New insights into the biological properties of Crocus sativus L.: Chemical modifications, human monoamine oxidases inhibition and molecular modeling studies

Article abstract of DOI:10.1016/j.ejmech.2014.05.048

Although there are clinical trials and in vivo studies in literature regarding the anxiolytic and antidepressant activities of the components of Crocus sativus L., their effects on the human monoamine oxidases (hMAO-A and hMAO-B), enzymes which are involved in mental disorders and neurodegenerative diseases, have not yet been investigated. We have thus examined the hMAO inhibitory activities of crocin and safranal (the most important active principles in saffron) and, subsequently, designed a series of safranal derivatives to evaluate which chemical modifications confer enhanced inhibition of the hMAO isoforms. Docking simulations were performed in order to identify key molecular recognitions of these inhibitors with both isoforms of hMAO. In this regard, different mechanisms of action were revealed. This study concludes that safranal and crocin represent useful leads for the discovery of novel hMAO inhibitors for the clinical management of psychiatric and neurodegenerative disorders.

Full text of DOI:10.1016/j.ejmech.2014.05.048

European Journal of Medicinal Chemistry 82 (2014) 164e171
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
New insights into the biological properties of Crocus sativus L.:
chemical modifications, human monoamine oxidases inhibition and
molecular modeling studies
,
Celeste De Monte a, Simone Carradori a *, Paola Chimenti a, Daniela Secci a,
,
d
b
c
c
Luisa Mannina a , Francesca Alcaro , Anel Petzer , Clarina I. N'Da ,
ꢀ
b
b
b
c, *
ꢁ
Maria Concetta Gidaro , Giosue Costa , Stefano Alcaro , Jacobus P. Petzer
a Dipartimento di Chimica e Tecnologie del Farmaco, Sapienza University of Rome, P.le A. Moro 5, 00185 Rome, Italy
b Dipartimento di Scienze della Salute, “Magna Graecia” University of Catanzaro, Campus Universitario “S. Venuta”, Viale Europa Loc. Germaneto, 88100
Catanzaro, Italy
c Pharmaceutical Chemistry and Centre of Excellence for Pharmaceutical Sciences, School of Pharmacy, North-West University, Private Bag X6001,
Potchefstroom 2520, South Africa
d Istituto di Metodologie Chimiche, Laboratorio di Risonanza Magnetica “Annalaura Segre”, CNR, via Salaria km 29.300, 00015 Monterotondo, Rome, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Although there are clinical trials and in vivo studies in literature regarding the anxiolytic and antide-
pressant activities of the components of Crocus sativus L., their effects on the human monoamine oxi-
dases (hMAO-A and hMAO-B), enzymes which are involved in mental disorders and neurodegenerative
diseases, have not yet been investigated. We have thus examined the hMAO inhibitory activities of crocin
and safranal (the most important active principles in saffron) and, subsequently, designed a series of
safranal derivatives to evaluate which chemical modifications confer enhanced inhibition of the hMAO
isoforms. Docking simulations were performed in order to identify key molecular recognitions of these
inhibitors with both isoforms of hMAO. In this regard, different mechanisms of action were revealed. This
study concludes that safranal and crocin represent useful leads for the discovery of novel hMAO in-
hibitors for the clinical management of psychiatric and neurodegenerative disorders.
Received 29 January 2014
Received in revised form
13 May 2014
Accepted 21 May 2014
Available online 22 May 2014
Keywords:
Crocin
Crocus sativus L.
hMAO inhibitors
Depression
© 2014 Elsevier Masson SAS. All rights reserved.
Neurodegenerative disorders
Safranal
1. Introduction
extracts has revealed the presence of more than 150 components in
the stigmas such as crocetin (a carotenoid derivative), crocins (di-
Traditional herbal medicines have received significant attention
for the treatment of specific illnesses, although in many instances
limited evidence exist for their mechanisms of action [1]. In
particular, a better understanding of the mechanisms of action of
these components would greatly facilitate the possibility of
discovering new therapeutic agents with fewer side effects, spe-
cifically for the treatment of mental and neurodegenerative dis-
eases [2].
glycosidic esters of crocetin), picrocrocin (responsible for the bitter
taste), safranal (the volatile oil responsible for the saffron aroma),
anthocyanins, flavonoids, vitamins, amino acids, proteins, starch,
mineral substances and gums [3]. With regard to their pharmaco-
logical properties, crocin (crocetin di-gentiobiose ester) was found
to enhance the memory and learning abilities in ethanol-induced
learning behavior impaired mice and rats [4]. Saffron, in turn,
was found to be effective in reducing the effects of aluminum-
toxicity in adult mice [5]. Components of saffron extracts and cro-
cetin have been shown to bind to the PCP (phenylcyclidine) binding
Crocus sativus L. (saffron spice) has attracted much attention in
the field of medicine, especially with regard to the discovery of
psychoactive agents. In fact, chemical analysis of C. sativus L.
site of the NMDA receptor as well as to the s1 receptor. Since s
receptor ligands could inhibit the ischemic-induced presynaptic
release of excitotoxic amino acids, these saffron extracts may play a
role in the management of mental disorders [6].
Moreover, in a double-blind and randomized trial, depressed
outpatients were randomly assigned to receive extracts of C. sativus
* Corresponding authors.
E-mail addresses: simone.carradori@uniroma1.it (S. Carradori), jacques.petzer@
nwu.ac.za (J.P. Petzer).
http://dx.doi.org/10.1016/j.ejmech.2014.05.048
0223-5234/© 2014 Elsevier Masson SAS. All rights reserved.

C. De Monte et al. / European Journal of Medicinal Chemistry 82 (2014) 164e171
165
Scheme 1. Structures of bioactive compounds (crocin and safranal) of Crocus sativus L. and synthesis of safranal derivatives 3e9. Reagents and conditions: (a) EtOH, acetic acid (cat.),
rt; (b) (if X ¼ S and R ¼ R1 ¼ H) MeOH, CH3COONa, rt; (c) dry acetone, anhydrous K2CO3, rt; (d) (if X ¼ S and R ¼ R1 ¼ H) EtOH, rt.
L. or fluoxetine and it was found that the efficacies of these treat-
ments are comparable [7]. Similarly, in 6-week double-blind and
randomized trials, the effect of a hydro-alcoholic extract of stigmas
of C. sativus L. was comparable to those of fluoxetine [8] and
imipramine [9] in the treatment of mild to moderate depression.
Another clinical study that supported the efficacy of saffron active
principles as antidepressant agents was a 6-week double-blind,
placebo-controlled and randomized trial: at the end of the study
patients receiving ethanolic extracts of C. sativus L. gave improved
outcomes on the Hamilton Depression Rating Scale [10]. It is
particularly noteworthy that treatment with saffron extracts is not
associated with sexual dysfunction in men and women, a side effect
often encountered with antidepressant drugs [11,12]. Evidence
suggests that both safranal and crocin contribute to the antide-
pressant effect of C. sativus L. extracts in mice, probably via the
inhibition of dopamine and norepinephrine re-uptake by crocin,
and the inhibition of serotonin reuptake by safranal [13]. Further-
more, saffron and its active constituents provide neuroprotection in
a rat model of Parkinson's disease (PD), possess antioxidant effects
in an in vitro model of Alzheimer's disease (AD), anxiolytic activity
in rodents [14] and also neuroprotective effects in animal models of
other neurodegenerative conditions such as the retinal dystrophies
[15]. The ethanolic extract of saffron stigmas also exhibits antioxi-
study evaluated new AChE inhibitors which were derived from
saffron [17].
Although there are many studies and clinical trials reported in
literature regarding the anxiolytic and antidepressant effects of
safranal and crocin [2], no detailed pharmacological studies are
available to explain the mechanism of action of saffron in the
treatment of depression. We are particularly interested in the
possibility that saffron components may modulate the activities of
the human monoamine oxidase (hMAO) enzymes, which are tar-
gets for the treatment of both depression and neurodegeneration.
hMAO-A and hMAO-B catalyze the oxidative deamination of
endogenous and dietary amines: the hMAO-A isoform is mainly
expressed in catecholaminergic neurons and regulates the meta-
bolism of serotonin, epinephrine and norepinephrine, while
hMAO-B mostly metabolizes benzylamine and phenethylamine
and is expressed in those areas, such as the substantia nigra, where
dopamine neurons are predominant. Both hMAO isoforms are
involved in the metabolism of dopamine [18,19]. The excessive
activity of hMAO-A could lead to psychiatric disorders such as
depression and dysthymia, while hMAO-B overexpression in-
creases with age and plays a pivotal role in neurodegenerative
diseases such as PD and AD [20,21]. Therefore, in our continued
interest in the discovery of new leads/scaffolds for the design of
hMAO inhibitors [22e25], we first assessed the hMAO inhibitory
activities of two main active principles of saffron (Scheme 1): crocin
(1) and safranal (2). These two compounds are characterized by
dant properties by inhibiting amyloid ß-peptide (Aß) fibril forma-
tion and deposition in the human brain, which characterize AD [16].
Even with the limitations of cholinesterase inhibitors, a recent

166
C. De Monte et al. / European Journal of Medicinal Chemistry 82 (2014) 164e171
opposite physicochemical properties, are present in different
amounts in the plant, and are usually reported as analytical/phar-
macological markers. We also designed and synthesized a new
series of safranal derivatives (3e9, Scheme 1) in order to establish
which chemical modifications of the safranal unsaturated nucleus
could enhance inhibition potency towards hMAO-A and/or
hMAO-B.
2.1.4. 1-((2,6,6-Trimethylcyclohexa-1,3-dienyl)methylene)
thiosemicarbazide (4)
Thiosemicarbazide (1 eq) was added to a stirring solution of
safranal (1 eq) in ethanol (50 mL) at room temperature. After 48 h,
the reaction mixture was filtered and the solid was washed with n-
hexane to give product 4 as a yellow powder (mp 188e192 ꢀC, 85%
yield); 1H NMR (400 MHz, CDCl3):
CH3), 2.16 (s, 2H, CH2), 5.91 (d, J ¼ 9.6 Hz, 1H, CH]), 5.95e5.99 (m,
1H, CH]), 6.29 (bs, 1H, NH2, D2O exch.), 6.95 (bs, 1H, NH2, D2O
exch.), 7.89 (s, 1H, CH]), 9.51 (bs, 1H, NH, D2O exch.); 13C NMR
d
1.20 (s, 6H, 2 ꢂ CH3), 1.98 (s, 3H,
2. Materials and methods
2.1. Chemistry
(101 MHz, DMSO-d6)
d
19.38 (CH3), 27.16 (2 ꢂ CH3), 33.41 (C),
128.99 (CH]), 130.14 (CH]), 133.60 (]CCH3), 135.61 (C]), 143.95
(C]N), 177.58 (C]S), (CH2 signal missing due to overlap with
Commercial samples of crocin (crocin-1, crocetin di-gentiobiose
ester) and safranal (>88%) were purchased from SigmaeAldrich
(Milan, Italy). Safranal was further purified by column chroma-
tography on silica gel (high purity grade, pore size 60 Å, 230e400
mesh particle size) using ethyl acetate:n-hexane (1:3) as the eluant.
1H NMR and IR spectra of the purified product were in agreement
with those reported in the literature. The chemicals, solvents for
synthesis, and spectral grade solvents were purchased from Sig-
maeAldrich and were used without further purification. All
melting points were measured on a Stuart® melting point appa-
ratus SMP1, and are uncorrected. 1H and 13C NMR spectra were
recorded, respectively, at 400 and 101 MHz on a Bruker spec-
trometer using CDCl3, CD3OD and DMSO-d6 as the solvents.
DMSO-d6); IR (neat): cmꢁ1 3427 (
y
NeH), 3152 (
y
¼ CeH), 1542
(yC]N), 1289 (yC]S). Anal. Calcd. for C11H17N3S: C, 59.16; H, 7.67;
N, 18.81. Found: C, 58.87; H, 7.45; N, 19.07.
2.1.5. 2-Methyl-1-((2,6,6-trimethylcyclohexa-1,3-dienyl)
methylene)thiosemicarbazide (5)
2-Methylthiosemicarbazide (1 eq) was added to a stirring so-
lution of safranal (1 eq) in ethanol (50 mL) with catalytic amounts
of acetic acid at room temperature. After 48 h, the reaction mixture
was filtered, the liquid was concentrated in vacuo and the obtained
solid was filtered and purified by column chromatography (ethyl
acetate:light petroleum, 1:2) to give product 5 as a yellow powder
Chemical shifts are expressed as
d
units (parts per millions) relative
(mp 108e110 ꢀC, 77% yield); 1H NMR (400 MHz, DMSO-d6):
d 1.15 (s,
to the solvent signal. Coupling constants J are valued in Hertz (Hz).
IR spectra were recorded on a FT-IR PerkineElmer SpectrumOne
equipped with ATR system. Elemental analyses for C, H, and N were
recorded on a PerkineElmer 240 B microanalyzer and the analytical
results were within 0.4% of the theoretical values for all com-
pounds. All reactions were monitored by TLC performed on 0.2 mm
thick silica gel plates (60 F254, Merck).
6H, 2 ꢂ CH3), 1.96 (s, 3H, CH3), 2.11 (s, 2H, CH2), 3.72 (s, 3H, CH3),
5.95 (bs, 2H, 2 ꢂ CH]), 7.39 (bs, 1H, NH2, D2O exch.), 7.64 (s, 1H,
CH]), 8.38 (bs, 1H, NH2, D2O exch.); 13C NMR (101 MHz, DMSO-d6)
d
19.66 (CH3), 27.10 (2 ꢂ CH3), 32.59 (C), 33.53 (NCH3), 128.87
(CH]), 130.24 (CH]), 133.59 (]CCH3), 135.16 (C]), 141.53 (C]N),
180.63 (C]S), (CH2 signal missing due to overlap with DMSO-d6);
IR (neat): cmꢁ1 3429 (
y
NeH), 3235 (
y
¼ CeH), 1571 (
yC]N), 1356 (y
CeN). Anal. Calcd. for C12H19N3S: C, 60.72; H, 8.07; N, 17.70. Found:
2.1.1. Bis[(2S,3R,4S,5S,6R)-3,4,5-trihydroxy-6-({[(2R,3R,4S,5S,6R)-
3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl]oxy}
methyl)tetrahydro-2H-pyran-2-yl] (2E,4E,6E,8E,10E,12E,14E)-
2,6,11,15-tetramethyl-2,4,6,8,10,12,14-hexadecaheptaenedioate (1)
C, 61.03; H, 7.79; N, 18.01.
2.1.6. 1-((2,6,6-Trimethylcyclohexa-1,3-dienyl)methylene)-4-
phenylthiosemicarbazide (6)
Red powder; mp 186 ꢀC; IR: cmꢁ1 3240 (
1014 ( CeO).
y
OeH), 1694 (
y
C]O),
4-Phenylthiosemicarbazide (1 eq) was added to a stirring solu-
tion of safranal (1 eq) in ethanol (50 mL) with catalytic amounts of
acetic acid at room temperature. After 48 h, the reaction mixture
was filtered, the solid was washed with n-hexane and light petro-
leum and purified by column chromatography (ethyl acetate:light
petroleum, 1:2) to give product 6 as a yellow powder (mp
y
2.1.2. 2,6,6-Trimethylcyclohexa-1,3-dienecarbaldehyde (2)
Light yellow oil; 1H NMR (400 MHz, CD3OD):
1.19 (s, 6H,
d
2 ꢂ CH3), 2.16e2.18 (m, 2H, CH2), 2.19 (s, 3H, CH3), 5.99e6.03 (m,
1H, CH]), 6.20e6.25 (m, 1H, CH]), 10.12 (s, 1H, CHO); IR (neat):
161e163 ꢀC, 80% yield); 1H NMR (400 MHz, DMSO-d6):
d 1.18 (s, 6H,
cmꢁ1 2865 and 2749 (
y
CeH), 1661 (
y
C]O). Anal. Calcd. for
2 ꢂ CH3), 1.97 (s, 3H, CH3), 2.11 (s, 2H, CH2), 5.97 (bs, 2H, 2 ꢂ CH]),
7.18 (t, J ¼ 7.2 Hz, 1H, Ar), 7.36 (t, J ¼ 7.8 Hz, 2H, Ar), 7.60 (d,
J ¼ 8.4 Hz, 2H, Ar) 8.22 (bs, 1H, NH, D2O exch.), 9.37 (s, 1H, CH]),
C
10H14O: C, 79.96; H, 9.39. Found: C, 79.59; H, 9.01.
2.1.3. 1-((2,6,6-Trimethylcyclohexa-1,3-dienyl)methylene)
semicarbazide (3)
11.54 (bs, 1H, NH, D2O exch.); 13C NMR (101 MHz, DMSO-d6)
d 19.63
(CH3), 27.23 (2 ꢂ CH3), 33.44 (C), 124.68 (CH-benzene), 125.44
(2 ꢂ CH-benzene), 128.72 (CH]), 129.36 (CH]), 130.22 (2 ꢂ CH-
benzene),133.46 (]CCH3),136.24 (C-benzene),139.28 (C]),144.42
(C]N), 175.15 (C]S), (CH2 signal missing due to overlap with
Semicarbazide hydrochloride (1 eq) was solubilized in the
minimum amount of water in presence of sodium acetate (1 eq)
and added to a stirring solution of safranal (1 eq) in ethanol (50 mL)
at room temperature. After 48 h, the reaction mixture was filtered
and the solid was purified by column chromatography (ethyl ace-
tate:n-hexane, 1:1) to obtain product 3 as a white powder (mp
DMSO-d6); IR (neat): cmꢁ1 3306 (
C]N), 1195 (
y
NeH), 3149 (
y
¼ CeH), 1552
(y
yC]S). Anal. Calcd. for C17H21N3S: C, 68.19; H, 7.07;
N, 14.03. Found: C, 67.96; H, 7.31; N, 14.29.
170e175 ꢀC, 77% yield); 1H NMR (400 MHz, CDCl3):
d 1.19 (s, 6H,
2 ꢂ CH3), 1.96 (s, 3H, CH3), 2.15 (s, 2H, CH2), 5.85e5.94 (m, 2H,
2 ꢂ CH]), 6.96 (bs, 1H, NH2, D2O exch.), 7.67 (s, 1H, CH]), 7.89 (bs,
1H, NH2, D2O exch.), 8.12 (bs, 1H, NH, D2O exch.); 13C NMR
2.1.7. 2-(2-((2,6,6-Trimethylcyclohexa-1,3-dienyl)methylene)
hydrazono)thiazolidin-4-one (7)
The intermediate thiosemicarbazone 4 (1 eq) reacted with
ethyl-bromoacetate (1 eq), in methanol (50 mL) and sodium acetate
(1 eq) at room temperature under magnetic stirring for 24 h. The
reaction mixture was poured on ice, filtered and the solid was
washed with light petroleum to give compound 7 as an orange
powder (mp 146e151 ꢀC, 81% yield); 1H NMR (400 MHz, DMSO-d6):
(101 MHz, CDCl3)
(CH2), 128.27 (CH]), 129.76 (CH]), 133.17 (]CCH3), 134.51 (C]),
141.69 (C]N), 158.24 (C]O); IR (neat): cmꢁ1 3437 (
NeH), 1683
C]O), 1666 ( C]N). Anal. Calcd. for C11H17N3O: C, 63.74; H, 8.27;
N, 20.27. Found: C, 64.02; H, 8.49; N, 20.49.
d
19.18 (CH3), 26.90 (2 ꢂ CH3), 33.49 (C), 40.64
y
(y
y

C. De Monte et al. / European Journal of Medicinal Chemistry 82 (2014) 164e171
167
d
1.15 (s, 6H, 2 ꢂ CH3), 1.94 (s, 3H, CH3), 2.11 (s, 2H, CH2), 3.85 (s, 2H,
kynuramine while for the for hMAO-B activity measurements the
kynuramine concentration was 30 M. The reactions were initiated
CH2-thiazolidinone), 5.97e5.98 (m, 2H, 2 ꢂ CH]), 8.30 (s, 1H,
m
CH]), 11.87 (bs, 1H, NH, D2O exch.); 13C NMR (101 MHz, DMSO-d6)
with the addition of hMAO-A or hMAO-B (0.0075 mg protein/mL)
d
19.22 (CH3), 27.09 (2 ꢂ CH3), 33.56 (C), 33.67 (CH2-thiazolidinone)
and, following a 20 min incubation at 37 ꢀC, the reactions were
129.72 (2 ꢂ CH]), 130.17 (]CCH3), 134.16 (C]), 137.08 (C]N),
terminated by the addition of 400 mL NaOH (2 N) and 1000 mL
155.58 (C]N-thiazolidinone), 174.98 (C]O), (CH2 signal missing
water. The concentrations of 4-hydroxyquinoline were subse-
quently measured spectrofluorometrically at excitation and emis-
sion wavelengths of 310 nm and 400 nm, respectively (Varian Cary
Eclipse fluorescence spectrophotometer). Quantitative estimations
of 4-hydroxyquinoline were made with a linear calibration curve
due to overlap with DMSO-d6); IR (neat): cmꢁ1 3448 (
y
NeH), 1717
(yC]O), 1622 (yC]N), 1243 (yC]S). Anal. Calcd. for C13H17N3OS: C,
59.29; H, 6.51; N, 15.96. Found: C, 59.57; H, 6.24; N, 16.19.
2.1.8. 3-Benzyl-2-(2-((2,6,6-trimethylcyclohexa-1,3-dienyl)
constructed with known amounts (0.047e1.56 mM) of the authentic
methylene)hydrazono)thiazolidin-4-one (8)
metabolite. The necessary control samples were included to
confirm that the test inhibitors do not fluoresce or quench the
fluorescence of 4-hydroxyquinoline under these assay conditions.
IC50 values were calculated by plotting the initial rate of kynur-
amine oxidation versus the logarithm of the inhibitor concentra-
tion to obtain a sigmoidal doseeresponse curve. For this purpose,
the kinetic data were fitted to the one site competition model
incorporated into the Prism 5 software package (GraphPad). The
IC50 values are reported as the mean standard deviation (SD) of
triplicate determinations (Table 1).
The thiazolidinone 7 (1 eq) was suspended in 50 mL of anhy-
drous acetone in the presence of anhydrous potassium carbonate
(1 eq), and reacted with equimolar amounts of benzyl bromide for
48 h at room temperature. The reaction mixture was poured on ice
and extracted with chloroform (3 ꢂ 50 mL). The organics were
reunited, dried over anhydrous sodium sulfate, concentrated in
vacuo and purified by column chromatography (ethyl acetate:n-
hexane, 1:2) to give compound 8 as an orange oil (65% yield); 1H
NMR (400 MHz, DMSO-d6):
d
1.18 (s, 6H, 2 ꢂ CH3), 1.96 (s, 3H, CH3),
2.11e2.12 (m, 2H, CH2), 4.02 (s, 2H, CH2-thiazolidinone), 4.90 (s, 2H,
CH2-benzyl), 5.97e6.00 (m, 2H, 2 ꢂ CH]), 7.29e7.36 (m, 5H, Ar),
2.3. In silico methods
8.36 (s, 1H CH]); 13C NMR (101 MHz, CDCl3)
d 19.45 (CH3), 26.90
(2 ꢂ CH3), 32.43 (C), 33.74 (CH2-thiazolidinone), 40.98 (CH2), 46.56
(CH2), 127.89 (2 ꢂ CH]), 128.48 (2 ꢂ CH-benzene), 128.84 (2 ꢂ CH-
benzene), 129.72 (]CCH3), 129.98 (C]), 134.22 (CH-benzene),
135.77 (C-benzene), 137.80 (C]N), 157.86 (C]N-thiazolidinone),
The X-ray coordinates of hMAO-A and hMAO-B, in complex with
harmine and safinamide, respectively, were extracted from the PDB
[26], codes 2Z5X [27] and 2V5Z [28]. The protein structures were
prepared using the Protein Preparation Wizard [29] of the graphical
user interface of Maestro 9.4 [30]. In silico experiments were per-
formed with selected ligands, among which are compounds which
exhibited potent hMAO inhibition (Table 1). The compounds
selected for the in silico study were: crocin 1, semicarbazone 3,
thiosemicarbazone 4 and hydrazothiazole 9. In particular, the
modeling study proved to be useful to establish that crocin 1 and
semicarbazone 3, which possess similar IC50 values for the inhibi-
tion of the hMAOs while having very different structures, bind to
distinct sites on the hMAO proteins. In addition, the in silico study
also provided insight into how the isosteric substitution S / O
causes a significant increase of the hMAO inhibitory activity of the
thiosemicarbazone 4 compared to the semicarbazone 3. To study
the interactions of the hMAO isoforms with the test ligands, all
water molecules and non-protein residues (except for the FAD co-
factor) were removed, hydrogen atoms were added, and finally,
energy minimization was performed until the RMSD of all heavy
atoms was within 0.3 Å of the original PDB model.
172.14 (C]O); IR (neat): cmꢁ1 1722 (
yC]O), 1601 (yC]N). Anal.
Calcd. for C20H23N3OS: C, 67.96; H, 6.56; N,11.89. Found: C, 68.20; H,
6.81; N, 11.53.
2.1.9. 1-(4-(3-Nitrophenyl)thiazol-2-yl)-2-((2,6,6-
trimethylcyclohexa-1,3-dienyl)methylene)hydrazine (9)
The intermediate thiosemicarbazone 4 (1 eq) reacted with 2-
bromo-3’-nitroacetophenone (1 eq) in ethanol (50 mL) at room
temperature under magnetic stirring for 24 h. The reaction mixture
was concentrated in vacuo, and the resulting solid was purified by
column chromatography (ethyl acetate:light petroleum,1:2) to give
product 9 as a brown powder (mp 120e125 ꢀC, 87% yield); 1H NMR
(400 MHz, DMSO-d6):
d 1.09 (s, 3H, CH3), 1.21 (s, 3H, CH3), 1.93 (s,
3H, CH3), 2.10e2.12 (m, 2H, CH2), 5.92e5.94 (m, 2H, 2 ꢂ CH]), 7.59
(s, 1H, C5H-thiazole), 7.70e7.80 (m, 1H, Ar), 8.04 (s, 1H, Ar),
8.13e8.22 (m, 1H, Ar), 8.27e8.33 (m, 1H, Ar), 8.66 (s, 1H, CH]),
11.91 (bs, 1H, NH, D2O exch.); 13C NMR (101 MHz, DMSO-d6)
d 15.59
(CH3), 27.16 (2 ꢂ CH3), 33.52 (C), 107.02 (C5H-thiazole), 120.53 (CH-
benzene), 122.15 (CH-benzene), 130.51 (2 ꢂ CH]), 130.58 (CH-
benzene), 131.95 (CH-benzene), 133.43 (C-benzene), 139.62 (]
CCH3), 141.50 (C]), 145.25 (C]N), 148.67 (C4-thiazole), 148.89 (C-
benzene), 170.03 (C2-thiazole), (CH2 signal missing due to overlap
with DMSO-d6); IR (neat): cmꢁ1 1531 (yas NeO), 1347 (ys NeO).
Anal. Calcd. for C19H20N4O2S: C, 61.94; H, 5.47; N, 15.21. Found: C,
62.22; H, 5.19; N, 14.98.
Docking studies were carried out using the Glide [31] SP soft-
ware. The maximum number of poses for ligands in the receptor
cavity was set to 10. The experiments employed full ligand flexi-
bility in the 3D structures of rigid receptor active sites. The default
settings of Glide 5.9 were used for the remaining parameters. For
each docking run the binding site was defined as a cube of 25 Å.
This cube was centered on the former positions of the removed co-
Table 1
2.2. hMAO inhibition studies
The IC50 values for the inhibition of hMAO-A and hMAO-B by compounds 1e9.
Compound
IC50 MAO-A (
m
M)
IC50 MAO-B (
128 10.9
mM)
As enzyme sources, commercially available (SigmaeAldrich,
Milan, Italy) microsomes from insect cells, containing recombinant
hMAO-A and hMAO-B, were employed. The enzyme activity mea-
surements were based on the hMAO-A or hMAO-B catalyzed
oxidation of kynuramine to 4-hydroxyquinoline. The incubations
were conducted in 500 mL potassium phosphate buffer (100 mM,
pH 7.4, made isotonic with KCl 20.2 mM) in the presence of various
concentrations of the test inhibitors and 4% DMSO as the cosolvent.
1 (crocin)
2 (safranal)
170 22.4
No inhibition
164 12.8
56.8 2.59
93.2 3.07
19.5 2.65
132 8.59
86.6 5.25
9.93 1.51
No inhibition
111 6.26
74.5 5.68
117 12.7
224 30.7
168 6.90
215 16.5
0.100 0.010
3
4
5
6
7
8
9
For hMAO-A activity measurements the reactions contained 45 mM

168
C. De Monte et al. / European Journal of Medicinal Chemistry 82 (2014) 164e171
crystallized ligands in the active sites. The natural (1) and semi-
synthetic compounds (3, 4 and 9) were prepared by the LigPrep
[32] module, taking into account additional protonated and
tautomeric forms calculated at pH 5e9. In order to verify the
docking precision of Glide SP, harmine and safinamide were
redocked into the hMAO active sites. The docking protocol
possessed the ability to reproduce the binding modes of harmine
and safinamide, with a very low RMSD values between the ligand
position in the experimental X-ray structure and the best docked
pose (harmine: 0.08 Å; safinamide: 0.16 Å). After each docking
simulation, a positive result was gained when the G-score of the
ligand was better than the respective X-ray G-score, with a toler-
ance of 10%.
neuropharmacological actions of saffron extracts. Safranal (2), the
major component of the volatile fraction of saffron, did not show
any inhibition of the hMAO isoforms. In contrast, the semisynthetic
derivatives of safranal possess interesting biological properties. For
example, although relatively weak inhibitors, both semicarbazone
3 [IC50(MAO-A) ¼ 164
none 7 [IC50(MAO-A) ¼ 132
found to be inhibitors of hMAO-A and hMAO-B. Similarly, 2-
m
M; IC50(MAO-B) ¼ 111
m
M] and thiazolidi-
m
M; IC50(MAO-B) ¼ 168
mM] were
methylthiosemicarbazone derivative 5 [IC50(MAO-A) ¼ 93.2
m
M;
8
IC50(MAO-B)
¼
117
m
M] and N-benzylated thiazolidinone
[IC50(MAO-A) ¼ 86.6
mM; IC50(MAO-B) ¼ 215
mM] also proved to be
relatively weak inhibitors of hMAO-A and hMAO-B. Thio-
semicarbazone 4 possessed higher potencies for the inhibition of
With this experimental protocol, crocin (1), due to steric hin-
drance, is unable to fit into the active sites of hMAO-A and hMAO-B.
It was thus decided to perform a blind-docking procedure which
allows for binding to alternative pockets with allosteric functions.
For this ligand, the site partitioning approach was followed to
sample different parts of the binding cavity. In the second phase,
docking between the receptor and ligand was performed consid-
ering only one specific site for both isoforms. The MM-GBSA
hMAO-A and hMAO-B with IC50 values of 56.8
respectively. Interestingly, 4-phenylthiosemicarbazone
found to be a moderately potent hMAO-A inhibitor with an IC50
value of 19.5 M. This compound also represents a relatively se-
lective hMAO-A inhibitor since it is 11-fold more potent as a hMAO-
A inhibitor compared to its hMAO-B inhibition potency [IC50(MAO-
m
M and 74.5
m
was
M,
6
m
B) ¼ 224
mM]. The most notable inhibitor of the series, however, is
the hydrazothiazole derivative 9, which is the most potent inhibitor
of both hMAO-A and hMAO-B among the compounds evaluated. In
fact, 9 is a highly potent hMAO-B inhibitor with an IC50 value of
€
method available in Schrodinger Software Prime 3.3 [33] was
used to calculate ligand binding and strain energies (MM-GBSA dG
Bind). The Pymol 1.5.0.4 [34] molecular graphics system was used
to visualize the results of the docking experiments.
0.100
m
M. Even though 9 also is a notable hMAO-A inhibitor
M], it is 99-fold more potent as a hMAO-B
[IC50(hMAO-A) ¼ 9.93
m
inhibitor, and thus the most isoform selective inhibitor of the se-
ries. For comparison, the reversible hMAO-A inhibitor, toloxatone,
3. Results and discussion
is reported to inhibit hMAO-A with an IC50 value of 3.92
the well known reference hMAO-B inhibitor lazabemide exhibits an
IC50 of 0.091 M for the inhibition of hMAO-B [36]. The IC50 value
for the inhibition of hMAO-B recorded for compound 9 is 0.100 M,
mM, while
Starting from the commercially available natural product
safranal (2), which was purified by column chromatography, we
designed and synthesized (thio)semicarbazone (3e6), thiazolidi-
none (7e8) and hydrazothiazole (9) derivatives. For these de-
rivatives, the safranal nucleus was retained in order to evaluate the
effects of an increasing steric hindrance and different substituents
on hMAO inhibition potency. The general approach for the syn-
thesis of these derivatives is outlined in Scheme 1. Safranal was
reacted in ethanol with semicarbazide, thiosemicarbazide, 2-
methylthiosemicarbazide and 4-phenylthiosemicarbazide to yield
products 3e6 [25]. Compound 4 was subsequently reacted with
ethyl-bromoacetate in a mixture of methanol and sodium acetate to
obtain the 1,3-thiazolidin-4-one derivative 7. The lactamic nitrogen
of thiazolidinone 7 was functionalized using benzyl bromide in
anhydrous acetone and potassium carbonate to yield the N-ben-
zylated derivative 8. To obtain the hydrazothiazole 9, thio-
semicarbazone 4 was reacted with 2-bromo-3'-nitroacetophenone
in ethanol. All the synthesized compounds were washed with n-
hexane and light petroleum and purified by column chromatog-
raphy before characterization by spectroscopic methods and
elemental analysis.
m
m
a value that is similar to the reported IC50 value for the inhibition of
hMAO-B by lazabemide. Compound 9 is therefore approximately
equipotent to lazabemide as a hMAO-B inhibitor. Although no clear
structureeactivity relationships are apparent, this study shows
that, with the appropriate substitution, the structure of safranal
may be converted into highly potent hMAO-B inhibitors, and
moderately potent hMAO-A inhibitors. In this respect, compound 9
is a highly promising and novel scaffold for the design of future
hMAO inhibitors. Further structural modification of 9 should be
undertaken to derive useful structureeactivity relationships for the
inhibition of the hMAOs by this class of inhibitors, and to further
optimize the structure for potent and selective hMAO inhibition.
During in silico molecular docking studies, the goal is to estab-
lish the general binding orientations of the ligands, to determine
the nature of the interactions between the ligands and receptor and
to estimate affinities of the ligands for the receptor by calculating
binding energies. Stabilization of ligands is mainly attributed to the
formation of productive contacts between ligands and the target
protein. For the binding of 1, 3, 4 and 9 to hMAO-A and hMAO-B,
these contacts are given in Table S1 of Supporting Information.
The hMAO inhibitory properties of derivatives 1e9 (Table 1)
were investigated using commercially available microsomes from
insect cells containing recombinant hMAO-A and hMAO-B with
kynuramine as the substrate [35].
The results of the hMAO inhibition experiments show that the
natural product crocin (1) display IC50 values in the higher micro-
4. Mechanism of action
molar range for the inhibition of hMAO-A and hMAO-B (170
m
M and
In this study, the binding of crocin (1) and safranal derivatives 3,
4 and 9 to the hMAO enzymes were examined. The global energy of
interactions (in kcal/mol) for each docking experiment is reported
in Table S2 (Supporting Information). As will be discussed below,
derivatives 3, 4 and 9 fit within the active site cavities of the hMAOs
with acceptable G-score values (<X-ray G-score). Since crocin (1)
does not fit within the active sites of hMAO-A and hMAO-B, a blind-
docking simulation was carried out and a potential allosteric site
where crocin may bind was identified. This site is situated in the
area near the gate of the entrance tunnel to the binding pocket.
128 M, respectively). Crocin is therefore a relatively weak inhibitor
m
of the hMAOs, and unless high brain levels are attained after the
administration of saffron extracts to humans and experimental
animals, the inhibition of the hMAOs is unlikely to represent a
mechanism by which saffron extracts confer antidepressant and
neuroprotective effects. Further investigation into the in vivo MAO
inhibitory properties, the pharmacokinetic properties and in
particularly brain tissue levels of crocin, are necessary to
completely elucidate the contribution of crocin to the

C. De Monte et al. / European Journal of Medicinal Chemistry 82 (2014) 164e171
169
Fig. 1. (A) Best pose of crocin (1) and its polar contacts with respective distances on the hMAO-A surface. The natural ligand is rendered as cyan sticks, pocket amino acids are shown
as blue lines, the superimposed reference drug (harmine) as yellow sticks, FAD co-factor is highlighted with green sticks, while the protein target is shown as a blue surface. All
molecules are colored according to atom type. (B) Best pose of crocin (1) and its polar contacts with respective distances on the hMAO-B surface. The natural ligand is rendered as
cyan sticks, pocket amino acids are shown as blue lines, the superimposed reference drug (safinamide) as yellow sticks, the FAD co-factor is highlighted with green sticks, while the
protein target is shown as a blue surface. All molecules are colored according to atom type. (For interpretation of the references to color in this figure legend, the reader is referred to
the web version of this article.)
Fig. 2. (A) Semicarbazone 3 best pose and its polar contacts with respective distances in the hMAO-A binding cavity. The ligand is rendered as cyan sticks, the GLN215 pocket amino
acid is shown as blue lines, the superimposed reference drug (harmine) as yellow lines, and the protein target is shown as a blue cartoon. All molecules are colored according to
atom type. (B) Semicarbazone 3 best pose and its polar contacts with respective distances in the hMAO-B binding cavity. The ligand is rendered as cyan sticks, the GLN206 pocket
amino acid is shown as blue lines, the superimposed reference drug (safinamide) as yellow lines, and the protein target is shown as a blue cartoon. All molecules are colored
according to atom type. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
A computational study, carried out following our recently pub-
lished protocol [37,38], reveals that crocin (1) inhibits both hMAO
isoforms with a non-competitive mechanism of action. As shown in
Fig. 1, crocin does not fit into the binding pocket of either hMAO-A
or hMAO-B. The modeling study suggests that crocin binds to the
protein surface, to an allosteric site, and may thus inhibit the
catalysis of amine substrates. In agreement with the in vitro inhi-
bition data, which shows that crocin is a more potent hMAO-B
inhibitor, crocin better interacts with the hMAO-B than the
hMAO-A isoform, as shown by the MM-GBSA dG Bind values
(Table S2), and by the higher number of interactions (H-bonds)
which it establishes with the amino acids, to stabilize the pose
(Fig. 1 and Table S1 of the Supporting Information). In the case of
isoform A, in its best pose, crocin fits well on the target surface,
blocking with its central portion the access to the active site; in the
case of hMAO-B isoform, a part of the natural ligand is placed into a
cavity on the surface. In both poses, H-bonds interactions occur
between the polar OH groups located at the terminal ends of crocin
and enzyme surface residues.
Fig. 3. (A) Thiosemicarbazone 4 best pose and its polar contacts with respective distances in the hMAO-A binding site. The ligand is rendered as cyan sticks, the GLN215 pocket
amino acid is shown as blue lines, the superimposed reference drug (harmine) as yellow lines, and the protein target is shown as a blue cartoon. All molecules are colored according
to atom type. (B) Thiosemicarbazone 4 best pose and its polar contacts with respective distances in the hMAO-B binding cavity. The ligand is rendered as cyan sticks, the GLU84 and
PRO102 pocket amino acids are shown as blue lines, the superimposed reference drug (safinamide) as yellow lines, and the protein target is shown as a blue cartoon. All molecules
are colored according to atom type. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

170
C. De Monte et al. / European Journal of Medicinal Chemistry 82 (2014) 164e171
Fig. 4. (A) Best pose of hydrazothiazole 9 and its polar contacts with respective distances in the hMAO-A binding cavity. The ligand is rendered as cyan sticks, the TYR197 and
GLY443 pocket amino acids are shown as blue lines, the superimposed reference drug (harmine) as yellow lines, and the protein target is shown as a blue cartoon. All molecules are
colored according to atom type. (B) Best pose of hydrazothiazole 9 and its polar contacts with respective distances in the hMAO-B binding cavity. The ligand is rendered as cyan
sticks, the LEU164 and GLN206 pocket amino acids are shown as blue lines, the superimposed reference drug (safinamide) as yellow lines, while the protein target is shown as a
blue cartoon. All molecules are colored according to atom type. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this
article.)
Despite the similar values of the hMAO inhibition potencies of
crocin and the semicarbazone derivative 3, computational studies
suggest that 3 may act as a competitive inhibitor, with its best pose
located in the binding pocket (Fig. 2). In accordance to the hMAO
inhibition data (IC50 values in Table 1), the G-score values obtained
in the docking simulations (Table S2) also show that thio-
semicarbazone 4 is a more potent hMAO inhibitor than is semi-
carbazone 3. While the isosteric substitution S / O does not affect
the general binding orientation of 4 within hMAO-A (compared to
3) and the H-bond interaction with GLN215 in hMAO-A, it results in
significantly different interactions of 4 with hMAO-B compared to
the interactions observed between 3 and hMAO-B (Fig. 3).
Finally, based on the G-score values, hydrazothiazole 9 is a more
potent hMAO-B inhibitor than compounds 3 and 4, an observation
supported by the hMAO inhibition data. The docking simulation
further shows that 9 binds within the hMAO-B active site, and is
therefore most likely a competitive inhibitor. Interestingly, the G-
score values calculated for binding of 9 to hMAO-A and hMAO-B
suggest that this compound is a selective inhibitor of the hMAO-B
isoform, which is in accordance to the results of the hMAO inhi-
bition studies (Fig. 4).
relationships for the inhibition of the hMAOs by this promising
class of compounds, and possibly to further enhance isoform
selectivity.
Molecular docking studies were carried out on natural crocin (1)
and the semisynthetic compounds 3, 4 and 9. The results suggest
that crocin may inhibit both hMAO isoform with non-competitive
mechanisms by binding to allosteric sites on the surfaces of the
proteins. In contrast, the semisynthetic compounds 3, 4 and 9 bind
within the active sites of the hMAOs and may thus act competi-
tively. The results of the docking studies were also in agreement
with the finding that thiosemicarbazone 4 is a more potent hMAO
inhibitor than semicarbazone 3, and that hydrazothiazole 9 acts as
a selective inhibitor of the hMAO-B enzyme. The results of this
study show that, with the appropriate chemical modification,
safranal may serve as scaffold for the discovery of hMAO inhibitors
for the clinical management of mental and neurodegenerative
disorders.
Conflict of interest
The authors state no conflict of interest and that they have
received no payment in preparation of this manuscript.
5. Conclusion
Acknowledgments
A variety of studies have suggested that saffron may have a
potential role in the treatment of mental disorders such as
depression, or may enhance the beneficial effects of conventional
drugs with fewer side effects. Based on this consideration, we
investigated the hMAO-A and hMAO-B inhibitory properties of two
natural components of C. sativus L., crocin (1) and safranal (2), as
well as those of newly designed compounds (3e9) derived from
chemical modifications of safranal. hMAO-A and hMAO-B are two
important enzymes which are targets for the treatment of neuro-
psychiatric and neurodegenerative diseases. The results document
that, while crocin is a relatively weak inhibitor of the hMAOs,
safranal is not a hMAO inhibitor. Based on this, we conclude that the
inhibition of the MAOs by crocin and safranal are unlikely to
represent mechanisms by which saffron extracts confer their
beneficial central effects. The designed chemical derivatives of
safranal, however, displayed much improved inhibitory activities
against both hMAO enzymes. In this respect, compound 9 was
found to be an exceptionally potent hMAO-B inhibitor, and thus a
highly promising and novel scaffold for the design of future hMAO-
B inhibitors. Compound 9 also displayed relatively good selectivity
for the hMAO-B isoform. Further structural modification of 9 should
be undertaken to derive more useful structureeactivity
This work was supported by “Ateneo 2012” project (P. Chimenti
and L. Mannina) using the facilities of the “Unit of Metabolomics:
Studies on Food, Nutraceuticals and Biological Fluids” and by
Interregional Research Center for Food Safety and Health at the
Magna Græcia University of Catanzaro (MIUR PON a3_00359).
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2014.05.048.
References
[1] (a) S. Carradori, M. D'Ascenzio, P. Chimenti, D. Secci, A. Bolasco, Mol. Divers. 18
(2014) 219e243;
(b) S. Carradori, D. Secci, A. Bolasco, P. Chimenti, M. D'Ascenzio, Expert Opin.
Ther. Patents 22 (2012) 759e801;
(c) A. Bolasco, R. Fioravanti, S. Carradori, Expert Opin. Ther. Patents 15 (2005)
1763e1782;
(d) A. Bolasco, S. Carradori, R. Fioravanti, Expert Opin. Ther. Patents 20 (2010)
909e939.
[2] J. Sarris, A. Panossian, I. Schweitzer, C. Stough, A. Scholey, Eur. Neuro-
psychopharmacol. 21 (2011) 841e860.

C. De Monte et al. / European Journal of Medicinal Chemistry 82 (2014) 164e171
171
[3] S.Z. Bathaie, S.Z. Mousavi, Crit. Rev. Food Sci. Nutr. 50 (2010) 761e786.
[21] M. Bortolato, K. Chen, J.C. Shih, Adv. Drug. Deliv. Rev. 60 (2008) 1527e1533.
[22] (a) S. Carradori, M. D'Ascenzio, C. De Monte, D. Secci, M. Yanez, Arch. Pharm.
ꢀ~
[4] B.A. Wani, A.K.R. Hamza, F.A.J. Mohiddin, Med. Plant. Res.
5
(2011)
2131e2135.
Chem. Life Sci. 346 (2013) 17e22;
[5] Z.I. Linardaki, M.G. Orkoula, A.G. Kokkosis, F.N. Lamari, M. Margarity, Food
Chem. Toxicol. 52 (2013) 163e170.
(b) M. D'Ascenzio, S. Carradori, D. Secci, L. Mannina, A.P. Sobolev, C. De Monte,
R. Cirilli, M. Yanez, S. Alcaro, F. Ortuso, Bioorg. Med. Chem. 22 (2014)
ꢀ~
[6] M. Lechtenberg, D. Schepmann, M. Niehues, N. Hellenbrand, B. Wünsch,
A. Hensel, Planta Med. 74 (2008) 764e772.
[7] A.A. Basti, E. Moshiri, A.A. Noorbala, A.H. Jamshidi, S.H. Abbasi,
S. Akhondzadeh, Prog. Neuro-psychopharmacol. Biol. Psychiatry 31 (2007)
439e442.
2887e2895.
[23] (a) P. Chimenti, A. Petzer, S. Carradori, M. D'Ascenzio, R. Silvestri, S. Alcaro,
F. Ortuso, J.P. Petzer, D. Secci, Eur. J. Med. Chem. 66 (2013) 221e227;
(b) S. Carradori, D. Secci, P. D'Ascenzio, A. Bolasco, J. Heterocycl. Chem. (2014),
http://dx.doi.org/10.1002/jhet.1856.
ꢀ~
[8] A.A. Noorbalaa, S. Akhondzadeh, N. Tahmacebi-Pour, A.H. Jamshidi,
J. Ethnopharmacol. 97 (2005) 281e284.
[24] D. Secci, A. Bolasco, S. Carradori, M. D'Ascenzio, R. Nescatelli, M. Yanez, Eur. J.
Med. Chem. 58 (2012) 405e417.
[9] S. Akhondzadeh, H. Fallah-Pour, K. Afkham, A.H. Jamshidi, F. Khalighi-Cigar-
oudi, BMC Complement. Altern. Med. 4 (2004) 12e16.
[25] F. Chimenti, D. Secci, A. Bolasco, P. Chimenti, A. Granese, S. Carradori,
ꢀ~
M. Yanez, F. Orallo, F. Ortuso, S. Alcaro, Bioorg. Med. Chem. 18 (2010)
[10] E. Moshiria, A.A. Bastib, A.A. Noorbalac, A.H. Jamshidid, S.H. Abbasie,
S. Akhondzadehc, Phytomedicine 13 (2006) 607e611.
[11] L. Kashani, F. Raisi, S. Saroukhani, H. Sohrabi, A. Modabbernia, A.-A. Nasehi,
A. Jamshidi, M. Ashrafi, P. Mansouri, P. Ghaeli, S. Akhondzadeh, Hum. Psy-
chopharmacol. Clin. Exp. 28 (2013) 54e60.
5715e5723.
[26] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig,
I.N. Shindyalov, P.E. Bourne, Nucleic Acids Res. 28 (2000) 235e242.
[27] S.Y. Son, J. Ma, Y. Kondou, M. Yoshimura, E. Yamashita, T. Tsukihara, Proc. Natl.
Acad. Sci. U. S. A. 105 (2008) 5739e5744.
[12] A. Modabbernia, H. Sohrabi, A.-A. Nasehi, F. Raisi, S. Saroukhani, A. Jamshidi,
M. Tabrizi, M. Ashrafi, S. Akhondzadeh, Psychopharmacology 223 (2012)
381e388.
[13] M. Schmidt, G. Betti, A. Hensel, Wien. Med. Wochenschr. 157 (2007) 315e319.
[14] G. Georgiadoua, P.A. Tarantilis, N. Pitsikas, Neurosci. Lett. 528 (2012) 27e30.
[15] F. Di Marco, S. Romeo, C. Nandasena, S. Purushothuman, C. Adams, S. Bisti,
J. Stone, Am. J. Neurodegener. Dis. 2 (2013) 208e220.
[16] M.A. Papandreou, C.D. Kanakis, M.G. Polissiou, S. Efthimiopoulos,
P. Cordopatis, M. Margarity, F.N. Lamari, J. Agric. Food Chem. 54 (2006)
8762e8768.
[28] C. Binda, J. Wang, L. Pisani, C. Caccia, A. Carotti, P. Salvati, D.E. Edmondson,
A. Mattevi, J. Med. Chem. 50 (2007) 5848e5852.
[29] Schrodinger Suite 2013 Protein Preparation Wizard; Epik Version 2.4,
€
€
€
Schrodinger, LLC, N. Y. N. Y., 2013; Impact Version 5.9, Schrodinger, LLC, New
€
York, NY, 2013; Prime Version 3.2, Schrodinger, LLC, New York, NY, 2013.
[30] Maestro, Version 9.4, Schrodinger, LLC, New York, NY, 2013.
[31] Glide, Version 5.9, Schrodinger, LLC, New York, NY, 2013.
[32] LigPrep, Version 2.6, Schrodinger, LLC, New York, NY, 2013.
[33] Prime, Version 3.3, Schrodinger, LLC, New York, NY, 2013.
[34] The PyMOL Molecular Graphics System, Version 1.5.0.4 Schrodinger, LLC.
€
€
€
€
€
[17] G.D. Geromichalos, F.N. Lamari, M.A. Papandreou, D.T. Trafalis, M. Margarity,
A. Papageorgiou, Z. Sinakos, J. Agric. Food Chem. 60 (2012) 6131e6138.
[18] K.N. Westlund, R.M. Denney, R.M. Rose, C.W. Abell, Neuroscience 25 (1988)
439e456.
[19] D.E. Edmondson, C. Binda, J. Wang, A.K. Upadhayay, A. Mattevi, Biochemistry
48 (2009) 4220e4230.
[35] L. Novaroli, M. Reist, E. Favre, A. Carotti, M. Catto, P.-A. Carrupt, Bioorg. Med.
Chem. 13 (2005) 6212e6217.
[36] A. Petzer, A. Pienaar, J.P. Petzer, Drug Res. 63 (2013) 462e467.
[37] F. Chimenti, R. Fioravanti, A. Bolasco, F. Manna, P. Chimenti, D. Secci, O. Befani,
P. Turini, F. Ortuso, S. Alcaro, J. Med. Chem. 50 (2007) 425e428.
ꢀ~
[38] A. Gaspar, T. Silva, M. Yanez, D. Vina, F. Orallo, F. Ortuso, E. Uriarte, S. Alcaro,
[20] M.B.H. Youdim, D.E. Edmondson, K.F. Tipton, Nat. Rev. 7 (2006) 295e309.
F. Borges, J. Med. Chem. 54 (2011) 5165e5173.