Biotransformation of LASSBio-579 and pharmacological evaluation of p-hydroxylated metabolite a N-phenylpiperazine antipsychotic lead compound

Article abstract of DOI:10.1016/j.ejmech.2012.08.011

Using a combination of docking and molecular dynamics simulations, we predicted that p-hydroxylation by CYP1A2 would be the main metabolic pathway for the 1-[1-(4-chlorophenyl)-1H-4pyrazolylmethyl] phenylhexahydropiperazine, LASSBio-579 (3). As the result of a screening process with strains of filamentous fungi, Cunninghamella echinulata ATCC 9244 was chosen to scale up the preparation of the p-hydroxylated metabolite (4). About 30 min after i.p. administration of (3) to rats was identified as the p-hydroxylated metabolite, confirming our in silico previsions. Chemical synthesis of the metabolite was performed and allowed its pharmacological evaluation in binding assays revealing its high affinity for D2 and D4 receptors, indicating that this metabolite should participate to the antipsychotic effect of (3) in vivo. Furthermore, we report here that both (3) and its p-hydroxylated metabolite (4) have a much lower affinity than clozapine for two receptors involved in adverse reactions. Voltammetric assays were useful to understand the redox profile of (3).

Full text of DOI:10.1016/j.ejmech.2012.08.011

European Journal of Medicinal Chemistry 62 (2013) 214e221
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Biotransformation of LASSBio-579 and pharmacological evaluation
of p-hydroxylated metabolite a N-phenylpiperazine antipsychotic
lead compound
Tatiana F. Gomes a, Thais E.T. Pompeu b, Daniel A. Rodrigues a, François Noël b,
Ricardo Menegatti a, Carolina H. Andrade a, José R. Sabino c, Eric S. Gil a,
Teresa Dalla Costa d, Andresa H. Betti d, Camila B. Antonio d, Stela M.K. Rates d,
,
Carlos A.M. Fraga e, Eliezer J. Barreiro e, Valéria de Oliveira a
*
a Laboratório de Bioconversão, LabMol and Laboratório de Química Farmacêutica Medicinal, Faculdade de Farmácia, Universidade Federal de Goiás,
Goiânia, GO, Brazil
b Laboratório de Farmacologia Bioquímica e Molecular, Departamento de Farmacologia Básica e Clínica, Universidade Federal do Rio de Janeiro,
Rio de Janeiro, RJ, Brazil
c Instituto de Física, Universidade Federal de Goiás, Goiânia, GO, Brazil
d Programa de Pós-Graduação em Ciências Farmacêuticas da Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil
e Laboratório de Avaliação e Síntese de Substâncias Bioativas (LASSBio), Faculdade de Farmácia, Universidade Federal do Rio de Janeiro,
Rio de Janeiro, RJ, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history:
Using a combination of docking and molecular dynamics simulations, we predicted that p-hydroxylation
by CYP1A2 would be the main metabolic pathway for the 1-[1-(4-chlorophenyl)-1H-4pyrazolylmethyl]
phenylhexahydropiperazine, LASSBio-579 (3). As the result of a screening process with strains of fila-
mentous fungi, Cunninghamella echinulata ATCC 9244 was chosen to scale up the preparation of the
p-hydroxylated metabolite (4). About 30 min after i.p. administration of (3) to rats was identified as the
p-hydroxylated metabolite, confirming our in silico previsions. Chemical synthesis of the metabolite was
performed and allowed its pharmacological evaluation in binding assays revealing its high affinity for D2
and D4 receptors, indicating that this metabolite should participate to the antipsychotic effect of (3)
in vivo. Furthermore, we report here that both (3) and its p-hydroxylated metabolite (4) have a much
lower affinity than clozapine for two receptors involved in adverse reactions. Voltammetric assays were
useful to understand the redox profile of (3).
Received 27 June 2012
Received in revised form
6 August 2012
Accepted 8 August 2012
Available online 20 August 2012
Keywords:
Biotransformation
Filamentous fungi
Pharmacological evaluation
LASSBio-579
Schizophrenia
Ó 2012 Elsevier Masson SAS. All rights reserved.
1. Introduction
effects at therapeutic doses [4]. The introduction of clozapine (1) for
treatment-resistant schizophrenia gave rise to a new group of atyp-
Schizophrenia is a chronic psychiatric disorder that affects
approximately 1% of the world’s population. Diagnostic features of
schizophrenia include positive symptoms (e.g., auditory hallucina-
tions, disorganized thoughts, delusions), negative symptoms (e.g.,
social withdrawal, avolition, anhedonia), and cognitive dysfunction
(e.g., attention impairment, memory deficits) [1e3]. Schizophrenia
conventionally has been treated with dopamine D2 receptor antag-
onists such as haloperidol. However, these drugs have little effects on
negative and cognitive symptoms and elicit extrapyramidal side
ical antipsychotics. These drugs exhibit potent antagonism at
multiple receptor subtypes, including dopamine and serotonin
receptors [4e6]. However, a significant population of patients is still
refractory to treatment, and these new drugs also induce serious side
effects [6e8], thus underscoring the importance of developing more
effective and safer antipsychotic drugs. In a research program aiming
the development of new atypical antipsychotic drugs, with similar
clozapine therapeutic profile, high-affinity to D4 receptor with
conformational limited flexibility, but devoid of the hematological
side effects, a series of heterocyclic N-phenylpiperazine derivatives
were planned through molecular hybridization between clozapine
(1) and L-741 (2) [9]. In vitro and behavioral assays showed that three
of these compounds [LASSBio-579 (3), LASSBio-580 and LASSBio-
581] act on dopaminergic and serotonergic neurotransmission [9e
11]. Among them, LASSBio-579 (3) or 1-[1-(4-chlorophenyl)-1H-4-
* Corresponding author. Faculdade de Farmácia, Universidade Federal de Goiás,
1a Avenida esquina c/Praça Universitária, S/N. Caixa Postal 131, Setor Universitário,
CEP: 74.605-220, Goiânia, GO, Brazil. Tel.: þ55 62 3209 6432; fax: þ55 62 3209
6037.
E-mail address: valeria@farmacia.ufg.br (V. de Oliveira).
0223-5234/$ e see front matter Ó 2012 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2012.08.011

T.F. Gomes et al. / European Journal of Medicinal Chemistry 62 (2013) 214e221
215
Fig. 1. Compound (3) design, metabolite (4) biosynthesis and organic synthesis.
pyrazolylmethyl] phenylhexahydropiperazine, Fig. 1, presented the
2. Results and discussion
best pharmacological profile [10]. As this compound has a very poor
oral bioavailability [12], its activity could depend on until now
unknown metabolite(s) that should deserve our attention. In the
present study, we have used a structure-based metabolism predic-
tion methodology combining docking and molecular dynamics
(MD) simulations to predict the metabolism of (3) by CYP1A2. Using
our experience with microbial models of mammalian metabolism
[10e14], we produced the p-hydroxylated (4) metabolite of (3) by
bioconversion and used it as a reference standard for identifying the
metabolite present in plasma after i.p. administration of (3) to rats.
The pharmacological activity of the p-hydroxylated (4) metabolite
was assayed in binding assays for determining its affinity to relevant
receptors for the treatment of schizophrenia. Since the metabolism
process involves bioelectrochemical mechanisms, we also carried out
voltammetric studies in order to understand the electrochemical
behavior of (3) and its metabolite (4).
2.1. In silico metabolism prediction
Computational methods based on a combination of docking and
molecular dynamics simulations can help to understand drug
metabolism and to predict drugedrug interactions [15,16]. The
putative binding poses of (3) in the active site of CYP1A2 were
estimated using the MOE-Dock software allowing side chains
flexibility [17,18]. The analysis was restricted to CYP1A2 because
this is the major isoform responsible for the metabolism of cloza-
pine (1) and olanzapine [19,20], two N-phenylpiperazines struc-
turally related to (3) [9].
The best pose obtained from docking was submitted to coupled
energy minimization and MD relaxation studies, to provide addi-
tional structural relaxation and to establish a reasonable model for
hydrogen bonding. Fig. 2 illustrates the energetically favored
Fig. 2. Binding pose of (3) in the active site of CYP1A2 predicted by docking and MD simulations. (A) Schematic representation of active site residues and interactions with (3) (B) 3D
representation including the heme group (magenta), the active site residues (gray), water (red sphere) and (3) (yellow). The hydrogen atoms were removed for better visualization.
The distance between (3) and heme iron is w3.7 Å. Orientation consistent with aromatic hydroxylation. (For interpretation of the references to colour in this figure legend, the
reader is referred to the web version of this article.)

216
T.F. Gomes et al. / European Journal of Medicinal Chemistry 62 (2013) 214e221
docking mode for (3) into the CYP1A2. The binding energy of the
best pose was e 9.01 kcal/mol.
2.3. Synthesis of compound (4)
As we can see from Fig. 2B, the benzene ring of (3) is oriented
perpendicularly to the porphyrin plane, within a distance of
approximately 3.7 Å from the heme iron, suggesting that metabo-
lism may occur by aromatic hydroxylation.
Compound (4) was also synthetized through conditions of
reductive amination. The reaction was carried out between the
aldehyde (5) and 4-hydroxyphenylpiperazine (6) in methanol at
room temperature, using zinc chloride as catalyst and sodium
cyanoborohydride as reducing agent which produced (4) in yields
of 90.0% after 1 h (Fig. 1) [24]. The compound (4) structure was
confirmed by comparing with its standard.
Single crystals of (3) and compound (4) were obtained by the
slow evaporation of the solvent using methanol at room tempera-
ture. ORTEP [25] views and the characteristics of these crystals of
LASSBio-579 (3) and its p-hydroxylated metabolite (4) are shown in
Fig. 4A and B, respectively.
2.2. Comparing of biotransformation in vitro by microorganisms
and in vivo by rodents
The use of microorganisms as models of mammalian metabo-
lism was introduced in the early 1970s [21] and represents a very
useful tool in the production of compounds with improved or
different activity as well as less toxicity originated by the fungal
enzymatic biodiversity [22]. For our screening procedure [23], the
following fifteen strains were used in order to find a microor-
ganism that would be able to hydroxylate (3). Absídia blakesleana
ATCC 26617, A. blakesleana ATCC 10148b, Aspergillus candidus ATCC
2023, Aspergillus ochraceus ATCC 1009, Beauveria bassiana ATCC
7149, Chaetomium indicum LCP 984200, Cunninghamella echinu-
lata ATCC 9244, C. echinulata ATCC 9245, Cunninghamella elegans
ATCC 36112, C. elegans ATCC 26169, Curvularia lunata NRRL 2380,
Fusarium roseum ATCC 14717, Mortierella isabellina NRRL 1757,
Mucor griseocyanus ATCC 1207a, Rhizopus arrhizus ATCC 11145. As
metabolite (4) was produced exclusively by C. echinulata ATCC
9244 (Fig. 3), this microorganism was selected for the preparative-
scale production of (4), necessary for structural elucidation
through IR, LC-MS and NMR besides preliminary pharmacological
assays. Fig. 3A shows the time-course of (3) microbial metabolism
together with metabolite (4) formation. After 48 h incubation, the
(3) remaining in the supernatant corresponded to less than 10% of
the original amount and the concentration of metabolite (4) had
reached a plateau. In order to verify if metabolite (4) was produced
in vivo in rodents, we administered (3) to rats, by i.p. Fig. 3B shows
that 30 min after administration of (3) at 50 mg/kg, its plasmatic
concentration was already much lower than of its p-hydroxylated
metabolite (4), confirming our in silico previsions. In fact, using UV
detection at 257 nm, only one main metabolite was detected, with
a retention time in HPLC very close to the one of metabolite (4)
produced by C. echinulata used as internal standard. This
compound was already detected in the first sample of rat plasma,
30 min after (3) administration, at a concentration close to its
maximum. Differently from (3), the plasma concentration of this
metabolite did not decreased at least until 24 h, when the last
sample of plasma was collected, indicating that it has a very long
half-life.
2.4. Redox profile determination
As a radical scavenger profile could be responsible for an indi-
rect neuroprotective effect [26] of these compounds in vivo, we
decided to investigate the redox profile. Voltammetric assays were
carried out using hydromethanolic solutions in the presence and
absence of a surfactant (LSS) due to the poor hydrosolubility of our
compounds. The use of surfactant improved the homogeneity of
the medium as well as the peak resolutions for both compounds (3)
and (4) as illustrated in Fig. 5, respectively. Analysis of the vol-
tammograms presented in Fig. 5 indicates that (3) and its metab-
olite (4) do not present a radical scavenger profile because they
have only oxidative potentials. Compound (3) presented one
oxidative potential near 0.6 V (Fig. 5A) whereas metabolite (4)
showed one near 0.2 V and other near 1.2 V, indicating that an
overpotential is needed to complete oxidation (Fig. 5B).
2.5. Pharmacologic evaluation
Since our results indicated that (4) is the main metabolite of (3)
in vivo, in a rodent model, we decided to determine its affinity for
dopamine and 5-HT receptors that have been reported as putative
molecular targets for the antipsychotic effect of clozapine (1), our
reference drug. Fort this purpose, we first repeated and extended
our previous binding assays with (3) and (1), initially restricted to
the D2, 5-HT2A and 5-HT1A receptors [27]. Such assays, performed
in exactly the same experimental conditions, are necessary to allow
reliable comparison of (4) to its parent (3) and reference (1) drug.
Table 1 indicates that (3), like (1), has a high affinity for the D4
receptor, but not for the 5-HT2C receptor that has been implicated
in the mechanism of action of (1) by some authors [28,29]. With
respect to (4), our data indicate that it has a binding profile very
80
A
B
500
60
40
20
0
400
300
200
100
0
0
24
48
72
0
6
12
18
24
Time (h)
Time (h)
Fig. 3. (A). Time course of (3, -) metabolism by Cunninghamella echinulata ATCC 9244 and of compound (4, C) formation. (B). Plasma level-time curve for (3, -) and metabolite
(4, C) after i.p. administration in rat. Each value is the mean of three different experiments, and the standard deviation was less than 5%.

T.F. Gomes et al. / European Journal of Medicinal Chemistry 62 (2013) 214e221
217
Fig. 4. ORTEP representations of compounds (3) (A) and (4) (B) with the labeling schemes. Ellipsoids are drawn at 30% of probability level, hydrogen atoms are represented as
spheres of arbitrary radius. Dihedral angles and double bonds for compounds (3) and (4) of the asymmetric unit.
similar to (3), but with a 6-fold higher affinity for the D4 receptor
and a 27-fold lower affinity for the 5-HT1A receptor, so that it could
be considered as a dual D2-D4 ligand.
3. Conclusions
The computational methods show how this antipsychotic lead-
compound interacts in active site of CYP1A2. The voltammetric
techniques indicated that (3) and compound (4) present one and
two oxidative potentials respectively, so that they have not a radical
scavenger profile. The compound (4) is a main metabolite of (3) in
rat and that it has a high affinity for D2 and D4 receptors indicating
that it could participate to the antipsychotic effect of (3) in vivo.
Furthermore, the binding studies revealed a more favorable profile
than (1) for both (3) and its metabolite (4) regarding two receptors
involved in adverse effects.
As side-effects are to be considered when tailoring the drug
treatment to the individual schizophrenic patient [30], we also
compared the affinity of (3) and its metabolite (4) for two receptors
that have been considered as source of adverse effects of (1), ie the
a1B and the muscarinic receptors. Table 1 indicates that both (3) and
its main metabolite (4) have a low affinity for the a1B receptor and
no affinity for the muscarinic receptor, indicating that these lead
compounds could have much less propensity that clozapine to
produce adverse effects like postural hypotension and constipation.
Fig. 5. Cyclic voltammograms for hydromethanolic solution of 1 mM (3). 100 mV sꢀ1, -0,25 a 1.25 V in the presence (A) or absence (B) of 5% LSS, methanolic solution of 1 mM (4).
100 mV sꢀ1, ꢀ0.25 at 1.25 V in the presence (C) or absence (D) of 5% LSS. blank (d),1stscan (- - -), 2nd scan (/) and 3rd scan ($- $- $).

218
T.F. Gomes et al. / European Journal of Medicinal Chemistry 62 (2013) 214e221
Table 1
Affinities of (1), (3) and compound (4) for different receptors.
Ki (mM) (95% confidence interval)
Receptor
(1)
0.13 [3] (0.037e0.470)
(3)
(4)
#
D2
D4
5-HT1A
5-HT2A
5-HT2C
a2
0.38 [4] (0.20e0.74)
0.17 [4] (0.052e0.580)
0.21 [4] (0.12e0.37)
6.82 [3]
8.44 [4]
2.49 [3]
0.53 [3] & (0.22e1.31)
0.034 [3] * (0.018e0.066)
0.06 [2] (0.042e0.085)
0.26 [3] (0.140e0.503)
0.02 [2] (0.0037e1.12)
0.024 [2] (0.055e0.100)
0.10 [3] (0.043e0.250)
0.029 [3] (0.0068e0.12)
0.022 [2] (0.0076e0.064)
&&&,
5.87 [3]
*** (3.03e11.3)
###
(0.84e55.2)
(6.34e11.2)
(2.02e3.09)
8.22 [3]&&& (3.32e20.3)
12.3 [3]&&& (9.7e15.7)
###
###
&&&,
6.45 [3]
1.15 [3]
> 30
** (3.72e11.2)
* (0.88e1.49)
###
&&&,
a1B
Muscarinic
3.0 [3]
> 30
(0.99e9.08)
$
$
Ki values are geometric means (with their 95% confidence interval) of [n] independent experiments. $Ki values above 30
compounds (3) and (4), respectively).
mM (only 22% and 15% inhibition at 30 mM for
*p < 0.05; **p < 0.01; ***p < 0.001: (4) vs (3); #p < 0.05; ##p < 0.01; ###p < 0.001: (3) vs (1); &p < 0.05; &&p < 0.01; &&&p < 0.001: (3) vs (1); ANOVA followed by Student-
NewmaneKeuls test for multiple comparison.
4. Experimental section
the active sites search in the enzyme structure and dummy atoms
were created from the obtained alpha spheres. The ligand place-
ment was performed using alpha PMI method, with the Affinity dG
scoring function. The top 30 poses were retained and refined using
MMFF94x force field energy minimization with Generalized Born
solvation model, allowing the receptor side chain residues within
6 Å to relax around the mobile ligand. The receptor side chains were
tethered with a force constant of 1.0 kcal/(mol Å2). Energy mini-
mization was stopped when the root-mean-square (RMS) gradient
cutoff of 0.01 kcal/(mol Å) was reached. Final poses were ranked
using the Affinity dG scoring method to calculate the free energy of
binding. In order to determine the possible metabolic sites of the
substrate, the distances between the heme iron of CYP1A2 and the
atoms of (3) were measured for all docking results. One confor-
mation with a relatively low energy CYP1A2-(3) was selected as the
starting conformation for the subsequent molecular dynamics
(MD) simulations. The MD simulations were carried out using the
Desmond MD package version 2.2 [32] and the OPLS-AA 2005 force
field [33]. The complex was solvated with SPC water molecules
generated via an orthorhombic box. The number of solvated water
molecules in each system is about 17,000 and the initial MD
simulation cell dimension was about 95 Å ꢁ 80 Å ꢁ 100 Å and
involved the complex being solvated by a layer of water molecules
of at least 10 Å in all directions. By assuming normal charge states of
ionizable groups corresponding to pH 7, sodium (Naþ) and chloride
(Clꢀ) counter-ions at physiological concentration of 0.015 mol/L
were added in the box in random positions to ensure the global
charge neutrality. The complex was relaxed by performing equili-
bration dynamics at constant temperature (300 K) and constant
pressure (1 bar). The constant pressure and temperature were
controlled via Langevin dynamics method. The simulation time was
120 ps and the coordinates were stored every 1.2 ps. The system
was equilibrated and the average structure was obtained and
minimized, which was considered as the stable binding metabo-
lism mode of (3) in the active site of CYP1A2.
4.1. Substrate
LASSBio-579(3) was synthesized at the Laboratório de Avaliação
e Síntese de Substâncias Bioativas (LASSBio, Rio de Janeiro, Brazil).
4.2. General experimental procedures
IR spectra were recorded on a Mattson Genesis spectrophotom-
eter, series FTIR. The 1H, 13C NMR measurements were done using
a Bruker Avance III 500 instrument (operating at 500.13 MHz for 1H)
equipped with a 5-mm tunable multinuclear triple resonance probe
head equipped with z gradient with SiMe4 as the internal reference.
Mass spectra were recorded on a GCeMS Thermoquest spectrometer
(model: Voyager), and a VG Autospec-Q spectrometer. HPLC was
performed with a Hitachi/Merck L-6270 apparatus equipped with
a UVeVIS detector (L 6200) and a differential refractometer detector
(RI-71). CD spectra were recorded on a JobineYvon Mark III Dichro-
graph in ethanol as solvent. TLC was performed on Merck Kieselgel
60F254, 0.2 mm thick. Silica gel (Merck) was used for column chro-
matography. Purification by means of HPLC was accomplished with
a silica gel column (Hibar 60, 7 m,1 cm wide, 25 cm long). Chemicals
were products of Fluka or Aldrich. All solvents used were freshly
distilled. Baker’s yeast was obtained from a local store. The following
enzymes were used in this work: Candida rugosa Lipase (Sigma, Type
VII, 950 U mgꢀ1), Pseudomonas cepacia Lipase (Amano Pharmaceu-
ticals Co., Japan) and Porcine Pancreas Lipase (Sigma, Type II). Dis-
tillated water was purified by a Milli-Q system (MilliporeÒ). Solvents
were HPLC grade; all other chemicals were reagent grade and of the
highest purity available.
4.3. In silico metabolism prediction
Docking studies were performed using Molecular Operating
Environment (MOE) version 2008.10 (Chemical Computing Group
Inc., Montreal, Canada) allowing side chains flexibility. The
substrate structure (3) was built manually with Spartan for Linux v.
O8 software (Wavefunction, Inc.). Geometry optimization was
performed using Density Functional Model (hybrid functional
B3LYP, basis set 6-31G*). The X-ray crystal structure of CYP1A2
4.4. Microorganisms
The fungal strains were purchased from American Type Culture
Collection (ATCC strains), Rockville, MD, USA and Northern Utili-
zation Research and Development Division (NRRL strains), Peoria,
IL, USA and were kindly donated by Dr. Robert Azerad, Université
René Descartes, Paris V, France.
complexed with a-naphthoflavone at 1.95 Å resolutions (PDB code
2HI4) was obtained in the Protein Data Bank [31]. Hydrogen atoms
were added using the MOE software. The iron metal atom from the
heme group was set to a þ3 charge. Protonation states according to
a pH of 7 were assigned using the “Protonate 3D” option in MOE.
Water molecules were removed, except for the active site water
molecule (wat733), as it has been described to be important for
CYP1A2 substrate binding [31]. MOE Alpha Site Finder was used for
4.5. Conditions of cultivation and biotransformation
The microorganisms were maintained on agar potato slants
(Acumedia e Baltimore, Maryland), stored at 2e4 ꢂC. Fungi were
grown at 27 ꢂC in liquid peptone dextrose soybean meal medium e

T.F. Gomes et al. / European Journal of Medicinal Chemistry 62 (2013) 214e221
219
PDSM (peptone 5 g, dextrose 20 g, soybean meal 5 g, dihydrogen
potassium phosphate 5 g, sodium chloride 5 g, yeast extract 3 g in
1 L of distilled water). The fungal mycelia were transferred from the
agar slant to liquid potato dextrose soy medium (PDSM). Cultures
were incubated in 250 mL flasks containing 100 mL of medium at
27 ꢂC with rotary shaking at 200 rpm. The substrate was dissolved
in ethanol/dimethylformamide (DMF) 1:1 to the concentration of
10 mg/mL. After the culture had been incubated for 65 h, 1 mL of
substrate was added to each flask. The incubation was allowed to
continue for 4 days. Aliquots were withdrawn at 24, 48, 72 and 96 h
of incubation, centrifuged and the supernatants analyzed by TLC
and reverse-phase HPLC. Control flasks consisted of culture broth
without substrate to exclude components of cells walls fungi
possibly detected by HPLC. The experiments without microorgan-
isms were carried out to verify the stability of the substrate by
addition of 1 mL of a solution of ethanol/dimethylformamide (1:1).
No oxidation products could be observed under these conditions.
at 200 ꢂC, nebulizer pressure 23e24 psi, capillary voltage 3.5 kV,
needle 5000 V, and shield 600 V, using the reverse-phase chro-
matographic method described under chromatographic procedures
and substituting the gradient system with water by methanol in
isocratic conditions at flow rate 0.5 mL/min. The flow rate of the
electrospray solutions to the ion source was 5 mL/min and nitrogen
was used as collision gas. All mass spectra were carried out in the
positive ionization mode and were recorded in the scan mode
(from 250 to 1000 Th). A VARIANPÒ Workstation system software
was used for data acquisition and processing. HPLC (n-hexane/i-
PrOH 95:5, 0.7 mL minꢀ1): tR 12 min (S,R) and 17 min (R,S).
4.9. Metabolite (4) synthesis
To a mixture of the starting amine (6) (354 mg, 2.0 equiv),
aldehyde (5) (411 mg, 2.0 equiv) and ZnCl2 (136 mg, 1.0 equiv) in
5 mL MeOH was added NaCNBH3 (62 mg, 1.0 equiv) in one portion.
The mixture was stirred at room temperature for 1 h. MeOH was
then evaporated and the residue was partitioned between 1 N
NaOH and CH2Cl2. The phases were separated and the aqueous
layer was extracted with CH2Cl2. The combined organic layers were
dried (Na2SO4), concentrated in vacuo, and the residue purified
through column chromatography to afford (4) in 90% yield, as
a white solid, mp 180e185 ꢂC.
4.6. Scale up incubation of (3) with C. echinulata
C. echinulata ATCC 9244 was chosen to scale up the production
of the p-hydroxylated metabolite of (3). C. echinulata from agar
slant was transferred to the 250 mL flasks containing 100 mL of
PDSM. The fungus was incubated for 65 h and the substrate was
added to the medium (50 mg for each flask) and incubated for 4
days at 27 ꢂC with rotary shaking at 200 rpm. The culture was
filtered with celite and the filtrate saturated with sodium chloride
and extracted three times with an equal volume (250 mL) of ethyl
acetate. After drying with MgSO4, the organic extract was evapo-
rated in a rotary evaporator. The remaining biomass was extracted
with acetone and the ketonic extract was evaporated and analyzed
as the organic fraction.
4.10. Crystallography analysis
A prismatic sample of (3) was mounted on an Enraf Nonius CAD-
4 diffractometer employing Cu-Ka radiation at room temperature.
Two reflections were monitored hourly and registered 6% of
intensity decay. The unit cell was indexed in the monoclinic system
using 25 reflections. Data reduction was performed using XCAD-4
[34]. The structure was solved in the polar P21 space group with
the Direct Methods. A prismatic sample of (4) was mounted on
4.7. Separation, purification and identification of metabolite (4)
a Bruker Smart ApexII diffractometer [35] employing Mo-Ka radi-
The organic fraction was separated by flash-chromatography on
a silica gel column eluted with ethyl acetateemethanol. Fractions
with similar components were combined under the guidance of
TLC analysis. ESI-MS and NMR analysis of compound (4) was
carried out after purification: MS (m/z, %) 371.1 (M þ 16.0), (100),
192.2 (78), 178 (40). dH(500 MHz; DMSOd6) 2.43e2.63 (4 H, m, H-9
and 11), 2.70e2.83 (4 H, m, H-8 and 12), 3.47 (2 H, s, H-6), 6.63 (2 H,
d, J ¼ 9.0, H-200 and 600), 6.75 (2 H, d, J ¼ 9.0, H-300 and 500), 7.53 (2 H,
d, J ¼ 9.0, H-30 and 50), 7.68 (1 H, s, H-3), 7.86 (2 H, d, J ¼ 9.0, H-20 and
60), 8.44 (1 H, s, HeOH), 8.78 (1 H, s, H-5); dC(125 MHz; DMSO-d6)
49.6 (C-9 and 11), 51.2 (C-6), 52.2 (C-8 and 12), 115.1 (C-200 and 600),
117.4 (C-300 and 500), 119.4 (C-20 and 60), 119.1 (C-10), 119.7 (C-4), 127.2
(C-5), 129.1 (C-30 and 50), 138.6 (C-40), 141.6 (C-3), 144.2 (C-100), 150.8
(C-400).
ation at room temperature. The unit cell was indexed in the triclinic
system using 5414 reflections. Data reduction was performed using
SAINT [35]. The structure was solved in the P-1 space group with
the Direct Methods. Structure refinements were performed with
full matrix least squares method using the package SHELX. Details
of the crystal, data collection and structure refinement are provided
as supplementary material. The crystal structure of (3) lacks
a center of symmetry and (3) crystallizes with two independent
compound (suffixes a and b) in the asymmetric unit, that are
distinguished by the C3/C4/C6/N7 and C60/C10/N1/N2 dihedral
angles. The metabolite (4) crystal structure is center symmetric
with eight independent compounds (suffixes a, b, c and d) located
in the asymmetric unit, being distinguished by the dihedral angles
C600/C100/N10/C11 and C60/C10/N1/N2. In all molecules the C4 atom
of the pyrazole moiety is close to gauche conformation with N7 as
noted by the dihedral C4/C6/N7/C12 values. The dihedral C3/C4/C6/
N7 shows the asymmetric nature of the structure, is positive for
both compounds a and b of (3), and changes sign in two of the four
compounds of the metabolite. The selected dihedral parameters are
summarized in Fig. 5. Single bond distances are as expected for
formal bonds, and the bonds C4eC5 and N2eC3 of average
distances 1.358(9) Å and 1.318(9) Å, respectively, are both double
bonds. Compounds a and b of the compound (4) have N2eC3
distance slightly longer than the average by 0.02 Å, which might
be caused by a pronounced resonance.
4.8. HPLC/ESI-MS apparatus and chromatographic conditions
Analytical reverse-phase HPLC was carried out using Gilson
pumps, a Gilson 321 sample injector equipped with a 20
ml loop and
a
UV detector. Lichrospher 100 RP 18 column (0.5 m,
A
250 mm ꢁ 4.6 mm) was used. The analyses were developed at room
temperature. Detection was set at 257 nm. The flow rate was
1.0 mL/min. A gradient elution of methanol-buffer (65:35) was used
over 9 min. The mobile phase consisted of sodium monohydrate
phosphate (pH 4.5, 0.02 M) and methanol (35:65, v/v). TLC was
carried out with the ascending method using AlugramÒ silica gel
60F254 precoated plates (MachereyeNagel, Germany) and EtOAc-
MeOH (95:05) as solvent. Spots were detected under UV light at
366 nm. ESI-MS analysis were performed in a pressure ionization
1200 L quadrupole MS/MS instrument (MICROMASS QUATTRO LC.)
Electroanalytical assays: The cyclic voltammetric experiments
were carried out on a m
AutolabÒ potentiostat/galvanostat from Eco
Chemie (Holland) coupled to PGSTAT 20 software version 4.9 for
acquisition of the data, connected to a electrochemical cell and
three electrode system. The working electrode was a glassy carbon

220
T.F. Gomes et al. / European Journal of Medicinal Chemistry 62 (2013) 214e221
electrode (B 2 mm2), the reference electrode was a calomelane
electrode, and a platinum ring was used as auxiliary electrode. Both
sample solutions were prepared in methanol PA purchased from
Merck S/A in order to reach 1 mmol Lꢀ1. The electrolyte support was
a hydromethanolic solution with the surfactant lauryl sulfate of
sodium (LSS) at 5% of methanolephosphate buffer 0.1 mmol Lꢀ1 pH
7.0 (1:1).
centrifuged twice at 900 gmax at 4 ꢂC for 10 min. The resulting
supernatants were combined and ultracentrifuged at 48000 gav for
10 min. The pellet was resuspended in buffer and incubated at 37 ꢂC
during 10 min for endogenous neurotransmitters removal. This
suspension was cooled on ice and ultracentrifuged twice at 48000
gav for 10 min at 4 ꢂC. The final pellet was resuspended in buffer
yielding a proportion of 1.5 mL/g tissue and stored in liquid
nitrogen until use. The protein concentration was determined by
the method described by Lowry and co-workers [19] using bovine
serum albumin as standard.
4.11. Animal and biological sample collection
Male Wistar rats (200e300 g) were provided by the State
Foundation for the Research and Production in Health (FEPPS, Porto
Alegre, Brazil). The rats were housed under standard conditions
with room temperature 22 ꢃ 2 ꢂC and had free access to food and
water. The compound (3) was administered by the intraperitoneal
(i.p.) route (50 mg/kg, injection volume of 1 mL/kg), suspended in
saline with the aid of no more than 5% (v/v) polysorbate 80. The
experimental groups consisted of 5 animals each. Samples of 2.0 mL
plasma were collected at 0.5, 2, 6, 12, 24 h after intraperitoneal
administration and frozen until ready to be analyzed. Samples were
4.12.3. Binding assays
For the binding to D2-like receptors, striatal membranes from
rat brain (50
m
g protein) and [3H]-YM-09151-2 (0.1 nM) were
incubated at 37 ꢂC for 60 min in a solution containing 120 mM NaCl,
5 mM KCl, 5 mM MgCl2, 1.5 mM CaCl2, 1 mM EDTA and TriseHCl
50 mM (pH 7.4) in a final volume of 500
was estimated in the presence of 30
M (ꢀ)-sulpiride. For 5-HT1A
receptors, hippocampal membranes from rat brain (50 g protein)
and 1 nM [3H]-8eOHeDPAT were incubated at 37 ꢂC for 15 min in
a solution containing 1 mM CaCl2, 1 mM MnCl2, 10 M pargyline
and TriseHCl 50 mM (pH 7.4), in a final volume of 500 L. Non-
specific binding was estimated in the presence of 10 M 5-HT. For
mL. Non-specific binding
m
m
deproteinized by the addition of 200
25 s followed by centrifugation at 13000 rpm for 30 s. The super-
natant was filtered in a MillexÒ system and 20
L was injected into
mL of methanol and mixed for
m
m
m
m
HPLC for analysis at the same conditions. The protocols for animal
studies were approved by the Ethics Committee of the Federal
University of Rio Grande do Sul (UFRGS, Porto Alegre, Brazil).
5-HT2A receptors, cortical membranes from rat brain (150 mg
protein) and [3H]-ketanserin (1 nM) were incubated at 37 ꢂC for
15 min in a solution containing 100 nM prazosin and TriseHCl
50 mM (pH 7.4), in a final volume of 500
was estimated in the presence of 1 M ketanserin. For the binding
to D4 receptors, membrane preparations of cells transfected with
mL. Non-specific binding
4.12. Pharmacologic evaluation
m
4.12.1. Radioligands and drugs
human receptors (ChemiscreenÔ, Millipore) (5 mg protein) and
[3H]-YM-09151-2 (82.7 Ci/mmol), [3H]-8eOHeDPAT (187 Ci/
mmol), [3H]-ketanserin (60 Ci/mmol) [3H]-RX821002 (60 Ci/mmol),
[3H]-QNB (48 Ci/mmol), [3H]-prazosin (60 Ci/mmol) and [3H]-
mesulergine (84.3 Ci/mmol) were purchased from New England
Nuclear Life Science Products, PerkinElmer, USA. Ketanserin
tartrate, pargyline, prazosin hydrochloride, 5-HT creatinine sulfate,
(ꢀ)-sulpiride, (ꢀ)-epinephrine and atropine sulfate were
purchased from Sigma, São Paulo, Brazil. In the binding experi-
ments, we used the metabolite (4) obtained by chemical synthesis.
The test substances and (ꢀ)-sulpiride were dissolved in DMSO,
prazosin was dissolved in ethanol. Other agents were dissolved in
purified water to obtain stock solutions. Subsequent dilutions were
prepared in the incubation buffer. At the final concentration used,
DMSO and ethanol had no effect in our assays.
[3H]-YM-09151-2 (0.1 nM) were incubated at room temperature for
120 min in a solution containing 120 mM NaCl, 5 mM KCl, 5 mM
MgCl2, 1.5 mM CaCl2, 1 mM EDTA and TriseHCl 50 mM (pH 7.4) in
a final volume of 500
presence of 10 M clozapine (1).
For 5-HT2C receptors, membrane preparations of cells trans-
mL. Non-specific binding was estimated in the
m
fected with human receptors (ChemiscreenÔ, Millipore) (5
mg
protein) and 1 nM [3H]-mesulergine were incubated at 37 ꢂC for
60 min in a solution containing 50 mM HEPES (pH 7.4), 5 mM MgCl2
and 1 mM CaCl2 in a final volume of 500
mL. Non-specific binding
was estimated in the presence of 100 M 5-HT.
m
For a2A receptors, cortical membranes from rat brain (150 mg
protein) and 1 nM [3H]-RX821002 were incubated at 30 ꢂC for 45 min
in a solution containing TriseHCl 50 mM (pH 7.4), in a final volume of
500
100
(150
m
m
L. Non-specific binding was estimated in the presence of
M (ꢀ)-epinephrine. For a1B receptors, rat liver membranes
m
g protein) and 0.1 nM [3H]-prazosinwere incubated at 30 ꢂC for
4.12.2. Tissue preparation
Adult male Wistar rats (200e300 g) were killed by decapitation.
The brains were immediately removed on ice and hippocampus,
striatum and cortex were dissected, weighted and stored in liquid
nitrogen until use. This procedure was approved by the Institu-
tional Ethical Committee for Animal Care from the Federal
University of Rio de Janeiro. Striatum were homogenized in a Potter
apparatus with a motor-driven Teflon pestle at 4 ꢂC in 20 volumes
per gram of tissue of ice-cold TriseHCl 50 mM buffer (pH 7.4)
containing MgCl2 8 mM and EDTA 5 mM. The resulting suspension
was ultracentrifuged at 48000 gav at 4 ꢂC for 20 min. The pellet was
resuspended in 20 volumes of buffer and incubated at 37 ꢂC during
10 min for endogenous neurotransmitters removal. This suspension
was cooled on ice and ultracentrifuged twice at 48000 gav for
20 min at 4 ꢂC. The final pellet was resuspended in buffer yielding
a proportion of 1.5 mL/g tissue and stored in liquid nitrogen until
use. Hippocampus and cortex were homogenized in a Potter
apparatus with a motor-driven Teflon pestle at 4 ꢂC in 20 volumes
(hippocampus) or 10 volumes (cortex) of ice-cold TriseHCl 50 mM
(pH 7.4) buffer per gram of tissue. The resulting suspension was
45 min in a solution containing EDTA 1 mM and TriseHCl 50 mM (pH
7.4), in a final volume of 500 L. Non-specific binding was estimated
in the presence of 1 M prazosin. For the muscarinic receptors,
cortical membranes from rat brain (150
g protein) and 0.1 nM [3H]-
QNB were incubated at room temperature for 60 min in a solution
containing TriseHCl 50 mM (pH 7.4), in a final volume of 500 L. Non-
specific binding was estimated in the presence of 1 M atropine
sulfate. In all assays, different concentrations of (3) and (4) were used
and a 30 M value was assumed as the cut off for evaluation. After
incubation, samples were rapidly diluted with 3 ꢁ 4 mL TriseHCl
5 mM and immediately filtered under vacuum on glass fiber filters
(GMF 3, Filtrak, Germany) previously soaked in 0.5% poly-
ethyleneimine (only for the binding to 5-HT1A, 5-HT2A and a2
receptors). Filters were then dried and immersed in a scintillation
mixture (POPOP (1,4-bis-[2-(5-phenyloxazolyl)]-benzene) 0.1 g/L
and POP (2,5-diphenyloxazole) 4.0 g/l in toluene). The radioactivity
retained in the filters was counted with a Packard Tri-Carb 1600 TR
liquid scintillation analyzer.
m
m
m
m
m
m

T.F. Gomes et al. / European Journal of Medicinal Chemistry 62 (2013) 214e221
221
[11] E.O. Carneiro, V. de Oliveira, R. Menegatti, C.A.M. Fraga, E.J. Barreiro, Braz. J.
Pharm. Sci. 41 (2005). pp. 392e392.
[12] E.M.M.B. Costa, F.C. Pimenta, W.C. Luz, V. de Oliveira, Braz. J. Microbiol. 39
(2008) 405e408.
[13] F. Pazini, R. Menegatti, J.R. Sabino, C.H. Andrade, G. Neves, S.M.K. Rates,
F. Noël, C.A.M. Fraga, E.J. Barreiro, V. de Oliveira, Bioorg. Med. Chem. Lett. 20
(2010) 2888e2891.
[14] E.O. Carneiro, C.H. Andrade, R.C. Braga, A.C. Tôrres, R.O. Alves, L.M. Lião,
C.A.M. Fraga, E.J. Barreiro, V. de Oliveira, Bioorg. Med. Chem. Lett. 15 (2010)
3734e3736.
4.12.4. Statistical analysis
The median inhibitory concentrations (IC50) were estimated
from the competition curves, using a computerized non-linear
regression analysis of the untransformed data (Prism 4.0, Graph-
Pad Software Inc.), assuming a single population of binding sites.
The Ki values were calculated using the ChengePrusoff equation:
Ki ¼ IC50/(1 þ [radioligand]/Kd).
[15] J.-Y. Park, D. Harris, J. Med. Chem. 46 (9) (2003) 1645e1660.
[16] H. Sun, D.O. Scott, Chem. Biol. Drug Des. 75 (1) (2010) 3e17.
[17] A.M. Clark, P. Labute, J. Chem. Inf. Model. 47 (5) (2007) 1933e1944.
[18] P. Labute, J. Comput. Chem. 29 (10) (2008) 1693e1698.
[19] O.V. Olesen, K. Linnet, J. Clin. Pharmacol. 41 (2001) 823e832.
[20] C. Knox, V. Law, T. Jewison, P. Liu, S. Ly, A. Frolkis, A. Pon, K. Banco, C. Mak,
V. Neveu, Y. Djoumbou, R. Eisner, A.C. Guo, D.S. Wishart, Nucleic Acids Res. 39
(Database issue) (2011) D1035eD1041.
[21] R.V. Smith, J.P. Rosazza, J. Pharm. Sci. 11 (1975) 1737e1758.
[22] D. Zhang, H. Zhang, N. Aranibar, R. Hanson, Y. Huang, P.T. Cheng, S. Wu,
S. Bonacorsi, M. Zhu, A. Swaminathan, W.G. Humphreys, Drug Metab. Dispos.
34 (2006) 267e280.
[23] J.D. Moody, J.P. Freeman, P.P. Fu, C.E. Cerniglia, Drug Metab. Dispos. 30 (2002)
1274e1279.
[24] G. Luo, G.K. Mattson, M.A. Bruce, H. Wong, B.J. Murphy, D. Longhi, I. Antal-
Zimanyi, G.S. Poindexter, Bioorg. Med. Chem. Lett. 14 (24) (2004) 5975e
5978.
[25] L.J. Farrugia, N. Mac Donald, Polyhedron 17 (14) (1998) 2877e2887.
[26] W. Ji Yuan, T. Yasuhara, T. Shingo, K. Muraoka, T. Agari, M. Kameda, T. Uozumi,
N. Tajiri, T. Morimoto, M. Jing, T. Baba, F. Wang, H. Leung, T. Matsui,
Y. Miyoshi, Isao Date, BMC Neurosci. 9 (2008) 75.
[27] G. Neves, R. Menegatti, C.B. Antonio, L.R. Grazziottin, R.O. Vieira,
S.M.K. Rates, F. Noël, E.J. Barreiro, C.A.M. Fraga, Bioorg. Med. Chem. 18
(2010) 1925e1935.
[28] K. Herrick-Davis, E. Grinde, M.J. Teitler, Pharmacol. Exp. Ther. 295 (2000)
226e232.
[29] M.D. Wood, C. Heidbreder, C. Reavill, C.R. Ashby Jr., D.N. Middlemiss, 5-HT2C,
Supplementary material
Crystallographic data for the structural analysis have been
deposited at the Cambridge Crystallographic Data Centre, under
CCDC No. CCDC 838359 and CCDC 838358, for (3) and compound
(4), respectively. Copies of this information may be obtained free of
charge from The Director, CCDC, 12 Union Road, Cambridge, CB2
1EZ, UK (fax: þ44-1223-336-033; e-mail: deposit@ccdc.cam.ac.uk
or http://www.ccdc.cam.ac.uk).
Acknowledgments
The authors thank FAPEG, CAPES, CNPq and INCT-INOFAR
(#573.564/2008-6) for financial support and fellowships.
References
[1] D.M. Gardner, R.J. Baldessarini, P. Warraich, Can. Med. Assoc. J. 172 (2005)
1703e1711.
[2] H.Y. Meltzer, Clin. Neurosci. 3 (1995) 64e75.
[3] A. Farah, J. Clin. Psychiatry 7 (2005) 268e274.
[4] D.G.C. Owens, Drugs 51 (1996) 895e930.
[5] T. Hirose, Y. Uwahodo, S. Yamada, T. Miwa, T. Kikuchi, H. Kitagawa, K.D. Burris,
C.A. Altar, T. Nabeshima, J. Psychopharmacol. 18 (2004) 375e383.
[6] R. Menegatti, A.C. Cunha, V.F. Ferreira, E.F.R. Perreira, A. El-Nabawi,
A.T. Eldefrawi, E.X. Albuquerque, G. Neves, S.M.K. Rates, C.A.M. Fraga,
E.J. Barreiro, Bioorg. Med. Chem. 11 (2003) 4807e4813.
[7] G. Neves, R. Fenner, A.P. Heckler, A.F. Viana, L. Tasso, R. Menegatti,
C.A.M. Fraga, E.J. Barreiro, T. Dalla Costa, S.M.K. Rates, Braz. J. Med. Biol. Res. 36
(2003) 625e629.
Drug Dev. Res. 54 (2001) 88e94.
[30] S. Leucht, C. Corves, D. Arbter, R.R. Engel, C. Li, J.M. Davis, The Lancet. 373
(2009) 31e41.
[31] S. Sansen, J.K. Yano, R.L. Reynald, G.A. Schoch, K.J. Griffin, C.D. Stout,
E.F. Johnson, J. Biol. Chem. 282 (2007) 14348e14355.
[32] D.E. Shaw, P. Maragakis, K. Lindorff-Larsen, S. Piana, R.O. Dror, M.P. Eastwood,
J.A. Bank, J.M. Jumper, J.K. Salmon, Y. Shan, W. Wriggers, Science 330 (6002)
(2010) 341e346.
[33] W.L. Jorgensen, D.S. Maxwell, J. Tirado-Rives, Development and testing of the
OPLS all-atom force field on conformational energetics and properties of
organic liquids, J. Am. Chem. Soc. 118 (45) (1996) 11225e11236.
[34] K. Harms, S. Wocadlo, XCAD4, University of Marburg, Germany, 1995.
[35] Bruker, SMART, APEX2, SAINT, Bruker AXS Inc., Madison, Wisconsin, USA,
2007.
[8] G. Neves, M. Kliemann, A.H. Betti, D.J. Conrado, L. Tasso, C.A.M. Fraga, E.J. Barreiro,
T. Dalla Costa, S.M.K. Rates, Pharmacol. Biochem. Behav. 89 (2008) 23e30.
[9] D.J. Conrado, H. Verli, G. Neves, C.A. Fraga, E.J. Barreiro, S.M. Rates, T. Dalla
Costa, J. Pharm. Pharmacol. 60 (2008) 699e707.
[10] L.E.S. Dias, C.H. Andrade, F. Pazini, V. De Oliveira, Braz. J. Pharm. Sci. 41 (2005).
pp. 133e133.