QSAR classification model
Sub-set
Train
Statisticsa
%
Groups
Ci(mj)pred = 0
Ci(mj)pred = 1
Specificity
Sensitivity
Accuracy
Specificity
Sensitivity
Accuracy
81.3
98.0
89.5
81.0
97.7
89.1
Ci(mj)obs = 0
Ci(mj)obs = 1
Total
Ci(mj)obs = 0
Ci(mj)obs = 1
Total
1533
36
352
1762
CV
513
14
120
585
a
Sensitivity = Sn = positive correct/positive total; specificity = Sp = negative cor-
rect/negative total; accuracy = Ac = total correct/overall total.
type of ontology allows a clear definition of the multiplex condi-
tions for one assay in our dataset following the same line of think-
ing used for other ontology-like datasets in the literature.47
The first parameter li5 = p(c)ꢁdipli5 codifies the influence of the
chemical structure of the compound on the biological activity. It is
known that the spectral moment of order 5 codifies information
about all types of structural fragments with five or less bonds in
the molecule. In addition to the topological information, wli5 also
codifies information about the physicochemical properties of the
atoms and bonds in the molecule. It depends on the type of atomic
or bond weights wij used. In our equation, we set wij equal to the val-
ues of standard bond dipole moments (dip) in order to incorporate
both geometrical and electronic charge information.33–37 Conse-
quently, ⁄li5 codifies the effect of the drug structure on the biological
activity, but depending on the confidence of assays reported in
CHEMBL. In this sense, we have pre-multiplied li5 by the parameter
p(cl). The parameter p(cl) is a probability (a priori) of confidence for a
given data value into the CHEMBL dataset studied. Next, we have de-
fined p(c) = 1, 0.75, or 0.5 for data values reported as being curated at
expert, intermediate, or auto-curation level, respectively.
2. Results and discussion
2.1. TOPS-MODE multiplexing model of drug neuroprotective
effects
The outcome of multiplexing neuroprotective assays depend
both on drug structure and the set of assay conditions selected
(mj).37 In this work, we report the first mx-QSAR model capable
of predict whether a drug with a determined molecular structure
may give or not a positive result in different multiplexing assay
conditions mj. These models are expected to give different classifi-
cation probabilities of the compound for different: organisms (ot),
biological assays (au), molecular or cellular targets (te), or standard
type of activity measure (sx). It is also desirable to use an algorithm
that takes into consideration the different degrees of accuracy or
level of curation (cl) in the experimental data. The best mx-QSAR
model found was:
The other three terms in the equation express the structural
dissimilarity between one specific compound and a group of active
compounds that have been assayed in specific multiplex
conditions defined by the sub-ontology mj P (ot, te, sx). We have
quantified this effect in terms of the deviation
D
li5(mj) = dipli5
ꢀ
p1(mj)ꢁhdipli5(mj)i. In general, we have defined p1(mj) = n1(mj)/
ntot(mj); where n1(mj) and ntot(mj) are the number of positive or to-
tal results for compounds assayed in the jth condition in the
CHEMBL dataset. These moving average or deviation terms repre-
sent the hypothesis: H0 the structural dissimilarity between one
compound, with respect to the average of all compounds in a
group, predicts the final behavior of the compound. For instance,
SiðmjÞ ¼ ꢀ7:01 ꢁ 10ꢀ4
ꢁ
D
li5 ꢀ 7:84 ꢁ 10ꢀ4
ꢁ
D
li5ðsÞ ꢀ 2:93 ꢁ 10ꢀ4
ꢁ
ꢁ
D
li5ðaÞ
þ1:16 ꢁ 10ꢀ4
ꢁ
li5ðoÞ þ 2:84 ꢁ 10ꢀ4
D
li5ðtÞ þ 4:198684
N ¼ 3683 Rc ¼ 0:7 Sn ¼ 98:0 Sp ¼ 81:3 Ac ¼ 89:5
D
li5(ot) = stdli5 ꢀ p1(ot)hstdli5(ot)i measures the deviation from
ð2Þ
the average value hli5(te)i of li5 for all active compounds (C = 1) as-
sayed in the organism ot P t = 1, 2,. . . for Human, or other organ-
isms, respectively. This type of model able to model/interpret
cross-species activity is of major importance in order to reduce as-
S(mj) = S(di, au, cl, ot, te, sx) is a real-valued variable that scores the
propensity of the drug to be active in multiplex pharmacological as-
says of the drug di carried out on the conditions selected mj P au, cl,
ot, te, and sx. The statistical parameters for the above equation in
training are: Number of cases used to train the model (N), Canonical
Regression Coefficient (Rc), Sensitivity (Sn), Specificity (Sp), and
Accuracy (Ac).15 The probability cut-off for this LDA model is
ip1(mj) > 0.5 P Ci(mj) = 1. It means that the ith drug (di) predicted
by the model with probability >0.5 is expected to give a positive
outcome in the jth assays carry out under the given set of conditions
mj. This linear equation presented good results both in training and
external validation series with overall Accuracy in training series
above 90% (see Table 1). According to previous reports38–46 values
accuracy higher than 75% are acceptable for LDA-QSAR models.
The reader should be aware that N here is not number of com-
pounds but number of statistical cases. One compound may lead
to 1 or more statistical cases because it may give different outcomes
for alternative biological assays carried out in diverse sets of multi-
plex conditions defined by the ontology mj P (au, cl, ot, te, sx). This
says in humans.38 By analogy,
D
li5(te) = stdli5 ꢀ hstdli5(te)i is the
dissimilarity between the structure of compound ith (expressed
by stdli5) with respect to all compounds active against the molec-
ular or cellular target te. In Table 2 we give some examples of aver-
age values for the different targets, organisms, or standard
measure types. Online Supplementary data files contain detailed
lists of values of these parameters.
2.2. Experimental-theoretical study of new neuroprotective
drugs
2.2.1. Experimental assay of neuroprotective effects of new 1,3-
rasagilines
The second aim of this work is the exemplification of the use of
the new mx-QSAR model with a practical case of study. To this end,
we obtained again by organic synthesis and reported by the first

F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
1873
Table 2
Selected examples of average values for different targets, measures, and organisms
CHEMBLID
h
l5(mj)i
Name or units
CHEMBLID
hl
5(mj)i
Name or units
Protein targets
3599
3916
319
Protein targets
1907608
384
614585
3772
1907609
4573
4980
1907592
4802
614205
612556
612405
1899
4888.44
5304.48
6020.28
8533.75
5316.33
5234.11
6982.89
7518.82
14682.75
7333.74
6480.94
7956.01
5233.51
4098.58
8841.88
5383.23
Acetylcholine receptor protein alpha chain
Acetylcholine receptor protein delta chain
Alpha-1a adrenergic receptor
Alpha-1b adrenergic receptor
Alpha-2a adrenergic receptor
Alpha-2b adrenergic receptor
Beta-1 adrenergic receptor
Betaine transporter
Caspase-1
Caspase-10
DOPA decarboxylase
Dopamine D2 receptor
Dopamine transporter
7569.51
5731.6
Glutamate NMDA receptor
HT-29 (Colon adenocarcinoma cells)
IMR-32 (Neuroblastoma cells)
Metabotropic glutamate receptor 1
Muscarinic acetylcholine receptor
Neurokinin 1 receptor
Neuronal acetylcholine receptor protein alpha-7
Neuronal acetylcholine receptor; alpha2/beta2
Nitric-oxide synthase, endothelial
7436.96
7395.55
6108.82
5542.08
5161.98
6023.39
4874.72
3838.14
7046.1
9075.71
7502.62
4329.83
9118.42
7273.42
315
1867
1942
249
3715
4801
5037
1843
5102
2799
4054
5208
5102
NT2 (Teratocarcinoma cells)
PC-12 (Adrenal phaeochromacytoma cells)
SCG (Superior cervical ganglion neurones)
Serotonin 3a (5-HT3a) receptor
Sodium channel protein type I alpha
Sodium channel protein type VIII alpha
Transient receptor potential cation channel V2
GABA transporter 1
GABA transporter 3
Vanilloid receptor
4906
2752
2863
Organisms
Organisms
Organism-01
Organism-03
Organism-04
Organism-06
5775.6
Homo sapiens
Organism-07
Organism-08
Organism-09
Organism-10
6006.98
16162.12
7775.19
8776.43
Rattus norvegicus
Heliothis virescens
Mus musculus
4932.01
5680.88
7534.91
Caenorhabditis elegans
Cavia porcellus
Felis catus
Tropedo californica
Standard type measures
Standard type measures
Activity
Decrease
nNOS activity
Dopamine release
ED50
Efficacy
ED50
Ki
7208.42
6334.59
6514.67
4584.1
4753.33
5779.68
8002.48
5524.2
nM
%
%
Ratio
4211.43
nM
Inhibition
Concentration
Activity
%max
3862.29
2781.19
11283.7
14405.49
11160.96
11419.52
8060.06
6031.43
nM
% dose gꢀ1
%
%
%
nM
%
%
l
%
l
g mlꢀ1
g kgꢀ1
ED50
Morbidity
Survived
Inhibition
nM
nM
EC50
4077.11
%
time, experimental measures of the Neuroprotection capacity
(NP) of the new 2,3-rasagiline derivatives in 3 different assays:
assay (1) in absence of neurotoxic agents, (2) in the presence of
glutamate, and (3) in the presence of H2O2. In general, the synthe-
sized compounds (7a, 7b, 8a, 8b, 9a, 9b, 10a, and 10b) were sub-
jected to an initial study to determine their neuroprotective
capacity in both the absence and presence of neurotoxic agents,
using the reduction method bromide 3-(4,5-dimethyl-2-thia-
zoyl)-2,5-difeniltetrazólico (MTT) method used to determine cell
viability, given by the number of cells present in culture. The abil-
ity of cells to reduce MTT is an indicator of mitochondrial integ-
rity and its functional activity is interpreted as a measure of cell
viability. There were three types of assays in cultured neurons of
embryonic motor cortex of Sprague–Dawley rats of 19 days. All
results are expressed as mean SEM of at least three independent
experiments (Table 3). First, assay (1), we studied the ability to
induce a neuroprotective effect in the absence of any neurotoxic
stimulus. Secondly, in assay (2), the neuroprotective effect was
studied in the presence of glutamate excitotoxicity compound
that causes a pathological process in which the neurons are dam-
aged sobreactivarse receptors (NMDA and AMPA/kainate) the
excitatory neurotransmitter glutamate, ultimately leading to
apoptosis. Finally, in assay (3), we examined the ability of the
compounds synthesized to protect neurons from damage caused
by H2O2 neuronal death caused by oxidative stress. The results
indicate that all compounds prepared, except 10b, have a high
neuroprotective capacity against damage caused by H2O2, assay
(3). Especially the compound 7a highlights with a NP = 46.2%,
with the remainder of compounds level average NP = 16.5%.
Moreover, the compound 7b showed a high neuroprotective
capacity in all three types of assays performed with values of
NP = 23.4% in assay (1), and NP = 15.2% in assay (2), and
NP = 14.5% in assay (3), see Table 3.
2.2.2. Prediction of multiplexing outcomes of new
neuroprotective drugs in other assays
After that, we used the mx-QSAR model to predict the more
probable results for all these compounds in >500 assays not carried
out experimentally in this work. Notably, the mx-QSAR model has
predicted a very high probability p1(mj) of activity for compound 7
(both isomers 7a and 7b) against different receptors related to glu-
tamate (see Table 4). This result may be in coincidence with the
experimental value of protection = 46.2% for 7a in the presence of
glutamate. Nuritova and Frenguelli,48 a novel neuroprotective
strategy involving retrograde release of glutamate.
3. Materials and methods
3.1. Computational methods
3.1.1. CHEMBL dataset
This dataset includes Nd = 2217 unique drugs and/or organic
compounds previously assayed in different multiplexing assay
conditions (mj). We assigned a value of the observed (obs) class
variable Ci(mj)obs = 1 (active compound) or Ci(mj)obs = 0 (non-active
compounds) to every ith drug biologically assayed in different mj
conditions. One compound may lead to 1 or more statistical cases
because it may give different outcomes (statistical cases) for alter-
native biological assays carried out in diverse sets of multiplex
conditions. In this work, we defined mj according to the ontology
mj P (au, cl, ot, te, sx). A general data set composed of >10,000 mul-
tiplexing assay endpoints was downloaded from the public data-
base CHEMBL.9,49 In any case, after a carefully curation of the
dataset we retain 4915 multiplexing assay endpoints (statistical
cases) after elimination of all cases with missing information or
very low representation. The different conditions that may change
in the dataset are the following: organisms (ot), biological assays

1874
F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
Table 3
a
Neuroprotective ability of the new 1,3-rasagiline derivatives tested in the absence of neurotoxic agents, in the presence of glutamate and in the presence of H2O2
d
Compound
Formula
Protectionb (%)
e.s.m.
Glutamatec (%)
e.s.m.
H2O2 (%)
e.s.m.
5.6
OH
7a
0
3.8
0
4.0
46.2
HN
OH
7b
8a
23.4
4.6
6.2
15.2
2.0
3.0
14.5
17.8
2.1
4.8
HN
OH
ꢀ5
0
N
OH
8b
ꢀ1
4.3
ꢀ0.5
4.0
19.6
2.6
N
OAc
9a
0
4.6
3.9
5.6
ꢀ1.5
ꢀ0.5
1.6
2.0
1.8
4.0
18.2
12.3
16.4
3.1
2.9
5.2
N
OAc
9b
0
N
OBz
10a
ꢀ5.1
N
OBz
10b
6.9
8.0
ꢀ1.3
4.7
ꢀ5.8
3.0
N
a
b
c
Highest neuroprotective ability is highlighted in bold.
% Protection (compd 5
% Protection (compd 5
% Protection (compd 5
l
l
l
M).
M) against Glutamate 100 lM.
M) against H2O2 100 lM.
d
X
X
4
SiðmjÞ ¼ b0 þ b1 ꢁ li5
¼ b0 þ b1 ꢁ li5
þ
j¼2bj ꢁ
D
li5ðmjÞ
(au), molecular or cellular targets (te), or standard type of activity
measure (sx). In closing, we analyzed N = 4915 statistical cases con-
formed by the above-mentioned Nd = 2217 drugs; which have been
assayed each one in at least one out of Na = 338 possible assays. For
each one of these assays the dataset studied present for each drug
at least one out of Ns = 35 standard types of biological activity mea-
sures in turn carried out in at least one out of Nt = 148 molecular or
cellular targets. We withdrawn these values from CHEMBL as re-
sults of experiments carried out on at least 1 out of 11 model
organisms and with 3 different levels of data curation Nc = 3 (ex-
pert, intermediate, or auto-curation level). Please, see details on
the assignation of cases to different classes in Section 2.
4
þ
bj ꢁ ðli5 ꢀ p1ðmjÞ ꢁ hl5i ðmjÞiÞ
ð3Þ
j¼2
Where, S(mj) = S(di, au, cl, ot, te, sx) is a real-valued variable that scores
the propensity of the drug to be active in multiplex pharmacological
assays of the drug depending on the conditions selected mj. The sta-
tistical parameters used to corroborate the model were: Number of
cases in training (N), and overall values of Specificity (Sp), Sensitivity
(Sn), and Accuracy (Ac).15 In this model, stdl5i is the spectral moment
or order k = 5 calculated with Modeslab. In this study, we used only
standard bond distance as entries of the main diagonal of the bond
adjacency matrix. The parameter p1(mj) is a probability, calculated
a priori, with which any drug is expected to give a positive results,
Ci(mj) = 1, in the test carried out under the mj conditions. The param-
eter p(cl) is a probability, calculated a priori, of confidence for a given
data value into the CHEMBL dataset studied. The structural deviation
3.1.2. The moving averages model
In order to seek the mx-QSAR model we used the technique LDA
implemented in the software package STASTICA 6.0.50 The LDA
model studied here based on the concept of moving averages has
the following general form:
terms
D
li5(mj) = li5 ꢀ p1(mj)ꢁhli5(mj)i represent the hypothesis H0.

F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
1875
Table 4
Prediction of endpoints in multiplexing assays for compound 7
Assay ID
Target namea
p1(mj)b
Measure (U)
Cutoffc
Organism
715720
715721
715722
712468
712469
835748
835748
835748
835748
835748
872629
714496
713191
717250
682643
876081
859484
866501
911164
748592
617201
669461
682643
876081
748592
874089
859484
866501
911164
752536
859484
866501
911164
874089
682643
876081
874089
682643
876081
748592
859484
866501
911164
682643
859484
682643
876081
682643
876081
682643
876081
752536
859484
866501
911164
1613870
682643
876081
679291
751927
755291
883402
671419
617201
682643
876081
MGluR 5
MGluR 5
MGluR 5
MGluR 8
MGluR 8
Glu[NMDA]R 3B
Glu[NMDA]R epsilon 4
Glu[NMDA]R zeta 1
Glu[NMDA] 3A
Glu[NMDA] epsilon 3
MGluR 8
MGluR 4
MGluR 2
MGluR 3
GABA receptor gamma-2
GABA receptor gamma-2
NAChR protein alpha-2
NAChR protein alpha-2
NAChR protein alpha-2
MAChR M4
Serotonin 2b (5-HT2b) receptor
DOPA decarboxylase
GABA receptor beta-3
GABA receptor beta-3
MAChR M3
Alpha-2a adrenergic receptor
NAChR protein alpha-4
NAChR protein alpha-4
NAChR protein alpha-4
NAChR protein beta-2
NAChR protein beta-2
NAChR protein beta-2
NAChR protein beta-2
Alpha-2c adrenergic receptor
GABA receptor beta-2
GABA receptor beta-2
Alpha-2b adrenergic receptor
GABA receptor alpha-1
GABA receptor alpha-1
MAChR M5
NAChR protein alpha-9
NAChR protein alpha-9
NAChR protein alpha-9
GABA receptor alpha-4
NAChR protein alpha-7
GABA receptor gamma-3
GABA receptor gamma-3
GABA receptor theta
GABA receptor theta
GABA receptor alpha-3
GABA receptor alpha-3
NAChR protein alpha-3
NAChR protein alpha-3
NAChR protein alpha-3
NAChR protein alpha-3
Nuclear factor NF-kappa-B p105
GABA receptor pi
0.9999
1.0000
1.0000
1.0000
0.9998
0.9502
0.9294
0.9344
0.9489
0.9294
0.9939
0.9999
0.9715
1.0000
0.9950
0.9651
0.9448
0.9446
0.9925
0.9958
0.9764
0.9796
0.9568
0.9497
0.9934
0.9819
1.0000
0.9571
0.9925
1.0000
0.9949
1.0000
0.9573
0.9776
0.9994
0.9418
0.9711
0.9995
0.9497
0.9974
0.9584
1.0000
0.9874
0.9991
0.9584
0.9249
0.9418
0.9925
0.9418
0.9925
0.9497
0.9866
1.0000
0.9571
0.9329
0.9824
0.9991
0.9497
0.9937
0.9901
0.9998
0.9838
0.9232
0.9428
0.9992
0.9368
Activity (nM)
Activity (nM)
Activity (nM)
Activity (nM)
Activity (nM)
Selectivity ratio
Selectivity ratio
Selectivity ratio
Selectivity ratio
Selectivity ratio
EC50 (nM)
EC50 (nM)
EC50 (nM)
EC50 (nM)
Ki (nM)
100,000
17,000
100,000
45,000
100,000
158
15.3
158
1774
15.3
31
320
300
600
10,000
10,000
7
3
30
501
18
100
10,000
10,000
501
783
7
3
30
1990
7
3
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
R. norvegicus
R. norvegicus
R. norvegicus
R. norvegicus
R. norvegicus
R. norvegicus
R. norvegicus
R. norvegicus
R. norvegicus
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
H. sapiens
Ki (nM)
Activity (%)
Activity (%)
Inhibition (%)
Ki (nM)
Ki (nM)
Inhibition (%)
Ki (nM)
Ki (nM)
Ki (nM)
Ki (nM)
Activity (%)
Activity (%)
Inhibition (%)
Ki (nM)
Activity (%)
Activity (%)
Inhibition (%)
Ki (nM)
Ki (nM)
Ki (nM)
Ki (nM)
Ki (nM)
30
783
10,000
10,000
263
10,000
10,000
501
Ki (nM)
Ki (nM)
Activity (%)
Activity (%)
Inhibition (%)
Ki (nM)
Activity (%)
Ki (nM)
Ki (nM)
Ki (nM)
Ki (nM)
Ki (nM)
7
3
30
10,000
7
10,000
10000
10,000
10,000
10,000
10,000
22.5
7
Ki (nM)
Ki (nM)
Activity (%)
Activity (%)
Inhibition (%)
EC50 (nM)
Ki (nM)
Ki (nM)
Selectivity
Ki (nM)
EC50 (nM)
EC50 (nM)
Ki (nM)
Ki (nM)
Ki (nM)
3
30
346
10,000
10,000
10
1250
4
1300
20
112
GABA receptor pi
Nitric-oxide synthase, brain
Nitric-oxide synthase, brain
Neurokinin 1 receptor
Neurokinin 1 receptor
Dopamine D2 receptor
Serotonin 2a (5-HT2a) receptor
GABA receptor alpha-5
GABA receptor alpha-5
10,000
10,000
Ki (nM)
a
mGluR = metabotropic glutamate receptor, NAChR = neuronal acetylcholine receptor, MAChR = muscarinic acetylcholine receptor, Glu[NMDA]R = glutamate [NMDA]
receptor.
b
p1(mj) is the probability with which the model predict a value of the measure higher than the cutoff for compound 7.
Cutoff is the threshold value for this assay (average value for all compounds in CHEMBL for this assay).
c
H0: the different deviations of the ith drug (di) with respect to the
average of all positive drugs for different multiplexing assay condi-
tions (mj) predict the final behavior of the compound. See a detailed
discussion of terms and mj conditions in Section 2 and also in the
introductory part. This type of movingaverage or deviation-like mod-
els has been used before to solve different problems.22,34

1876
F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
3.2. Experimental methods
catalytic amounts of DMAP, in MeCN.55 In the same way and
using benzoyl chloride instead of acetic anhydride, the corre-
3.2.1. General chemistry
sponding benzoates 11a and 11b were prepared.
The compounds (7a–b, 8a–b, 9a–b and 10a–b) were synthe-
sized according to the strategy given in Figure 2. As shown in
3.2.2. Synthesis and identification of 1,3-rasagilines
this Scheme, N-(1-oxo-1H-3-indanyl)trifluoroacetamide
5
was
All compounds were obtained again in sufficient amounts for
pharmacological assays. Synthesis and characterization was car-
ried out following the know how published in the previous liter-
ature.56 Melting points are uncorrected and were determined in
Reichert Kofler Thermopan or in capillary tubes on a Büchi 510
apparatus Infrared spectra, recorded on a Perkin–Elmer 1640-
FT spectrophotometer. 1H NMR spectra (300 MHz) and 13C
NMR spectra (75 MHz) were recorded in a Bruker AMX spec-
trometer, using TMS as internal reference (chemical shifts in d
values, J in Hz). Mass spectra were recorded on a HP5988A spec-
trometer. FABMS were obtained using MICROMASS AUTOSPEC
mass spectrometer. Microanalyses were performed in a Perkin–
Elmer 240B elemental analyzer by the Microanalysis Service of
the University of Santiago de Compostela. X-ray diffraction data
were collected with an Enraf-Nonius CDAD4 automatic diffrac-
tometer using the program CAD4-EXPRESS. We monitored most
reactions by TLC on pre-coated silica gel plates (Merck 60
F254, 0.25 mm). Synthesized products were purified by flash col-
umn chromatography on silica gel (Merck 60, 230–240 mesh)
and crystallized if necessary. Solvents were dried by distillation
prior use. Further details such as figures of different spectra
are available online in the Supplementary data of our previous
work.56
prepared by intramolecular cyclization of 3-amino-3phenylprop-
anoic acid, 2, following procedures previously described in the
literature.51 Compound 2 was prepared by reaction of Rodio-
now-Johnson of an ethanolic solution of benzaldehyde, malonic
acid and ammonium acetate, in 61% yield. After protection of
the amino group with trifluoroacetic anhydride at room temper-
ature, and treatment with thionyl chloride, compound 4 was in-
volved in an intramolecular cyclization reaction using excess
aluminium chloride in refluxing dichloromethane. After treat-
ment with a saturated solution of NaHCO3, the ketoamide 5
was obtained. The partial reduction of the ketone group and
deprotection of amino group of compound 5 were performed
simultaneously with NaBH4 in methanol.52 This reaction affords
a mixture of epimers aminoalcohols 6a–6b (cis/trans 60:40) in
70% yield. The alkylation of the mixture 6a–6b with propargyl
bromide and potassium carbonate in hot acetonitrile was carried
out.53 It provided in a global yield of 82%, a mixture of the cor-
responding mono- and dipropargylated derivatives (7a–7b and
8a–8b), which were easily separated by flash column chromatog-
raphy using hexane/EtOAc (4:1) as eluent.54 Compounds 8a and
8b were converted to the corresponding acetates 10a and 10b, in
good yields, by treatment with acetic anhydride, Et3N and
Figure 2. Synthesis of compounds 7a–b, 8a–b, 9a–b, and 10a–b.

F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
1877
3.2.3. Compound 2
(5 ꢂ 20 mL). The combined organic layers were dried (Na2SO4). Re-
moval of solvent left 5 (1.6 g, yield 62%) as a white solid. Mp 121–
123 °C. IR
(300 MHz, DMSO-d) d = 9.98 (d, 1H, J = 6.4 Hz, D2O exch., NH),
7.77–7.52 (m, 4H, Harom), 5.57 (t, 1H, J = 7.8 Hz, 3-H), 3.12 (ddd,
Synthesis and characterization of ( )-3-amino-3-phenylpropa-
noic acid. To a solution of benzaldehyde (15.00 g, 141.3 mmol) in
EtOH (60 mL), ammonium acetate (21.90 g, 283.0 mmol) and
malonic acid (14.70 g, 141.3 mmol) were added and the resulting
mixture was heated at 82 °C for 15 h. The precipitated was filtered
off, washed with hot EtOH (3 ꢂ 20 mL), to give 2 (14.02 g, yield
m .
= 3295, 1698, 1551, 1145, 770 cmꢀ1 1H NMR
1H, J = 18.7 Hz, 8.0 Hz, 2.5 Hz, 2a-H), 2.61 (dt, 1H, J = 18.7 Hz,
3.3 Hz, 2b-H) ppm. 13C NMR (75 MHz, DMSO-d) d = 203.45 (CO),
159.15 (COCF3), 153.10 (C-3a), 136.44 (C-7a), 135.49, 129.20,
125.70 and 122.89 (CHarom), 117.90 (COCF3), 47.64 (CH2), 42.67
(CH) ppm. MS (EI): m/z (%): 243 (43) [M+], 215 (100), 202 (30),
146 (47) [M+–COCF3], 104 (53), 77 (55). Anal. calcd for C11H8F3NO2
(243.2): C 54.33, H 3.32, N 5.76; found C 54.67, H 3.09, N 5.98.
61%) as a white solid. Mp 230–232 °C. IR
m
= 2863, 1579, 1509,
1387, 1359 cmꢀ1
.
1H NMR (300 MHz, TFA-d): d = 11.03 (d, 3H,
J = 4.1 Hz, D2O exch., OH + NH2), 6.96–6.88 (m, 5H, Harom), 4.41-
4.38 (m, 1H, CH), 2.95 (ddd, 1H, J = 18.4 Hz, 10.1 Hz, 3.4 Hz, CH2),
2.7 (dd, 1H, J = 18.4 Hz, 3.8 Hz, CH2) ppm. 13C NMR (75 MHz,
TFA-d) d = 175.07 (CO), 130.68 (C), 129.02, 127.90 and 124.59
(CHarom), 51.84 (CH), 34.27 (CH2) ppm. MS (EI): m/z (%): 165 (4)
[M+], 164 (2) [Mꢀ1]+, 119 (3) [M+–COOH], 106 (100), [M+–CH2–
COOH], 104 (17), 79 (34), 77 (22). Anal. calcd for C9H11NO2
(165.19): C 65.44, H 6.71, N 8.48; found C 65.59, H 6.51, N 8.42.
3.2.7. Compounds 6a and 6b
Synthesis and characterization of compounds ( )-cis- and ( )-
trans-3-amino-1H-indan-1-ol, 6a and 6b. To
a solution of 5
(1.00 g, 4.11 mmol) in dry MeOH (10 mL), under argon, NaBH4
(0.47 g; 12.33 mmol) was added and stirred at room temperature
for 72 h. The excess of solvent was evaporated to give a residue
that was triturated in H2O (20 mL) and extracted with EtOAc
(15 ꢂ 20 mL). The combined organic layers were dried (Na2SO4).
The solvent was removed and the yellow oil residue was purified
by flash column chromatography using DCM/MeOH (40:1) as elu-
ent, to give a mixture 60:40 of isomers cis/trans 6a/6b (0.47 g, yield
3.2.4. Compound 3
Synthesis and characterization of ( )-3-(2,2,2-trifluoroacetami-
do)-3-phenylpropanoic acid. A solution of 2 (4.00 g, 24.2 mmol) in
trifluoroacetic anhydride (15 mL), under argon, was stirred at room
temperature for 24 h. After evaporation of solvent and trituration
in Et2O, the white solid obtained was washed with Et2O
(3 ꢂ 15 mL), to give 3 (5.35 g, yield 85%) as a white solid. Mp
77%), as a yellow oil. IR
m .
= 3277, 1686, 1459, 1054, 761 cmꢀ1 1H
124–125 °C. IR
m
= 3315, 1823, 1695, 1551, 1169, 1029 cmꢀ1
.
1H
NMR (300 MHz, DMSO-d) d = 7.36–7.16 (m, 8H, Harom(c+t)), 5.07
(dd, 1H, J = 6.4 Hz, 3.3 Hz, 1-Ht), 4.82 (t, 1H, J = 7.7 Hz, 1-Hc), 4.36
(t, 1H, J = 6.4 Hz, 3-Ht), 3.91–3.96 (m, 1H, 3-Hc), 3.10 (br s, 6H,
NMR (300 MHz, DMSO-d) d = 12.41 (br s, 1H, D2O exch., COOH),
9.92 (d, 1H, J = 8.1 Hz, D2O exch., NH), 7.35–7.24 (m, 5H, Har-
om),5.24 (dt, 1H, J = 8.8 Hz, 5.6 Hz, CH), 2.95–2.74 (AB part of an
ABM system, 2H, JAB = 16.2 Hz, JAM = 9.4 Hz, JBM = 5.7 Hz, CH2)
ppm. 13C NMR (75 MHz, DMSO-d) d = 171.65 (COOH), 156.15
(COCF3), 141.08 (C), 128.78, 127.81 and 126.81 (CHarom), 118.14
(COCF3), 50.67 (CH), 40.05 (CH2) ppm. MS (EI): m/z (%): 262 (4)
[M+1]+, 261 (3) [M+], 244 (12) [M+–OH], 215 (60) [M+–COOH],
202 (100) [M+–CH2COOH], 132 (38), 104 (36), 77 (32). Anal. calcd
for C11H10F3NO3 (261.20): C 50.58, H 3.86, N 5.36; found C 50.43,
H 3.97, N 5.39.
D2O exch., OH + NH2(c+t)), 2.67 (dt, 1H, J = 13.7 Hz, 7.0 Hz, 2
a-Hc),
2.19 (ddd, 1H, J = 13.2 Hz, 7.0 Hz, 3,5 Hz, 2 -Ht), 1.90–1.81 (m,
a
1H, 2b-Ht), 1.46–1.36 (m, 1H, 2b-Hc) ppm. 13C NMR (75 MHz,
DMSO-d) d = 147.65 (C-7ac), 146.76 (C-7at), 145.63 (C-3ac), 145.21
(C-3at), 127.71, 127.13, 126.92, 126.66, 124.44, 123.80, 123.41
and 123.26 (CHarom(c+t)), 72.22 (C-1c), 71.12 (C-1t), 54.29 (C-3c),
52.99 (C-3t), 47.55 (C-2c), 46.26 (C-2t) ppm. MS (EI): m/z (%): 149
(29) [M+], 148 (23) [Mꢀ1]+, 132 (100) [M+ꢀH2O], 117 (18), 104
(98), 77 (19).
3.2.5. Compound 4
3.2.8. Compounds 7a, 7b, 8a, and 8b
Synthesis and characterization of ( )-3-(2,2,2-trifluoroacetami-
do)-3-phenylpropanoyl chloride. A solution of 3 (3.0 g, 11.5 mmol)
in thionyl chloride (15 mL) was heated at 82 °C for 24 h. The result-
ing mixture was evaporated and the residue obtained was tritu-
rated in cyclohexane, to give a brown solid, that was filtered off,
washed with cyclohexane (5 ꢂ 15 mL), to give 4 (3.15 g, yield
Synthesis and characterization of compounds ( )-cis and ( )-
trans-3-(N-propargylamino)-1-indanol, 7a and 7b and ( )-cis and
( )-trans-3-(N,N-dipropargylamino)-1-indanol, 8a and 8b. A mix-
ture of 6a/6b (0.50 g, 3.35 mmol), K2CO3 (0.46 g, 3.35 mmol) and
MeCN (15 mL) was stirred at room temperature under argon for
5 min. A solution of propargyl bromide (0.3 mL, 2.7 mmol) dis-
solved in MeCN (2 mL) was added dropwise with stirring. After
being stirred for 24 h, the solvent was evaporated and the residue
was dissolved in EtOAc (20 mL). The organic layer was washed
with NaOH 2 N (3 ꢂ 25 mL) and dried (Na2SO4). The excess of sol-
vent was removed to give a brown oil, that was purified by flash
column chromatography using hexane/EtOAc (4:1) as eluent to
give, in first place 8a (80 mg, yield 18%) as a white solid, then 8b
(60 mg, yield 12%) as a yellow solid, in third place 7a (150 mg, yield
30%) as a brown solid and finally 7b (110 mg, yield 22%) as a brown
oil.
98%) as a yellow solid. Mp 60–62 °C. IR
m = 3306, 1703, 1556,
1275, 1153 cmꢀ1 1H NMR (300 MHz, DMSO-d) d = 9.93 (d, 1H,
.
J = 8.1 Hz, D2O exch., NH), 7.34–7.23 (m, 5H, Harom), 5.23 (dt, 1H,
J = 8.8 Hz, 5.8 Hz, CH),), 2.94–2.72 (AB part of an ABM system,
2H, JAB = 16.3 Hz, JAM = 9.4 Hz, JBM = 5.6 Hz, CH2) ppm 13C NMR
(75 MHz, DMSO-d) d = 171.67 (COCl), 156.12 (COCF3), 141.14 (C),
128.88, 127.89 and 126.84 (CHarom), 118.18 (COCF3), 50.69 (CH),
40.07 (CH2) ppm. MS (EI): m/z (%): 261(5), 243 (7) [M+–Cl], 215
(100) [M+–COCl], 202 (32) [M+–CH2–COCl], 146 (24), 104 (40), 79
(43). Anal. calcd for C11H9ClF3NO2 (279.64): C 47.25, H 3.24, N
5.01; found C 47.02, H 3.19, N 4.92.
3.2.9. ( )-cis-7a
3.2.6. Compound 5
Mp 111–112 °C. IR m .
= 3248, 1441, 1332, 1058, 768 cmꢀ1 1H
Synthesis and characterization of ( )-N-(1-oxo-1H-3-indanyl)-
2,2,2trifluoroacetamide. A solution of 4 (3.0 g, 10.8 mmol) in
Cl2CH2 (20 mL) was added, at 0 °C, dropwise and under argon, to
a solution of AlCl3 (2.8 g, 21.4 mmol) in Cl2CH2 (15 mL), and was
heated at 42 °C for 24 h. The excess of solvent was removed to give
a brown solid that was triturated in H2O, filtered off and washed
with H2O (3 ꢂ 30 mL). Then, the solid obtained was dispersed in
NMR (300 MHz, CDCl3) d = 7.49–7.27 (m, 4H, Harom), 5.02 (dd, 1H,
J = 6.1 Hz, 3.2 Hz, 1-H), 4.35 (dd, 1H, J = 5.9 Hz, 3.2 Hz, 3-H), 3.48
(d, 2H, J = 2.3 Hz, CH2), 2.55 (dt, 1H, J = 13.5 Hz, 6.5 Hz, 2a-H),
2.40 (br s, 2H, D2O exch., OH + NH), 2.29 (t, 1H, J = 2.6 Hz, CH),
1.88 (dt, 1H, J = 13.5 Hz, 2.9 Hz, 2b-H) ppm. 13C NMR (75 MHz,
CDCl3) d = 145.72 (C-7a), 143.24 (C-3a), 128.69, 128.44, 125.07
and 124.70 (CHarom), 81.32 (C„CH), 74.54 (C-1), 72.40 (C„CH),
59.04 (C-3), 42.74 (CH2), 35.92 (C-2) ppm. MS (EI): m/z (%):186
Et2O and extracted with
a saturated solution of NaHCO3

1878
F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
(5) [Mꢀ1]+, 168 (26) [M+–H2O], 148 (100) [M+–propargyl], 130
[M+1]+, 225 (2) [M+–acetyl], 208 (5), 171 (5), 154 (91), 137 (100).
Anal. calcd for C17H17NO2 (267.3): C 76.38, H 6.41, N 5.24; found
C 76.65, H 6.12, N 5.02.
(71), 116 (72), 77 (80). HRMS (EI): (187):
186.0919, found 186.0921.
C12H13NO calcd
3.2.10. ( )-trans-7b
3.2.14. Compound 10a
IR
m
= 3285, 1455, 1332, 1051, 755 cmꢀ1
.
1H NMR (300 MHz,
Synthesis and characterization of ( )-cis-3-(N,N-dipropargyla-
mino)-1-indanyl benzoate. To a solution of 8a (0.08 g, 0.36 mmol),
DMAP (a catalytic amount) in MeCN (5 mL), at 0 °C and under ar-
CDCl3) d = 7.44–7.31 (m, 4H, Harom), 5.42 (t, 1H, J = 5.9 Hz, 1-H),
4.63–4.59 (m, 1H, 3-H), 3.49 (d, 2H, J = 2.3 Hz, CH2), 2.32–2.26
(m, 1H, 2
a-H), 2.04 (s, 1H, CH), 1.84 (br s, 2H, D2O exch., OH + NH),
gon, was added dropwise a solution of benzoyl chloride (82
lL,
1.26(t, 1H, J = 7.3 Hz, 2b-H) ppm. 13C NMR (75 MHz, CDCl3)
d = 145.20 (C-7a), 143.52 (C-3a), 128.52, 128.48, 124.67 and
124.60 (CHarom), 81.83 (C„CH), 74.53 (C-1), 71.89 (C„CH), 59.37
(C-3), 43.94 (CH2), 36.12 (C-2) ppm. MS (EI): m/z (%): 186 (6)
[Mꢀ1]+, 168 (23) [M+–H2O], 148 (100) [M+–propargyl], 130 (63),
116 (49), 103 (70), 77 (67). HRMS (EI): C12H13NO calcd 186.0919,
found 186.0920.
0.72 mmol) and Et3N (100 L, 0.72 mmol). The mixture was stirred
l
at room temperature for 2 h. The solvent was evaporated and the
residue was dissolved in CH2Cl2 (10 mL). The layer organic was
washed with a saturated solution of NaCl (3 ꢂ 10 mL), dried
(Na2SO4) and evaporated, to give a yellow oil that was purified
by flash column chromatography using hexane/EtOAc/CH2Cl2
(30:1:1) as eluent to give 10a (0.065 g, yield 65%) as a clear oil.
IR
m .
= 3291, 1711, 1265, 1108, 1069, 769 cmꢀ1 1H NMR
3.2.11. ( )-cis-8a
(300 MHz, CDCl3) d = 8.10–8.07 (m, 2H, 20-H, 30-H), 7.60–7.32 (m,
7H, 40-H, 50-H, 60-H, 4 ꢂ Harom), 6.33 (t, 1H, J = 7.0 Hz, 1-H), 4.66
(t, 1H, J = 7.1 Hz, 3-H), 3.64–3.50 (AB system, 2H, J = 17.0 Hz,
CH2), 3.63–3.49 (AB system, 2H, J = 17.0 Hz, CH2), 2.93 (dt, 1H,
Mp 91–92 °C. IR
m
= 3283, 3198, 1309, 1123, 1054, 767 cmꢀ1. 1H
NMR (300 MHz, CDCl3) d = 7.50–7.31 (m, 4H, Harom), 5.02 (t, 1H,
J = 5.4 Hz, 1-H), 4.31 (t, 1H, J = 6.1 Hz, 3-H), 3.63–3.50 (AB system,
2H, J = 16.9 Hz, CH2), 3.62–3.49 (AB system, 2H, J = 16.9 Hz, CH2),
2.81 (br s, 1H, D2O exch., OH), 2.56 (dt, 1H, J = 13.7 Hz, 6.1 Hz,
J = 14.9 Hz, 7.6 Hz, 2a-H), 2.13 (dt, 1H, J = 14.1 Hz, 6.7 Hz, 2b-H),
2.24 (t, 2H, J = 2.3 Hz, 2 ꢂ CH) ppm. 13C NMR (75 MHz, CDCl3)
d = 166.69 (CO), 143.23 (C-7a), 140.97 (C-3a), 133.29 (C0-4),
130.49 (C0-1), 129.96, 129.46, 128.77, 128.62, 125.39 and 125.32
2
a
-H), 2.26 (t, 2H, J = 2.5 Hz, 2 ꢂ CH), 2.13 (dt, 1H, J = 13.4 Hz,
5.0 Hz, 2b-H) ppm. 13C NMR (75 MHz, CDCl3) d = 145.35 (C-7a),
142.03 (C-3a), 128.58, 128.30, 125.45 and 124.47 (CHarom), 79.85
(2 ꢂ C„CH), 74.01 (C-1), 73.22 (2 ꢂ C„CH), 64.41 (C-3), 39.53
(2 ꢂ CH2), 38.31 (C-2) ppm. MS (EI): m/z (%): 225 (3) [M+], 224
(4) [Mꢀ1]+, 207 (6) [M+–H2O], 186 (16) [M+–propargyl], 141 (7),
116 (88), 92 (100), 77 (30). HRMS (EI): C15H15NO calcd 225.1154;
found 225.1152.
(4 ꢂ CHarom
,
4 ꢂ C0-H), 80.65 (2 ꢂ C„CH), 76.05 (C-1), 73.08
(2 ꢂ C„CH), 65.28 (C-3), 39.31 (2 ꢂ CH2), 33.52 (C-2) ppm. MS
(FAB): m/z (%): 331 (12) [M+2]+, 330 (37) [M+1]+, 231 (59), 186
(5), 154 (92), 137 (100), 105 (34). Anal. calcd for C22H19NO2
(329.39): C 80.22, H 5.81, N 4.25; found C 80.56, H 5.45, N 4.39.
3.2.15. Compound 9b
3.2.12. ( )-trans-8b
Synthesis and characterization of ( )-trans-3-(N,N-dip-
ropargylamino)-1-indanyl acetate. The same procedure as de-
scribed for 9a was used to prepare compound 9b from 8b by
reaction with acetic anhydride, Et3N, DMAP in MeCN; yield: 67%.
Mp 64–65 °C. IR
m
= 3271, 2957, 1369, 1137, 1002, 763 cmꢀ1. 1H
NMR (300 MHz, CDCl3) d = 7.48–7.30 (m, 4H, Harom), 5.35 (dd, 1H,
J = 6.4 Hz, 3.8 Hz, 1-H), 4.73 (dd, 1H, J = 7.0 Hz, 5.1, 3-H), 3.53–
3.40 (AB system, 2H, J = 16.9 Hz, CH2), 3.52–3.39 (AB system, 2H,
Mp 70–71 °C. IR
m .
= 3284, 3252, 1720, 1243, 1134 cmꢀ1 1H NMR
J = 16.9 Hz, CH2), 2.60 (ddd, 1H, J = 14.0 Hz, 6.7 Hz, 4.8 Hz, 2
a
-H),
(300 MHz, CDCl3) d = 7.47–7.28 (m, 4H, Harom), 6.24 (d, 1H,
J = 4.4 Hz, 1-H), 4.84 (t, 1H, J = 6.4 Hz, 3-H), 3.57–3.41 (AB system,
2H, J = 16.9 Hz, CH2), 3.56–3.40 (AB system, 2H, J = 16.7 Hz, CH2),
2.24 (t, 2H, J = 2.5 Hz, 2 ꢂ CH), 2.10–2.02 (m, 1H, 2b-H), 1.9 (br s,
1H, D2O exch., OH) ppm. 13C NMR (75 MHz, CDCl3) d = 145.17 (C-
7a), 142.39 (C-3a), 128.73, 128.62, 125.75 and 124.42 (CHarom),
80.09 (2 ꢂ C„CH), 74.58 (C-1), 72.94 (2 ꢂ C„CH), 65.49 (C-3),
39.30 (2 ꢂ CH2), 37.62 (C-2) ppm. MS (EI): m/z (%): 225 (6) [M+],
224 (4) [Mꢀ1]+, 207 (6) [M+–H2O], 186 (25) [M+–propargyl], 141
(9), 116 (94), 92 (100), 77 (51). HRMS (EI): C15H15NO calcd
225.1154; found 225.1148.
2.64–2.55 (m, 1H, 2
a
-H), 2.25–2.21 (m, 3H, 2b-H, 2 ꢂ CH), 2.04
(s, 3H, CH3) ppm. 13C NMR (75 MHz, CDCl3) d = 171.01 (COCH3),
143.96 (C-7a), 140.84 (C-3a), 129.42, 128.46, 125.88 and 125.22
(CHarom), 80.18 (2 ꢂ C„CH), 76.55 (C-1), 72.91 (2 ꢂ C„CH),
66.06 (C-3), 38.98 (2 ꢂ CH2), 33.63 (C-2), 21.25 (CH3) ppm. MS
(FAB): m/z (%): 269 (2) [M+2]+, 268 (9) [M+1]+, 230 (70), 186 (5),
154 (100), 137 (97). Anal. calcd for C17H17NO2 (267.3): C 76.38, H
6.41, N 5.24; found C 76.01, H 6.83, N 5.11.
3.2.13. Compound 9a
Synthesis and characterization of ( )-cis-3-(N,N-dipropargyla-
mino)-1-indanyl acetate. A mixture of 8a (0.08 g, 0.36 mmol), ace-
3.2.16. Compound 10b
tic anhydride (69
l
L, 0.72 mmol), Et3N (100
l
L, 0.72 mmol), DMAP
Synthesis and characterization of ( )-trans-3-(N,N-dip-
ropargylamino)-1-indanyl benzoate. The same procedure as de-
scribed for 10a was used to prepare compound 10b from 8b by
reaction with benzoyl chloride, Et3N, DMAP in MeCN; yield: 65%.
(a catalytic amount) in MeCN (5 mL), under argon, was stirred at
room temperature for 3 h. The solvent was removed and the resi-
due was partitioned between EtOAc and H2O, and the organic layer
was washed with a saturated solution of NaCl (3 ꢂ 15 mL), dried
(Na2SO4) and evaporated, to give 9a (0.092 g, yield 96%) as a white
Mp 74–75 °C, IR
m .
= 3281, 3248, 1700, 1267, 1109, 763 cmꢀ1 1H
NMR (300 MHz, CDCl3) d = 8.03–8.00 (m, 2H, 20-H, 30-H), 7.57–
7.31 (m, 7H, 40-H, 50-H, 60-H, 4 ꢂ Harom), 6.51 (dd, 1H, J = 7.0 Hz,
3.2 Hz, 1-H), 4.92 (t, 1H, J = 6.7 Hz, 3-H), 3.61–3.46 (AB system,
2H, J = 16.7 Hz, CH2), 3.60–3.45 (AB system, 2H, J = 16.7 Hz, CH2),
solid. Mp 72–73 °C. IR
m .
= 3266, 2917, 1728, 1235, 1031, 775 cmꢀ1
1H NMR (300 MHz, CDCl3) d = 7.48–7.32 (m, 4H, Harom), 6.06 (t, 1H,
J = 13.5 Hz, 1-H), 4.58 (t, 1H, J = 14.5 Hz, 3-H), 3.58–3.44 (AB sys-
tem, 2H, J = 16.8 Hz, CH2), 3.57–3.43 (AB system, 2H, J = 16.8 Hz,
2.76 (dt, 1H, J = 14.4 Hz, J = 6.7 Hz,
2a-H), 2.58 (ddd, 1H,
CH2), 2.81 (dt, 1H, J = 15.2 Hz, 7.4 Hz,
2
a
-H), 2.24 (t, 2H,
J = 14.4 Hz, 7.1 Hz, 2.9 Hz, 2b-H), 2.27 (t, 2H, J = 2.3 Hz, 2 ꢂ CH)
ppm. 13C NMR (75 MHz, CDCl3) d = 166.56 (CO), 143.95 (C-7a),
141.00 (C-3a), 132.95 (C0-4), 130.29 (C0-1), 129.66, 129.46,
128.56, 128.29, 126.02 and 125.35 (4 ꢂ CHarom, 4 ꢂ C0-H), 80.17
(2 ꢂ C„CH), 77.27 (C-1), 72.98 (2 ꢂ C„CH), 66.13 (C-3), 39.07
(2 ꢂ CH2), 33.85 (C-2) ppm. MS (FAB): m/z (%): 331 (12) [M+2]+,
J = 2.5 Hz, 2 ꢂ CH), 2.19–2.13 (m, 4H, 2b-H, CH3) ppm. 13C NMR
(75 MHz, CDCl3) d = 171.01 (COCH3), 142.84 (C-7a), 140.65 (C-3a),
129.16, 128.50, 125.10 and 124.88 (CHarom), 80.39 (2 ꢂ C„CH),
75.26 (C-1), 72.81 (2 ꢂ C„CH), 64.91 (C-3), 39.03 (2 ꢂ CH2),
33.12 (C-2), 21.24 (CH3) ppm. MS (FAB): m/z (%): 268 (26)

F. Luan et al. / Bioorg. Med. Chem. 21 (2013) 1870–1879
1879
330 (48) [M+1]+, 230 (62), 186 (5), 154 (99), 137 (100), 105 (23).
Anal. calcd for C22H19NO2 (329.39): C 80.22, H 5.81, N 4.25; found
C 80.45, H 5.32, N 4.60.
23. Helguera, A. M.; Combes, R. D.; González, M. P.; Cordeiro, M. N. D. S. Curr. Top.
Med. Chem. 2008, 8, 1628.
24. Tetko, I. V.; Gasteiger, J.; Todeschini, R.; Mauri, A.; Livingstone, D.; Ertl, P.;
Palyulin, V. A.; Radchenko, E. V.; Zefirov, N. S.; Makarenko, A. S.; Tanchuk, V. Y.;
Prokopenko, V. V. J. Comput. Aided Mol. Des. 2005, 19, 453.
25. Vilar, S.; Cozza, G.; Moro, S. Curr. Top. Med. Chem. 2008, 8, 1555.
26. Marrero-Ponce, Y. C.-G. J.; Olazabal, E.; Serrano, H. S.; Morales, A.; Castañedo,
N.; Ibarra-Velarde, F.; Huesca-Guillen, A.; Jorge, E.; DelValle, A.; Torrens, F.;
Castro, E. A. J. Comput. Aided Mol. Des. 2004, 18, 615.
27. Marrero-Ponce, Y.; Casanola-Martin, G. M.; Khan, M. T.; Torrens, F.; Rescigno,
A.; Abad, C. Curr. Pharm. Des. 2010, 16, 2601.
28. Katritzky, A. R.; Perumal, S.; Petrukhin, R.; Kleinpeter, E. J. Chem. Inf. Comput.
Sci. 2001, 41, 569.
3.2.17. Biological assay of neuroprotective effects of 1,3-
rasagilines
The compounds were evaluated in this work following essen-
tially the same protocol reported in a previous work by Yañez
et al.8
29. Katritzky, A. R.; Oliferenko, A.; Lomaka, A.; Karelson, M. Bioorg. Med. Chem. Lett.
2002, 12, 3453.
Acknowledgements
30. Katritzky, A. R.; Kulshyn, O. V.; Stoyanova-Slavova, I.; Dobchev, D. A.; Kuanar,
M.; Fara, D. C.; Karelson, M. Bioorg. Med. Chem. 2006, 14, 2333.
31. González-Diaz, H.; Duardo-Sánchez, A.; Ubeira, F. M.; Prado-Prado, F.; Perez-
Montoto, L. G.; Concu, R.; Podda, G.; Shen, B. Curr. Drug Metab. 2010, 11,
379.
32. González-Diaz, H.; Romaris, F.; Duardo-Sánchez, A.; Perez-Montoto, L. G.;
Prado-Prado, F.; Patlewicz, G.; Ubeira, F. M. Curr. Pharm. Des. 2010, 16, 2737.
33. González-Diaz, H.; Prado-Prado, F.; Ubeira, F. M. Curr. Top. Med. Chem. 2008, 8,
1676.
The authors acknowledge the Portuguese Fundação para a Ciên-
cia e a Tecnologia (FCT) and the European Social Fund for financial
support (SFRH/BPD/63666/2009). Moreover, this work has been
further supported by FCT through Project PTDC/QUIQUI/113687/
2009 and Grant No. Pest-C/EQB/LA0006/2011 and Xunta de Galicia
through Project 07CSA008203PR.
34. Tenorio-Borroto, E.; PenuelasRivas, C. G.; VasquezChagoyan, J. C.; Castanedo,
N.; Prado-Prado, F. J.; García-Mera, X.; González-Diaz, H. Bioorg. Med. Chem.
2012.
Supplementary data
35. Lecht, S.; Haroutiunian, S.; Hoffman, A.; Lazarovici, P. Ther. Clin. Risk Manag.
2007, 3, 467.
36. Youdim, M. B.; Wadia, A.; Tatton, W.; Weinstock, M. Ann. N. Y. Acad. Sci. 2001,
939, 450.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmc.2013.01.035.
37. Gerets, H. H.; Dhalluin, S.; Atienzar, F. A. Methods Mol. Biol. 2011, 740, 91.
38. Patankar, S. J.; Jurs, P. C. J. Chem. Inf. Comput. Sci. 2003, 43, 885.
39. García-García, A.; Galvez, J.; de Julián-Ortiz, J. V.; García-Domenech, R.; Muñoz,
C.; Guna, R.; Borrás, R. J. Antimicrob. Chemother. 2004, 53, 65.
40. Marrero-Ponce, Y.; Castillo-Garit, J. A.; Olazabal, E.; Serrano, H. S.; Morales, A.;
Castanedo, N.; Ibarra-Velarde, F.; Huesca-Guillen, A.; Sánchez, A. M.; Torrens,
F.; Castro, E. A. Bioorg. Med. Chem. 2005, 13, 1005.
References and notes
1. Liu, R.; Yuan, H.; Yuan, F.; Yang, S. H. Neurol. Res. 2012, 34, 331.
2. Xing, C.; Hayakawa, K.; Lok, J.; Arai, K.; Lo, E. H. Neurol. Res. 2012, 34, 325.
3. Howells, D. W.; Sena, E. S.; O’Collins, V.; Macleod, M. R. Int. J. Stroke 2012, 7,
371.
41. Marrero-Ponce, Y.; Machado-Tugores, Y.; Pereira, D. M.; Escario, J. A.; Barrio, A.
G.; Nogal-Ruiz, J. J.; Ochoa, C.; Aran, V. J.; Martinez-Fernández, A. R.; Sánchez, R.
N.; Montero-Torres, A.; Torrens, F.; Meneses-Marcel, A. Curr. Drug Discov.
Technol. 2005, 2, 245.
42. Casanola-Martin, G. M.; Marrero-Ponce, Y.; Khan, M. T.; Ather, A.; Sultan, S.;
Torrens, F.; Rotondo, R. Bioorg. Med. Chem. 2007, 15, 1483.
43. Casanola-Martin, G. M.; Marrero-Ponce, Y.; TareqHassanKhan, M.; Torrens, F.;
Perez-Gimenez, F.; Rescigno, A. J. Biomol. Screening 2008, 13, 1014.
44. Casanola-Martin, G. M.; Marrero-Ponce, Y.; Khan, M. T.; Khan, S. B.; Torrens,
F.; Perez-Jimenez, F.; Rescigno, A.; Abad, C. Chem. Biol. Drug Des. 2010, 76,
538.
45. Rodríguez-Soca, Y.; Munteanu, C. R.; Dorado, J.; Pazos, A.; Prado-Prado, F. J.;
González-Díaz, H. J. Proteome Res. 2010, 9, 1182.
46. González-Díaz, H.; Muiño, L.; Anadón, A. M.; Romaris, F.; Prado-Prado, F. J.;
Munteanu, C. R.; Dorado, J.; PazosSierra, A.; Mezo, M.; González-Warleta, M.;
Gárate, T.; Ubeira, F. M. Mol. Biosyst. 1938, 2011, 7.
4. Xing, C.; Arai, K.; Lo, E. H.; Hommel, M. Int. J. Stroke 2012, 7, 378.
5. Capettini, L. S.; Savergnini, S. Q.; da Silva, R. F.; Stergiopulos, N.; Santos, R. A.;
Mach, F.; Montecucco, F. Mediators Inflamm. 2012, 2012, 824093.
6. Dudas, B.; Semeniken, K. Handb. Exp. Pharmacol. 2012, 325.
7. Allegri, R. F.; Guekht, A. Drugs Today 2012, 48, 25.
8. Yañez, M.; Galán, L.; Matias-Guiu, J.; Vela, A.; Guerrero, A.; García, A. G. Brain
Res. 2011, 1423, 77.
9. Gaulton, A.; Bellis, L. J.; Bento, A. P.; Chambers, J.; Davies, M.; Hersey, A.; Light,
Y.; McGlinchey, S.; Michalovich, D.; Al-Lazikani, B.; Overington, J. P. Nucleic
Acids Res. 2012, 40, D1100.
10. Nurisso, A.; Simoes-Pires, C.; Martel, S.; Cressend, D.; Guillot, A.; Carrupt, P. A.
Chimia (Aarau) 2012, 66, 286.
11. Mok, N. Y.; Brenk, R. J. Chem. Inf. Model. 2011, 51, 2449.
12. Riera-Fernández, I.; Martín-Romalde, R.; Prado-Prado, F. J.; Escobar, M.;
Munteanu, C. R.; Concu, R.; Duardo-Sánchez, A.; González-Díaz, H. Curr. Top.
Med. Chem. 2012.
47. Martínez-Romero, M.; Vázquez-Naya, J. M.; Rabuñal, J. R.; Pita-Fernández, S.;
Macenlle, R.; Castro-Alvariño, J.; López-Roses, L.; Ulla, J. L.; Martínez-Calvo, A.
V.; Vázquez, S.; Pereira, J.; Porto-Pazos, A. B.; Dorado, J.; Pazos, A.; Munteanu, C.
R. Curr. Drug Metab. 2010, 11, 347.
48. Nuritova, F.; Frenguelli, B. G. Neuroscience 2012, 222, 159.
49. Heikamp, K.; Bajorath, J. J. Chem. Inf. Model. 1831, 2011, 51.
50. StatSoft. Inc., 2002.
51. Guillon, J.; Hebert, G.; Dallemagne, P.; Leger, J. M.; Vidaillac, C.; The, C.;
Lisowski, V.; Rault, S.; Demotes-Mainard, J.; Jarry, C. J. Enzyme Inhib. Med. Chem.
2003, 18, 147.
13. Prado-Prado, F.; García-Mera, X.; Escobar, M.; Sobarzo-Sánchez, E.; Yañez, M.;
Riera-Fernández, P.; González-Díaz, H. Eur. J. Med. Chem. 2011, 46, 5838.
14. García, I.; Fall, Y.; Gómez, G.; González-Díaz, H. Mol. Diversity 2011, 15, 561.
15. Hill, T.; Lewicki, P. STATISTICS Methods and Applications. A Comprehensive
Reference for Science, Industry and Data Mining; Tulsa: StatSoft, 2006.
16. Concu, R.; Dea-Ayuela, M. A.; Pérez-Montoto, L. G.; Prado-Prado, F. J.; Uriarte,
E.; Bolás-Fernández, F.; Podda, G.; Pazos, A.; Munteanu, C. R.; Ubeira, F. M.;
González-Díaz, H. BBA-Proteins Proteomics 2009, 1794, 1784.
17. Speck-Planche, A.; Kleandrova, V. V.; Luan, F.; Cordeiro, M. N. D. S. Mol. BioSyst.
2012, 8, 2188.
52. Gordaliza, M.; Dallemagne, P.; Rault, S.; Robba, M. An. Quim. 1989, 85, 66.
53. Youdim, M. B. H.; Finberg, J. P. M.; Levy, R.; Sterling, J.; Lerner, D.; Berger-
Paskin, T.; Yellin, F. U.S. Patent 1995, U.S. 5453446.
54. Herzig, Y.; Lerman, L.; Goldenberg, W.; Lerner, D.; Gottlieb, H. E.; Nudelman, A.
J. Org. Chem. 2006, 71, 4130.
55. Sukeda, M.; Ichikawa, S.; Matsuda, A.; Shuto, S. J. Org. Chem. 2003, 68, 3465.
56. González-Díaz, H.; Prado-Prado, F.; García-Mera, X.; Alonso, N.; Abeijón, P.;
Caamaño, O.; Yañez, M.; Munteanu, C. R.; Pazos, A.; Dea-Ayuela, M. A.; Gómez-
Muñoz, M. T.; Garijo, M. M.; Sansano, J.; Ubeira, F. M. J. Proteome Res. 2011, 10,
1698.
18. Speck-Planche, A.; Kleandrova, V. V.; Luan, F.; Cordeiro, M. N. D. S. Eur. J. Pharm.
Sci. 2012, 47, 273.
19. Speck-Planche, A.; Kleandrova, V. V.; Luan, F.; Cordeiro, M. N. D. S. Comb. Chem.
High Throughput Screening 2012, 15, 666.
20. Speck-Planche, A.; Kleandrova, V. V.; Luan, F.; Cordeiro, M. N. D. S. Ecotoxicol.
Environ. Saf. 2012, 80, 308.
21. Speck-Planche, A.; Kleandrova, V. V.; Luan, F.; Cordeiro, M. N. D. S. Bioorg. Med.
Chem. 2012, 20, 4848.
22. Speck-Planche, A.; Kleandrova, V. V.; Luan, F.; Cordeiro, M. N. D. S. Bioorg. Med.
Chem. 2011, 19, 6239.