Effect of phosphodiesterase 7 (PDE7) inhibitors in experimental autoimmune encephalomyelitis mice. Discovery of a new chemically diverse family of compounds.

Abstract

Phosphodiesterase (PDE) 7 is involved in proinflammatory processes, being widely expressed both on lymphocytes and on certain brain regions. Specific inhibitors of PDE7 have been recently reported as potential new drugs for the treatment of neurological disorders because of their ability to increase intracellular levels of cAMP and thus to modulate the inflammatory process, as a neuroprotective well-established strategy. Multiple sclerosis is an unmet disease in which pathologies on the immune system, T-cells, and specific neural cells are involved simultaneously. Therefore, PDE7 inhibitors able to interfere with all these targets may represent an innovative therapy for this pathology. Here, we report a new chemically diverse family of heterocyclic PDE7 inhibitors, discovered and optimized by using molecular modeling studies, able to increase cAMP levels in cells, decrease inflammatory activation on primary neural cultures, and also attenuate the clinical symptoms in the experimental autoimmune encephalomyelitis (EAE) mouse model. These results led us to propose the use of PDE7 inhibitors as innovative therapeutic agents for the treatment of multiple sclerosis.

Full text

Page 1

Effect of Phosphodiesterase 7 (PDE7) Inhibitors in Experimental
Autoimmune Encephalomyelitis Mice. Discovery of a New Chemically
Diverse Family of Compounds
Miriam Redondo,†José Brea,‡Daniel I. Perez,†Ignacio Soteras,†Cristina Val,‡Concepción Perez,†
Jose A. Morales-García,§,⊥Sandra Alonso-Gil,§,⊥Nuria Paul-Fernandez,∥,⊥Rocío Martin-Alvarez,∥,⊥
María Isabel Cadavid,‡María Isabel Loza,‡Ana Perez-Castillo,§,⊥Guadalupe Mengod,∥,⊥
Nuria E. Campillo,†Ana Martinez,†and Carmen Gil*,†
†Instituto de Química Médica (CSIC), Juan de la Cierva 3, 28006, Madrid, Spain
‡Instituto de Farmacia Industrial, Facultad de Farmacia, Universidad de Santiago de Compostela, Campus Universitario Sur s/n,
15782 Santiago de Compostela, Spain
§Instituto de Investigaciones Biomédicas (CSIC-UAM), Arturo Duperier 4, 28029, Madrid, Spain
∥Instituto de Investigaciones Biomédicas de Barcelona (CSIC-IDIBAPS), Rosselló 161, 08036 Barcelona, Spain
⊥Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Spain
*
S Supporting Information
ABSTRACT: Phosphodiesterase (PDE) 7 is involved in proinflammatory processes,
being widely expressed both on lymphocytes and on certain brain regions. Specific
inhibitors of PDE7 have been recently reported as potential new drugs for the treatment of
neurological disorders because of their ability to increase intracellular levels of cAMP and
thus to modulate the inflammatory process, as a neuroprotective well-established strategy.
Multiple sclerosis is an unmet disease in which pathologies on the immune system, T-cells,
and specific neural cells are involved simultaneously. Therefore, PDE7 inhibitors able to
interfere with all these targets may represent an innovative therapy for this pathology.
Here, we report a new chemically diverse family of heterocyclic PDE7 inhibitors, dis-
covered and optimized by using molecular modeling studies, able to increase cAMP levels
in cells, decrease inflammatory activation on primary neural cultures, and also attenuate the
clinical symptoms in the experimental autoimmune encephalomyelitis (EAE) mouse
model. These results led us to propose the use of PDE7 inhibitors as innovative
therapeutic agents for the treatment of multiple sclerosis.
■INTRODUCTION
The nucleotides cyclic adenosine 3′,5′-monophosphate (cAMP)
and cyclic guanosine 3′,5′-monophosphate (cGMP) are two
ubiquitous second messengers that mediate a variety of cellular
responses. Two enzymatic processes control the intracellular
levels of these nucleotides: the first one by regulation of their
synthesis, achieved through the action of adenylate or guanylate
cyclase, and the second one through their hydrolysis catalyzed
by phosphodiesterases (PDEs).1Up to now there are 11 known
families of PDEs that control the levels of cAMP and cGMP by
catalyzing their hydrolysis to AMP and GMP, respectively. The
family members are classified according to their sequence
identity, cellular distribution, substrate specificity, and sensitivity to
different PDE inhibitors,2,3being good targets for pharmacological
intervention. In fact, one of the greatest advances in the PDE field
in the past 15 years has been the increased availability and more
recently the clinical use of family selective inhibitors.4Therefore,
PDE inhibitors may have considerable therapeutic utility as anti-
inflammatory agents, antiasthmatics, vasodilators, smooth muscle
relaxants, cardiotonic agents, antidepressants, antithrombotics, and
agents for improving memory and other cognitive functions.5The
members of the PDE superfamily are well placed to be targets for
pharmacological intervention. As elevation of intracellular cAMP
level shows immunosuppressive and anti-inflammatory proper-
ties,1,6selective inhibitors of cAMP-specific PDEs have been
widely studied as therapeutic agents for the treatment of human
diseases.5The PDEs responsible for controlling specifically the
intracellular levels of this nucleotide are PDE4, PDE7, and PDE8.
More information is known about PDE4 being involved in the
hydrolysis of cAMP within immune and central nervous system
(CNS) cells. Thus, PDE4 inhibitors have been widely studied as
efficient anti-inflammatory agents for different diseases.7However,
a major drawback of these compounds is the significant emetic
effects. To overcome these adverse effects, several strategies have
been developed.8An alternative approach is to target different
cAMP specific PDEs families, as PDE79is also expressed in
Received:
Published: March 5, 2012
December 21, 2011
Article
pubs.acs.org/jmc
© 2012 American Chemical Society
3274
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284

Page 2

T-cells and CNS and is a good target for the control of
neuroinflammation.10
The PDE7 family is composed of two genes, PDE7A and
PDE7B. High mRNA concentrations of both PDE7A and
PDE7B are expressed in rat brain and in numerous peripheral
tissues, although the distribution of these enzymes at the
protein levels has not been reported. Within the brain PDE7A
mRNA is abundant in the olfactory bulb, hippocampus, and
several brain-stem nuclei.11The highest concentrations of
PDE7B transcripts in the brain are found in the cerebellum,
dentate gyrus of the hippocampus, and striatum.12,13There is
very little information regarding the physiological functions
regulated by PDE7. It has been shown that PDE7 is involved in
proinflammatory processes and is necessary for the induction of
T-cell proliferation.14Specific inhibitors of PDE7 have been
recently reported as potential new drugs for the treatment of
neurological disorders15by modulation of the inflammation
process, a well-established neuroprotective strategy. Moreover,
multiple sclerosis simultaneously involved pathologies on the
immune system, T-cells, and specific neural cells, such as microglia
and oligodendrocytes. Current pharmacological treatments for
multiple sclerosis have severe drawbacks such as lack of efficacy
and pharmacokinetics properties. PDE7 inhibitors may represent a
well targeted and innovative therapy for this pathology.
Several years ago, our research group was the first to report
the first PDE7 selective inhibitors.16Since then, a lot of effort
has been made to increase potency and selectivity of these com-
pounds, providing a great variety of diverse chemical com-
pounds with interesting pharmacological profiles.17
Following our ongoing research on this field, we have
recently demonstrated that PDE7 inhibitors belonging to the
quinazoline family enhance neuroprotection and decrease neuro-
inflammation in well-characterized cellular and animal models of
Parkinson’s disease,18spinal cord injury,19and stroke.20
With the aim to design potent and specific PDE7 inhibitors,
we developed a pharmacophore model for PDE7A1 inhibitors
using Catalyst/Hypogen program to identify the chemical
features that are responsible for the inhibitory activity.21Four
pharmacophore features, such as one hydrogen bond acceptor,
one aromatic ring, and two hydrophobic aliphatic points, were
identified to be involved in inhibitor−PDE7 interaction (Chart 1).
This pharmacophore model was able to predict the activity of
an external test set of PDE7 inhibitors with a correlation
coefficient of 0.96. New leads identification was carried out by
performing virtual screening using validated pharmacophoric
queries and four chemical databases (Maybridge, Chemical
Diversity, Specs, and National Cancer Institute). Further data
reduction was done employing virtual filters based on distances
(T = 1.5 Å), angles (10°), and Lipinski’s rules of five. According
to this procedure, three new hits (1−3) were found as potential
PDE7A1 inhibitors structurally diverse from those previously
known22(Table 1). Then the biological activity of the hits
was calculated using the catalytic domain of PDE7A1 and a
radiometric assay.
These three hits were selected for further medicinal chemistry
programs aimed to increase not only efficacy but also the ADME
profile to be considered for further drug development in the
neurodegenerative field. Here, the results obtained in the
development of the family derived from the furan 1 and its thera-
peutic potential for multiple sclerosis are reported.
■RESULTS AND DISCUSSION
Chemistry. The 5-(2-methyl-4-nitrophenyl)-2-furylmethyl
3,4,5-triethoxybenzoate (1) has a furan heterocycle between
two substituted phenyl rings moieties. In particular, the
structural modifications proposed are based on the modifica-
tion of the number and nature of substituents attached to both
phenyl rings and the variation of the linker between the furan
and the alkyloxyphenyl ring (Chart 2).
The synthetic routes for the preparation of the furan
derivatives are summarized in Schemes 1 and 2. The 5-phenyl-
2-furylmethyl benzoate derivatives (Scheme 1) were obtained
in one synthetic step after the coupling reaction of the
alkyloxybenzoic acid and the corresponding 5-phenyl-2-furylme-
thanol. These furan alcohols were commercially available with the
only exception being 5-(2,4-dichlorophenyl)-2-furylmethanol (4),
which was synthesized by reduction of the 5-(2,4-dichlorophenyl)-
2-furoic acid with lithium aluminum hydride as described in the
Supporting Information.
Chart 1. Pharmacophore Features and Their Angle and
Distance (Å) Relations
Table 1. PDE7A1 Activities for Derivatives 1−3
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3275

Page 3

The synthesis of benzyl 5-phenylfuroate and benzyl-5-
phenylfuramide derivatives (Scheme 2) involved the coupling
of alkyloxybenzyl alcohol or alkyloxybenzylamine with the
furoic acid derivative. Among different coupling reagents pre-
viously used, benzotriazol-1-yloxytripyrrolidinophosphonium
hexafluorophosphate (PyBOP) was chosen as the most con-
venient one.
The structures of all the synthesized products were unequivocally
confirmed by mass spectrometry, elemental analysis, and1H and
13C NMR spectroscopic data.
In Vitro Evaluation of PDE7 Inhibition. The new
derivatives here synthetized (5−32) were tested for their
inhibitory potencies against PDE7A1 using recombinant human
isoenzyme as described in the Experimental Section. All the
compounds were tested at a fixed concentration (10 μM),
performed in duplicate. The percentage of inhibition obtained
on PDE7A1 for all the compounds at this concentration is
shown in Table 2. From these data we can conclude that several
new compounds are inhibitors of PDE7A1.
As the biological assay conditions differ from those
previously used,22the lead compound 1 was also here evaluated
to compare the data with the new synthesized derivatives.
Noteworthy is the fact that when the lead compound 1 was
evaluated in the new experimental conditions using the
complete enzyme instead of the catalytic domain of PDE7A1,
only 29% of PDE7A1 inhibition at 10 μM was found. Regarding
these new data, modifications performed on the chemical
structure of 1 increase in general the inhibition of PDE7A1.
When the percentage of PDE7A1 inhibition was greater than
50%, the dose−response curve was determined and the IC50
values were calculated (Table 3). Some of these new com-
pounds present IC50values for PDE7A1 inhibition in the low
micromolar range, being suitable candidates to be further explored.
A kinetic study varying cAMP concentration was performed with
two of these compounds (13 and 31) to better characterize the
enzymatic inhibition. Double-reciprocal plotting of the data is
depicted in Figure 1. The intercept of the plot in the horizontal
axis (1/[cAMP]) changes when the 13 and 31 concentrations
increase, whereas the intercept in the vertical axis (1/V) does
not change. These results suggest that furan derivatives act as
competitive inhibitors of the cAMP binding site and allow us to
determine the Ki of these inhibitors (5.91 and 7.22) for
compounds 13 and 31, respectively.
Chart 2. Selected Lead Compound 1 for Further
Optimization and Areas Where Modifications Have Been
Done
Scheme 1a
aReagents: (i) coupling reagent, TEA, rt, CH2Cl2.
Scheme 2a
aReagents: (i) PyBOP, TEA, rt, CH2Cl2.
Table 2. PDE7A1 Inhibition of Compounds (5−32)
compd% inh PDE7A1 @ 10 μM
24.4 ± 13.4
49.5 ± 3.0
48.8 ± 11.1
3.7 ± 2.0
39.9 ± 8.6
16.4 ± 1.1
10.0 ± 6.9
−9.9 ± 1.5
57.9 ± 0.1
54.9 ± 3.6
34.6 ± 2.3
38.3 ± 7.8
8.0 ± 0.1
28.8 ± 0.2
compd% inh PDE7A1 @ 10 μM
49.1 ± 9.5
14.0 ± 1.1
48.0 ± 7.7
14.0 ± 7.8
30.7 ± 14.8
30.8 ± 0.5
31.1 ± 8.7
29.2 ± 0.2
28.0 ± 4.5
44.8 ± 1.2
41.6 ± 6.5
23.5 ± 0.4
70.5 ± 4.6
−10.3 ± 0.6
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3276

Page 4

Docking Studies. In order to gain some insight in the
binding mode of our family of compounds, docking experi-
ments were carried out with help from the Surflex method
implemented in SYBYL according to the procedure described
in the Experimental Section. Considering the kinetic studies
results, the docking studies were focused on the cAMP binding
site. After careful examination of docking solutions two binding
modes that indicate the chemical structure−activity relation-
ships (SARs) in this series of compounds were proposed. In
fact, the different substituents and their location (ortho, meta,
or para) when R3is a phenyl ring, determine which binding
mode is selected.
The mainly binding mode (namely, “A” from now) is
characterized by a hydrophobic interaction of the phenyl group
in R3with Val380 and a NH−π interaction23,24with Gln413,
which is in fact, a conserved residue among PDEs families.25
A clear example of this binding mode is compound 13 (see
Figure 2). In addition to these interactions, the furan moiety
contributes to an increase in inhibition by means of a stacking
interaction with another conserved residue as Phe416.
Furthermore, the 3,4,5-trimethoxyphenyl ligand’s moiety is
located in a pocket where the ligand can benefit from a
hydrogen bond between the methoxy group in R2and the
proton Hεof His256 group (dHε···O= 2.7 Å) and a hydrophobic
interaction with Phe384.
When R1and R2are ethoxy groups (compound 24), the
binding mode remains mode “A” (see Figure 3). The presence
of these bigger substituents forces the ligand to reorient the
ester group, keeping a hydrophobic interaction with Phe384.
This new location causes a weakening of the parallel stacking
interaction between the furan ring and Phe416 (found for
compound 13) because of the displacement of the furan ring in
order to keep the hydrophobic interaction between ligand’s
benzene ring in R3and Val380. As a result, compound 24
shows a lower percentage of PDE7A1 inhibition than com-
pound 13 (Table 2).
Regarding the substituent located in the para position in the
phenyl attached in R3such as a methyl group (compound 23),
the ligand orients this substituent facing it toward the Leu401
adopting thus the binding mode “B” (see Figure 4). This new
orientation avoids potential steric clashes and allows the
establishment of a new hydrophobic interaction between the
methyl group and the side chain of Leu401. Meanwhile the
previous stacking between the furan and Phe416 is mostly
conserved. Furthermore, this new orientation forces the second
phenyl ring to a new position where two new interactions with
Ile323 are established which can help to stabilize the ligand in
this new binding mode.
Finally, when the ester group linker is replaced by an amide
one (compounds 30−32), the conformation of the ligands
Figure 1. Studies of the PDE7A1 inhibition in the absence (control)
and in the presence of two different concentrations of the furan
derivatives 13 and 31 incubating different cAMP concentrations and
measuring AMP formation. Data represent the mean ± SEM (vertical
bars) of triplicate measurements.
Figure 2. Proposed binding mode for compound 13 (PDB code
1ZKL) showing relevant interactions with nearby residues. Distances
are given in angstroms.
Figure 3. Proposed binding mode for compound 24 (for PDB code
1ZKL) showing relevant interactions with nearby residues. Distances
are given in angstroms.
Table 3. IC50for PDE7A1 of Selected Compounds and BRL50481 as Standard Referencea
67 13 141921 31
PDE7A1 IC50(μM)
aIC50(BRL50481) = 0.09 μM.
33.37 ± 8.112.34 ± 3.22 5.17 ± 1.112.63 ± 0.92 7.31 ± 3.5714.07 ± 2.93.20 ± 1.23
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3277

Page 5

dramatically changes by the formation of an intramolecular
hydrogen bond between NH group and furan’s oxygen. This
conformation allows new stabilizing interactions between the
phenyl ring with Leu401’s side chain where it can interact
hydrophobically (Figure 5).
In Vitro Antiinflamatory Studies. Our next step was to
explore whether these new PDE7 inhibitors are able to reduce
the inflammatory response in different cell based assays. Thus,
we used primary cultures of astrocytes and microglia treated
with lipopolysaccharide (LPS), a potent inflammatory agent.
The potential anti-inflammatory activity of the selected
PDE7 inhibitors (7, 13, 14, 19, 21, and 31) was measured by
evaluating the production of nitrites from primary cultured glial
cells (astrocytes and microglia). To assess the safety of the
compounds on cell culture, the astrocytes and microglia
viability were checked using the 3-[4,5-dimethylthiazol-2-yl]-
2,5-diphenyltetrazolium bromide (MTT) test at two different
concentrations of the compounds. As both concentrations used
do not affect the cell survival, they were selected for the
antiinflammatory experiment.
Cultures were incubated with the indicated concentrations of
compounds 7, 13, 14, 19, 21, and 31 for 1 h, and then cells
were treated for 24 h with LPS. When primary astrocytes and
microglial cells were stimulated with LPS (Figures 6 and 7), a
significant induction of nitrite production in the culture
medium was produced, while if the PDE7 inhibitors were in
the medium, a significant decrease of nitrite production was
observed as a consequence of a reduction of the inflamatory cell
response by the drug treatment. We used as standard references
the well-known PDE4 inhibitor rolipram and the PDE7
inhibitor BRL50481.26These compounds showed a neuro-
protective effect that was in accordance with our previously
reported results,20confirming that these cell permeable PDE7
inhibitors are able to decrease neural inflammatory activation.
Candidate Selection for in Vivo Studies. The final aim
of our study was to proof the potential efficacy of these new
PDE7 inhibitors in a multiple sclerosis murine model. The
selection of the candidate for in vivo studies was done based on
pharmacological and pharmacokinetic properties.
Recently, potential cardiotoxic effect of long-term PDE3A
inhibition was reported.27Thus, to avoid adverse effects in
further development steps, we measured the inhibition on
PDE3A of our selected compounds (Table 4).
Considering these new data, almost every derivative tested
inhibit PDE3A at the same concentration that PDE7A1 was
also inhibited, showing a lack of selectivity and opening safety
concerns in further pharmacological development. However,
derivative 13 shows an IC50value 10-fold higher on PDE3A
regarding PDE7A1 (ratio PDE7A1/PDE3A = 0.09), resulting
in selection of this candidate for the next studies.
Figure 4. Proposed binding mode for compound 23 (PDB code
1ZKL) showing relevant interactions with nearby residues. Distances
are given in angstroms.
Figure 5. Proposed binding mode for compound 31 (PDB code
1ZKL) showing relevant interactions with nearby residues. Distances
are in angstroms.
Figure 6. Cell viability and nitrite production. Glial cultures were treated during 16 h with rolipram (30 μM), BRL50481 (BRL, 30 μM), and
compounds 7, 13, 14, 19, 21, 31 (5 or 10 μM), and cell viability was measured. Some cultures were treated with lipopolysaccharide (LPS) in the
presence or absence of compound at indicated concentrations, and the production of nitrites was measured as indicated in the Experimental Section.
Values represent the mean ± SD of six replications in three different experiments: (∗∗∗) p < 0.001 versus LPS-treated cells.
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3278

Page 6

To evaluate isoenzyme specificity of 13, inhibition of
different cAMP isoenzymes was determined (Table 5). While
marginal inhibition was found on PDE4B2, no activity was
found on PDE7B and PDE4D3. The partial inhibition of
PDE4B2 may increase the biological action of 13 in vivo
because cAMP levels might be enhanced. Moreover, it is worth
mentioning that inhibition of PDE4D3 is associated with the
adverse emetic effects in the development of PDE4 inhibitors.28
As derivative 13 does not inhibit this specific isoenzyme, its
safety potential regarding PDEs inhibition is reinforced.
We also check the ability of derivative 13 to regulate
intracellular cAMP levels in cell cultures. Two different
compound concentrations (10 and 30 μM) were used, and in
both cases we observed an increase of cAMP even more effective
than the reported PDE7 inhibitor BRL5048126(Figure 8).
Finally, to decipher if this compound has the drug profile to
be administered in vivo, we determined its ability to cross the
blood−brain barrier (BBB), one of the major obstacles for
the treatment of diseases in the central nervous system. The
majority of compounds enter the brain by transcellular passive
diffusion, which is driven by a concentration gradient between
the blood and the brain.29Parallel artificial membrane
permeability assay (PAMPA) is a high throughput technique
developed to predict passive permeability through biological
membranes. Here, we used the PAMPA-BBB method described
by Di et al.,30employing a brain lipid porcine membrane to
determine the ability of compound 13 to penetrate into the
brain. The in vitro permeabilities (Pe) of commercial drugs
through lipid membrane extract together with compound 13
were determined and described. A good correlation between
the experimental and described values, Pe(exp) = 1.1512
(bibl) − 0.8973 (R2= 0.9779), was obtained (see Supporting
Information). From this equation, the Pe(exp) for 13 is (8.1 ±
0.1) × 10−6cm s−1. Following the pattern established in the
literature for BBB permeation prediction31that classifies
compounds as CNS+ when they present a permeability of
>3.71 × 10−6cm s−1, we can consider that compound 13 is able
to cross the BBB by passive permeation.
In Vivo Studies. Experimental Autoimmune Encepha-
lomyelitis Model. The results found in cell cultures for the
new PDE7 inhibitor 13 together with their ability to cross the
BBB prompted us to evaluate it in chronic experimental
autoimmune encephalomyelitis (EAE) mice, a well establish
murine model for multiple sclerosis.
EAE was induced in C57BL/6J mice by immunization with
MOG35−55in complete Freund’s adjuvant on day 0. Clinical
signs and score were monitored up to day 41. Mice began to
show neurological deficits on day 12, reaching a maximum
Figure 7. Cell viability and nitrite production. Astrocytes cultures were treated during 16 h with rolipram (30 μM), BRL50481 (BRL, 30 μM), and
compounds 7, 13, 14, 19, 21, 31 (5 or 10 μM), and cell viability was measured. Some cultures were treated with lipopolysaccharide (LPS) in the
presence or absence of compound at indicated concentrations, and the production of nitrites was measured as indicated in the Experimental Section.
Values represent the mean ± SD of six replications in three different experiments: (∗∗∗) p < 0.001 versus LPS-treated cells.
Table 4. PDE3A Inhibition for Selected Compounds
compd
% inh PDE3A @
10 μM
3.0 ± 12.7
46.1 ± 3.0
74.9 ± 5.4
11.0 ± 4
57.1 ± 12
49.6 ± 1.2
IC50(PDE3A)
(μM)
104.2 ± 36.6
54.6 ± 12.1
1.8 ± 0.85
nd
3.4 ± 1.3
3.9 ± 0.7
ratio PDE7A1/
PDE3A
7
13
14
19
21
31
0.12
0.09
1.44
nd
4.15
0.82
Table 5. IC50(μM) on Different PDEs for Compound 13
compd PDE7A1
5.1±1.11
PDE7BPDE3A
54.6 ± 8.1 12.4 ± 4.2 66.3 ± 13.9
PDE4B2 PDE4D3
1323.2 ± 4.4% @ 10 μM
Figure 8. Intracellular cAMP level in raw cells treated with compound
13 at 30 and 10 μM.
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3279

Page 7

score around day 16. A therapeutic regimen of administration
was chosen to test the PDE7 inhibitor in the EAE model. Thus,
a daily intraperitoneal (i.p.) administration of compound 13 for
26 days started at day 5 after disease onset. The dose was
selected considering the IC50on the target and the cellular anti-
inflammatory activity. As we can see in Figure 9A, a clear and
significant attenuation of clinical symptoms during the first 16
days of treatment (day 32 after immunization) was observed.
However, this behavior is lost in the last 10 days of treatment.
The body weight curve (Figure 9B), reflects these findings, and
the compound 13-treated mice start to lose weight in the same
timeframe as clinical scores decrease (described above).
This experiment showed for the first time efficacy of PDE
inhibitors on a well established model of multiple sclerosis. Com-
pound 13 administrated to the EAE when the worst neurological
score was measured showed a recovery of the clinical symptoms
reaching a statistically significant plateau of maximum effect from
day 21 to 31 post immunization. However, this attenuation of the
symptoms was counterbalanced by the worsening of the animals
during the last 10 days of treatment until day 42, when the animals
were sacrificed. To decipher if this effect is due to the
administrated compound (stability, metabolism, etc.) or to the
animal model, further studies are in progress.
■CONCLUSIONS
PDE7 inhibitors represent a new class of innovative drugs with
great potential for several neurological disorders. Their efficacy
in animal models of Parkinson disease, stroke, and spinal cord
injury has been recently reported, while their value for the
treatment of multiple sclerosis is here disclosed. We presented
the medicinal chemistry program around one hit found by
using a pharmacophore model. The work led to a novel series
of PDE7 inhibitors, chemically diverse from those known and
containing a furan ring in their chemical structures.
The biological profile has been well characterized showing an
increase of cAMP level and a reduction of the inflammatory
response in primary neural cell cultures. Moreover, a clear
strategy to select the best compound for in vivo testing, using
pharmacodynamic and pharmacokinetic criteria, was followed.
Finally, compound 13 was chosen because of its PDE7 selectivity
and its ability to cross the BBB. This compound is able to reverse
clinical symptoms in an EAE mice model. Together, these data
supported the potential of PDE7 inhibitors for the therapeutic
pharmacological treatment of multiple sclerosis.
■EXPERIMENTAL SECTION
Chemical Procedures. Substrates were purchased from commer-
cial sources and used without further purification. Melting points were
determined with a Mettler Toledo MP70 apparatus. Flash column
chromatography was carried out at medium pressure using silica gel
(E. Merck, grade 60, particle size 0.040−0.063 mm, 230−240 mesh
ASTM) with the indicated solvent as eluent. Compounds were
detected with UV light (254 nm).1H NMR spectra were obtained on
the Bruker AVANCE-300 spectrometer working at 300 MHz or on a
Varian INOVA 400 spectrometer working at 400 MHz. Typical
spectral parameters were as follows: spectral width 10 ppm, pulse
width 9 μs (57°), data size 32 K.13C NMR experiments were carried
out on a Bruker AVANCE-300 spectrometer operating at 75 MHz or
on a Varian INOVA 400 spectrometer working at 100 MHz. The
acquisition parameters were as follows: spectral width 16 kHz,
acquisition time 0.99 s, pulse width 9 μs (57°), data size 32 K.
Chemical shifts are reported in values (ppm) relative to internal Me4Si,
and J values are reported in Hz. Elemental analyses were performed by
the Analytical Department at CENQUIOR (CSIC), and the results
obtained were within ±0.4% of the theoretical values.
5-(4-Phenyl)-2-furylmethyl 3,4,5-Triethoxybenzoate (7).
PyBOP (358 mg, 0.68 mmol) was added as coupling reagent to a
solution of 3,4,5-triethoxybenzoic acid (175 mg, 0.68 mmol) with TEA
(158 μL, 1.15 mmol) in CH2Cl2(10 mL) followed by 5-phenyl-2-
furylmethanol (100 mg, 0.57 mmol), and the resulting mixture was
stirred at 25 °C during 12 h. The reaction mixture was evaporated in
vacuo and the solid was purified by column chromatography on silica
using hexane/ethyl acetate (8:1) as eluent to afford a yellow solid
(246 mg, 87% yield), mp 71.1 °C.1H NMR (300 MHz, CDCl3): δ
7.70−7.29, 6.64, 6.56, 5.33, 4.10, 1.42, 1.35.
CDCl3): δ 166.5, 153.0, 149.5, 149.1, 129.2, 125.7, 128.1, 125.7, 124.3,
124.4, 113.7, 108.8, 106.3, 69.4, 65.2, 59.2, 16.0, 15.2. MS (ES, [M + K]+):
m/z = 449. Anal. (C25H28O5) C, H, O.
3,4,5-Trimethoxybenzyl 5-Phenyl-2-furoate (13). PyBOP
(639 mg, 1.23 mmol) was added as coupling reagent to a solution
of 5-phenyl-2-furoic acid (200 mg, 1.03 mmol) with TEA (284 μL,
2.06 mmol) in CH2Cl2(10 mL) followed by 3,4,5-trimethoxybenzyl
alcohol (210 μL, 1.27 mmol), and the resulting mixture was stirred at
25 °C during 23 h. The reaction mixture was evaporated in vacuo and
the solid was purified by column chromatography on silica using
hexane/ethyl acetate (8:1) as eluent to afford a white solid (90 mg,
24% yield), mp 98.4 °C.1H NMR (300 MHz, CDCl3): δ 7.79, 7.45−
7.26, 6.69, 5.29, 3.93−3.85.13C NMR (75 MHz, CDCl3): δ 159.0,
158.2, 153.7, 143.9, 138.5, 131.7, 129.8, 129.4, 129.2, 125.3, 120.7,
107.3, 106.1, 67.1, 61.2, 56.6. MS (ES, [M + Na]+): m/z = 391. Anal.
(C21H20O6) C, H, O.
13C NMR (75 MHz,
Figure 9. Therapeutic effect of compound 13 in MOG35-55-induced
EAE in C57BL6J mice. Fourteen C57BL6J mice immunized with
MOG35-55 at 100 μg/mouse in the presence of complete Freund’s
adjuvant developed clinical signs on day 12. Animals were divided into
two groups. One group was treated with compound 13 starting from
day 5 after the onset of the disease (arrowhead) for 26 days. Mice in
the control group were administered vehicle only: (A) EAE scores in
drug treatment; (B) percentage of body weight variation. Results were
expressed as the mean ± standard error of the mean, and statistical
differences in EAE scores between the vehicle and the treatment group
were calculated by Mann−Whitney test (∗, p < 0.0002): (●) vehicle
(n = 7); (▲) 10 mg/kg compound 13 (n = 7) i.p. daily.
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3280

Page 8

3,4,5-Trimethoxybenzyl 5-(2-Trifluoromethylphenyl)-2-
furoate (14). PyBOP (393 mg, 0.76 mmol) was added as coupling
reagent to a solution of 5-(2-trifluoromethylphenyl)-2-furoic acid (200
mg, 0.76 mmol) with TEA (174 μL, 1.26 mmol) in CH2Cl2(10 mL)
followed by 3,4,5-trimethoxybenzyl alcohol (129 μL, 0.65 mmol), and
the resulting mixture was stirred at 25 °C during 12 h. The reaction
mixture was evaporated in vacuo and the solid was purified by column
chromatography on silica using hexane/ethyl acetate (6:1) as eluent to
afford a white solid (123 mg, 36% yield), mp 89.8 °C.1H NMR (300
MHz, CDCl3): δ 7.80, 7.68, 7.53, 7.30, 6.79, 6.69, 5.30, 3.89, 3.86.13C
NMR (75 MHz, CDCl3): δ 158.8, 154.5, 153.7, 144.8, 138.5, 132.3,
131.6, 131.1, 129.5, 128.8, 127.3, 124.1, 120.2, 112.4, 106.1, 67.2, 61.3,
56.5. MS (ES, [M + Na]+): m/z = 459. Anal. (C22H19F3O6) C, H, O.
3,4,5-Trimethoxybenzyl 5-(4-Chloro-2-nitrophenyl)-2-
furoate (19). PyBOP (369 mg, 0.71 mmol) was added as coupling
reagent to a solution of 5-(4-chloro-2-nitrophenyl)-2-furoic acid (200 mg,
0.71 mmol) with TEA (164 μL, 1.19 mmol) in CH2Cl2(10 mL)
followed by 3,4,5-trimethoxybenzyl alcohol (101 μL, 0.61 mmol), and
the resulting mixture was stirred at 25 °C during 12 h. The reaction
mixture was evaporated in vacuo and the solid was purified by column
chromatography on silica using hexane/ethyl acetate (6:1) as eluent to
afford a white solid (130 mg, 38% yield), mp 147.3 °C.1H NMR (300
MHz, CDCl3): δ 8.37−7.87, 7.86, 7.30, 5.87, 4.48, 4.44.13C NMR (75
MHz, CDCl3): δ 158.5, 153.8, 151.1, 148.4, 145.6, 138.5, 136.1, 132.8,
131.4, 131.2, 124.9, 122.0, 124.8, 120.2, 112.2, 106.0, 67.4, 61.3, 56.6.
MS (ES, [M + Na]+): m/z = 470. Anal. (C25H28O5) C, H, O.
3,4,5-Trimethoxybenzyl 5-(4-Nitrophenyl)-2-furoate (21).
PyBOP (509 mg, 0.98 mmol) was added as coupling reagent to a
solution of 5-(4-nitrophenyl)-2-furoic acid (200 mg, 0.82 mmol) with
TEA (227 μL, 1.64 mmol) in CH2Cl2(10 mL) followed by 3,4,5-
trimethoxybenzyl alcohol (168 μL, 0.98 mmol), and the resulting
mixture was stirred at 25 °C during 24 h. The reaction mixture was
evaporated in vacuo and the solid was purified by column
chromatography on silica using hexane/ethyl acetate (4:1) as eluent
to afford a white solid (34 mg, 10% yield), mp 161.2 °C.1H NMR
(300 MHz, CDCl3): δ 7.52, 7.93, 7.31, 6.95, 6.69, 5.30, 3.93−3.86.13C
NMR (75 MHz, CDCl3): δ 158.6, 155.4, 153.8, 147.9, 145.6, 143.2,
135.4, 131.4, 125.7, 124.8, 120.6, 110.5, 106.3, 67.5, 61.3, 56.6. MS
(ES, [M + Na]+): m/z = 436. Anal. (C21H19NO8) C, H, N, O.
3,4,5-Trimethoxybenzyl-5-(4-methyl-2-nitrophenyl)-2-furamide
(31). PyBOP (618 mg, 1.19 mmol) was added as coupling reagent to a
solution of 5-(4-methyl-2-nitrophenyl)-2-furoic acid (300 mg, 1.19
mmol) with TEA (274 μL, 1.98 mmol) in CH2Cl2(10 mL) followed
by 3,4,5-trimethoxybenzylamine (173 μL, 0.99 mmol), and the
resulting mixture was stirred at 25 °C during 12 h. The reaction
mixture was evaporated in vacuo and the solid was purified by column
chromatography on silica using hexane/ethyl acetate (3:1) as eluent to
afford a yellow solid (56 mg, 14% yield), mp 114.2 °C.1H NMR (300
MHz, CDCl3): δ 7.49, 7.34, 7.15, 6.65, 6.51, 4.48, 3.81, 3.75, 2.39.13C
NMR (75 MHz, CDCl3): δ 158.2, 153.9, 150.6, 148.4, 148.2, 141.1,
137.3, 133.9, 133.2, 125.2, 124.8, 120.7, 116.6, 111.5, 105.1, 61.3, 56.6,
43.8, 21.5. MS (ES, [M + H]+): m/z = 427. Anal. C22H22N2O7(C, H,
N, O).
Radiometric Phosphodiesterase Inhibition Assay. The
methodology used for measuring human recombinant PDE7A1,
PDE7B, and PDE3A activity was based on a scintillation proximity
assay (SPA) from Perkin-Elmer (TRKQ7090). The activity of the
phosphodiesterase is measured by co-incubating the enzyme with
[3H]cAMP, and the hydrolysis of the nucleotide is quantified by
radioactivity measurement after binding of [3H]AMP to scintillation
binding bead.
Either 0.02 unit of PDE7A1 (Calbiochem no. 524751), 0.02 unit of
PDE3A (Calbiochem no. 524742), or 0.5 unit of PDE7B (Abcam no.
ab79800) was incubated in a 96-well Flexiplate with 5 nCi [3H]cAMP
and inhibitors in 100 μL of assay buffer (contained in the kit) for
20 min at 30 °C. After the incubation time, an amount of 50 μL of a
solution of SPA beads (approximately 1 mg per well) was added to
each well and the plate was shaken for 1 h at room temperature.
Finally, beads were settled for 30 min and radioactivity was detected in
a Microbeta Trilux reader.
IC50values were calculated by nonlinear regression fitting using
GraphPad Prism. Data (radioactivity vs log concentration) were fitted
to a sigmoidal dose−response equation: Y = Bottom + (Top −
Bottom)/(1 + 10((log IC50−X)n)), where Bottom and Top were the
minimum and maximal inhibition for PDE, respectively, IC50was the
concentration of compound that inhibited the PDE activity at 50%,
and n was the slope of the concentration−response curve.
The mode of inhibitory action for compounds 13 and 31 was
determined by varying the concentration of unlabeled cAMP in the
reaction cocktail within the range of 5 nM to 2 μM in the presence of a
fixed concentration of [3H]cAMP tracer and inhibitor concentrations.
Enzyme activity data were analyzed using Lineweaver−Burk plots
using GraphPad Prism.
Fluorescence Polarization Phosphodiesterase Inhibition
Assay. The ability of compounds to inhibit PDE4B2 (human
recombinant) and PDE4D3 (human recombinant) was determined
by IMAP fluorescence polarization (FP) assay (Molecular Devices
R8175).
0.05 U of PDE4B2 (Calbiochem no. 524736) and PDE4D3
(Calbiochem no. 524733) were incubated in a 96-well black half-area
plate with 1 nM fluorescein adenosine 3′,5′-cyclic phosphate
(contained in the kit) and inhibitors in 40 μL of assay buffer. Plates
were mixed on a shaker for 10 s and incubated at ambient temperature
for 60 min. IMAP binding reagent was added (60 μL of a 1 in 600
dilution in binding buffer of the kit stock solution) to terminate the
assay. Plates were allowed to stand at ambient temperature for 1 h.
The FP ratio was measured with a Tecan Ultra Evolution reader.
IC50values were calculated by nonlinear regression fitting using
GraphPad Prism. Data (fluorescence polarization vs log concen-
tration) was fitted to a sigmoidal dose−response equation: Y = Bottom +
(Top − Bottom)/(1 + 10((log IC50−X)n)), where Bottom and Top were
the minimum and maximal inhibition for PDE, respectively, IC50was
the concentration of compound that inhibited the PDE activity at 50%,
and n was the slope of the concentration−response curve.
Molecular Modeling. To carry out docking analysis, chain A of
protein structures was processed with help of SYBYL, version 8.0,
software,32adding hydrogens and capping terminal residues with
neutral ends. Then newly added hydrogen positions were optimized
with MMFF94 force field until a 0.01 kcal/mol gradient was reached.
Ligand geometries were also minimized with MMFF94 force field until
a 0.01 kcal/mol gradient was reached.
The docking studies were performed with Surflex (SYBYL). An
“Automatic” protocol was generated, and dockings were carried out
using 20 initial conformations for each ligand allowing flexibility.
Preminimization and postminimization were used by considering 100
solutions in each case. Cscorescoring function was used with all four
options, claiming a relaxed structure.
Visual inspection of solutions of selected ligands was carried out in
order to find which ones allow the best agreement with SAR data.
Finally the selected ligand−complex geometries were minimized with
the MMFF94 force field until a 0.01 kcal/mol gradient was reached.
To test our docking protocol and validate the procedure, we carried
out docking experiments for some selected 3D structures The first one
was for the complex of 3,5-dimethyl-1-(3-nitrophenyl)-1H-pyrazole-4-
carboxylic acid ethyl ester bound to PDE4D (PDB code 1Y2K), as this
ligand shares a high similarity with some of our inhibitors. The high
similarity between the docked solution (only one cluster within 2.0 Å
rmsd of heavy atoms was found) and the experimental one gave us
confidence in our protocol (rmsd of 0.2 Å of heavy atoms vs
experimental results; see Figure S2 in the Supporting Information). In
fact, the main difference between both orientations is related to the
benzene with supports of a nitro group. This difference can be
explained because of the lack of water W1016 from the PDB, which
was not present in the model of PDE4D used for docking where only
the metal coordination waters (W1003−W1008) were retained. On
the other hand, in order to test our protocol specifically in PDE7, we
carried out the docking experiment with the 3D structures of PDB
codes 1ZKL and 3G3N. In the case of 1ZKL, the reproduction of the
experimental binding mode of IBMX inhibitor was only possible (rmsd
of 0.1 Å of heavy atoms vs experimental; see Figure S3 in the
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3281

Page 9

Supporting Information) when waters W573 and W542 were added
aside from the metal coordination waters (W504−W509). This fact is
not surprising, as those two additional water molecules are within
3.0 Å from the ligand. As only those two structures of PDE7A were
available at the moment of this study (PDB codes 1ZKL and 3G3N)
and given the redocking results obtained for 3G3N, we decided to use
structure 1ZKL for docking purposes. Finally, as the best redocking
results were obtained for structure 1Y2K, which is the ligand that is the
most similar to our inhibitors, we decided to retain only the metal
coordination waters present in PDB code 1ZKL for docking.
Primary Cell Cultures. Glial cells were prepared from neonatal rat
cerebral cortex, as previously described by Luna-Medina et al.33Briefly,
after removal of the meninges the cerebral cortex was dissected,
dissociated, and incubated with 0.25% trypsin/EDTA at 37 °C for 1 h.
After centrifugation, the pellet was washed 3 times with HBSS (Gibco)
and the cells were plated on noncoated flasks and maintained in
HAMS/DMEM (1:1) medium containing 10% FBS. After 15 days the
flasks were agitated on an orbital shaker for 4 h at 240 rpm at 37 °C.
The supernatant was collected, centrifuged, and the cellular pellet
containing the microglial cells was resuspended in complete medium
(HAMS/DMEM (1:1) containing 10% FBS) and seeded on uncoated
96-well plates. Cells were allowed to adhere for 2 h, and the medium
was removed to eliminate nonadherent oligodendrocytes. New fresh
medium containing 10 ng/mL GM-CSF was added. The remaining
astroglial cells adhered on the flasks were then trypsinized, collected,
centrifuged, and plated onto 96-well plates with complete medium.
The purity of cultures obtained by this procedure was >98% as
determined by immunofluorescence with the OX42 (microglial
marker) and the GFAP (astroglial marker) antibodies. After 1 week
in culture, cells were treated with rolipram, BRL50481, and the
different compounds (7, 13, 14, 19, 21, and 31) at several
concentrations. Cell viability was then measured after 16 h in culture.
For nitrite release quantification some cultures were also treated with
lipopolysaccharide (LPS, 10 μg/mL) alone or in combination with
compounds.
Cell Viability Assay. Cell viability was measured using the MTT
assay from Roche, based on the ability of viable cells to reduce yellow
MTT to blue formazan. Briefly, cells cultured in 96-well plates and
treated with the indicated compounds for 16 h were incubated with
MTT (0.5 mg/mL, 4 h) and subsequently solubilized in 10% SDS/
0.01 M HCl for 12 h in the dark. The extent of reduction of MTT was
quantified by absorbance measurement at 595 nm according to the
manufacturer's protocol.
Nitrites Measurement. Accumulation of nitrites in medium was
assayed by the standard Griess reaction. After stimulation of cells with
the different compounds, supernatants were collected and mixed with
an equal volume of Griess reagent (Sigma). Samples were then
incubated at room temperature for 15 min and absorbance read using
a plate reader at 492/540 nm.
cAMP Measurements in Raw Cells. Quantification of cAMP was
carried out using the EIA (enzyme immunoassay) kit from GE
Healthcare. Briefly, raw cells were seeded at 3 × 104/well in 96-well
dishes and incubated overnight before the assay. After 60 min of
incubation with compound 13, cAMP intracellular levels were
determined following the manufacturer’s instructions.
CNS Penetration: In Vitro Parallel Artificial Membrane
Permeability Assay (PAMPA). Blood−Brain Barrier (BBB).
Prediction of the brain penetration was evaluated using a parallel
artificial membrane permeability assay (PAMPA).30Ten commercial
drugs, phosphate buffer saline solution at pH 7.4 (PBS), ethanol, and
dodecane were purchased from Sigma, Acros Organics, Merck,
Aldrich, and Fluka. The porcine polar brain lipid (PBL) (catalog no.
141101) was from Avanti Polar Lipids. The donor plate was a 96-well
filtrate plate (Multiscreen IP sterile plate PDVF membrane, pore size
of 0.45 μm, catalog no. MAIPS4510) and the acceptor plate was an
indented 96-well plate (Multiscreen, catalog no. MAMCS9610), both
from Millipore. Filter PDVF membrane units (diameter 30 mm, pore
size 0.45 μm) from Symta were used to filter the samples. A 96-well
plate UV reader (Thermoscientific, Multiskan spectrum) was used for
the UV measurements. Test compounds [(3−5 mg of caffeine,
enoxacine, hydrocortisone, desipramine, ofloxacine, piroxicam, testo-
sterone), 12 mg of promazine, and 25 mg of verapamile and atenolol]
were dissolved in EtOH (1000 μL). Then 100 μL of this compound
stock solution was taken and 1400 μL of EtOH and 3500 μL of PBS
pH 7.4 buffer were added to reach 30% of EtOH concentration in the
experiment. These solutions were filtered. The acceptor 96-well
microplate was filled with 180 μL of PBS/EtOH (70/30). The donor
96-well plate was coated with 4 μL of porcine brain lipid in dodecane
(20 mg mL−1), and after 5 min, 180 μL of each compound solution
was added. An amount of 1−2 mg of every compound, their ability to
pass the brain barrier to be determined, was dissolved in 1500 μL of
EtOH and 3500 μL of PBS, pH 7.4 buffer, filtered, and then added to
the donor 96-well plate. Then the donor plate was carefully put on the
acceptor plate to form a “sandwich”, which was left undisturbed for 2 h
and 30 min at 25 °C. During this time the compounds diffused from
the donor plate through the brain lipid membrane into the acceptor
plate. After incubation, the donor plate was removed. UV plate reader
determined the concentration of compounds and commercial drugs in
the acceptor and the donor wells. Every sample was analyzed at three
to five wavelengths, in three wells, and in two independent runs.
Results are given as the mean [standard deviation (SD)], and the
average of the two runs is reported. Ten quality control compounds
(previously mentioned) of known BBB permeability were included in
each experiment to validate the analysis set.
EAE Induction and Treatment. Six-week-old female C57BL6
mice (15−20 g) were purchased from Harlan (Spain). All
experimental procedures followed the European Communities Council
Directive of November 24, 1986 (86/609/EEC). The protocol was
approved by the Ethics Committee of the University of Barcelona and
of the Generalitat de Catalunya. The mice were maintained on a 12 h
light/dark cycle at a constant environmental temperature with free
access to food and water for 1 week prior to experimentation.
EAE was induced by subcutaneous immunization with 100 μg of
MOG35−55peptide (EspiKem S.r.l., Italy) in 100 μL of complete
Freund’s adjuvant (CFA) (Sigma-Aldrich) enriched with Mycobacte-
rium tuberculosis (H37Ra strain, Difco, Detroit, MI, U.S.). Mice were
immediately intraperitoneally injected with 200 ng of Bordetella
pertussis toxin (Sigma-Aldrich) and again 48 h after the immunization.
Animals (n = 13) were weighed and examined for clinical signs on a
daily basis. Disease severity of EAE was graded according to a five-
point scale: grade 0 = no disability; grad 1 = a flaccid tail; grade 2 =
a mild but definite weakness of one or both hind legs; grade 3 =
moderate paraparesis of one hind leg; grade 4 = no hind leg
movement; grade 5 = a moribund state with little or no spontaneous
movement and impaired respiration.34
Stock solution of the compound (100 mg/mL in DMSO) was
diluted 1:50 in a solution of 5% Tocrisolve (Tocris, U.K.) in distilled
water. Mice were treated through daily intraperitoneal (ip) injection
starting on day 5 after the onset of the disease at a dose of 10 mg/kg of
animal (n = 7) or with only vehicle (n = 7).
■ASSOCIATED CONTENT
*
Elemental analysis results of compounds 5−32; experimental
procedures for compounds 4−6, 8−12, 15−18, 20, 22−30, and
32; linear correlation between experimental and described
values in the PAMPA-BBB assay; Figures S1 and S2. This
material is available free of charge via the Internet at http://
pubs.acs.org.
■AUTHOR INFORMATION
Corresponding Author
*Phone: +34 91 5622900. Fax: +34 91 5644853. E-mail: cgil@
iqm.csic.es.
Notes
The authors declare no competing financial interest.
S Supporting Information
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3282

Page 10

■ACKNOWLEDGMENTS
The authors gratefully acknowledge the financial support of
Ministry of Science and Innovation (MICINN), Projects Nos.
SAF2009-13015-C02-01, SAF2009-13015-C02-02, SAF2010-
16365, SAF2009-1152, CTQ2009-07664, and PI10-01874;
Instituto de Salud Carlos III (ISCiii), Project No. RD07/
0060/0015 (RETICS program) and CIBERNED; Fundación
Española para la Ciencia y la Tecnología (FECYT), Project No.
FCT-09-INC-0367. M.R. and D.I.P. acknowledge pre- and
postdoctoral fellowships from the CSIC (JAE program),
respectively. BRAINco Biopharma is acknowledged.
■ABBREVIATIONS USED
ADME, absorption, distribution, metabolism, and excretion; BBB,
blood−brain barrier; cAMP, cyclic adenosine 3′,5′-monophos-
phate; CFA, complete Freund’s adjuvant; cGMP, cyclic guanosine
3′,5′-monophosphate; CNS, central nervous system; DCC, N,N′-
dicyclohexylcarbodiimide; DMEM, Dulbecco’s modified Eagle’s
medium; DMSO, dimethyl sulfoxide; EAE, experimental auto-
immune encephalomyelitis; EDTA, ethylenediaminetetraacetic
acid; ES, electrospray; FBS, fetal bovine serum; FP, fluorescence
polarization; GFAP, glial fibrillary acidic protein; GM-CSF,
granulocyte-macrophage colony-stimulating factor; HAMS,
Ham’s F-12 nutrient mixture; HBSS, Hank’s balanced salt
solution; IC50, inhibitory concentration 50; IMAP, immobilized
metal ion affinity-based fluorescence polarization, MOG, myelin
oligodendrocyte glycoprotein; mRNA, messenger ribonucleic acid;
MS, mass spectrometry; MTT, 3-[4, 5-dimethylthiazol-2-yl]-2,5-
diphenyltetrazolium bromide; NMR, nuclear magnetic resonance;
LPS, lipopolysaccharide; PAMPA, parallel artificial membrane
permeability assay; PBL, porcine polar brain lipid; PBS, phosphate
buffer saline; PDE, phosphodiesterase; PDVF, polyvinylidene
difluoride; PyBOP, benzotriazol-1-yloxytripyrrolidinophospho-
nium hexafluorophosphate; SAR, structure−activity relationship;
SD, standard deviation; SDS, sodium dodecyl sulfate; SPA,
scintillation proximity assay; TEA, triethylamine; THF, tetrahy-
drofuran; UV, ultraviolet light
■REFERENCES
(1) Essayan, D. M. Cyclic nucleotide phosphodiesterase (PDE)
inhibitors and immunomodulation. Biochem. Pharmacol. 1999, 57,
965−973.
(2) Conti, M.; Jin, S. L. The molecular biology of cyclic nucleotide
phosphodiesterases. Prog. Nucleic Acid Res. Mol. Biol. 1999, 63, 1−38.
(3) Bender, A. T.; Beavo, J. A. Cyclic nucleotide phosphodiesterases:
molecular regulation to clinical use. Pharmacol. Rev. 2006, 58, 488−
520.
(4) Francis, S. H.; Conti, M.; Houslay, M. D. Phosphodiesterases as
Drug Targets; Springer-Verlag: Berlin and Heidelberg, Germany, 2011.
(5) Lugnier, C. Cyclic nucleotide phosphodiesterase (PDE) super-
family: a new target for the development of specific therapeutic agents.
Pharmacol. Ther. 2006, 109, 366−398.
(6) Allison, A. C. Immunosuppressive drugs: the first 50 years and a
glance forward. Immunopharmacology 2000, 47, 63−83.
(7) Houslay, M. D.; Schafer, P.; Zhang, K. Y. Keynote review:
phosphodiesterase-4 as a therapeutic target. Drug Discovery Today
2005, 10, 1503−1519.
(8) Giembycz, M. A. Life after PDE4: overcoming adverse events
with dual-specificity phosphodiesterase inhibitors. Curr. Opin.
Pharmacol. 2005, 5, 238−244.
(9) Conti, M.; Beavo, J. Biochemistry and physiology of cyclic
nucleotide phosphodiesterases: essential components in cyclic
nucleotide signaling. Annu. Rev. Biochem. 2007, 76, 481−511.
(10) Giembycz, M. A.; Smith, S. J. Phosphodiesterase 7A: a new
therapeutic target for alleviating chronic inflammation? Curr. Pharm.
Des. 2006, 12, 3207−3220.
(11) Miro, X.; Perez-Torres, S.; Palacios, J. M.; Puigdomenech, P.;
Mengod, G. Differential distribution of cAMP-specific phosphodies-
terase 7A mRNA in rat brain and peripheral organs. Synapse 2001, 40,
201−214.
(12) Sasaki, T.; Kotera, J.; Omori, K. Novel alternative splice variants
of rat phosphodiesterase 7B showing unique tissue-specific expression
and phosphorylation. Biochem. J. 2002, 361, 211−220.
(13) Reyes-Irisarri, E.; Perez-Torres, S.; Mengod, G. Neuronal
expression of cAMP-specific phosphodiesterase 7B mRNA in the rat
brain. Neuroscience 2005, 132, 1173−1185.
(14) Nakata, A.; Ogawa, K.; Sasaki, T.; Koyama, N.; Wada, K.;
Kotera, J.; Kikkawa, H.; Omori, K.; Kaminuma, O. Potential role of
phosphodiesterase 7 in human T cell function: comparative effects
of two phosphodiesterase inhibitors. Clin. Exp. Immunol. 2002, 128,
460−466.
(15) Gil, C.; Campillo, N. E.; Perez, D. I.; Martinez, A.
Phosphodiesterase 7 (PDE7) inhibitors as new drugs for neurological
and inflammatory disorders. Expert Opin. Ther. Pat. 2008, 18, 1127−
1139.
(16) Martinez, A.; Castro, A.; Gil, C.; Miralpeix, M.; Segarra, V.;
Domenech, T.; Beleta, J.; Palacios, J. M.; Ryder, H.; Miro, X.; Bonet,
C.; Casacuberta, J. M.; Azorin, F.; Piña, B.; Puigdomenech, P. Benzyl
derivatives of 2,1,3-benzo- and benzothieno[3,2-a]thiadiazine 2,2-
dioxides: first phosphodiesterase 7 inhibitors. J. Med. Chem. 2000, 43,
683−689.
(17) Castro, A.; Jerez, M. J.; Gil, C.; Martinez, A. Cyclic nucleotide
phosphodiesterases and their role in immunomodulatory responses:
advances in the development of specific phosphodiesterase inhibitors.
Med. Res. Rev. 2005, 25, 229−244.
(18) Morales-Garcia, J.; Redondo, M.; Gil, C.; Alonso-Gil, S.;
Martinez, A.; Santos, A.; Perez-Castillo, A. Phosphodiesterase 7
inhibition preserves dopaminergic neurons in cellular and rodent
models of Parkinson disease. PLoS One 2011, 6, e17240.
(19) Paterniti, I.; Mazzon, E.; Gil, C.; Impllizzari, D.; Palomo, V.;
Redondo, M.; Perez, D. I.; Esposito, E.; Martinez, A.; Cuzzocrea, S.
PDE 7 inhibitors: new potential drugs for the therapy of spinal cord
injury. PLoS One 2011, 6, e15937.
(20) Redondo, M.; Zarruk, J. G.; Ceballos, P.; Perez, D. I.; Perez, C.;
Perez-Castillo, A.; Moro, M. A.; Brea, J.; Val, C.; Cadavid, M. I.; Loza,
M. I.; Campillo, N. E.; Martinez, A.; Gil, C. Neuroprotective efficacy of
quinazoline type phosphodiesterase 7 inhibitors in cellular cultures and
experimental stroke model. Eur. J. Med. Chem. 2012, 47, 175−185.
(21) Jerez, M. J.; Castro, A.; Gil, C.; Martinez, A. Development of a
pharmacophoric model for specific PDE7 inhibitors. Drugs Future
2004, 29 (Suppl. A), 129.
(22) Gil, C.; Castro, A.; Jerez, M. J.; Ke, H.; Wang, H.; Ballester, S.;
González-García, C.; Martínez, A. New PDE7 inhibitors leads for
neurodegenerative diseases discovered by using a pharmacophoric
model. Drugs Future 2008, 33 (Suppl. A), 228.
(23) Vaupel, S.; Brutschy, B.; Tarakeshwar, P.; Kim, K. S. Character-
ization of weak NH−pi intermolecular interactions of ammonia with
various substituted pi-systems. J. Am. Chem. Soc. 2006, 128, 5416−5426.
(24) Mohan, N.; Vijayalakshmi, K. P.; Koga, N.; Suresh, C. H.
Comparison of aromatic NH···pi, OH···pi, and CH···pi interactions of
alanine using MP2, CCSD, and DFT methods. J. Comput. Chem. 2010,
31, 2874−2882.
(25) Ke, H.; Wang, H. Crystal structures of phosphodiesterases and
implications on substrate specificity and inhibitor selectivity. Curr. Top.
Med. Chem. 2007, 7, 391−403.
(26) Smith, S. J.; Cieslinski, L. B.; Newton, R.; Donnelly, L. E.;
Fenwick, P. S.; Nicholson, A. G.; Barnes, P. J.; Barnette, M. S.;
Giembycz, M. A. Discovery of BRL 50481 [3-(N,N-dimethylsulfona-
mido)-4-methyl-nitrobenzene], a selective inhibitor of phosphodies-
terase 7: in vitro studies in human monocytes, lung macrophages, and
CD8+ T-lymphocytes. Mol. Pharmacol. 2004, 66, 1679−1689.
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3283

Page 11

(27) Movsesian, M. A.; Kukreja, R. C. Phosphodiesterase inhibition
in heart failure. Handb. Exp. Pharmacol. 2011, 237−249.
(28) Robichaud, A.; Stamatiou, P. B.; Jin, S. L.; Lachance, N.;
MacDonald, D.; Laliberte, F.; Liu, S.; Huang, Z.; Conti, M.; Chan,
C. C. Deletion of phosphodiesterase 4D in mice shortens alpha(2)-
adrenoceptor-mediated anesthesia, a behavioral correlate of emesis.
J. Clin. Invest. 2002, 110, 1045−1052.
(29) Di, L.; Kerns, E. H.; Carter, G. T. Strategies to assess blood−
brain barrier penetration. Expert Opin. Drug Discovery 2008, 3, 677−
687.
(30) Di, L.; Kerns, E. H.; Fan, K.; McConnell, O. J.; Carter, G. T.
High throughput artificial membrane permeability assay for blood-
brain barrier. Eur. J. Med. Chem. 2003, 38, 223−232.
(31) Crivori, P.; Cruciani, G.; Carrupt, P. A.; Testa, B. Predicting
blood−brain barrier permeation from three-dimensional molecular
structure. J. Med. Chem. 2000, 43, 2204−2216.
(32) SYBYL, version 8.0; Tripos International (1699 South Hanley
Road, St. Louis, MO, 63144, U.S.).
(33) Luna-Medina, R.; Cortes-Canteli, M.; Alonso, M.; Santos, A.;
Martinez, A.; Perez-Castillo, A. Regulation of inflammatory response
in neural cells in vitro by thiadiazolidinones derivatives through
peroxisome proliferator-activated receptor gamma activation. J. Biol.
Chem. 2005, 280, 21453−21462.
(34) McFarlin, D. E.; Blank, S. E.; Kibler, R. F. Recurrent
experimental allergic encephalomyelitis in the Lewis rat. J. Immunol.
1974, 113, 712−715.
Journal of Medicinal Chemistry
Article
dx.doi.org/10.1021/jm201720d | J. Med. Chem. 2012, 55, 3274−3284
3284