New insights into selective PDE4D inhibitors: 3-(Cyclopentyloxy)-4-methoxybenzaldehyde O-(2-(2,6-dimethylmorpholino)-2-oxoethyl) oxime (GEBR-7b) structural development and promising activities to restore memory impairment

Article abstract of DOI:10.1016/j.ejmech.2016.08.018

Phosphodiesterase type 4D (PDE4D) has been indicated as a promising target for treating neurodegenerative pathologies such as Alzheimer's Disease (AD). By preventing cAMP hydrolysis, PDE4 inhibitors (PDE4Is) increase the cAMP response element-binding protein (CREB) phosphorylation, synaptic plasticity and long-term memory formation. Pharmacological and behavioral studies on our hit GEBR-7b demonstrated that selective PDE4DIs could improve memory without causing emesis and sedation. The hit development led to new molecule series, herein reported, characterized by a catechol structure bonded to five member heterocycles. Molecular modeling studies highlighted the pivotal role of a polar alkyl chain in conferring selective enzyme interaction. Compound 8a showed PDE4D3 selective inhibition and was able to increase intracellular cAMP levels in neuronal cells, as well as in the hippocampus of freely moving rats. Furthermore, 8a was able to readily cross the blood-brain barrier and enhanced memory performance in mice without causing any emetic-like behavior. These data support the view that PDE4D is an adequate molecular target to restore memory deficits in different neuropathologies, including AD, and also indicate compound 8a as a promising candidate for further preclinical development.

Full text of DOI:10.1016/j.ejmech.2016.08.018

European Journal of Medicinal Chemistry 124 (2016) 82e102
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
New insights into selective PDE4D inhibitors: 3-(Cyclopentyloxy)-4-
methoxybenzaldehyde O-(2-(2,6-dimethylmorpholino)-2-oxoethyl)
oxime (GEBR-7b) structural development and promising activities to
restore memory impairment
,
,
,
, 1
Chiara Brullo a 1, Roberta Ricciarelli b 1, Jos Prickaerts c 1, Ottavio Arancio d
,
Matteo Massa a, Chiara Rotolo a, Alessia Romussi a, Claudia Rebosio e, Barbara Marengo b,
Maria Adelaide Pronzato b, Britt T.J. van Hagen c, Nick P. van Goethem c,
Pasqualina D'Ursi f, Alessandro Orro f, Luciano Milanesi f, Sara Guariento a, Elena Cichero a,
Paola Fossa a, Ernesto Fedele e 2, Olga Bruno a
,
g,
, *, 2
a Department of Pharmacy, Section of Medicinal Chemistry, School of Medical and Pharmaceutical Sciences, University of Genoa, Viale Benedetto XV, 3,
16132 Genoa, Italy
b Department of Experimental Medicine, Section of General Pathology, School of Medical and Pharmaceutical Sciences, University of Genoa, Via L.B. Alberti,
2, 16132 Genoa, Italy
c Department of Psychiatry and Neuropsychology, School for Mental Health and Neuroscience (MHeNS), Maastricht University, Universiteitssingel 50, 6229
ER, Maastricht, The Netherlands
d Department of Pathology & Cell Biology, Taub Institute for Research on Alzheimer's Disease and the Aging Brain, Columbia University, New York, NY
10032, USA
e Department of Pharmacy, Section of Pharmacology and Toxicology, School of Medical and Pharmaceutical Sciences, University of Genoa, Viale Cembrano,
4, 16147 Genoa, Italy
f Institute for Biomedical Technologies and National Research Council (ITB-CNR), Via Fratelli Cervi 93, 20090 Segrate, MI, Italy
g Center of Excellence for Biomedical Research, University of Genoa, Viale Benedetto XV, 9, 16132 Genoa, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Phosphodiesterase type 4D (PDE4D) has been indicated as a promising target for treating neurodegen-
erative pathologies such as Alzheimer's Disease (AD). By preventing cAMP hydrolysis, PDE4 inhibitors
(PDE4Is) increase the cAMP response element-binding protein (CREB) phosphorylation, synaptic plasticity
and long-term memory formation. Pharmacological and behavioral studies on our hit GEBR-7b demon-
strated that selective PDE4DIs could improve memory without causing emesis and sedation. The hit
development led to new molecule series, herein reported, characterized by a catechol structure bonded to
five member heterocycles. Molecular modeling studies highlighted the pivotal role of a polar alkyl chain in
conferring selective enzyme interaction. Compound 8a showed PDE4D3 selective inhibition and was able
to increase intracellular cAMP levels in neuronal cells, as well as in the hippocampus of freely moving rats.
Furthermore, 8a was able to readily cross the blood-brain barrier and enhanced memory performance in
mice without causing any emetic-like behavior. These data support the view that PDE4D is an adequate
molecular target to restore memory deficits in different neuropathologies, including AD, and also indicate
compound 8a as a promising candidate for further preclinical development.
Received 21 June 2016
Received in revised form
9 August 2016
Accepted 10 August 2016
Available online 13 August 2016
Keywords:
PDE4D inhibitors
cAMP enhancers
Memory behavior test
Pharmacokinetic analyses
Molecular dynamics simulation
In silico ADMET properties
© 2016 Elsevier Masson SAS. All rights reserved.
Abbreviations: AD, Alzheimer Disease; cAMP, cyclic adenosine monophosphate; CR3, conserved region 3; CREB, cAMP response element-binding protein; cGMP, cyclic
guanosine monophosphate; COPD, chronic obstructive pulmonary disease; COSY, correlation spectroscopy; GPCRs, G protein coupled receptors; LTP, long term potentiation;
MD, molecular dynamics simulations; PDE, phosphodiesterase; PDE4Is, PDE4 inhibitors; PKA, protein kinase A; SAR, structure-activity-relationship; UCR1 and UCR2, up-
stream conserved region 1 and 2.
* Corresponding author.
E-mail address: obruno@unige.it (O. Bruno).
These authors contributed equally to this work.
These authors also contributed equally to this work.
1
2
http://dx.doi.org/10.1016/j.ejmech.2016.08.018
0223-5234/© 2016 Elsevier Masson SAS. All rights reserved.

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
83
1. Introduction
monomers. Many studies have been recently focused on the role of
UCR1-UCR2 oligomers and also of the C terminal helix conserved
region 3 (CR3) as negative regulatory domains [21,15].
Alzheimer's disease (AD) and other types of dementia involve
today approximately 35 million people worldwide, a number ex-
pected to increase exponentially in the near future. The main AD
symptoms consist of a progressive brain disorder leading to
cognitive failure, neuropsychiatric alterations and memory loss.
The cAMP/PKA/CREB transcription is a key pathway in memory
formation [1,2] and its impairment is thought to be involved in
memory loss occurring during neurodegenerative diseases char-
acterized by cognitive disorders [3,4]. Therefore, the impaired CREB
signaling is now considered a possible alternative target for AD
pharmacological treatment [5,6].
As the increase of cAMP is a key step in the CREB phosphory-
lation pathway, the intracellular levels of this second messenger are
critical and tightly regulated by the adenylate cyclase synthesizing
activity and type 4 phosphodiesterase (PDE4) enzyme, which is
responsible for cAMP degradation to 50-AMP. Accordingly, PDE4
inhibitors (PDE4Is) have been found to facilitate long term poten-
tiation (LTP) and to improve memory function in a variety of
cognitive tasks, both under physiological and pathological condi-
tions [7e12].
Crystal structures of PDE4B and PDE4D with fragments of UCR1,
UCR2 and CR3, obtained by differently engineered constructs, have
been recently solved in presence and absence of inhibitors [15,22].
Unfortunately, these structures do not fully recapitulate the dy-
namic process of the interaction between the UCR2/CR3 and the
PDE4 active site. Consequently, any detailed consideration about
the PDE4B/PDE4D selectivity will be addressed only when com-
plete X-ray structures of these isoforms in complex with their in-
hibitors become available.
PDE4 inhibitors may be classified as “typical” or “atypical” based
on whether they interact with the catalytic domain alone or also
with the UCR2 domain [23]. Roflumilast and apremilast (Fig. 1), two
recently approved drugs for the treatment of chronic obstructive
pulmonary disease (COPD) and psoriatic arthritis, respectively, are
“typical” site-directed competitive inhibitors that show preferential
activity toward PDE4B and PDE4D isoforms [24,25]. On the con-
trary, rolipram (Fig. 1) may be classified as an “atypical non selec-
tive” inhibitor, being able to bind only the catalytic domain or to
extend its interaction to UCR2. These different interactions may
cause weak or strong rolipram binding and could explain the so-
called low- and high-affinity-rolipram binding sites (LARBS and
HARBS) [26]. More recently, Burgin and co-workers, by exploiting
the Y/F difference between human PDE4B and PDE4D in the UCR2
domain, reported the atypical selective inhibitors D159404 and D-
28 (Fig. 1) [15,27].
Recent insights in this field, including the studies on the
memory rescue ability of our lead GEBR-7b (1, Fig. 1) [13,14],
showed that selective inhibition of the PDE4D isoform results in
memory improvements without side effects (i.e. emesis and seda-
tion), thus validating PDE4D as an important target for neurode-
generative pathologies such as AD [15,16].
Focusing on PDE4 family, 4 genes encoded for as many members
(PDE4A-B-C-D) and 25 splice variants [17,18].
During the last ten years, we reported the synthesis of several
series of small molecules bearing a catecholic moiety (typical of
rolipram-related PDE4Is) and an amino function, linked to the ar-
omatic portion by an iminoether chain. The chain length and its
spatial direction have been investigated in order to obtain infor-
mation about the selectivity toward the different PDE4 isoforms
[28e31]. Among the numerous compounds, in addition to 1 also
compound 2 (Fig. 1), characterized by a hydroxypropyl chain
ending with a hydroxy-piperidine moiety, showed an interesting
enzymatic profile. Indeed, it inhibited PDE4D3 with an IC50 value of
Although the four PDE4 isoforms (A, B, C and D) show a high
sequence similarity (78%) in their catalytic domain, they can be
distinguished on the basis of two regulatory domains (UCR1 and
UCR2), inserted between the N-terminal portion and the catalytic
domain, as well as the C terminal helix conserved region 3 (CR3). It
has been reported that PDE4 long forms are functional dimers and
that dimerization is mediated by UCR1 and UCR2 interactions
[19,20]. On the other side, PDE4 lacking UCR1 (short forms) are
Fig. 1. Structures of compounds 1 (GEBR-7b), 2, and some selective and non selective PDE4 inhibitors.

84
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
1.79
mM (versus 1.91
mM for 1), and was almost completely inactive
2. Results and discussion
toward the PDE4A4, PDE4B2, PDE4C2 isoforms (% inhibition at
10 M, less than 10%) [30]. This activity and selectivity improve-
m
2.1. Chemistry
ment is justified by the fact that 2 is more polar than 1, and makes
three extra H-bonds with two water molecules and the backbone of
residue N209, involving the OH group on the 4-hydroxypiperidine
moiety, the OH positioned on the linker and the 4-
hydroxypiperidine moiety nitrogen.
The
starting
building
block
3-(cyclopentyloxy)-4-
methoxybenzaldehyde 12 was obtained by isovanillin alkylation
with bromocyclopentane, as reported in the literature [33]. The
reactants chloroacetyl- or chloropropionylamines were prepared
by condensation of chloroacetyl chloride or chloropropionyl chlo-
ride with the appropriate amines, according to our recently re-
ported procedure [30].
For the synthesis of compounds 3e7, the starting 3-(cyclo-
pentyloxy)-4-methoxybenzaldehyde oxime 13 was obtained by
reaction of the substituted benzaldehyde 12 with hydroxylamine in
ethanol [29].
More recently, an accurate computational protocol, consisting of
docking and molecular dynamics simulation (MD), has been
applied to analyse in depth the interaction of compound 1, and
some of its derivatives, with the available crystal structure of the
PDE4D catalytic site [32]. These computational results highlighted a
new binding mode within the human PDE4D, in which the com-
pounds were able to form a wide network of stable, direct and
water-mediated H-bond interactions through their carbonyl or
iminoether moieties, mainly in the nearby of Q369 and Y329 resi-
dues. Taken together, those MD simulation and our previous SAR
considerations evidenced the pivotal role of the iminoether linker,
and suggested to further investigate on this molecule portion to
modulate activity and increase selectivity toward PDE4D isoforms.
Therefore, we designed new series of catechol-based derivatives
where the iminoether function has been replaced by five
membered heterocycles bearing the same alkyl chains used for
previous compounds (Fig. 2). In particular, in compounds 3e4 and
5e7, the iminoether function is inserted in an isoxazoline or iso-
xazole ring, respectively; in compounds 8e11, the use of a pyrazole
ring allowed the isosterism ]NeOe/]NeNe and the contextual
alkyl chain displacement in respect to the catechol portion. The
heterocycle insertion, by reducing the molecules flexibility, gives
fewer conformers and, therefore, should provide more information
about pharmacophore features.
All synthesized compounds were screened on recombinant
human PDE4D3 enzyme and, then, only active compounds (>50%
inhibition) were tested on PDE4A4, PDE4B2 and PDE4C2 isoforms
to assess their selectivity. An extensive in silico evaluation of
pharmacokinetic and toxicity properties was performed to gain
preliminary information on the most promising drug-like de-
rivatives. Cytotoxic and genotoxic effects, and ability to increase
cAMP in vitro were then evaluated for the most interesting com-
pounds, selected on the basis of their enzymatic profile.
From the results of the in silico and in vitro studies, we chose
compounds 8a and 11b for the in vivo pharmacokinetic analysis.
Given its better pharmacokinetic profile, compound 8a was further
investigated in vivo to evaluate its ability to enhance cAMP levels in
the hippocampus and to improve memory without side effects.
Finally, docking analyses and molecular dynamics (MD) simu-
lation were performed on 4b and 8a, as representatives of struc-
tural diversity, to investigate the molecule-enzyme binding modes
and support our SAR considerations.
The oxime 13 was treated with N-chlorosuccinimide to obtain
the
3-(cyclopentyloxy)-N-hydroxy-4-
methoxybenzenecarboximidoyl chloride 14, which was then
cyclized with ethylacrilate, following a published procedure [34].
The intermediate dihydroisoxazole-5-carboxylate ethylester 15
was then treated with the appropriate amine to give compounds
3aef (Scheme 1).
Compounds 4aeg were prepared by cyclization of the oxime-
chloride 14 with 3-butenoic acid, obtaining the intermediate
dihydroisoxazol-5-acetic acid 16 [34] that, in turn, was treated with
the appropriate amine, in the presence of diphenylphosphorylazide
(DPPA) (Scheme 1).
Also compounds 4a,b have been prepared by an alternative
convergent synthetic strategy, which involves the cyclization of
oxime-chloride 14 with the suitable 1-amino-3-buten-1-ones
17a,b. The latter were, in turn, prepared by a novel and simple
one-pot reaction, starting from 3-butenoic acid, which was firstly
transformed in the corresponding acyl chloride and subsequently
condensed as a crude with the appropriate cycloamine (Scheme 1).
Compounds 5a,b and 6a,b were prepared by the procedure
already reported in the literature for similar compounds [35]. The
isoxazole intermediate 18 was obtained by a 1,3-dipolar addition of
14 with 3-butyn-2-ol. Then, the secondary alcohol group was
oxidized with a mixture of acetic anhydride and dimethyl sulfoxide,
yielding the corresponding keton 19 that was then a-brominated to
derivative 20. The bromoketone was treated with two equivalents
of the appropriate cycloamine, obtaining the aminoketones 5a,b
that were immediately converted into their corresponding hydro-
chlorides to avoid decomposition. Then, the aminoalcohols 6a,b
were obtained by reducing 5a,b with sodium borohydride and so-
dium methoxide in anhydrous methanol (Scheme 2). The cycliza-
tion of 14 with methylpropiolate ester gave the intermediate
isoxazole ester 21 that, in turn, was treated with an excess of the
suitable cycloamine to give compounds 7a,b (Scheme 2).
In order to synthesize the pyrazole derivatives 8e11, we first
Fig. 2. Development of new catechol-based heterocyclic derivatives.

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
85
Scheme 1. Synthesis of compounds 3a-f, 4a-g and 17a,b. Reagents and conditions: (a) NH2OH$HCl, NaHCO3, H2O/EtOH, rt, 4 h, 61%; (b) N-chlorosuccinimide, an. DMF, 40e50 ꢂC, 1 h,
65%; (c) acrylic acid ethyl ester, DCE, TEA, 0 ꢂC, nitrogen atmosphere, then 60 ꢂC, 1 h, then rt, 48 h, 50%; (d) excess cycloamine or cycloamine in an. DMF, 60 ꢂC, overnight, 22e85%;
(e) 3-butenoic acid, DCE, TEA, added at 0 ꢂC under nitrogen atmosphere, then 60 ꢂC, 1 h, then rt, 48 h, 35%; (f) cycloamine, an. DMF, TEA, DPPA, 80 ꢂC, 24 h, 21e70%; (g) thionyl
chloride, 60 ꢂC, 30 min; (h) cycloamine, an. THF, rt, 2 h, 67% (17a) or 72% (17b); (i) 17a or 17b, an. DMF, TEA, 0 ꢂC under nitrogen atmosphere, then rt, 72 h, 41% (4a) or 40% (4b).
prepared
methoxyphenyl]-1H-pyrazole 22 by a one-pot 1,3-dipolar cyclo-
addition [36]. In details, the 3-(cyclopentyloxy)-4-
the
building
block
3-[3-(cyclopentyloxy)-4-
In addition, we have also analysed their selectivity toward
PDE4A4, PDE4B2 and PDE4C2 using the strategy of the single
concentration (10 mM) assay (Table 2).
methoxybenzaldehyde 12 was treated with p-toluenesulfonyl hy-
drazide, NaOH and 1-vinylimidazole in acetonitrile (Scheme 3).
The pyrazole intermediate 22 was then treated with an excess of
epichlorohydrin in the presence of TEA. This reaction could lead to
1,5 and/or 1,3 di-substituted pyrazole isomers. The 1H and 13C NMR
spectra analyses showed a 1,3-disubstituted pyrazole structure for
the oxirane derivative (23), as reported in the experimental section.
The reaction of the above mentioned 23 with an excess of the
suitable cycloamine gave compounds 8aec in good yields. The
acetyl derivative 9a was obtained from 8a with acetic anhydride
(Scheme 3).
Most of newly synthesized compounds showed low activity
toward PDE4D3 (inhibition less than 40%). In the case of the iso-
xazoline series, only compounds 4b, 4d, and 4g showed a PDE4D3
inhibition higher than 50% with IC50 values of 5.58
mM, 1.22 mM and
1.11 M, respectively. On the contrary, all derivatives of the series 3,
m
which are apparently the most similar to our lead 1, were inactive.
Assuming that the same interaction between the enzyme and the
catechol scaffold, as well as the C]NeOeCe moiety, occurs for
compounds 1, 2, 3 and 4, it is likely that the inclusion of the latter
group in the more planar and rigid ring is detrimental for the
proper enzyme-interaction of the terminal amino group. This
observation is strongly supported by the comparison between
compounds 3b (20% of inhibitory activity), 4b (51% inhibition, IC50
The treatment of the above pyrazole derivative 22 with 4-
(chloroacetyl)morpholine
or
4-(chloroacetyl)-2,6-
dimethylmorpholine, in the presence of triethylamine (TEA), led
to the amide derivatives 10a,b (Scheme 3).
5.58 mM) and 1 (67% inhibition, IC50 1.91 mM). On the other hand,
although the activity of dimethylmorpholino and diethanolamino
derivatives 4b and 4g was predictable on the basis of our previous
SAR considerations (highlighting the essential role for oxygen atom
in the chain end), the potency of the pyrrolidino derivative 4d, as
well as the inactivity of 4a and 4c, were unexpected and are difficult
to explain. In addition, in this series, only 4b resulted in PDE4D3
selective inhibition, while both 4d and 4g were significantly active
also on PDE4A4 and PDE4B2 (inhibition higher than 50%).
In agreement with the above observations, also the more planar
and rigid isoxazole derivatives 5, 6, and 7 were inactive. Moreover,
the lack of activity for 5a and 5b matches the inactivity of the
analogue open-chain aminoketone derivatives previously reported
[30].
The same procedure was used to obtain 11a, starting from the
pyrazole derivative 22 and 4-(3-chloropropanoyl)morpholine
(Scheme 3). On the other hand, in order to obtain the analogue 11b,
the pyrazole 22 was treated with an excess of methyl 3-
bromopropanoate, yielding the intermediate 24 that was then
hydrolized to the corresponding acid 25. The latter was treated
with 2,6-dimethylmorpholine, in the presence of TEA and DPPA, to
obtain the desired product 11b (Scheme 3).
2.2. Enzymatic assays
A preliminary screen was performed by testing our compounds
on recombinant human PDE4D3 activity, at the concentration of
10 mM in duplicate (Table 1), as previously reported [29e31]. To
Among the pyrazole isoster derivatives, the most active ones
were 8a, 8b and 11b with IC50 values of 7.6 mM, 6.0 mM and 1.8 mM,
respectively, while 11a showed only 42% of PDE4D3 inhibition and
10a and 10b were inactive. Once again, only the molecules bearing
the longest linker (e.g. 8 and 11) may take the required amino
terminal spatial arrangement. Interestingly, 9a, the acetyl deriva-
tive of 8a, showed very low activity, thus confirming that the
hydroxypropyl chain needs a free eOH group to strongly interact
obtain information about the inhibitory potency on PDE4D3, the
most active compounds (i.e. those showing inhibition >50%) have
been further tested in the same assay at five different concentra-
tions (5 ꢀ 10ꢁ8e10ꢁ4 M) and IC50s were determined by nonlinear
regression analysis of the inhibition curve, using Hill equation
curve fitting (Graph Pad Prism software) (Table 2).

86
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
Scheme 2. Synthesis of compounds 5a,b, 6a,b, 7a,b. Reagents and conditions: (a) 3-butyn-2-ol, an. toluene, TEA, 60 ꢂC, 18 h, 60%; (b) acetic anhydride, DMSO, rt, 24 h, 80%; (c)
glacial acetic acid, DCM, 50 ꢂC, then bromine in DCM, 60e70 ꢂC, 1 h, 60%; (d) cycloamine, DCM, ꢁ10 ꢂC (for 5a) or ꢁ30 ꢂC (for 5b), 5 min, then sat. HCl sol. in an. diethyl ether, 72%
(5a) or 25% (5b); (e) NaBH4, an. methanol, then sodium methoxide in an. methanol, rt, 4 h, 65% (6a) or 30% (6b); (f) methyl propiolate, an. toluene, TEA, 60 ꢂC, 12 h, 33%; (g)
cycloamine, 60 ꢂC, 12 h, 53% (7a) or 69% (7b).
Scheme 3. Synthesis of compounds 8a-c, 9a, 10a,b and 11a,b. Reagents and conditions: (a) p-toluenesulfonyl hydrazide, an. CH3CN, rt, 1 h, then 5 M NaOH and 1-vinylimidazole,
50 ꢂC, 48 h, 65%; (b) epichlorohydrin, TEA, 70 ꢂC, 3 h, 73%; (c) morpholine or 2,6-dimethylmorpholine excess, or 4-hydroxypiperidine in an. DMF, 50e60 ꢂC, 18 h, 40e61%; (d) acetic
anhydride, sodium acetate, 40e50 ꢂC, 5 h, 52%; (e) chloroacetylmorpholine or chloroacetyl-2,6-dimethylmorpholine, TEA, an. DMF, 60 ꢂC, 24 h, 53% (8a) or 58% (8b); (f) chlor-
opropionylmorpholine, TEA, DMF an., 120 ꢂC, 48 h, 38%; (g) methyl-3-bromopropionate, TEA, an. CH3CN, 70 ꢂC, 24 h, 67%; (h) NaOH, EtOH, 60 ꢂC, 4 h, 74%; (i) 2,6-
dimetylmorpholine, TEA, DPPA, an. DMF, 80 ꢂC, 15 h, 71%.
with PDE4D. Quite surprisingly, however, is the total lack of activity
of the 4-hydroxy-piperidine derivative 8c, considering the close
structural correlation with our previous potent and selective
inhibitor 2. Finally, among active compounds, 8a and 11b were the
most selective PDE4D3 inhibitors.
In conclusion, the pyrazole series 8aec and 11a,b showed the

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
87
Table 1
Molecular structure and inhibition activity of compounds 3aef, 4aeg, 5a,b, 6a,b, 7a,b, 8aec, 9a, 10a,b, 11a,b, 12aec, 13a,c, and 14aec toward PDE4D3a.
Compound
X
Y
NR2
PDE4D3
% Inhib (10 mM)
3a
CO
21
3b
CO
29
3c
3d
3e
3f
CO
CO
CO
CO
26
15
8
33
4a
4b
CH2CO
CH2CO
29
51
4c
4d
4e
4f
CH2CO
CH2CO
CH2CO
CH2CO
CH2CO
30
57
40
23
65
4g
5a
5b
COCH2
COCH2
39
13
6a
6b
CH(OH)CH2
CH(OH)CH2
38
15
(continued on next page)

88
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
Table 1 (continued )
Compound
X
Y
NR2
PDE4D3
% Inhib (10 mM)
7a
7b
CO
41
CO
35
8a
8b
CH2CH(OH)CH2
CH2CH(OH)CH2
68
78
8c
CH2CH(OH)CH2
CH2CH(OCOCH3)CH2
CH2CO
2
9a
28
27
23
10a
10b
CH2CO
11a
11b
CH2CH2CO
CH2CH2CO
42
53
1
CH2CO
67
79
rolipram
e
e
a
Data are expressed as % of inhibition with respect to controls. All drugs have been tested in duplicate at the concentration of 10
used as reference compounds.
m
M. Compound 1 and rolipram have been
Table 2
Inhibition activity of compounds 4b,d,g, 8a,b and 11b, toward PDE4D3, PDE4A4, PDE4B2 and PDE4C2a.
Compound
PDE4D3
PDE4A4
PDE4B2
PDE4C2
% Inhib 10
m
M (IC50 mM)
% Inhib 10
m
M
% Inhib 10
m
M
% Inhib 10 mM
4b
4d
4g
8a
51 (5.58)
57 (1.22)
65 (1.11)
68 (7.6)
78 (6.0)
53 (1.8)
45
66
69
22
52
28
34
79
27
59
67
18
53
28
23
65
23
47
40
11
45
42
25
57
8b
11b
1b
67 (1.91)
79 (0.09)
Rolipramb
a
Data are expressed as % inhibition with respect to controls. All drugs have been tested in duplicate at the concentration of 10 mM. Values for PDE4D3 inhibition are taken
from Table 1. IC50 values for PDE4D3 are reported in parenthesis. Compound 1 and rolipram have been used as reference compounds.
b
Data taken from Ref. [29].
most interesting enzymatic profile. Compound 11b was the most
potent on PDE4D3, whereas 8a showed the highest selectivity
(Tables 1 and 2).
2.3. Computational studies
2.3.1. Molecular dynamics and docking
A recent molecular dynamics simulation (MD) study was per-
formed to analyse in depth the interaction of compound 1 with the

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
89
available crystal structure of the PDE4D catalytic site [32]. The re-
sults revealed a new binding mode of 1, and some of its derivatives,
within the human PDE4D, in which the compounds formed a wide
and stable H-bonded network, mainly in the nearby of Q369 and
Y329 residues, making direct and/or water-mediated H-bonds
through their carbonyl and iminoether moieties.
On this basis, we found it interesting to investigate the binding
mode of compounds 4b and 8a, where the iminoether function,
emerged as an interesting binding feature in compound 1 and its
derivatives, has been inserted in an isoxazoline ring or has been
substituted by a pyrazole ring. In addition, 4b and 8a well represent
the series with a high degree of molecular diversity bearing
different linkers, that are an amide carbonyl (similarly to 1) or a
hydroxypropyl chain (similarly to 2), respectively.
Starting from the most promising conformation obtained from
the docking, we performed MD studies. For each MD simulation,
the most represented complex conformation was evaluated. Com-
pound 8a conformation was stable, in agreement with the experi-
mental data assessing its relevant inhibitory activity on PDE4D. Its
occupancy of the binding site is different from that of compound 1
and is very similar to that of roflumilast (Figs. 3 and 4), being
located more closely to the metal ions. Compound 8a is anchored to
the enzyme by several hydrogen bonds, from the extremities all
along its structure, including the centre (Fig. 4, left).
In details, an H-bond involves one catechol oxygen and Q369,
while the alcoholic oxygen of 8a, the key features anchoring the
molecule into the catalytic site, is engaged in two H-bonds, one
with H160 and the other with a water molecule of the metal pocket.
Furthermore, the morpholino nitrogen makes an H-bond with
M273 backbone.
(see Fig. 1 in Supporting Information). In details, 4b does not
perform the bi-dentate H-bond with Q369 since the spatial orien-
tation of the carbonyl group forces Q369 side chain to turn differ-
ently, in order to avoid the clash. In addition, its morpholino ring
displaces the water molecule close to Y329, thus the inhibitor loses
that water mediated H-bond already observed for compound 1. The
ligand is anchored to the protein by two H-bonds: the amidic
carbonyl binds T333 and the morpholino oxygen binds N321. An
additional
Fig. 2 in Supporting Information).
p p interaction with F372 completes the pattern (see
ꢁ
These computational results are fully in agreement with the
experimental data and highlight, once more, the importance of the
iminoether linker to determine a high PDE4D inhibitory activity
(compound 1). On the other hand, also the alkyl linker plays a role
on the inhibitor arrangement in the catalytic site, as highlighted by
compound 4b interactions. In addition, the isosterism ]NeOe/]
NeN given by the pyrazole ring seems to guarantee a good inhib-
itory activity on PDE4D also for compound 8a, since the molecule
retains a favourable spatial orientation of the chain and of the
morpholino moiety into the catalytic pocket. Moreover, this study
emphasises a very important role of the alcoholic function to
address the rational design of new compounds and confirm the
above SAR considerations. In fact, in the case of 8a, its polarity
strongly drives the molecule toward the metal pockets where
several H-bonds stabilize the ligand. These interactions replace that
with Y329, and make the complex of 8a with the enzyme very
stable.
2.3.2. In silico evaluation of pharmacokinetic properties
The computational prediction of descriptors related to absorp-
tion, distribution, metabolism, excretion and toxicity (ADMET)
properties represents a useful in silico strategy accelerating the
lead compound discovery process [37].
As concerns compound 4b, it is worth noting that the substi-
tution of the flexible chain of 1 with an unsaturated isoxazole ring
determines the occupancy of the same binding area that we already
described for 1 [32]. However, the spatial orientation, imposed to
the chain and to the morpholino moiety by the unsaturated five-
membered ring, forces 4b to move in a region where the side
chains of Q369 and Y329 are normally located. Consequently, in a
sort of “domino effect”, these two interactions are lost by the ligand
In this work,, a series of ADMET properties was calculated for
the newly synthesized compounds 3b, 8a, 8b, 11b, 4b, 4d and 4g,
being the most active PDE4D3 inhibitors, and also for 1, rolipram
and roflumilast, as reference compounds. In particular, we evalu-
ated the logarithmic ratio of the octanol-water partitioning coef-
ficient (cLogP), the extent of blood-brain barrier permeation
(LogBB), the rate of passive diffusion-permeability (LogPS), the
human intestinal absorption (HIA), the volume of distribution (Vd),
the role played by plasmatic protein binding (%PPB) and by the
compound affinity toward the human serum albumin (LogKHa SA),
and the overall estimation of oral bioavailability (%F).
In addition, a number of descriptors predictive of compound
metabolism and toxicity were estimated, including the compound
potential behavior as P-glycoprotein inhibitor/substrate, the ability
to interact with the endocrine system or to act as a cytochrome
P450 3A4 inhibitor/substrate, and the median lethal dose (LD50
related to oral administration.
)
As shown in Table 3, all the compounds are characterized by a
favourable profile in terms of lipophilicity, being the calculated
cLogP below 5 (Lipinski rules), and also display the ability to be fully
adsorbed by the human intestinal membrane (HIA). Notably, all
derivatives show an adequate blood-brain barrier permeation,
showing LogBB and/or LogPS values falling in the recommended
ranges (0 < LogBB < 1.5; ꢁ3 < LogPS < ꢁ1.
The values of the calculated volume of distribution (Vd) and of
the potential binding to plasma proteins (%PPB) seem acceptable
and are within those displayed by rolipram and roflumilast. Finally,
the newly synthesized compounds show a very favourable oral
bioavailability profile (%F, column 8), especially being those of 8a
and 11b higher than that of 1.
Fig. 3. The proposed binding of compound 8a (magenta) in comparison with roflu-
milast (slate blue) into PDE4D active site. The main amino acid residues are shown in
sticks, the enzyme is shown in cartoon diagram. The residues forming the metal
binding, Q switch and P clamp and solvent filled side pockets displayed in cartoon
diagram are coloured orange, pink and green respectively. Zinc and Magnesium ions
and waters are represented as spheres. (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)
According to the descriptors listed in Table 4, none of our PDE4D
inhibitors should be substrate or inhibitor of the P-glycoprotein, or

90
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
Fig. 4. The putative binding mode of compound 8a (left) and roflumilast (right) within the PDE4D active site. Ligand carbon atoms are coloured in magenta and slate blue
respectively, protein carbon atoms are coloured in slate blue. The main amino acid residues are shown in sticks. Zinc and Magnesium ions are represented as spheres and the H-
bonds between each ligand and the active site amino acid are shown as dashed line. (For interpretation of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
Table 3
Calculated ADMET descriptors related to absorption and distribution properties.
Comp.
cLogP
LogBBa
LogPSb
HIA (%)c
Vd (l/kg)d
%PPB
LogKaHSA
%F (oral)
1
3b
8a
8b
11b
4b
4d
4g
3.45
3.79
2.54
3.75
3.26
2.98
2.73
1.23
1.72
3.42
0.76
0.94
0.25
0.48
0.32
0.48
0.35
ꢁ0.14
0.07
ꢁ0.13
ꢁ1.3
ꢁ1.2
ꢁ1.7
ꢁ1.5
ꢁ1.3
ꢁ1.4
ꢁ1.4
ꢁ2.5
ꢁ1.7
ꢁ1.3
100
100
100
100
100
100
100
100
100
100
1.9
2.1
2.0
2.4
2.1
2.0
1.8
1.4
1.4
2.0
76
77
76
80
89
77
77
66
3.66
3.82
4.02
4.08
4.75
4.16
4.09
3.37
3.69
4.20
97.5
97.1
99.6
99.6
99.2
98.2
99.4
99.6
99.6
42.2
Rolipram
Roflumilast
62.53
96.67
a
Extent of brain penetration based on ratio of total drug concentrations in tissue and plasma at steady-state conditions.
b
c
Rate of brain penetration. PS represents Permeability-Surface area product and is derived from the kinetic equation of capillary transport.
HIA represents the human intestinal absorption, expressed as percentage of the molecule able to pass through the intestinal membrane.
Volume of Distribution (Vd) of the compound in the body.
d
be involved in toxicity events such as binding with the endocrine
system. Indeed, all the compounds are characterized by acceptable
LogRBA values, being lower than ꢁ3 within high reliability indices.
Interestingly, none of our PDE4D inhibitors are able to inhibit
CYP3A4, whereas they may be its substrates.
different compounds. As a positive control, we used etoposide, a
topoisomerase II inhibitor that induces DNA double-stranded
breaks [39]. As shown in Fig. 5, the etoposide treatment of HTLA
cells led to a robust DNA damage, whereas compounds 8a, 11b, as
well as 1, did not produce the
g-H2AX immunoreactive band,
Finally, all compounds exhibit a favourable toxicity profile, being
the estimated LD50 in the range of 640e1400 mg/kg for mouse after
oral administration, values that are comparable with those calcu-
lated for rolipram and roflumilast.
typical marker of genotoxicity.
2.5. In vitro cAMP-enhancing potential
2.4. In vitro genotoxic and cytotoxic effect evaluation
To verify the capability of enhancing the intracellular accumu-
lation of cAMP, we used a cAMP-specific enzymatic immunoassay
(EIA) in HTLA cells exposed for 30 min to compound 8a and 11b,
which have been selected on the basis of their PDE4D3 inhibitory
activity, PDE4D3 selectivity, and lack of cytotoxic and genotoxic
effects. Compound 1 was used as a positive control. In order to
increase the basal level of cAMP, the cells received forskolin (FSK), a
specific activator of adenylyl cyclase, during the last 20 min of
incubation.
Cytotoxicity and genotoxicity assays on human neuronal cells
(HTLA) were performed on 8a and 11b, selected on the basis of their
enzymatic profile. For the cytotoxic potential, we analysed the
lactate-dehydrogenase release in cells exposed for 24 h to a single
high concentration (100 mM) of test compounds and 1 (as reference
compound). The results clearly indicate that none of the tested
compounds exerted cytotoxic effects (Table 5).
To evaluate genotoxicity, we analysed the phosphorylation of
the chromatin-bound histone H2AX (g-H2AX), which is a quanti-
tative marker for the DNA damage response at the site of double-
strand breaks [38]. To this end, we performed immunoblot anal-
ysis on protein extracts from HTLA cells exposed for 24 h to the
All tested compounds, including the reference compound 1,
were able to significantly increase the FSK-induced cAMP accu-
mulation, without affecting the basal levels of the cyclic nucleotide.
However, compared with 1, compounds 8a and 11b showed a
higher efficacy in increasing the intracellular cAMP level (Fig. 6).

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
91
Table 4
Calculated ADMET descriptors related to metabolism, excretion and toxicity properties.
Comp.
P-glycoprotein
Inhibitor
Endocrine system disruptiona
CYP3A4
LD50 (mg/kg)b
(R.I. ꢃ 0.30)
Substrate
(R.I. ꢃ 0.30)
LogRBA>ꢁ3
(R.I. ꢃ 0.60)
LogRBA >0
(R.I. ꢃ 0.75)
Inhibitor
Substrate
(R.I. ꢃ 0.30)
1
3b
8a
8b
11b
4b
4d
4g
Not
Not
Not
Not
Not
Not
Not
Not
Not
Not
Not
Not
Possible
Not
Not
Not
Not
Not
Not
0.08
0.10
0.07
0.06
0.09
0.09
0.07
0.03
0.05
0.07
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.00
0.00
Not
Not
Not
Not
Not
Not
Not
Not
Not
Not
0.90
0.89
0.85
0.89
0.89
0.94
0.88
0.82
0.53
0.58
1400
660
640
700
650
740
870
1400
660
1200
Rolipram
Roflumilast
Not
a
RBA represents the relative binding affinity with respect to that of estradiol. Compounds showing LogRBA > 0 are classified as strong estrogen binders, while those
showing LogRBA < ꢁ3 are considered as non-binders.
b
Lethal dose 50% (LD50) for mouse after oral administration (RI: reliability index. Borderline-allowed values for reliability parameter are ꢃ0.3, the most predictive fall in the
range 0.50e1.0).
Table 5
Cytotoxic potential of compounds 8a, 11b and 1a.
Compounds
Cytotoxicity %
Positive control
DMSO
8a
11b
1
100 8.0
3.73 1.2
3.30 0.4
0.79 0.8
0.70 0.9
a
Data are expressed as percent of positive control (100%
cytotoxicity). Cells were exposed to the different compounds
(100 mM) or to an equal volume of solvent (DMSO) for 24 h and
conditioned media were then analysed for lactate-
dehydrogenase release. Compound 1 has been used as refer-
ence compound. Data show mean SD for at least three in-
dependent experiments.
Fig. 5. Genotoxic potential of compounds 8a, 11b and 1. Western blot analysis of g-
Fig. 6. In vitro cAMP-enhancing potential of compounds 8a, 11b and 1 in cultured cells.
Neuronal cultured cells were pre-treated for 10 min with tested compounds or an
equal volume of DMSO. Then, 1 mM forskolin (FSK) was added, where indicated, for
20 min. At the end of the incubation periods, intracellular cAMP was measured with a
H2AX in HTLA cells treated for 1, 3 and 24 h with 100
(positive control) or with an equal volume of solvent (DMSO). The H2AX signals
represent the internal loading control. Compound has been used as reference
mM of 1, 8a, 11b, etoposide
1
compound. Each figure is representative of three independent experiments all showing
essentially similar results.
cAMP-specific EIA kit, accordingly to the manufacturer's instructions. Compound 1 has
been used as reference compound. Data show mean
SEM for 3 independent ex-
periments. *p < 0.05; **p < 0.01 vs FSK.
2.6. In vivo pharmacokinetic study
In conclusion, compounds 8a and 11b were rapidly absorbed
and were able to readily reach the brain. Compound 11b, however,
Pharmacokinetic analysis of compounds 8a and 11b was per-
formed in BALB/c mice following s.c. administration of a 10 mg/kg
dose. The main pharmacokinetic parameters of the two compounds
are summarized in Table 6, together with those previously reported
for compound 1 [31].
Table 6
Main pharmacokinetic parameters of compounds 8a, 11b and 1a.
Data showed that compounds 8a and 11b were rapidly absor-
bed, as indicated by the respective plasma Tmax of 0.17 h and 0.5 h,
and rapidly eliminated with a respective plasma half-life of 0.78 h
and 0.97 h. The brain distribution of compound 8a was also fast
(brain Tmax 0.17 h) and its brain/plasma (b/p) ratio ranged from
74.6 to 81.4% (calculated using AUC0-t or Cmax, respectively). The
concentration of compound 11b in mouse brain tissue peaked at
Compounds
8a
11b
1b
Parameters
Tmax (h)
Plasma
0.17
3650
0.78
Brain
0.17
2970
0.58
784
Plasma
0.5
1225
0.97
845
Brain
0.5
103
ndc
nd
Plasma
0.17
4406
0.7
Brain
0.17
1567
1.36
391
C
max (ng/mL or g)
t1/2 (h)
AUC0-t (ng$h/mL or g)
1051
1225
a
Analyses were performed in mice following s.c. administration of a 10 mg/kg
dose. Compound 1 has been used as reference compound.
0.5 h after administration and the b/p ratio (calculated using Cmax
was 8.4%
)
b
Results taken from Ref. [31].
c
nd ¼ not determined.

92
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
showed a rather poor brain penetration, whereas compound 8a
demonstrated a b/p ratio of 0.75e0.81, which is greatly improved
with respect to both compound 1 (b/p ratio 0.32e0.35) and other
recently reported analogues [31]. Thus, on the basis of its more
favourable selectivity and pharmacokinetic profiles, we choose
compound 8a for further in vivo studies.
intraperitoneally at the optimum time point of 3 h after the
learning trial [44], on natural forgetting in the ORT, using a 24 h
inter-trial interval. The results of the dose-response experiment
with this compound are summarized in Table 7. One-sample t-test,
comparing the d2 index of every condition to zero (chance level),
showed that vehicle, 0.001 and 0.01 mg/kg 8a did not significantly
differ from zero, implying no recognition of the familiar object after
a 24 h interval. On the contrary, 0.003 mg/kg of 8a showed a sig-
nificant difference compared to zero (P < 0.001). A one-way ANOVA
revealed a significant difference in the ORT performance of the
treatment conditions (F3,63 ¼ 2.988; P ¼ 0.0376). Post-hoc Dun-
nett's t-test, comparing every treatment to vehicle, showed that the
dose of 0.003 mg/kg could significantly enhance ORT memory
performance in rats (P < 0.05). In summary, from the doses tested
(0.001, 0.003 and 0.01 mg/kg i.p.), 0.003 mg/kg of 8a was able to
fully enhance memory performance compared to vehicle treat-
ment. Fig. 8 gives a visual representation of these results.
2.7. In vivo cAMP-enhancing potential
In these experiments, we have evaluated the in vivo effects of
compound 8a on extracellular cAMP levels in the hippocampus of
freely moving rats by means of intracerebral microdialysis. Intra-
hippocampal administration of compound 8a, at the concentration
of 30
extracellular cAMP (Fig. 7). However, when the PDE4D inhibitor
was infused at the concentration of 100 M, a marked cAMP
mM, caused a slight, though not significant, increase of
m
response could be observed (40e80% over basal values). The effects
of compound 8a were similar to those previously reported for the
PDE4D inhibitor 1 [31].
2.8.2. Effects of compound 8a on a scopolamine-induced memory
deficit of rats in the ORT
Compound 8a was also tested in a scopolamine-induced model
of memory impairment in the same paradigm to investigate its
effect on a memory deficit. Scopolamine (0.1 mg/kg) impairs
memory in the ORT at a 1 h inter-trial interval. Doses of 0.001, 0.003
and 0.01 mg/kg of 8a were administrated prior to trial 1 to evaluate
their effects on the scopolamine induced memory deficit.
2.8. Behavioral studies
2.8.1. Effects of compound 8a on natural forgetting of rats in the
ORT
The object recognition task (ORT) is a one-trial learning task for
episodic-like memory in rodents [40] and it measures if rodents
recognize/remember that an object from the learning trial has been
replaced with a different object. The specific timing of the treat-
ment, in relation to the learning or test trial, allows one to inves-
tigate acquisition, consolidation or retention memory processes
[41,42].
Untreated healthy rodents do not remember the familiar object
after a 24 h inter-trial interval [43] and, therefore, this ORT inter-
trial interval can be used to measure enhancements in episodic-
like memory.
The ORT measures of this study are given in Table 8. The animals
that received scopolamine and vehicle did not recognize the
familiar object after one hour, as measured with one-sample t-test
comparing the d2 index with zero. The vehicle þ vehicle treated
rats did show a significantly higher performance compared to zero
(P < 0.001), implying that the scopolamine model for memory
impairment worked in these animals. Furthermore, the
scopolamine þ0.003 mg/kg 8a treated animals also showed a
significantly higher performance than chance level (P < 0.001).
A one-way ANOVA, comparing the d2 index of all treatment
conditions, showed a significant effect (F4,71 ¼ 4.876; P ¼ 0.0016).
Dunnett's post-hoc test indicated that both the vehicle þ vehicle
(P < 0.01) and the scopolamine þ 0.003 mg/kg 8a (P < 0.001)
treated animals had a significantly higher ORT performance. These
results imply that the dose of 0.003 mg/kg of 8a is able to fully
reverse the scopolamine-induced memory deficit in rats. A graph-
ical representation of these results is depicted in Fig. 9.
The studies in healthy 4e5 months-old male Wistar rats were
designed to find the efficacious dose-range of 8a, administered
2.8.3. Effects of compound 8a on emetic-like behavior of rats in the
xylazine/ketamine anaesthesia test
The xylazine/ketamine-induced a2-adrenoceptor-mediated
anaesthesia test [45] was used to investigate the potential emetic
properties of compound 8a, administered at doses up to one hun-
dred times higher than that effective in the ORT (0.003, 0.03 and
0.3 mg/kg). The results of this test are summarized in Table 9. An
Table 7
Mean values of the different ORT measures on natural forgetting with compound
8a.a
Dose
e1 (s)
e2 (s)
d2
N
Vehicle
26.94 (3.07)
28.65 (2.95)
31.78 (4.10)
24.22 (2.81)
33.24 (2.48)
37.21 (3.49)
41.41 (2.94)
33.79 (3.21)
0.05 (0.05)
0.08 (0.09)
0.31 (0.06)###
0.09 (0.08)
22
15
15
15
0.001 mg/kg
0.003 mg/kg
0.01 mg/kg
Fig. 7. In vivo cAMP-enhancing effects of compounds 8a and 1in rat hippocampus.
Compounds 8a or
1 were administered through the microdialysis probe after 3
consecutive control samples had been collected and were present in the infusion fluid
for the time indicated by the horizontal empty bar. Data are expressed as percentages
of the mean basal value (defined 100%) that was determined by averaging the cAMP
content in the three fractions collected before drug treatment. Data represent
a
Data report the mean exploration time in T1 (e1) and T2 (e2) and discrimination
performance (d2) of the different treatment conditions in the 8a dose-response
study. The Standard Error of the Mean (SEM) is presented between brackets. One
sample t-tests were performed on the d2 measures. A significant difference from
zero (###p < 0.001) indicates that the rats remembered the object from T1.
mean
SEM of 5e6 different experiments. Data for compound 1 were taken from
Ref. [13]. *p < 0.05 vs mean basal values.

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
93
Fig. 8. The effect of compound 8a on ORT memory performance in healthy rats using a
24 h interval. Figure shows the average d2 value and SEM of each treatment condition
in the 8a dose-response study. The discrimination index (d2) is indicated on the y-axis
and the different treatment conditions are shown on the x-axis. ###p < 0.001 vs zero;
*p < 0.05 vs vehicle.
Fig. 9. The effect of compound 8a on ORT performance in rats with a scopolamine-
induced memory deficit. Figure shows the average d2 value ( SEM) of each treat-
ment condition in the scopolamine-induced memory deficit study. The discrimination
index (d2) is indicated on the y-axis and the different treatment conditions are shown
on the x-axis. ###p < 0.001 vs zero; **p < 0.01 and ***p < 0.001 vs scopolamine/vehicle.
Table 8
Mean values ( SEM) of the different ORT measures on a memory deficit model with
Table 9
compound 8aa.
Mean anaesthesia times in the xylazine/ketamine test on compound 8aa.
Dose
e1 (s)
e2 (s)
d2
n
Condition
Time anaesthetized
SEM
n
Vehicle þ vehicle
38.37 (5.58)
22.65 (4.88)
25.80 (4.50)
24.36 (5.60)
22.65 (2.48)
27.64 (3.44)
23.33 (4.23)
29.60 (3.22)
21.07 (2.10)
25.20 (2.17)
0.28 (0.07)###
ꢁ0.08 (0.09)
0.13 (0.07)
0.39 (0.05)###
0.14 (0.10)
17
15
15
14
15
Vehicle
100.0
6.85
8.67
7.42
4.22
15
12
10
15
Scopolamine þ vehicle
Scop þ 0.001 mg/kg
Scop þ 0.003 mg/kg
Scop þ 0.01 mg/kg
0.003 mg/kg
0.03 mg/kg
0.3 mg/kg
112.76
104.14
94.26
a
Data are reported as percentage variations ( SEM) induced by compound 8a
a
Data report the mean exploration time in T1 (e1) and T2 (e2) and discrimination
performance (d2) of the different treatment conditions in the 8a dose-response
study. The Standard Error of the Mean (SEM) is presented between brackets. One
sample t-tests were performed on the d2 measures. A significant difference from
zero (###p < 0.001) indicates that the animals remembered the object from T1.
relative to vehicle (set at 100%).
selectivity toward PDE4D in respect to the more lipophilic PDE4B,
as we have already observed in other analogue compounds [30].
Moreover, MD simulation also evidenced a different binding mode
of 8a in respect to our lead 1 [32]. Moreover, 8a is anchored to the
metal pocket in a similar manner to roflumilast suggesting that it
may be considered a “typical” PDE4D inhibitor. However, we can
not exclude a possible interaction of the terminal chain amino
residue outside the catalytic domain; in that case our compounds
may be also classified as “atypical”. New studies aiming at eluci-
dating the co-crystallized structure of 8a with PDE4D3 enzyme are
ongoing in order to definitively dispel any doubts in this regard.
Our in silico study evidenced good ADMET properties, then
confirmed by experimental data showing that the most active
compounds 11b and 8a lacked cytotoxic and genotoxic effects.
From a functional point of view, 11b and 8a were able to
significantly increase cAMP level in cultured cells, and compound
8a also increased cAMP levels in the hippocampus in vivo, similarly
to our lead 1 [13]
In the pharmacokinetic analysis, compounds 8a and 11b were
rapidly absorbed and were able to readily reach the brain. However,
11b showed a rather poor brain penetration in comparison with 8a
that was, therefore, selected for testing in the behavioral studies. Of
note, the brain/plasma ratios for roflumilast and rolipram are
approximately 1 and 2, respectively [46]. Thus, our compound 8a,
having a brain/plasma ratio of 0.8, is almost equal or half as brain
penetrant as roflumilast and rolipram, respectively.
anaesthesia time shorter than that of the vehicle condition would
implicate emetic-like effects [45]. Anaesthesia time was calculated
relatively to the vehicle condition, which was set at 100%. A one-
way ANOVA, comparing all conditions, did not show any signifi-
cant effect (F3,48 ¼ 1.643; n.s.), implying that 8a does not have
emetic-like effects in healthy adult rats.
3. Conclusions
Starting from our previous computational and pharmacological
studies on several series of our PDE4D inhibitors, we designed a
new library of catechol-based derivatives where the iminoether
function has been replaced by five membered heterocycles (iso-
xazole, isoxazolines, and pyrazole) bearing conserved alkyl chains.
The preliminary enzymatic screening performed on human
PDE4D3 isoform, evidenced a different behavior for each structural
subfamily. In particular, the inclusion of the C]NeOeCe moiety,
typical of previous compounds, in a more planar and rigid ring (e.g.
isoxazole derivatives 5e7) was detrimental for the inhibitory ac-
tivity. Among the most flexible isoxazolines (compounds 3 and 4),
only 4b, 4d and 4g showed satisfactory PDE4D3 inhibition, with 4b
being more selective toward the PDE4D3 isoform. The isoster pyr-
azole series 8aec and 11a,b showed the most interesting enzymatic
profile. The molecular dynamics simulation studies, performed on
the isoxazoline derivative 4b and on the pyrazole 8a, highlighted
the pivotal role of the iminoether function (but also of the alkyl
chain) in establishing a stable interaction with the enzyme. In
particular, the strong polarity of alkyl chain is responsible for 8a
In a natural forgetting paradigm, 8a was able to significantly
enhance long-term memory performance of rats in the ORT at the
dose of 0.003 mg/kg. The same dose was able also to fully reverse
the scopolamine induced short-term memory deficit in rats, up to
the level of the memory performance of control animals. For

94
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
comparison, the effective dose of 8a in these behavioral paradigms
is 10 times lower than that of roflumilast and rolipram [13,46,47].
Finally, the most interesting finding is that 8a did not have emetic-
like effects in the xylazine/ketamine anaesthesia test at doses 100
times higher than that improving memory.
In previous studies, roflumilast showed emetic-like effects at a
dose 100 times higher than that improving ORT performance in
rodents, while for rolipram this side effect was observed at the
same or at a10 times higher dose [46,13].
Based on our results, it is evident that, compared to rolipram,
our compound 8a is more selective for PDE4D3 than for the other
PDE4 isoforms, i.e. PDE4A4, PDE4B2 and PDE4C2. However, when
comparing IC50 values, 8a is less potent than rolipram on PDE4D.
Indeed, the IC50 values of all the synthesized compounds are in the
micromolar range, suggesting that they reflect binding to the low
affinity binding site of the enzyme (i.e. the catalytic domain).
Interestingly, 8a is 10 times more effective in improving memory
performance than roflumilast and rolipram, while it is less brain
penetrant and has a much higher IC50 value for the PDE4D3 isoform
than roflumilast and rolipram [46]. In addition, its emetic-like po-
tential is greatly reduced compared to roflumilast and rolipram
4.1.1. Synthesis of 3-(cyclopentyloxy)-N-hydroxy-4-
methoxybenzenecarboximidoyl chloride 14
N-chlorosuccinimide (1.02 g, 7.67 mmol) was added to a solu-
tion of 3-(cyclopentyloxy)-4-methoxybenzaldehyde oxime 13
(1.72 g, 7.31 mmol) in an. DMF (10 mL), and the mixture was stirred
at 40e45 ꢂC for 1 h. After cooling to room temperature, water
(65 mL) was added and the solution was put for 12 h in a refrig-
erator; afterward, water was removed and the oil residue was
solved in AcOEt (20 mL), the organic phase was washed with water
(3 ꢀ 20 mL), dried (MgSO4) and concentrated under reduced
pressure yielding a light brown oil. Yield: 65%; 1H NMR (CDCl3):
d
1.47e2.13 (m, 8H, 4CH2 cyclopent.), 3.87 (s, 3H, OCH3), 4.71e4.89
(m, 1H, OCH cyclopent.), 6.84 (d, J ¼ 8.4 Hz, 1H, H-5 Ar), 7.35 (dd,
J ¼ 8.4, 2.0 Hz, 1H, H-6 Ar), 7.38 (d, J ¼ 2.0 Hz, 1H, H-2 Ar); IR (film):
cmꢁ1 3000e3500 (OH); Anal. (C13H16ClNO3) calcd for C, H, N.
4.1.2. Synthesis of ethyl 3-[3-(cyclopentyloxy)-4-methoxyphenyl]-
4,5-dihydroisoxazole-5-carboxylate 15
Ethyl acrylate (2.31 mL, 21.14 mmol) and triethylamine (TEA)
(3.54 mL, 25.37 mmol) were added, under nitrogen atmosphere, to
[46]. Thus, 8a displays
a
dissociation between its memory
a
solution
of
3-(cyclopentyloxy)-N-hydroxy-4-
improving effect and reduced emetic-like potential. One possible
explanation could be that the IC50 values of the catalytic site/full
enzyme are not strictly related to behavior. Linked to this, it might
be possible that 8a causes inhibition of specific PDE4D isoforms
involved in memory processes and not those in emesis. Lastly, since
PDE4B is also emetic and 8a is less selective for this isoform, this
might also contribute to a reduced emetic potential of this com-
pound [48]. All these assumptions clearly need attention in future
research.
In conclusion, compound 8a is a selective PDE4D inhibitor that is
able to enhance memory performance without toxic and emetic
effects, thus representing a promising candidate for further lead
optimization and preclinical evaluations.
methoxybenzenecarboximidoyl chloride 14 (2.85 g, 10.57 mmol)
in dichloroethane (DCE) (16 mL) cooled at 0 ꢂC.
The yellow solution was stirred at 60 ꢂC for 1 h and at room
temperature for further 48 h. Water (20 mL) was added and the
mixture was extracted with dichloromethane (DCM) (3 ꢀ 10 mL).
The organic phases were washed with brine (3 ꢀ 10 mL), dried
(MgSO4) and concentrated under reduced pressure, yielding an oil
which was purified by silica gel (100e200 mesh) column chro-
matography using a mixture of diethyl ether/petroleum ether
(boiling point 40e60 ꢂC) (1:1) as the eluent. The final product
crystallized as a light yellow solid, which was re-crystallized by
diethyl ether. Yield: 50%; mp: 60 ꢂC; 1H NMR (CDCl3):
d 1.36 (t, 3H,
OCH2CH3), 1.50e2.10 (m, 8H, 4 CH2 cyclopent.), 3.60e3.65 (m, 2H,
CH2 isoxaz.), 3.89 (s, 3H, OCH3), 4.28 (q, 2H, OCH2CH3), 4.78e4.90
(m, 1H, OCH cyclopent.), 5.16 (t, J ¼ 8.6 Hz 1H, H-5 isoxaz.), 6.86 (d,
J ¼ 8.0 Hz, 1H, H-5 Ar), 7.05 (dd, J ¼ 8.2, 2.0 Hz, 1H, H-6 Ar), 7.37 (d,
J ¼ 2.0 Hz, 1H, H-2 Ar); Anal. (C18H23NO5) calcd for C, H, N.
4. Experimental
4.1. Chemistry
All chemicals were purchased from Chiminord and Aldrich
Chemical, Milan, Italy. Solvents were reagent grade. Unless other-
wise stated, all commercial reagents were used without further
purification.
Aluminium backed silica gel plates (Merck DC-Alufolien Kie-
selgel 60 F254, Darmstad, Germany), were used in thin-layer
chromatography (TLC) for routine monitoring the course of re-
actions. Detection of spots was made by UV light. Merck silica gel,
230e400 mesh, was used for chromatography.
All the amines were purified by distillation before the use.
Melting points are not “corrected” and were measured with a Buchi
M-560 instrument. IR spectra were recorded with a Perkin-Elmer
398 spectrophotometer. 1H NMR spectra were recorded on a Var-
ian Gemini 200 (200 MHz) or a Bruker DPX300 (300 MHz) in-
4.1.3. General procedure for the synthesis of 4-({3-[3-
(cyclopentyloxy)-4-methoxyphenyl]-4,5-dihydroisoxazol-5-yl}
carbonyl)amines 3a-f
A mixture of ethyl 3-[3-(cyclopentyloxy)-4-methoxyphenyl]-
4,5-dihydroisoxazole-5-carboxylate 15 and the suitable amine was
heated at 60 ꢂC overnight. If the amine was a solid, the reaction was
carried on using anhydrous dimethylformamide (DMF) (10 mL) as
the solvent. After cooling to room temperature, the reaction
mixture was poured into water (20 mL), yielding a crude solid that
was filtered and re-crystallized from diethyl ether. If no precipitate
was formed, the water suspension was extracted with DCM
(3 ꢀ 10 mL), washed with 1 N HCl solution (10 mL) and brine
(2 ꢀ 10 mL), dried (MgSO4) and evaporated under reduced pres-
sure. The solid obtained was re-crystallized from diethyl ether.
struments and chemical shifts are reported as
d (ppm) relative to
tetramethylsilane (TMS) as internal standard; signals were char-
acterized as s (singlet), d (doublet), t (triplet), m (multiplet), br s
(broad signal); J in Hz. MS spectra were recorded on HP 6890-5973
apparatus: injection temperature 250 ꢂC, HP5 poly(-
methylphenylsilxane) column 30 m x 0,25 mm, He flux 1 mL/min.
Elemental analyses were determined with an elemental
analyzer EA 1110 (Fison-Instruments, Milan, Italy). Analyses indi-
cated by the symbols of the elements or functions were within
0.4% of the theoretical values.
4.1.3.1. 4-({3-[3-(Cyclopentyloxy)-4-methoxyphenyl]-4,5-
dihydroisoxazol-5-yl}carbonyl)morpholine (3a). White solid; Yield:
76%; mp: 115e116 ꢂC; 1H NMR (CDCl3):
d
1.55e2.10 (m, 8H, 4CH2
cyclopent.), 3.30e4.30 (m, 13H, CH2 isoxazol. 4CH2
þ
morph. þ OCH3), 4.78e4.90 (m, 1H, OCH cyclopent.), 5.30e5.40 (m,
1H, H-5 isoxaz.), 6.88 (d, J ¼ 8.0 Hz, 1H, H-5 Ar), 7.13 (dd, J ¼ 6.6,
2.0 Hz, 1H, H-6 Ar), 7.37 (d, J ¼ 2.0 Hz, 1H, H-2 Ar); IR (CHCl3): cmꢁ1
1650 (CO); Anal. (C20H26N2O5) calcd for C, H, N.

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
95
4.1.4. Synthesis of {3-[3-(cyclopentyloxy)-4-methoxyphenyl]-4,5-
dihydroisoxazol-5-yl}acetic acid 16
4.1.6.2. 4-({3-[3-(Cyclopentiloxy)-4-methoxyphenyl]-4,5-
dihydroisoxazol-5-yl}acetyl)(2,6-dimethyl)morpholine
Yield: 40%; mp: 104e105 ꢂC.
(4b).
3-butenoic acid (5 mL, 58.58 mmol) and TEA (9.8 mL, 70 mmol)
were added, under nitrogen atmosphere, to a solution of 3-
(cyclopentyloxy)-N-hydroxy-4-methoxybenzenecarboximidoyl
chloride 14 (7.9 g, 29.29 mmol) in DCE (46 mL) cooled at 0 ꢂC and
the solution was stirred at 60 ꢂC for 1.5 h and at room temperature
for further 48 h. Water was added (20 mL) and the mixture was
extracted with DCM (3 ꢀ 10 mL). The organic phases were washed
with a NaHCO3 saturated solution (2 ꢀ 10 mL); then, the aqueous
phase was treated with 1 N HCl until precipitation of the final
product as a white solid. Yield: 35%; mp: 126e128 ꢂC; 1H NMR
4.1.7. General procedure for the synthesis of 1-(4-morpholinyl)-3-
buten-1-one and 1-(2,6-dimethyl-4-morpholinyl)-3-buten-1-one
17a,b
Thionyl chloride (9.68 g, 81.31 mmol, 5.9 mL) was added drop-
wise to 3-butenoic acid (5 g, 58.08 mmol, 4.9 mL) cooled at 0 ꢂC. The
mixture was refluxed for 30 min, then cooled at 0 ꢂC. The suitable
cycloamine (114 mmol), solved in anydrous THF (25 mL), was added
and the mixture was stirred at room temperature for 2 h. The solid
was filtered off, washed with an. THF and the collected organic
phases were evaporated under reduced pressure. A NaHCO3 satu-
rated solution (100 mL) was added to the crude and the mixture
was extracted with DCM (3 ꢀ 20 mL), the organic phase was
washed with water (20 mL), dried (MgSO4) and concentrated under
reduced pressure. The crude was purified by high vacuum distil-
lation (95 ꢂC/1 mmHg) to afford the pure products as a light yellow
oil.
(CDCl3):
d 1.50e2.10 (m, 8H, 4CH2 cyclopent.), 2.65e3.05 (m, 2H,
CH2COOH), 3.05e3.65 (m, 2H, CH2 isoxaz.), 3.90 (s, 3H, OCH3),
4.78e4.95 (m, 1H, OCH cyclopent.), 5.05e5.20 (m, 1H, H-5 isoxaz.),
6.86 (d, J ¼ 8.0 Hz, 1H, H-5 Ar), 7.06 (dd, J ¼ 8.2, 2.0 Hz, 1H, H-6 Ar),
7.38 (d, J ¼ 2.0 Hz,1H, H-2 Ar), 8.05 (br s,1H, COOH, disappears with
D2O); Anal. (C17H21NO5) calcd for C, H, N.
4.1.5. General procedure for the synthesis of 4-({3-[3-
(cyclopentyloxy)-4-methoxyphenyl]-4,5-dihydroisoxazol-5-yl}
acetyl)amines 4aeg (method A)
4.1.7.1. 1-(4-Morpholinyl)-3-buten-1-one (17a). Yield: 67%; 1H NMR
TEA (0.31 mL, 2.25 mmol), the suitable amine (3 mmol) and
diphenylphosphorylazide (DPPA) (0.36 mL, 1.65 mmol) were added
to a solution of {3-[3-(cyclopentyloxy)-4-methoxyphenyl]-4,5-
dihydroisoxazol-5-yl}acetic acid 16 (0.48 g, 1.5 mmol) in anhy-
drous DMF (4 mL), cooled in an ice-water bath, and the reaction
mixture was heated at 80 ꢂC for 24 h. After cooling to room tem-
perature, the reaction mixture was poured into water (50 mL),
extracted with DCM (3 ꢀ 10 mL), washed with a NaHCO3 saturated
solution (10 mL), 1 N HCl solution (10 mL) and brine (2 ꢀ 10 mL),
dried (MgSO4) and evaporated under reduced pressure. The solid
obtained was purified by crystallization from diethyl ether.
(CDCl3): d 3.00e3.13 (m, 2H, CH2CO), 3.30e3.65 (m, 8H, 4CH2
morph.), 4.95e5.15 (m, 2H, CH2¼CH), 5.75e5.93 (m, 1H, CH2]CH):
IR (CHCl3) cmꢁ1: 1633 (CO); GC-MS m/z: 155 (Mþ), 56, 70, 86, 114.
4.1.8. Synthesis of 1-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]
isoxazol-5-yl}ethanol 18
3-butyn-2-ol (0.36 mL, 4.56 mmol) and, subsequently, TEA
(0.95 mL, 6.84 mmol) were added dropwise to a solution of 3-
(cyclopentyloxy)-N-hydroxy-4-methoxybenzenecarboximidoyl
chloride 14 (1.23 g, 4.56 mmol) in an. toluene (5 mL) cooled at
0e5 ꢂC; then, the mixture was heated at 60 ꢂC for 18 h. After cooling
to room temperature, the mixture was filtered and the solid ob-
tained was washed with an. toluene (3 ꢀ 10 mL). The collected
organic phases were washed with water (3 ꢀ 20 mL), dried (MgSO4)
and concentrated under reduced pressure yielding a yellow oil
which was purified by Florisil (100e200 mesh) column chroma-
tography using diethyl ether as the eluent. The crude crystallized as
a white solid by addition of a mixture of diethyl ether/petroleum
ether (bp 40e60 ꢂC) (1:1). Yield: 60%; mp: 72e73 ꢂC; 1H NMR
4.1.5.1. 4-({3-[3-(Cyclopentiloxy)-4-methoxyphenyl]-4,5-
dihydroisoxazol-5-yl}acetyl)(2,6-dimethyl)morpholine
White solid. Yield: 38%; mp: 104e105 ꢂC. 1H NMR (CDCl3):
1.00e1.43 (m, 6H, 2CH3 morph.), 1.48e2.13 (m, 8H, 4CH2 cyclo-
4b.
d
pent.), 2.20e4.20 (m, 10H, 2CHO morph. þ 2CH2N morph. þ CH2
isoxaz. þ CH2CO), 3.90 (s, 3H, OCH3), 4.65e4.96 (m, 1H, OCH
cyclopent.), 5.05e5.28 (m, 1H, H-5 isoxaz.), 6.87 (d, J ¼ 8.0 Hz, 1H,
H-5 Ar), 7.08 (dd, J ¼ 8.0, 2.0 Hz, 1H, H-6 Ar), 7.37 (d, J ¼ 1.6 Hz, 1H,
H-2 Ar). Anal. (C23H32N2O5) calcd for C, H, N.
(CDCl3):
d
1.62 (d, J ¼ 6.6 Hz, 3H, CH3), 1.75e2.10 (m, 8H, 4CH2
cyclopent.), 2.02 (s, 1H, OH, disappears with D2O), 3.87 (s, 3H,
OCH3), 4.75e4.91 (m, 1H, OCH cyclopent.), 4.95e5.11 (m, 1H,
CHOH), 6.44 (s, 1H, H-5 isoxaz.), 6.88 (d, J ¼ 8.2 Hz, 1H, H-5 Ar), 7.24
(dd, J ¼ 8.2, 2.0 Hz, 1H, H-6 Ar), 7.37 (d, J ¼ 2.0 Hz, H-2 Ar); IR (KBr)
cmꢁ1: 3398 (OH); Anal. (C17H21NO4) calcd for C, H, N.
4.1.6. General procedure for the synthesis of 4-({3-[3-
(cyclopentyloxy)-4-methoxyphenyl]-4,5-dihydroisoxazol-5-yl}
acetyl)amines 4a,b (method B)
The suitable 1-amino-3-buten-1-one 17a or 17b (22 mmol)
solved in anydrous DMF (10 mL) was added dropwise, at 0 ꢂC and
under nitrogen atmosphere, to a solution of 3-(cyclopentyloxy)-N-
hydroxy-4-methoxybenzenecarboximidoyl chloride 14 (2.97 g,
11 mmol) in anydrous DMF (20 mL); then, TEA (1.72 g, 17 mmol,
2.37 mL). The mixture was stirred at room temperature for 72 h,
cooled to room temperature and poured into water (100 mL). The
mixture was extracted with diethyl ether (3 ꢀ 20 mL), the organic
phase was washed with brine (3 ꢀ 20 mL), dried (MgSO4) and
concentrated under reduced pressure. Compound 4a was obtained
as a yellow solid that was re-crystallized by anhydrous methanol.
Compound 4b was purified by silicagel (100e200 mesh) column
chromatography using diethyl ether as the eluent to afford a pure
product as a white solid.
4.1.9. Synthesis of 1-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]
isoxazol-5-yl}ethanone 19
Acetic anhydride (13.78 mL) was added to 1-{3-[3-(cyclo-
pentyloxy)-4-methoxyphenyl]isoxazol-5-yl}ethanol18 (4.18 g,
13.78 mmol) in DMSO (21 mL), and the solution was stirred at room
temperature for 24 h. Then, water (230 mL) was added and the solid
obtained was filtered and re-crystallized by absolute ethanol
yielding a white solid. Yield: 80%; mp: 121e122 ꢂC; 1H NMR
(CDCl3): d 1.42e2.10 (m, 8H, 4CH2 cyclopent.), 2.63 (s, 3H, CH3), 3.88
(s, 3H, OCH3), 4.75e4.94 (m, 1H, OCH cyclopent.), 6.91 (d, J ¼ 8.2 Hz,
1H, H-5 Ar), 7.13 (s, 1H, H-5 isoxaz.), 7.26 (dd, J ¼ 8.2, 2.0 Hz 1H, H-6
Ar), 7.39 (d, J ¼ 2.0 Hz, H-2 Ar); IR (KBr) cmꢁ1: 1696 (CO); Anal.
(C17H19NO4) calcd for C, H, N.
4.1.6.1. 4-({3-[3-(Cyclopentyloxy)-4-methoxyphenyl]-4,5-
dihydroisoxazol-5-yl}acetyl)morpholine (4a). Yield: 41%; mp:
144 ꢂC.
4.1.10. Synthesis of 2-bromo-1-{3-[3-(cyclopentyloxy)-4-
methoxyphenyl]isoxazol-5-yl}ethanone 20
Glacial acetic acid (0.05 mL) was added to a solution of 1-{3-[3-

96
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
(cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-yl}ethanone
19
7.08 (d, J ¼ 2.0 Hz,1H, H-5 Ar), 7.24 (d, J ¼ 2.0 Hz,1H, H-6 Ar), 7.35 (s,
1H, H-2 Ar); IR (KBr) cmꢁ1: 3125 (OH); Anal. (C21H28N2O5) calcd for
C, H, N.
(0.56 g, 1.86 mmol) in DCM (3 mL), and the reaction mixture was
heated at 50 ꢂC. Afterward, a solution of bromine (0.1 mL,
1.95 mmol) in DCM (1 mL) was added dropwise and the mixture
was heated at reflux for 1 h. After cooling to room temperature, the
suspension was diluted with DCM (20 mL), washed once with ice-
water (20 mL), then with water (2 ꢀ 20 mL), dried (MgSO4) and
concentrated under reduced pressure yielding a yellow solid which
was re-crystallized by absolute ethanol to give the final product as a
white solid. Yield: 60%; mp: 120e122 ꢂC; 1H NMR (CDCl3):
4.1.13. Synthesis of methyl 3-[3-(cyclopentyloxy)-4-
methoxyphenyl]isoxazole-5-carboxylate 21
A solution of TEA (0.87 g, 8.64 mmol, 1.16 mL) in anhydrous
toluene (15 mL) was added dropwise to a suspension of 3-(cyclo-
pentyloxy)-N-hydroxy-4-methoxybenzenecarboximidoyl chloride
14 (2.33 g, 8.64 mmol) and methyl propiolate (0.73 g, 8.64 mmol,
0.71 mL) in anhydrous toluene (10 mL), (30 min) and the mixture
was stirred at 60 ꢂC for 12 h. After cooling to room temperature, the
solid residue was filtered off and washed with anhydrous toluene.
Then, the collected organic phases were dried (MgSO4) and
concentrated under reduced pressure, yielding a light brown oil
that crystallized after addition of a mixture of diethyl ether/pe-
troleum ether (bp 40e60 ꢂC) (1:1) and standing overnight in a
refrigerator. Yield: 33%; mp: 126e127 ꢂC; 1H NMR (CDCl3):
d
1.57e2.14 (m, 8H, 4CH2 cyclopent.), 3.89 (s, 3H, OCH3), 4.47 (s, 2H,
CH2Br), 4.78e4.93 (m, 1H, OCH cyclopent.), 6.93 (d, J ¼ 8.2 Hz, 1H,
H-5 Ar), 7.20e7.37 (m, 2H, H-6 Ar þ H-5 isoxaz.), 7.41 (d, J ¼ 1.8 Hz,
H-2 Ar); IR (KBr) cmꢁ1: 1698 (CO); Anal. (C17H18BrNO4) calcd for C,
H, N.
4.1.11. General procedure for the synthesis of 1-{3-[3-
(cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-yl}-2-
cycloaminoethanones 5a,b
d
1.55e2.15 (m, 8H, 4CH2 cyclopent.), 3.93 (s, 3H, OCH3 Ar), 4.02 (s,
The suitable cycloamine (2.91 mmol) was added to a solution of
2-bromo-1-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-
yl}ethanone 20 (0.54 g, 1.42 mmol) in DCM (3 mL), cooled at ꢁ10 ꢂC
(for compound 5a) or at ꢁ30 ꢂC (for compound 5b). After a few
minutes, the reaction was completed and a saturated solution of
anhydrous HCl in diethyl ether was immediately added to the re-
action mixture, yielding the hydrochloride salts of compounds 5a,b
as white solids.
3H, OCH3 ester), 4.82e4.96 (m, 1H, OCH cyclopent.), 6.96 (d,
J ¼ 8.2 Hz, 1H, H-5 Ar), 7.23 (s, 1H, H-5 isoxaz.), 7.30 (dd, J ¼ 8.2,
2.2 Hz, 1H, H-6 Ar), 7.45 (d, J ¼ 2.2 Hz, 1H, H-2 Ar); IR (KBr) cmꢁ1
:
1744 (CO); Anal. (C17H19NO5) calcd for C, H, N.
4.1.14. General procedure for the synthesis of 4-({3-[3-
(cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-yl}carbonyl)
cycloamines 7a,b
A
suspension
of
methyl
3-[3-(cyclopentyloxy)-4-
4.1.11.1. 1-{3-[3-(Cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-yl}-
methoxyphenyl]isoxazole-5-carboxylate 21 (0.96 g, 2.90 mmol)
and of an excess of the appropriate cycloamine (41.6 mmol) was
stirred for 12 h at 60 ꢂC. After cooling to room temperature, water
(30 mL) was added and the product crystallized standing overnight
in a refrigerator. The white solid obtained was filtered and re-
crystallized by ethanol.
2-morpholin-4-ylethanone hydrochloride (5a). Yield: 72%; mp:
175 ꢂC (dec.); 1H NMR (CDCl3):
d 1.60e2.20 (m, 8H, 4CH2 cyclo-
pent.), 3.60e3.80 (m, 4H, 2CH2N morph.), 4.07 (s, 3H, OCH3),
4.20e4.50 (m, 4H, 2CH2O morph.), 4.27 (s, 2H, CH2N), 4.95e5.10 (m,
1H, OCH cyclopent.), 7.08 (d, J ¼ 2.0 Hz, 1H, H-5 Ar), 7.35e7.52 (m,
3H, H-2 Ar þ H-5 isoxaz. þ NHþ, 1H disappears with D2O), 7.55 (d,
J ¼ 2.0 Hz, 1H, H-6 Ar); IR (KBr) cmꢁ1: 1710 (CO); Anal.
(C21H26N2O5.HCl) calcd for C, H, N.
4.1.14.1. 4-({3-[3-(Cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-yl}
carbonyl)morpholine (7a). Yield: 53%; mp: 100e101 ꢂC; 1H NMR
(CDCl3):
d 1.55e2.21 (m, 8H, 4CH2 cyclopent.), 3.92 (s, 3H, OCH3),
4.1.12. General procedure for the synthesis of 1-{3-[3-
(cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-yl}-2-
cycloaminoethanols 6a-b
3.65e3.97 (m, 8H, 4CH2 morph.), 4.80e4.92 (m, 1CH, OCH cyclo-
pent.), 6.96 (d, J ¼ 8 Hz,1H, H-5 Ar), 7.09 (s, 1H, H isoxaz.), 7.03e7.38
(dd, J ¼ 8.0, 1.8 Hz, 1H, H-6 Ar), 7.42 (d, J ¼ 1.8 Hz, 1H, H-2 Ar); IR
(KBr) cmꢁ1: 1636 (CO); Anal. (C20H24N2O5) calcd for C, H, N.
NaBH4 (0.041 g, 1.1 mmol) was added portion-wise, at room
temperature, to
methoxyphenyl]isoxazol-5-yl}-2-cycloaminoethanones
a
suspension of 1-{3-[3-(cyclopentyloxy)-4-
hydro-
4.1.15. Synthesis of 3-[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-
pyrazole 22
chloride 5a or 5b (1.1 mmol) solved in anydrous methanol, NaBH4
(0.041 g, 1.1 mmol). Afterward, a solution of sodium methoxide
(60 mg, 1.1 mmol) in anydrous methanol (10 mL) was added
dropwise and the mixture was stirred at room temperature for 4 h.
Then, the solvent was removed under reduced pressure and the
crude was solved in DCM (20 mL), washed with water (3 ꢀ 20 mL),
dried (MgSO4) and concentrated under reduced pressure. The
morpholino derivative 6a was obtained as a yellow solid, which was
washed with hot diethyl ether. The dimethylmorpholino derivative
6b was obtained as oil, which was purified by Florisil (100e200
mesh) column chromatography using diethyl ether as eluent. The
solids obtained were further purified by preparative TLC (Silicagel)
using diethyl ether as eluent, yielding the final pure products as
white solids.
3-(cyclopentyloxy)-4-methoxybenzaldehyde 12 (3.90 g,
17.74 mmol), solved in anhydrous acetonitrile (10 mL), was added
to a solution of p-toluenesulfonyl hydrazide (3.3 g, 17.74 mmol) in
anydrous acetonitrile (20 mL) and the mixture was stirred at room
temperature for 1 h. Then, 5 N NaOH solution (3.55 mL,17.75 mmol)
was added and the mixture, which became coloured in red, was
stirred at room temperature for 20 min. Afterward, 1-
vinylimidazole (8.34 g, 8.03 mL, 88.77 mmol) was slowly added
and the mixture was heated at 50 ꢂC for 48 h. After cooling to room
temperature, the solvent was removed under reduced pressure and
the crude was solved in ethylacetate (20 mL). The organic phase
was washed with brine (2 ꢀ 20 mL),1 N HCl solution (20 mL), water
(3 ꢀ 20 mL), dried (MgSO4), and concentrated under reduced
pressure to obtain a light yellow oil that crystallized by addition of
methanol. The pure product is obtained as a light yellow solid.
4.1.12.1. 1-{3-[3-(Cyclopentyloxy)-4-methoxyphenyl]isoxazol-5-yl}-
2-morpholin-4-ylethanol (6a). Yield 65%; mp: 136e137 ꢂC; 1H NMR
Yield: 65%; mp: 80e82 ꢂC; 1H NMR (CDCl3):
d 1.49e2.19 (m, 8H,
(CDCl3):
d
1.70e2.30 (m, 8H, 4CH2 cyclopent.), 2.55e2.75 (m, 2H,
4CH2 cyclopent.), 3.87 (s, 3H, OCH3), 4.72e4.97 (m, 1H, OCH
cyclopent.), 6.53 (d, J ¼ 2.2 Hz,1H, H-4 pyraz.), 6.84e7.38 (m, 3H, 3H
Ar), 7.59 (d, J ¼ 2.2 Hz, 1H, H-5 pyraz.); IR (KBr) cmꢁ1: 3562 (NH);
Anal. (C15H18N2O2) calcd for C, H, N.
CH2N), 2.80e3.00 (m, 4H, 2CH2N morph.), 3.85e4.10 (m, 5H, 2CH2O
morph. þ OH, 1H disappears with D2O), 4.05 (s, 3H, OCH3),
4.85e5.15 (m, 2H, OCH cyclopent. þ CHOH), 6.71 (s,1H, H-5 isoxaz.),

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
97
4.1.16. Synthesis of 3-[3-(cyclopentyloxy)-4-methoxyphenyl]-1-
4.1.18. Synthesis of 1-(3-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]-
(oxiran-2-ylmethyl)-1H-pyrazole 23
1H-pyrazol-1-yl}-2-hydroxypropyl)piperidin-4-ol 8c
A mixture of 3-[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-pyr-
azole 22 (0.77 g, 2.98 mmol) and epichlorohydrin (3 mL,
38.26 mmol) was cooled at 5 ꢂC. Then, TEA (4.47 mmol, 0.62 mL)
was added dropwise and the reaction mixture was stirred until the
temperature became 25ꢂ and, then, heated at 70 ꢂC for 3 h. After
cooling to room temperature, the mixture was poured into water
(100 mL), the aqueous phase was extracted with DCM (2 ꢀ 20 mL).
The organic phase was washed with water (20 mL) until the pH of
the washing solution became neutral, dried (MgSO4) and concen-
trated under reduced pressure, yielding an oil that was purified by
silicagel (100e200 mesh) column chromatography using diethyl
ether as the eluent. The desired compound crystallized at 4 ꢂC after
addition of diethyl ether. The 1H and 13C NMR spectra analyses
showed that the purified compound was a single isomer, and C and
H assignments were made on the basis of spin-spin decoupling,
COSY, HSQC and HMBC experiments (data on Supporting Infor-
mation). Furthermore, a long range correlation ™C ¼ 131.4 (C-5
pyrazole)/™H ¼ 4.19 and 4.48 (NeCH2e) in the HMBC spectrum
clearly showed a 1,3-disubstituted pyrazole structure for the oxir-
ane derivative 23. Yield: 73%; mp: 53e55 ꢂC; 1H NMR (CDCl3):
Piperidin-4-ol (0.25 g, 2.5 mmol) was added to a solution of 3-
[3-(cyclopentyloxy)-4-methoxyphenyl]-1-(oxiran-2-ylmethyl)-1H-
pyrazole 23 (0.38 g, 1.21 mmol) in anhydrous DMF (2 mL). The
mixture was stirred at 50e60 ꢂC for 18 h. After cooling to room
temperature, the solvent was removed under reduced pressure.
The residue was dissolved in diethyl ether (20 mL) and the organic
phase was washed with water (20 mL), dried (Na2SO4), and
concentrated under reduced pressure. The crude was purified by
silicagel (100e200 mesh) column chromatography using as eluent
first a mixture of diethyl ether/methanol (9:1) and then a mixture
of DCM/methanol (7:3), obtaining the pure product as a light yellow
oil. Yield: 40%, 1H NMR (CDCl3):
d 1.30e2.03 (m, 12H, 4CH2 cyclo-
pent. þ 2CH2 pip.), 2.03e2.99 (m, 8H, 3CH2N þ 2OH, 2H disappear
with D2O), 3.60e3.80 (m, 1H, CHeOH pip.), 3.88 (s, 3H, OCH3),
4.04e4.28 (m, 3H, CH2N pyraz. þ CHOH), 4.78e4.98 (m, 1H, OCH
cyclopent.), 6.49 (d, J ¼ 2 Hz, 1H, H-4 pyraz.), 6.82e7.48 (m, 3H, H-
6 þ H-5 þ H-2 Ar), 7.52 (d, J ¼ 2.0 Hz,1H, H-5 pyraz.); IR (KBr) cmꢁ1
:
3413 (OH); Anal. (C23H33N3O4) calcd for C, H, N.
4.1.19. Synthesis of 2-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]-
1H-pyrazol-1-yl}-1-(morpholin-4-ylmethyl)ethyl acetate 9a
d
1.50e2.04 (m, 8H, 4CH2 cyclopent.), 2.53 (dd, J ¼ 4.7, 2.6 Hz,1H, HA
of CH2 epox.), 2.85 (dd, J ¼ 4.7, 4.3 Hz, 1H, HB of CH2 epox.),
3.34e3.40 (m, 1H, CHO epox.), 3.86 (s, 3H, OCH3), 4.19 (dd, J ¼ 4.6,
5.6 Hz, 1H, HA of CH2), 4.48 (dd, J ¼ 14.6, 3.1 Hz, 1H, HB of CH2),
4.84e4.94 (m, 1H, OCH cyclopent.), 6.49 (d, J ¼ 2.4 Hz, 1H, H-4
pyraz.), 6.88 (d, J ¼ 8.3 Hz, 1H, H-5 Ar), 7,29 (dd, J ¼ 8.3, 2.0 Hz, 1H,
H-6 Ar), 7.36 (d, J ¼ 2.0 Hz, 1H, H-2 Ar), 7.46 (d, J ¼ 2.4 Hz, 1H, H-5
pyraz.); 13C NMR (CDCl3): 24.2 (2CH2 cyclopent.), 32.9 (2CH2
cyclopent.), 45.5 (2CH2 cyclopent. epox.), 50.8 (CH epox.), 53.8
(CH2N), 56.1 (OCH3), 80.5 (OCH cyclopent.), 103.0 (C-4 pyraz.), 112.0
(C-5 Ar), 112.7 (C-2 Ar), 118.3 (C-6 Ar), 126.5 (C-1 Ar), 131.4 (C-5
pyraz.), 147.8 (C-3 Ar), 150.0 (C-4 Ar), 151.8 (C-3 pyraz.); Anal.
(C18H22N2O3) calcd for C, H, N.
A mixture of 1-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-
pyrazol-1-yl}-3-morpholin-4-ylpropan-2-ol 8a (0.60 g, 1.63 mmol)
and sodium acetate (0.2 g, 2.4 mmol) in acetic anhydride (5 mL)
was heated at 40e50 ꢂC for 5 h. After cooling to room temperature,
the mixture was poured into water (100 mL), extracted with diethyl
ether (3 ꢀ 10 mL) and the organic phase was washed with water
(3 ꢀ 20 mL), brine (3 ꢀ 20 mL), dried (MgSO4) and concentrated
under reduced pressure. The pure product was obtained as a yellow
oil. Yield: 52%; 1H NMR (CDCl3):
d 1.56e2.06 (m, 8H, 4CH2 cyclo-
pent.), 2.12 (s, 3H, CH3), 2.61e2.92 (m, 6H, 2CH2N morph. þ CH2N),
3.70e4.00 (m, 7H, OCH3 þ 2CH2O morph.), 4.32e4.58 (m, 2H, CH2N
pyraz.), 4.81e4.99 (m, 1H, OCH cyclopent.), 5.40e5.57 (m, 1H,
CHOAc), 6.50 (d, J ¼ 1.6 Hz, 1H, H-4 pyraz.), 6.84e7.60 (m, 4H, H-
6 þ H-5 þ H-2 Ar þ H-5 pyraz.); IR (KBr) cmꢁ1: 1715 (CO); Anal.
(C24H33N3O5) calcd for C, H, N.
4.1.17. General procedure for the synthesis of 3-[3-(cyclopentyloxy)-
4-methoxyphenyl]-1H-pyrazol-1-yl]-3-cycloaminopropan-2-ols
8a,b
4.1.20. General procedure for the synthesis of 4-(2-{3-[3-
(cyclopentyloxy)-4-methoxyphenyl]-1H-pyrazol-1-yl}acetyl)amines
10a,b
An excess of the appropriate amine (2 mL) was added to 3-[3-
(cyclopentyloxy)-4-methoxyphenyl]-1-(oxiran-2-ylmethyl)-1H-
pyrazole 23 (1 g, 3.15 mmol) and the mixture was heated at
50e60 ꢂC for 18 h. After cooling to room temperature, the mixture
was diluted with diethyl ether (20 mL) and the organic phase was
washed with water (20 mL), dried (MgSO4) and concentrated under
reduced pressure. The crude was purified by Florisil (100e200
mesh) column chromatography using diethyl ether as the eluent.
The desired compounds were obtained as yellow oils that were
treated with a HCl saturated ethanol solution to obtain the corre-
sponding hydrochloride salts as white solids.
The suitable chloroacetylamine (15 mmol) was added to a so-
lution of 3-[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-pyrazole 22
(0.52 g, 2 mmol) in an. DMF (3 mL). Then, TEA (1 mL, 0.73 g,
7.19 mmol) was added dropwise and the mixture was heated at
100 ꢂC for 18 h. After cooling to room temperature, the mixture was
poured into water (100 mL) and extracted with diethyl ether
(3 ꢀ 25 mL). The organic phase was dried (MgSO4) and concen-
trated under reduced pressure. The crude crystallized by addition of
a mixture of diethyl ether/petroleum ether (boiling point 40e60 ꢂC)
(1:1) and the desired compounds were obtained as white solids,
which were recrystallized by absolute ethanol (10a) or diethyl ether
(10b).
4.1.17.1. 1-{3-[3-(Cyclopentyloxy)-4-methoxyphenyl]-1H-pyrazol-1-
yl}-3-morpholin-4-ylpropan-2-ol dihydrochloride (8a). Yield: 51%;
mp: 162e163 ꢂC; 1H NMR (CDCl3):
d 1.70e2.40 (m, 8H, 4CH2
cyclopent.), 3.20e3.60, 3.74e4.00, 4.10e4.25 and 4.25e4.50 (4 m,
10H, 4CH2 morph. þ CH2N), 4.05 (s, 3H, OCH3), 4.80e5.35 (m, 5H,
OCH cyclopent. þ CHOH þ CH2N pyraz. þ OH, 1H disappears with
D2O), 6.85 (br s, 1H, H-4 pyraz.), 7.09 (d, J ¼ 8.2 Hz, 1H, H-5 Ar), 7.60
(d, J ¼ 8.2 Hz, 1H, H-6 Ar), 7.75 (s, 1H, H-2 Ar), 8.24 (s, 1H, H-5
pyraz.), 11.70e11.95 (br s, 1H, NHþ disappears with D2O); IR (KBr):
cmꢁ1 3550, 3150 (OH); Anal. (C22H31N3O4.2HCl.2H2O) calcd for C, H,
N.
4.1.20.1. 4-(2-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-pyrazol-
1-yl}acetyl)morpholine (10a). Yield: 58%; mp: 131e132 ꢂC; 1H NMR
(CDCl3):
d 1.60e2.12 (m, 8H, 4CH2 cyclopent.), 3.60e4.00 (m, 11H,
OCH3 þ 4CH2 morph.), 4.83e5.10 (m, 3H, CH2CO þ OCH cyclopent.),
6.76 (d, J ¼ 2.8 Hz, 1H, H-4 pyraz.), 6.96 (d, J ¼ 8.2 Hz, 1H, H-5 Ar),
7.47 (dd, J ¼ 8.2, 1.6 Hz, 1H, H-6 Ar), 7.68 (d, J ¼ 1.6 Hz, 1H, H-2 Ar),
7.84 (d, J ¼ 2.8 Hz, 1H, H-5 pyraz.); IR (KBr) cmꢁ1: 1663 (CO); Anal.
(C21H27N3O4) calcd for C, H, N.

98
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
4.1.21. Synthesis of 4-(3-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]-
1H-pyrazol-1-yl}propanoyl)morpholine 11a
solution of 3-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-pyr-
azol-1-yl}-propanoic acid 25 (0.67 g, 2 mmol) in anhydrous DMF
(5 mL),. Then, the mixture was heated at 80 ꢂC for 15 h. After cooling
to room temperature, the mixture was poured into water (100 mL),
extracted with diethyl ether (3 ꢀ 25 mL) and the organic phase was
dried (MgSO4) and concentrated under reduced pressure. The
crude was purified by silicagel (100e200 mesh) column chroma-
tography using diethyl ether as the eluent to afford the pure
4-(3-chloropropanoyl)morpholine (3.20 g, 18 mmol), solved in
anhydrous DMF (6 mL), was slowly added at 0 ꢂC to a solution of 3-
[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-pyrazole 22 (0.77 g,
3 mmol) in anhydrous DMF (5 mL). Then, TEA (0.80 g, 7.91 mmol,
1.1 mL) was added dropwise and the mixture was heated at 120 ꢂC
for 48 h. After cooling to room temperature, the mixture was
poured into water (100 mL), extracted with diethyl ether
(3 ꢀ 25 mL) and the organic phase was dried (MgSO4) and
concentrated under reduced pressure. The crude was purified by
silicagel (100e200 mesh) column chromatography using a gradient
elution from diethyl ether to diethyl ether/methanol (1:1). The pure
product was obtained as a light yellow oil. Yield: 38%; 1H NMR
product as a light yellow oil. Yield: 71%; 1H NMR (CDCl3):
d 1.08 (d,
J ¼ 6.2 Hz, 3H, CH3 morph), 1.18 (d, J ¼ 6.2 Hz, 3H, CH3 morph.),
1.60e2.06 (m, 8H, 4CH2 cyclopent.), 2.23e2.39, 2.62e2.7, 2.82e3.19
and 3.23e3.70 (4 m, 8H, 2CH2 morph. þ CH2CO þ 2CH morph.),
3.89 (s, 3H, OCH3), 4.50e4.59 (m, 2H, CH2N), 4.86e4.97 (m,1H, CHO
cyclopent.), 6.43 (d, J ¼ 2.2 Hz, 1H, H-4 pyraz.), 6.93 (d, J ¼ 8.4 Hz,
1H, H-5 Ar), 7.33e7.41 (m, 2H, H-2 þ H-6 Ar), 7.48 (d, J ¼ 2.2 Hz, 1H,
H-5 pyraz.); IR (CHCl3) cmꢁ1: 1671 (CO); Anal. (C24H33N3O4) calcd
for C, H, N.
(CDCl3):
d 1.54e2.07 (m, 8H, 4CH2 cyclopent.), 2.91e3.00 (m, 2H,
CH2CO), 3.32e3.67 (m, 8H, 4CH2 morph.), 3.89 (s, 3H, OCH3), 4.52 (t,
J ¼ 6.4 Hz, 2H, CH2N), 4.89e5.00 (m, 1H, OCH cyclopent.), 6.44 (d,
J ¼ 2.4 Hz, 1H, H-4 pyraz.), 6.92 (d, J ¼ 8.2 Hz, 1H, H-5 Ar), 7.22e7.40
(m, 2H, H-6 þ H-2 Ar), 7.49 (d, J ¼ 2.4 Hz, 1H, H-5 pyraz.); IR (film)
cmꢁ1: 1654 (CO); Anal. (C22H29N3O4) calcd for C, H, N.
4.2. Computational studies methods
4.2.1. Molecular dynamics and docking
4.1.22. Synthesis of methyl 3-{3-[3-(cyclopentyloxy)-4-
methoxyphenyl]-1H-pyrazol-1-yl}-propanoate 24
Compounds 4a and 8a were built, parameterized (Gasteiger-
Hückel method) and energy minimized within OpenEye Scientific
Software [49] using MMFF94s force field. Regarding the protein,
according to our previous studies, 1XOQ [50] was selected. For all
calculations, only one molecular chain (chain A) was kept and
hydrogen atoms were added at a neutral pH. The six waters into the
catalytic site were maintained. Docking studies were performed
using AutoDock 4.0 [51]. Molecular dynamics simulations of the
complex PDE4D-inhibitor were performed using AMBER12 package
[52], starting the experiments from the docking solutions in order
to assess their stability along the trajectories. The ligands were
parameterized using the semiempirical quantum chemistry
method AM1-BCC to derive partial charges, while the other con-
stants were derived by the Antechamber module for the GAFF force
field [53,54]. The protein was parametrized with the AMBER ff03
force field. Mg2þ and Zn2þ ions were treated according to the “non-
bonded” model method [55]. In order to remove all the possible bad
contacts between atoms, preliminary minimization in vacuum of
the system was run, using a steepest descendent algorithm until
energy convergence of 0.0001 kcal/mol. The inhibitor-protein
complex was solvated in a truncated octahedral periodic box with
8 Å of perimetral solvent thickness using TIP3P water model. Naþ
ions were added to neutralize the whole system. The simulations
were performed at neutral pH, with histidines 164 and 200 pro-
tonated at delta position to coordinate the Zn2þ ion. The water
molecule between the two ions was treated as hydroxide ion as
suggested by studies of Li et al. [56] and by MD simulations studies
on PDE11 previously performed [57]. Since HIS 160, close to the
hydroxide ion, can readily capture a proton, it was protonated at
both delta and epsilon positions. All the bonds involving hydrogen
atoms were constrained by the SHAKE algorithm [58] and the time
step was set to 2 fs. The non-bonded cut-off distance was 8 Å and
long range electrostatic interactions were calculated using the
Particle Mesh Ewald (PME) method. Four steps of minimization
were performed, keeping the position of protein and ligand
restrained for four minimization steps by a force constant of
500 kcal/mol/Å2, 100 kcal/mol/Å2 in the second, 10 kcal/mol/Å2 in
the third and without ligand and protein restrained in the fourth.
All the minimization steps were run till the system energy
convergence of 0.0001 kcal/mol was achieved. Next, the system
was heated from 0 to 100 K in 30 ps, using Langevin dynamics at
constant volume, and from 100 to 300 K in 20 ps at constant
pressure. After that, the system was equilibrated for 7 ns at constant
pressure of 1 atm. Position restrains of 10 kcal/mol/Å were used on
Methyl-3-bromopropionate (1.91 g, 11.45 mmol) and TEA (1 mL,
0.73 g, 7.19 mmol) were added dropwise at 0 ꢂC to a solution of 3-
[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-pyrazole 22 (0.6 g,
2.25 mmol) in an. acetonitrile (15 mL); then, the mixture was
heated at reflux for 48 h. After cooling to room temperature, the
solvent was removed under reduced pressure and the crude was
solved in diethyl ether (30 mL). The organic phase was washed with
water (3 ꢀ 20 mL), 4 N NaOH solution (2 ꢀ 20 mL), brine
(2 ꢀ 20 mL), dried (MgSO4), and concentrated under reduced
pressure. The obtained oil was purified by silicagel (100e200 mesh)
column chromatography, using diethyl ether as the eluent to afford
the pure product as a light yellow oil. Yield: 67%, 1H NMR (CDCl3):
d
1.50e2.18 (m, 8H, 4CH2 cyclopent.), 3.00 (t, J ¼ 6.0 Hz, 2H, CH2CO),
3.73 (s, 3H, OCH3), 3.90 (s, 3H, OCH3-Ar), 4.49 (t, J ¼ 6.0 Hz, 2H,
CH2N), 4.77e5.08 (m, 1H, OCH cyclopent.), 6.45 (d, J ¼ 3.0 Hz,1H, H-
4 pyraz.), 6.95 (d, J ¼ 6.8 Hz,1H, H-5 Ar), 7.20e7.50 (m, 3H, H-6 þ H-
2 Ar þ H-5 pyraz.); IR (CHCl3) cmꢁ1: 1735 (CO); Anal. (C19H24N2O4)
calcd for C, H, N.
4.1.23. Synthesis of 3-{3-[3-(cyclopentyloxy)-4-methoxyphenyl]-
1H-pyrazol-1-yl}-propanoic acid 25
NaOH (0.2 g, 5.0 mmol) was added to a solution of methyl 3-{3-
[3-(cyclopentyloxy)-4-methoxyphenyl]-1H-pyrazol-1-yl}-prop-
anoate 24 (1 g, 3 mmol) in 96% ethanol (3 mL) and the mixture was
heated at 60 ꢂC for 4 h. Then, the solvent was removed under
reduced pressure, the crude was solved in water (20 mL) and a 1 N
HCl solution was slowly added until pH became ¼ 1. The light
yellow solid obtained was filtered, washed with water and re-
crystallized by a mixture of diethyl ether/dichloromethane (1:1).
Yield: 74%; mp: 131e133 ꢂC; 1H NMR (CDCl3):
d 1.52e2.11 (m, 8H,
4CH2 cyclopent.), 3.02e3.10 (m, 2H, CH2), 3.91 (s, 3H, OCH3),
4.42e4.54 (m, 2H, CH2N), 4.86e5.02 (m, 1H, OCH cyclopent.), 6.54
(d, J ¼ 2.3 Hz, 1H, H-4 pyraz.), 6.93 (d, J ¼ 8.4 Hz, 1H, H-5 Ar),
7.21e7.41 (m, 2H, H-6 þ H-2 Ar), 7.52 (d, J ¼ 2.3 Hz, 1H, H-5 pyraz.);
IR (CHCl3) cmꢁ1: 3515 (OH), 1721 (CO); Anal. (C18H22N2O4) calcd for
C, H, N.
4.1.24. Synthesis of 4-(3-{3-[3-(cyclopentyloxy)-4-
methoxyphenyl]-1H-pyrazol-1-yl}propanoyl)2,6-
dimethylmorpholine 11b
TEA (0.30 g, 6 mmol, 0.4 mL), 2,6-dimethylmorpholine (0.50 mL,
4 mmol) and DPPA (0.69 g, 2.5 mmol) were added at 0 ꢂC to a

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
99
the ligand and protein during heating and equilibration steps. After
4.6. Microdialysis experiments
4.6.1. Animals
For the in vivo studies performed at Genoa University, all the
experimental procedures were approved by the Italian Ministry of
Health (protocol no. 29823) and were in accordance with the Eu-
ropean (2010/63/UE) and Italian (D.L. 26/2014) guidelines on the
use and care of laboratory animals.
equilibration, production molecular dynamics phase was per-
formed at 300 K using constant pressure of 1 atm. The exploited
method was validated performing a 20 ns MD simulation of the
complex 1XOQ-Roflumilast. Then 70 ns MD run were performed for
the docking complexes. The resulting trajectories were analysed
using Amber Tools and VMD programs. The root mean square de-
viation (RMSD) was calculated for the protein backbone and ligand
atoms using least-squares fitting. The pmemd CUDA program of the
AMBER12 package was used for MD simulations running on a
cluster Tesla K20 Graphical Processing Unit (GPU).
4.6.2. Surgery and cAMP determination
Male Sprague-Dawley rats (250e300 g) were anaesthetized
with intraperitoneal zoletil (tiletamine and zolazepam, 30 mg/kg),
placed in a stereotaxic frame (David Kopf Instruments, West
Hempstead, NY, USA) and implanted with microdialysis probes into
the dorsal hippocampi according to the following coordinates:
AP ¼ þ3.8, H ¼ þ6.5 from the interaural line (Paxinos and Watson
atlas, 1986). A piece of dialysis fibre (co-polymer of acrylonitrile
sodium methallyl sulphonate; AN69HF Hospal S.p.A., Bologna,
Italy; 0.3 mm outer diameter, 40 kDa mol. wt. cut-off) was covered
with epoxy glue to limit dialysis to the area of interest (8 mm glue-
free length). The skull was exposed and two holes were drilled on
the lateral surfaces. The dialysis probe was inserted transversely
into the brain so that the glue-free zone was exactly positioned into
the target area. Stainless steel cannulae (22-gauge diameter,15-mm
long) were glued to the ends of the fibre, bent up and fixed to the
skull surface with dental cement and modified Eppendorf tips.
After a 24-h recovery period, freely-moving rats were placed into
observation cages and the probes infused with modified Ringer's
4.2.2. In silico evaluation of pharmacokinetic properties
The prediction of ADMET properties was performed using the
Advanced Chemistry Development (ACD) Percepta platform (www.
acdlabs.com).
Any pharmacokinetic and metabolism descriptor was evaluated
by Percepta on the basis of training libraries, implemented in the
software, which include a consistent number of molecules whose
pharmacokinetic and toxicity profile are experimentally known.
4.3. Cytotoxicity and genotoxicity assays
The cells used in this study (human neuroblastoma cell line
HTLA-230, HTLA) were grown in Roswell Park Memorial Institute
medium (RPMI), with 0.1 mM non-essential aminoacids and 10%
fetal bovine serum.
For the cytotoxicity assay, cells were treated for 24 h with
medium (flow rate 5 ml/min; CMA/100 microinjection pump, CMA
100
m
M of the different compounds dissolved in DMSO. Control
Microdialysis, Stockholm, Sweden) containing (in mM): NaCl 145,
KCl 3, MgCl2 1, CaCl2 1.26, buffered at pH 7.4 with 2 mM phosphate
buffer. Following a stabilization period of 1 h, consecutive samples
were collected every 20 min. Compound 8a was locally adminis-
tered through the dialysis probe (retrodialysis) after three control
samples had been collected to estimate cAMP basal values and was
present in the infusion solution for the time reported in Fig. 7. At
the end of the experiment, the correct position of the probe was
verified by optical examination of the fibre tract. Animals pre-
senting hemorrhagic lesions were rejected.
cAMP content in the dialysates was assayed by a commercially
available radioimmunoassay kit (IZOTOP, Institute of isotopes, Co.,
Ltd., Budapest, Hungary) using the acetylation protocol. Under
these experimental conditions, the detection limit of the assay is
2 fmol (standard curve range 2e128 fmol).
cells received the same volume of solvent (1
m
L/ml medium). At the
end of the incubation period, conditioned media were analysed for
lactate-dehydrogenase release using the Cytoxicity Detection Kit-
PLUS
(Roche, Germany), according to manufacturer protocols.
To evaluate genotoxicity, HTLA cells were treated for 24 h with
the different compounds dissolved in DMSO and then processed for
total protein extraction as described previously [59]. Immunoblots
were done according to standard methods, using the following
antibodies: mouse monoclonal [2F3] to gamma H2A.X (phospho
S139) and rabbit polyclonal to Histone H2A.X (Abcam, UK); anti-
mouse and anti-rabbit secondary antibodies coupled to horse-
radish peroxidase (GE Healthcare, UK). Proteins were visualized
with an enzyme-linked chemiluminescence detection kit according
to the manufacturer's instructions (GE Healthcare). Chem-
iluminescence was monitored by exposure to films and signals
were analysed under non-saturating condition with an image
densitometer (Bio-Rad, CA, USA).
4.7. Behavioral studies
4.7.1. Animals
For the studies performed at Maastricht University, the experi-
mental procedures were approved by the local ethics committee of
Maastricht University for animal experiments and met govern-
mental guidelines. For the 8a study, twenty-four male Wistar rats
were supplied by Charles River (Sulzfeld. Germany) and tested
between 4 and 5 months of age. Average body weight at the
beginning of the study was 436 g. All animals were housed indi-
vidually in standard green line Techiplast IVC cages on sawdust
bedding. The animals were housed on a reversed 12/12-h light/dark
cycle (lights on from 19:00 h to 07:00 h) and had free access to food
and water. The rats were housed and tested in the same room. A
radio, playing softly, provided background noise in the room.
Testing was performed between 09:00 h and 18:00 h, with a small
desk lamp providing light for the experimenter.
4.4. cAMP enzymatic immunoassay (EIA)
Quantification of intracellular cAMP was performed with
DetectX® Direct Cyclic AMP Enzyme Immunoassay Kit (Arbor Assay,
MI, USA), following the manufacturer's protocol. cAMP levels were
calculated according to the standard curves prepared on the same
EIA plates.
4.5. Pharmacokinetic analysis
The pharmacokinetic analysis for 8a and 11b has been carried
out by the Center for Drug Metabolism and Pharmacokinetics
Research, Shanghai Institute of Materia Medica (Zhangjiang Hi-
Tech Park, Pudong, Shanghai, China).
Detailed description of experimental procedure is reported on-
line as supporting information material.
4.7.2. Treatments
8a and scopolamine were both dissolved in saline. Vehicle or

100
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
doses of 0.001, 0.003 and 0.01 mg/kg 8a were given intraperitoneal
(IP) at an injection volume of 1 mL/kg, 3 h after the first trial with
the 24 h inter-trial interval ORT. For the 1 h inter-trial interval ORT
with a scopolamine induced memory deficit, vehicle or scopol-
amine (0.1 mg/kg IP) were administered 30 min before the start of
the first trial. 8a was tested at doses of 0.001, 0.003 and 0.01 mg/kg
(IP), administered 30 min before the first trial. Doses of, 0.003, 0.03
and 0.3 mg/kg of 8a were tested in the xylazine (10 mg/kg IP)/ke-
tamine (10 mg/kg IP) induced a2-adrenoceptor-mediated anaes-
thesia test. Vehicle or 8a were administered subcutaneously (SC)
15 min after the xylazine/ketamine injections.
index of the conditions to zero (i.e. chance level). However, com-
parison of the d2 value with the value zero with no variance is not
the most suitable way of analysing object recognition since there is
an increased chance of making a type I error. For this reason, a one-
way ANOVA was performed to find significant differences between
d2 values of the conditions. When significant, post-hoc Dunnett's t-
tests were performed to compare the experimental conditions to
the vehicle group.
Trial
Exploration time (s)
Discrimination index
4.7.3. Object recognition task (rats)
The ORT has been performed as previously described elsewhere
[41,44].
T1
T2
e1 ¼ a1 þ a2
e2 ¼ a3 þ b
e
d2 ¼ (b ꢁ a3)/e2
Detailed description of experiments is reported online as sup-
porting information material.
4.8.4. Xylazine/ketamine induced a2-adrenoceptor-mediated
anaesthesia test
4.7.4. Xylazine/ketamine induced a2-adrenoceptor-mediated
anaesthesia test
Total time of anaesthesia was used as an outcome measurement.
To control for the differences between test days, the vehicle con-
dition was set to 100% and the times of the 8a conditions were
expressed in relation to this. To test for statistical significance be-
tween conditions, a one-way ANOVA was performed with these
relative times. When significant, post-hoc Dunnett's t-tests were
performed to compare the experimental conditions to the vehicle
condition.
To investigate possible emetic-like effect of 8a the rats were
anaesthetized with an IP injection of 10 mg/kg ketamine and
10 mg/kg xylazine (injection volume of 1 mL/kg). 15 min after the
induction of the anaesthesia, vehicle or one of the concentrations of
8a solutions was administrated, after which the rats were put in
dorsal position. The outcome measurement was the total time the
animals were anaesthetized, measured from the induction of the
anaesthesia to the righting reflex when the rats were back on all
four paws. To get a sufficient group size, the experiments were
repeated three times with 2 recovery days in between. The animals
received the same condition only once.
Conflict of interest
OB, CB, AR, EF, RR, OA, JP have a proprietary interest in selective
PDE4D inhibitors, including 8a.
4.8. Statistical analysis
Authors contributions
4.8.1. In vitro cAMP evaluation
CB, OB: molecules design and SAR; MM, CR, AR: molecules
synthesis, purification and characterization; JP, BTJvH, NPvG:
behavior studies; OA: pharmacokinetic analysis; RR, EF, CR, BM,
MAP: cyto- and genotoxic tests; in vitro and in vivo cAMP mea-
sures; EC, PF, SG, PdU, AO, LM: in silico physico-chemical properties
evaluation, docking studies and molecular dynamics simulation.
OB, CB, PF, RR, EF, OA, JP wrote the article.
Data presented (mean S.E.M) are expressed as pmol/ml. Dif-
ferences were analysed by ANOVA followed by multiple compari-
son Dunnett's test. Differences were considered significant at the
level of p < 0.05, at least.
4.8.2. In vivo cAMP evaluation
Data presented (mean S.E.M) are expressed as percentages of
the mean basal value (defined as 100%) that was determined by
averaging the content of the first 3 samples collected before drug
treatment. Differences were analysed by ANOVA followed by Bon-
ferroni's post hoc comparison test. Differences were considered
significant at the level of p < 0.05, at least.
Acknowledgements
Alzheimer Association is gratefully acknowledged for financial
support to this study (IIRG 11-208306). We thank Dr. M. Anzaldi, R.
Raggio, Mr. O. Gagliardo, and F. Tuberoni for spectral recording and
Mr. V. Ruocco for the informatics support to calculations.
B.T.J. van Hagen is financially supported by the Human
Enhancement and Learning (HEaL) initiative of Maastricht Univer-
sity. This work has been supported by the Italian Ministry of Edu-
cation and Research through the Flagship “InterOmics” (PB05) and
European MIMOmics (305280) Projects.
4.8.3. ORT
The readout parameters of the ORT are the times that rodents
spent on exploring each object during T1 and T2 [42,60]. The
exploration time (in seconds) of each object during T1 is presented
as ‘a1’ and ‘a2’. The time spent exploring the familiar and the novel
object in T2 is represented as ‘a3’ and ‘b’, respectively. Using this
information, the following variables were calculated: e1, e2 and d2
(see Table below). The d2 index is a relative measure of discrimi-
nation corrected for exploratory activity and it has been shown not
to be correlated to e1 or e2 [60]. The d2 index can range from ꢁ1 to
1. A significant difference from zero indicates that the rats
remembered the object from T1, and a difference from the vehicle
condition signifies an actual memory improvement. Of note, rats
require a minimum amount of exploration in order to show reliable
memory performance [41,42]. Therefore, animals were removed
from the analysis if they spent less than 9 s exploring the objects
during T1 or T2. One sample t-tests were used to compare the d2
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2016.08.018.
References
[1] R. Bourtchuladze, B. Frenguelli, J. Blendy, D. Cioffi, G. Schutz, A.J. Silva, Defi-
cient long-term memory in mice with a targeted mutation of the cAMP-
responsive element-binding protein, Cell 79 (1994) 59e68.
[2] J.C.P. Yin, J.S. Wallach, M. Del Vecchio, E.L. Wilder, H. Zhou, W.G. Quinn,

C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
101
T. Tully, Induction of a dominant negative CREB transgene specifically blocks
long-term memory in Drosophila, Cell 79 (1994) 49e58.
[3] O.V. Vitolo, A. Sant'Angelo, V. Costanzo, F. Battaglia, O. Arancio, M. Shelanski,
Amyloid beta -peptide inhibition of the PKA/CREB pathway and long-term
potentiation: reversibility by drugs that enhance cAMP signaling, Proc. Natl.
Acad. Sci. U. S. A. 99 (2002) 13217e13221.
[4] B. Gong, O.V. Vitolo, F. Trinchese, S. Liu, M. Shelanski, O. Arancio, Persistent
improvement in synaptic and cognitive functions in an Alzheimer mouse
model after rolipram treatment, J. Clin. Investig. 114 (2004) 1624e1634.
[5] A.F. Teich, R.E. Nicholls, D. Puzzo, J. Fiorito, R. Purgatorio, M. Fa', O. Arancio,
Synaptic therapy in Alzheimer's disease: a CREB-centric approach, Neuro-
therapeutics 12 (2015) 29e41.
[6] M.E. Gurney, E.C. D'Amato, A.B. Burgin, Phosphodiesterase-4 (PDE4) molecular
pharmacology and Alzheimer's disease, Neurotherapeutics 12 (2015) 49e56.
[7] M. Barad, R. Bourtchouladze, D.G. Winder, H. Golan, E. Kandel, Rolipram, a
type IV-specific phosphodiesterase inhibitor, facilitates the establishment of
long-lasting long-term potentiation and improves memory, Proc. Natl. Acad.
Sci. U. S. A. 95 (1998) 15020e15025.
[27] J. Singh, M.E. Gurney. Preparation of substituted aza-bridged bicyclics as PDE4
inhibitors for treatment of cardiovascular and CNS disease and other disor-
ders. WO2010059836A1 (2010).
[28] O. Bruno, C. Brullo, N. Arduino, S. Schenone, A. Ranise, F. Bondavalli,
L. Ottonello, P. Dapino, F. Dallegri, Synthesis and biological evaluation of
neutrophilic inflammation inhibitors, Il Farm. 59 (2004) 223e235.
[29] O. Bruno, A. Romussi, A. Spallarossa, C. Brullo, S. Schenone, F. Bondavalli,
F.N. Vanthuyne, C. Roussel, New selective phosphodiesterase 4D inhibitors
differently acting on long, short, and supershort isoforms, J. Med. Chem. 52
(2009) 6546e6557.
[30] C. Brullo, M. Massa, M. Rocca, C. Rotolo, S. Guariento, D. Rivera, R. Ricciarelli,
E. Fedele, P. Fossa, O. Bruno, Synthesis, biological evaluation, and molecular
modeling of new 3-(Cyclopentyloxy)-4-methoxybenzaldehyde O-(2-(2,6-
Dimethylmorpholino)-2-oxoethyl) oxime (GEBR-7b) related phosphodies-
terase 4D (PDE4D) inhibitors, J. Med. Chem. 57 (2014) 7061e7072.
[31] C. Brullo, M. Massa, C. Villa, R. Ricciarelli, D. Rivera, M.A. Pronzato, E. Fedele,
E. Barocelli, S. Bertoni, L. Flammini, O. Bruno, Synthesis, biological activities
and pharmacokinetic properties of new fluorinated derivatives of selective
PDE4D inhibitors, Bioorg. Med. Chem. 23 (2015) 3426e3435.
[32] P. D'Ursi, S. Guariento, G. Trombetti, A. Orro, E. Cichero, L. Milanesi, P. Fossa,
O. Bruno, Further insights in the binding mode of selective inhibitors to hu-
man PDE4D enzyme combining docking and molecular dynamics, Mol. Inf.
(2016), http://dx.doi.org/10.1002/minf.201501033.
[8] R. Bourtchouladze, R. Lidge, R. Catapano, J. Stanley, S. Gossweiler,
D. Romashko, R. Scott, T. Tully, A mouse model of Rubinstein-Taybi syndrome:
defective long-term memory is ameliorated by inhibitors of phosphodies-
terase 4, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 10518e10522.
[9] B. Gong, O.V. Vitolo, F. Trinchese, S. Liu, M. Shelanski, O. Arancio, Persistent
improvement in synaptic and cognitive functions in an Alzheimer mouse
model after rolipram treatment, J. Clin. Investig. 114 (2004) 1624e1634.
[33] M.J. Ashton, D.C. Cook, G. Fenton, J.A. Karlsson, M.N. Palfreyman, D. Raeburn,
Selective type IV phosphodiesterase inhibitors as antiasthmatic agents. The
ꢀ
[10] C. Giampa, S. Middei, S. Patassini, A. Borreca, F. Marullo, D. Laurenti, Phos-
synthesis
and
biological
activities
of
3-(cyclopentyloxy)-4-
phodiesterase type IV inhibition prevents sequestration of CREB binding
protein, protects striatal parvalbumin interneurons and rescues motor deficits
in the R6/2 mouse model of Huntington's disease, Eur. J. Neurosci. 29 (2009)
902e910.
methoxybenzamides and analogues, J. Med. Chem. 37 (1994) 1696e1703.
[34] H.O.H. Kan, I. Adachi, R. Kido, K. Hirose, Isoxazoles. XVIII. Synthesis and
pharmacological properties of 5-Aminoalkyl- and 3-Aminoalkylisoxazoles and
related derivatives, J. Med. Chem. 10 (1967) 411e418.
[11] O.A. Reneerkens, K. Rutten, H.W. Steinbusch, A. Blokland, J. Prickaerts, Selec-
tive phosphodiesterase inhibitors: a promising target for cognition enhance-
ment, Psychopharmacol. Berl. 202 (2009) 419e443.
[35] H. Kano, I. Adachi, R. Kido, K. Hirose, Isoxazoles. XVIII. Synthesis and phar-
macological properties of 5-Aminoalkyl- and 3-Aminoalkylisosazoles and
related derivatives, J. Med. Chem. 10 (1967) 411e418.
[12] A. Garcia-Osta, M. Cuadrado-Tejedor, C. Garcia-Barroso, J. Oyarzabal, R. Franco,
Phosphodiesterases as therapeutic targets for Alzheimer's disease, ACS Chem.
Neurosci. 3 (2012) 832e844.
[36] V.K. Aggarwal, J. de Vicente, R.V. Bonnert, A novel one-pot method for the
preparation of pyrazoles by 1,3-dipolar cycloadditions of diazo compounds
generated in situ, J. Org. Chem. 68 (2003) 5381e5383.
[13] O. Bruno, E. Fedele, J. Prickaerts, L. Parker, E. Canepa, C. Brullo, A. Cavallero,
E. Gardella, A. Balbi, C. Domenicotti, E. Bollen, H. Gijselaers, T. Vanmierlo,
K. Erb, C. Limebeer, F. Argellati, U. Marinari, M. Pronzato, R. Ricciarelli, GEBR-
7b, a novel PDE4D selective inhibitor that improves memory in rodents at
non-emetic doses, Brit. J. Pharmacol. 164 (2011) 2054e2063.
[14] A.S.R. Sierksma, D.L.A. Van den Hove, F. Pfau, M. Philippens, O. Bruno,
E. Fedele, R. Ricciarelli, H.W.M. Steinbusch, T. Vanmierlo, J. Prickaerts,
Improvement of spatial memory function in APPswe/PS1dE9 mice after
chronic inhibition of phosphodiesterase type 4D, Neuropharmacology 77
(2014) 120e130.
[15] A.B. Burgin, O.T. Magnusson, J. Singh, P. Witte, B.L. Staker, J.M. Bjornsson,
M. Thorsteinsdottir, S. Hrafnsdottir, T. Hagen, A.S. Kiselyov, L.J. Stewart,
M.E. Gurney, Design of phosphodiesterase 4D (PDE4D) allosteric modulators
for enhancing cognition with improved safety, Nat. Biotechnol. 28 (2010)
63e70.
[16] Y.F. Li, Y.F. Cheng, Y. Huang, M. Conti, S.P. Wilson, J.M. O'Donnell, H.T. Zhang,
Phosphodiesterase-4D knock-out and RNA interference-mediated knock-
down enhance memory and increase hippocampal neurogenesis via increased
cAMP signaling, J. Neurosci. 31 (2011) 172e183.
[37] H. Waterbeemd, E. Gifford, ADMET in silico modelling: towards prediction
paradise? Nat. Rev. Drug Discov. 2 (2003) 192e204.
[38] W.M. Bonner, C.E. Redon, J.S. Dickey, A.J. Nakamura, O.A. Sedelnikova, S. Solier,
Y. Pommier, GammaH2AX and cancer, Nat. Rev. Cancer 8 (2008) 957e967.
[39] A. Montecucco, G. Biamonti, Cellular response to etoposide treatment, Cancer
Lett. 252 (2007) 9e18.
[40] A. Ennaceur, J. Delacour, A new one-trial test for neurobiological studies of
memory in rats. 1: behavioral data, Behav. Brain Res. 31 (1988) 47e59.
[41] J. Prickaerts, A. Sik, F.J. van der Staay, J. de Vente, A. Blokland, Dissociable
effects of acetylcholinesterase inhibitors and phosphodiesterase type 5 in-
hibitors on object recognition memory: acquisition versus consolidation,
Psychopharmacol. Berl. 177 (2005) 381e390.
[42] S. Akkerman, A. Blokland, O. Reneerkens, N.P. van Goethem, E. Bollen,
H.J. Gijselaers, C.K. Lieben, H.W. Steinbusch, J. Prickaerts, Object recognition
testing: methodological considerations on exploration and discrimination
measures, Behav. Brain Res. 232 (2012) 335e347.
[43] A. Sik, P. van Nieuwehuyzen, J. Prickaerts, A. Blokland, Performance of
different mouse strains in an object recognition task, Behav. Brain Res. 147
(2003) 49e54.
[17] M.D. Houslay, G.S. Baillie, D.H. Maurice, cAMP-Specific phosphodiesterase-4
enzymes in the cardiovascular system: a molecular toolbox for generating
compartmentalized cAMP signaling, Circ. Res. 100 (2007) 950e966.
[18] M.D. Houslay, R.D. Adams, PDE4 cAMP phosphodiesterases: modular enzymes
that orchestrate signalling cross-talk, desensitization and compartmentaliza-
tion, Biochem. J. 370 (2003) 1e18.
[19] W. Richter, M. Conti, The oligomerization state determines regulatory prop-
erties and inhibitor sensitivity of type 4 cAMP-specific phosphodiesterases,
J. Biol. Chem. 279 (2004) 30338e30348.
[20] W. Richter, M. Conti, Dimerization of type 4 cAMP-specific phosphodiester-
ases is mediated by the upstream conserved regions (UCRs), J. Biol. Chem. 277
(2002) 40212e40221.
[21] D. Fox, A.B. Burgin, M.E. Gurney, Structural basis for the design of selective
phosphodiesterase 4B inhibitors, Cell. Signal. 26 (2014), 657e633.
[22] P. Cedervall, A. Aulabaugh, K.F. Geoghegan, T.J. McLellan, J. Pandit, Engineered
stabilization and structural analysis of the autoinhibited conformation of
PDE4, Proc. Natl. Acad. Sci. U. S. A. 112 (2015) E1414.
[23] J. Pandit, PDE4: new structural insights into the regulatory mechanism and
implications for the design of selective inhibitors, in: S. Liras, A.S. Bell (Eds.),
Phophodiesterases and Their Inhibitors, Wiley-VCH, 2014, pp. 29e40.
[24] P.H. Schafer, A. Parton, A.K. Gandhi, L. Capone, M. Adams, L. Wu, J.B. Bartlett,
M.A. Loveland, A. Gilhar, Y.F. Cheung, G.S. Baillie, M.D. Houslay, H.W. Man,
G.W. Muller, D.I. Stirling, Apremilast, a cAMP phosphodiesterase-4 inhibitor,
demonstrates anti-inflammatory activity in vitro and in a model of psoriasis,
Br. J. Pharmacol. 159 (2010) 842e855.
[44] K. Rutten, E.L. Van Donkelaar, L. Ferrington, A. Blokland, E. Bollen,
H.W. Steinbusch, P.A. Kelly, J.H. Prickaerts, Phosphodiesterase inhibitors
enhance object memory independent of cerebral blood flow and glucose
utilization in rats, Neuropsychopharmacology 34 (2009) 1914e1925.
[45] A. Robichaud, C. Savoie, P.B. Stamatiou, N. Lachance, P. Jolicoeur, R. Rasori,
C.C. Chan, Assessing the emetic potential of PDE4 inhibitors in rats, Brit. J.
Pharmacol. 135 (2002) 113e118.
[46] T. Vanmierlo, P. Creemers, S. Akkerman, M. van Duinen, A. Sambeth, J. De Vry,
T. Uz, A. Blokland, J. Prickaerts, The PDE4 inhibitor roflumilast improves
memory in rodents at non-emetic doses, Behav. Brain Res. 303 (2016) 26e33.
[47] K. Rutten, J. Prickaerts, A. Blokland, Rolipram reverses scopolamine-induced
and time-dependent memory deficits in object recognition by different
mechanisms of action, Neurobiol. Learn. Mem. 85 (2006) 132e138.
[48] M.E. Gurney, A.B. Burgin, O.T. Magnusson, L.J. Stewart, Small molecule allo-
steric modulators of phosphodiesterase 4, Handb. Exp. Pharm. 204 (2011)
167e192.
[49] OpenEye Scientific Software, Santa Fe, New Mexico, USA http://www.
eyesopen.com/www.eyesopen.com.
[50] G.L. Card, B.P. England, Y. Suzuki, D. Fong, B. Powell, B. Lee, C. Luu,
M. Tabrizizad, S. Gillette, P.N. Ibrahim, D.R. Artis, G. Bollag, M.V. Milburn,
S.H. Kim, J. Sclessinger, K.Y. Zhang, Structural basis for the activity of drugs
that inhibit phosphodiesterases, Structure 12 (2004) 2233e2247.
[51] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,
A. Olson, AutoDock4 and AutoDockTools4: automated docking with selective
receptor flexibility, J. Comput. Chem. 16 (2009) 2785e2791.
[25] D.S. Bundschuh, M. Eltze, J. Barsig, L. Wollin, A. Hatzelmann, R. Beume, In vivo
efficacy in airway disease models of roflumilast, a novel orally active PDE4
inhibitor, J. Pharm. Exp. Ther. 297 (2001) 280e290.
[26] J.E. Souness, S. Rao, Proposal for pharmacologically distinct conformers of
PDE4 cyclic AMP phosphodiesterases, Cell. Signal. 9 (1997) 227e236.
[52] D.A. Case, T.A. Darden, T.E. Cheatham III, C.L. Simmerling, J. Wang, R.E. Duke,
R. Luo, R.C. Walker, W. Zhang, K.M. Merz, B. Roberts, S. Hayik, A. Roitberg,
G. Seabra, J. Swails, A.W. Goetz, I. Kolossvary, K.F. Wong, F. Paesani, J. Vanicek,
R.M. Wolf, J. Liu, X. Wu, S.R. Brozell, T. Steinbrecher, H. Gohlke, Q. Cai, X. Ye,
J. Wang, M.J. Hsieh, G. Cui, D.R. Roe, D.H. Mathews, M.G. Seetin, R. Salomon-

102
C. Brullo et al. / European Journal of Medicinal Chemistry 124 (2016) 82e102
Ferrer, C. Sagui, V. Babin, T. Luchko, S. Gusarov, A. Kovalenko, P.A. Kollman,
AMBER 12, University of California, San Francisco, 2012.
[53] A. Jakalian, D.B. Jack, C.I. Bayly, Fast, efficient generation of high-quality
atomic charges. AM1-BCC model: II. Parameterization and validation,
J. Comput. Chem. 23 (2002) 1623e1641.
[54] J. Wang, W. Wang, P.A. Kollman, D.A. Case, Development and testing of a
general amber force field, J. Comput. Chem. 25 (2004) 1157e1174.
[55] R.H. Stote, M. Karplus, Zinc binding in proteins and solution: a simple but
accurate nonbonded representation, Proteins 23 (1995) 12e31.
[56] Z. Li, Y.H. Cai, Y.K. Cheng, X. Lu, Y.X. Shao, X. Li, M. Liu, P. Liu, H.B. Luo,
Identification of novel phosphodiesterase-4D inhibitors prescreened by mo-
lecular dynamics-augmented modeling and validated by bioassay, J. Chem.
Inf. Model. 53 (2013) 972e981.
[57] E. Cichero, P. D'Ursi, M. Moscatelli, O. Bruno, A. Orro, C. Rotolo, L. Milanesi,
P. Fossa, Homology modeling, docking studies and molecular dynamics sim-
ulations using graphical processing unit architecture to probe the type-11
phosphodiesterase catalytic site: a computational approach for the rational
design of selective inhibitors, Chem. Biol. Drug Des. 82 (2013) 718e731.
[58] S. Miyamoto, P.A. Kollman, Settle: an analytical version of the SHAKE and
RATTLE algorithm for rigid water models, J. Comput. Chem. 8 (1992) 952e962.
[59] C. d'Abramo, R. Ricciarelli, M.A. Pronzato, P. Davies, Troglitazone, a peroxi-
some proliferator-activated receptor-gamma agonist, decreases tau phos-
phorylation in CHOtau4R cells, J. Neurochem. 98 (2006) 1068e1077.
[60] S. Akkerman, J. Prickaerts, H.W. Steinbusch, A. Blokland, Object recognition
testing: statistical considerations, Behav. Brain Res. 232 (2012) 317e322.