Design, synthesis, biological evaluation and structural characterization of novel GEBR library PDE4D inhibitors

Article abstract of DOI:10.1016/j.ejmech.2021.113638

Memory and cognitive functions depend on the cerebral levels of cyclic adenosine monophosphate (cAMP), which are regulated by the phosphodiesterase 4 (PDE4) family of enzymes. Selected rolipram-related PDE4 inhibitors, members of the GEBR library, have been shown to increase hippocampal cAMP levels, providing pro-cognitive benefits with a safe pharmacological profile. In a recent SAR investigation involving a subset of GEBR library compounds, we have demonstrated that, depending on length and flexibility, ligands can either adopt a twisted, an extended or a protruding conformation, the latter allowing the ligand to form stabilizing contacts with the regulatory domain of the enzyme. Here, based on those findings, we describe further chemical modifications of the protruding subset of GEBR library inhibitors and their effects on ligand conformation and potency. In particular, we demonstrate that the insertion of a methyl group in the flexible linker region connecting the catechol portion and the basic end of the molecules enhances the ability of the ligand to interact with both the catalytic and the regulatory domains of the enzyme.

Full text of DOI:10.1016/j.ejmech.2021.113638

European Journal of Medicinal Chemistry 223 (2021) 113638
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Design, synthesis, biological evaluation and structural characterization
of novel GEBR library PDE4D inhibitors
Chiara Brullo a, Federica Rapetti a, Sara Abbate b, Tommaso Prosdocimi b,
Archimede Torretta b, Marta Semrau c, Matteo Massa a, Silvana Alfei a, Paola Storici c,
,
d
,
, **
Emilio Parisini b *, Olga Bruno a
a Department of Pharmacy, School of Medical and Pharmaceutical Sciences, University of Genova, Viale Benedetto XV 3, 16132, Genova, Italy
b Center for Nano Science and Technology @ PoliMi, Istituto Italiano di Tecnologia, via Giovanni Pascoli 70/3, 20133, Milano, Italy
c Elettra Sincrotrone Trieste S.C.p.A., SS 14 - km 163,5 in AREA Science Park, 34149, Trieste, Italy
d Latvian Institute of Organic Synthesis, Aizkraukles 21, LV, 1006, Riga, Latvia
a r t i c l e i n f o
a b s t r a c t
Article history:
Memory and cognitive functions depend on the cerebral levels of cyclic adenosine monophosphate
(cAMP), which are regulated by the phosphodiesterase 4 (PDE4) family of enzymes. Selected rolipram-
related PDE4 inhibitors, members of the GEBR library, have been shown to increase hippocampal
cAMP levels, providing pro-cognitive benefits with a safe pharmacological profile. In a recent SAR
investigation involving a subset of GEBR library compounds, we have demonstrated that, depending on
length and flexibility, ligands can either adopt a twisted, an extended or a protruding conformation, the
latter allowing the ligand to form stabilizing contacts with the regulatory domain of the enzyme. Here,
based on those findings, we describe further chemical modifications of the protruding subset of GEBR
library inhibitors and their effects on ligand conformation and potency. In particular, we demonstrate
that the insertion of a methyl group in the flexible linker region connecting the catechol portion and the
basic end of the molecules enhances the ability of the ligand to interact with both the catalytic and the
regulatory domains of the enzyme.
Received 16 March 2021
Received in revised form
7 June 2021
Accepted 8 June 2021
Available online 13 June 2021
Keywords:
PDE4D
GEBR library
PDE4D inhibitors
Crystallographic structure
NMR studies
© 2021 Elsevier Masson SAS. All rights reserved.
1. Introduction
of phosphodiesterase 4 (PDE4) enzymes increases hippocampal
cAMP levels, potentiates the cAMP cascade and triggers the acti-
Alzheimer's disease (AD) is a widespread age-related neurode-
generative disorder that drastically impairs cognitive abilities and
memory functions, thus imposing a heavy physical, emotional and
economic burden on patients and their families. To date, there is no
effective cure for AD. Physiological intracellular levels of the second
messenger cyclic adenosine monophosphate (cAMP) are known to
be crucially involved in sustaining correct memory functions [1].
Such levels are regulated by the phosphodiesterase 4 (PDE4) family
of enzymes, which catalyze the hydrolysis of cAMP to 50-AMP. In
recent years, a number of PDE4 inhibitors have been shown to
effectively improve memory and cognitive functions in AD patients
[2,3]. Indeed, by reducing excessive cAMP hydrolysis, the inhibition
vation of transcription factors that regulate the expression of cen-
tral nervous system (CNS) genes connected to memory and to
learning processes. Of all the different existing PDE4 isoforms,
PDE4B and PDE4D are widely considered as the two best targets for
the treatment of cognition deficit, anxiety, depression and other
neuropsychological conditions. Whereas PDE4B inhibition provides
antidepressant-like effects, the pharmacological profile of PDE4D
inhibitors is more related to cognition improvement [4]. Interest-
ingly, the role of PDE4D in cognition has been further validated in
the context of the genetic disorder acrodysostosis, a rare congenital
malformation syndrome that causes mental retardation, brachy-
dactyly and facial dysplasia. Such disorder originates from PDE4D
missense mutations that are likely to impair the regulation mech-
anism of the enzyme [5e7]. PDE4 systemic inhibition has always
been associated with severe side effects such as emesis and diar-
rhea. Considering the existence of several PDE4 sub-types (A to D),
each of them produced as several splicing variant isoforms of
* Corresponding author. Center for Nano Science and Technology @ PoliMi, Isti-
tuto Italiano di Tecnologia, Via G. Pascoli 70/3, 20133, Milano, Italy.
** Corresponding author.
E-mail addresses: emilio.parisini@osi.lv (E. Parisini), obruno@unige.it (O. Bruno).
https://doi.org/10.1016/j.ejmech.2021.113638
0223-5234/© 2021 Elsevier Masson SAS. All rights reserved.

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
different length, the development of isoform specific drugs is
generally regarded as a viable option for the design of safer in-
hibitors [8]. In their seminal work, Burgin and coworkers demon-
strated that it is possible to exploit the stabilization of the UCR2
regulatory domain over the entrance of the catalytic pocket to
selectively target the PDE4D/B long isoforms [5,9,10].
The GEBR library is an established and well-known panel of
catechol-based PDE4 inhibitors. The members of this library feature
three general moieties (Fig. 1): 1) the catechol portion. Since the
development of rolipram, this scaffold has been one of the most
investigated in PDE4 drug development studies. In all the GEBR
library compounds, one of the hydroxy catechol moiety is
substituted with a cyclopentyl group, which interacts with the Q2
pocket of the enzyme's catalytic domain; the other, the Q1-directed
catechol substituent, is either a methoxy or a difluoromethoxy
group, the latter having been introduced with the expansion of the
library in 2015 [11]. 2) the linker region. This is the most variable
portion of the entire library. Indeed, we extensively modified the
chemical nature of the linker in order to explore how different
functionalities may result in the possible exploration of the M-
pocket. Six families of linkers have been developed within the GEBR
library, each comprising up to 10 members featuring alternative
chemical modifications [12,13]. 3) the basic end. All of the more
active compounds feature a morpholine, a dimethylmorpholine or
4-hydroxypiperidine as the terminal portion.
Several members of the GEBR library (particularly GEBR-7b [12],
GEBR-54 [13] and GEBR-32a [14], Fig. 2) showed interesting
PDE4D3 selective inhibition, and in vitro and in vivo assays evi-
denced their great potential as therapeutic agents in pathologies
characterized by a dysregulation of memory and cognitive perfor-
mance, such as Alzheimer's disease. Moreover, unlike other PDE4
inhibitors, these compounds showed only limited side-effects such
as sedation or emesis when tested in animal models [13e15].
To contribute to the elucidation of the molecular bases of se-
lective inhibition, we recently provided a structural and functional
characterization of several members of the GEBR family, leading to
the identification of three major ligand conformational classes,
which we classified as protruding, twisted, and extended [16].
Interestingly, our X-ray crystallography study highlighted the
ability of the protruding compounds to explore the S-pocket,
pointing towards the external portion of the catalytic cavity, in the
proximity of the area that in long isoforms is occupied by the UCR2
helix. This structural property is quite innovative for PDE4 in-
hibitors, as only a few other recently developed compounds have
been shown to be able to exploit this space [5]. Also, on the basis of
molecular dynamic simulations we postulated hydrophobic in-
teractions between the tail portion of the ligand and the capping
portion of the enzyme (Phe196). In addition, pharmacological an-
alyses showed that the shorter compounds (such as GEBR-7b and
its fluorinated analogue GEBR-20b, Fig. 2) do not inhibit
Fig. 2. Structures of memory enhancers GEBR-7b, GEBR-20b, GEBR-54 and GEBR-32a.
differentially the PDE4 catalytic domain and the full-length
PDE4D3, while with the lengthening and the enhanced flexibility
of the tail (GEBR-4a and GEBR-11b, Fig. 3) [16] a differential
behavior starts to appear. More recently, enantiomeric separation
was performed on lead GEBR-32a and IC50 values against both the
PDE4D catalytic domain and the long PDE4D3 isoform were
determined for both enantiomers [17]. The (S)-enantiomer was
more active than the (R)-configured one when measured against
the full-length protein, whereas this fold-difference was reduced
when measured against the catalytic domain. These data clearly
showed that the inhibitory activity of GEBR-32a is significantly
influenced by the configuration of the stereogenic center and
confirmed our hypothesis that the tail portion of GEBR-32a con-
tributes significantly to the inhibition of the full-length PDE4D
enzyme.
Altogether, these results suggested that longer and more flexible
chains might efficiently adapt to the chemical environment of the
active site, in particular when a regulatory domain of the enzyme is
capping the catalytic pocket. Based on these data, the subset of
protruding ligands appeared to be the most promising candidate
for further library development.
To enhance the ability of the ligand to interact with the regu-
latory capping portion of the enzyme, we designed and synthesized
a new series of compounds featuring the catechol portion, an un-
modified basic end, and a flexible linker region bearing an addi-
tional methyl group (compounds type A, Fig. 4). Based on our
previous SAR study on GEBR library inhibitors, we hypothesized
that a ligand chemical derivatization strategy could be adopted to
provide stabilizing hydrophobic interactions with the surrounding
residues, either downwards towards the bottom of the catalytic
pocket or upwards towards the capping helix. To this end, the
introduction of a methyl group on the linker appeared to be
appropriate and potentially effective given the availability of free
space and the nonpolar nature of the surrounding region. Indeed, it
is well known that methyl group can exert electronic and steric
effects on biologically active molecules leading to increase potency
and selectivity towards the target, or to enhance resistance to
metabolic transformation, as the numerous examples in the history
of medicinal chemistry demonstrate [18].
In addition, in order to confirm that a certain flexibility in the
chain is required to better interact within the catalytic pocket as
well as with the UCR2 portion, we designed and synthesized a
second group of compounds (type B, Fig. 4). In these constrained
derivatives, the methyl group is entered in a five membered cycle
condensed with the phenyl of the catechol moiety.
R
O
N
O
O
N
R = CH3 GEBR-4a
R = CHF2 GEBR-11b
H
OH
O
Fig. 1. General structure of GEBR library compounds.
Fig. 3. The more flexible GEBR-4a and its fluorinated analogue GEBR-11b
2

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
H3CO
CH2
O
R
H3CO
O
H3CO
O
H3CO
O
O
N
OH
N
CH3
N
O
N
CH3
CH3
O
R
R
Z-isomers compounds B
Z-isomers compounds A
Z-oxime
H3CO
O
H3CO
O
H3CO
H3CO
CH2
O
N
N
R
CH3
OH
O
O
R
N
CH3
R
O
CH3
N
O
E-isomers compounds B
E-oxime
E-isomers compounds A
Fig. 4. Design of new compounds type A and B.
Since the key intermediates for the synthesis of compounds A
and B are ketoximes that could be obtained in both conformation E/
Z (Fig. 4), we firstly hypothesized to obtain for each derivatives a
mixture of two isomers, as previously reported for similar com-
pounds [12]. In constrained compounds type B, by blocking the
methyl in the condensed cycle with the phenyl we would force the
chain to assume two different arrangements in space, one in which
the terminal amine is closer to the catechol ring (Z isomers) and
another in which it is further away (E isomers).
However, as shown in Table 1 and discussed in the following
chemistry section, type A compounds actually afforded only Z
isomers while type B compounds afforded only E isomers.
We evaluated the percentage inhibition of the PDE4D catalytic
observed at lower
d
values results from the Z isomer. In this study,
unlike other previously synthesized oxime derivatives, which were
obtained as E/Z mixtures [12], 9 was obtained as a single com-
pound; hence, comparative analyses between the chemical shift of
the carbon atoms of the oxime group was not possible. On the other
hand, since the difference between the d values of the C]N signals
of the two isomers is generally minimal, the configuration of the
single compound could not be assigned with certainty on the basis
of the chemical shift of its C]N group only. The configuration of the
C]N bond was therefore assigned on the basis of the methyl
chemical shift (2.27 ppm) in the 1H NMR spectrum, according to
what previously reported for the E (2.19 ppm) and Z (2.26 ppm)
isomers of acetophenone oxime derivatives [25]. In addition, in the
domain for all the synthesized compounds at 100
mM concentra-
13C NMR analyses, the CH3 signal was observed at
d values corre-
tion. Based on these end-point assays, we then obtained dose
dependent curves and measured the IC50 values against both the
PDE4D catalytic domain alone and the PDE4D3 isoform only for
those compounds showing >50% inhibition. Finally, we carried out
the X-ray crystallographic characterization of the complexes be-
tween the PDE4D catalytic domain and the most active inhibitors,
two of which (compounds 1b and 2a) will be discussed here in
detail.
sponding to those reported for analogues Z isomers [24], thus
confirming our hypothesis.
The synthesis of compounds 1aec and 2a,b involved the treat-
ment of the oxime
9 with the suitable chloroacetyl- or
chlropropionyl-amides [13], in anhydrous DMF, in the presence of
anhydrous K2CO3 (for 1a,b), sodium ethoxide (for 1c) or NaH (60%
dispersion in mineral oil) (for 2a,b) (Scheme 1). The oxime 9 was
then treated with epichlorohydrin in sodium ethoxide to obtain
(1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanone-O-
(oxiran-2ylmethyl)-oxime 10, which in turn was reacted with
excess of morpholine or 2,6-dimethylmorpholine (for 3a,b) or with
piperidin-1-ol in anhydrous DMF (for 3c) (Scheme 1).
2. Results and discussion
2.1. Chemistry
To obtain compounds 4 and 5, trans-ferulic acid 11 was firstly
reduced to 3-(4-hydroxy-3-methoxyphenyl) propanoic acid 12 [26]
(Scheme 2) using H-cube hydrogenator. The treatment of 12 with
methansulfonic acid yielded the indanone 13 [27], which was then
alkylated with cyclopentyl bromide in the presence of K2CO3 to give
6-(cyclopentyloxy)-5-methoxindan-1-one 14. The subsequent re-
action of the ketone group with hydroxylamine hydrochloride in
ethanol led to the corresponding oxime 15. The 1H NMR and 13C
spectra analysis enabled us to assign the E configuration to the C]
N bond of compound 15. In this regard, the chemical shift of the
signal belonging to the carbon atom of the C]N group in the 13C
NMR spectrum (164 ppm, see Supporting Material) was discrimi-
nant. According to what reported for similar oximes [28], E isomers
The new methyl derivatives compounds 1, 2 and 3 were syn-
thetized starting from 3-cyclopentyloxy-4-methoxybenzaldehyde
6 (Scheme 1) [19], which was reacted with the suitable Grignard
reagent to obtain the 1-[3-(cyclopentyloxy)-4-methoxyphenyl]
ethanol 7. The subsequent oxidation to methylketone 8 was per-
formed with pyridinchlorochromate (PCC) in cyclohexane; despite
compounds 7 and 8 are already reported in the literature [20,21],
we have optimized both the methods and reaction yields. Then, the
intermediate ketone 8 was transformed in the corresponding
oxime derivative 9; this compound also was already reported in
two patents, but the configuration of the C]N bond was never
assigned [22,23]. Therefore, we performed an in-depth analysis of
the 1H NMR and 13C NMR spectra and assigned the Z configuration
to the C]N bond (see Supporting Material).
provide signals at
lower than 160.
d values greater than 160, while Z ones at values
Derivatives 4aec were obtained by reacting oxime 15 with the
proper chloroacetyl amides [13] in the presence of anhydrous
K2CO3 in anhydrous DMF. Moreover, to obtain derivatives 5aec,
oxime 15 was converted in the corresponding oxirane 16 by reac-
tion with epichlorohydrin in absolute ethanol in the presence of
sodium ethoxide. Then, the latter was reacted with excess
Generally, a mixture of two Z/E isomers is obtained. Therefore, in
the 13C NMR spectra two signals very close together in the range
150e160 ppm and belonging to the oxime carbon atom (C]N) can
be observed and can be used to assign the exact configurations. In
particular, according to literature [24], the signal with higher
chemical shift belongs to the oxime carbon of the E form, while that
3

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
Table 1
100
m
M concentration, compounds 1c, 3c, 4a-c and 5a-c did not
Structures of the new series compounds 1aec, 2aeb, 3aec, 4aec and 5aec.
show significant inhibitory activity (inhibition <30% in all cases),
hence we classified them as inactive.
For compounds 1a-b, 2a-b and 3a-b, the inhibitory activity was
measured against both the catalytic domain only and against the
PDE4D3 isoform. Results expressed as IC50 (mM) values are reported
in Table 2, while the full set of the corresponding dose-response
curves are shown in Fig. 5. In all cases, a significant increase in
IC50 can be observed in going from the catalytic domain to the
PDE4D3 isoform, suggesting that contacts with the inhibitory
domain capping the entrance of the catalytic pocket are formed.
Interestingly, within the whole series, compounds 1a and 1b show
the lowest IC50 values, while, at the same time, also featuring a
sizeable (5e15-fold) decrease in IC50 in going from the catalytic
domain to the full length enzyme. Moreover, it is worth noting that
compound 1b is the methylated equivalent of GEBR-7b, whose IC50
Comp.
Type
A
R
1a
we measured previously (16 2 and 11
2 mM against the catalytic
1b
1c
2a
2b
A
A
A
A
domain and the full length enzyme, respectively) [16]. The decrease
in IC50 suggests that the introduction of a methyl group in the linker
region has improved significantly the overall potency of the com-
pound, confirming our hypothesis formulated in the series design.
Indeed, the methylation increased lipophilicity enhancing the
interaction of the linker with the catalytic domain, as confirmed by
X-ray crystallography. The presence of a morpholine instead of a
dimethylmorpholine in the tail end does not affect significantly the
potency of the ligand (see compound 1a).
An outlier in the list is represented by compound 2a, whose IC50
far exceeds those of the remaining compounds in the series.
Compounds series 2, which relative to 1 features a one extra carbon
atom in the chain connecting the morpholine ring and the catechol
moiety, are predicted to be more flexible; hence, they are in prin-
ciple capable of adopting a wider spectrum of conformations
within the catalytic pocket. Indeed, this is confirmed by X-ray
crystallography analysis of 2a, which provides evidence of two
alternate conformations in the catalytic pocket, one of them
featuring an unfavorable twisted arrangement, as we discuss more
carefully in section 2.3. Compound 2b, on the contrary, showed
potency in the same range observed for compounds 1 and 3. We can
suppose that, owing to steric hindrance, the dimethylmorpholine
ring, in this case, prevents the twisted arrangement and therefore
2b should assume only the most efficient protruding conformation.
As concerns the inactive derivatives, both compounds 1c and 3c
are characterized by the presence of 4-hydroxypiperidine moiety in
the terminal end of the tail. To justify their poor ability to inhibit
both forms of enzyme (PDE4D cat and full PDE4D3), we can hy-
pothesize that the polarity of the hydroxy group is detrimental for a
good interaction with the hydrophobic part of the catalytic domain
as well as with the capping domain at the entrance of the catalytic
pocket. The same considerations can be extended to compounds 3a
and 3b, which, bearing the hydroxy substituent in the chain,
showed a lower potency in comparison with linear compounds 1a
and 1b.
3a
3b
A
A
3c
4a
A
B
4b
4c
B
B
5a
5b
B
B
Finally, all the more rigid compounds 4 and 5 resulted to be
inactive, thus confirming that the tail flexibility is an essential
feature for GEBR-library molecules to enter the catalytic pocket,
settle on its hydrophobic floor and reach out towards the regulatory
domain and interact with it.
5c
B
morpholine or 2,6-dimethylmorpholine to give 5a,b or with
piperidin-1-ol in anhydrous DMF to obtain 2c (Scheme 2).
2.3. Crystallographic studies
The PDE active site, which is highly conserved across the
different isoforms of the enzyme, features three distinct portions: a
metal binding pocket (M), hosting two octahedrally-coordinated
metal ions (Zn and Mg), the so-called Q switch and P clamp
pocket (Q), where the residues Ile502 and Phe538 provide a tight
2.2. Enzymatic assays and SAR considerations
Based on preliminary end-point assays carried out on the PDE4D
catalytic domain with each of the newly synthesized species at
4

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
Scheme 1. Synthesis of derivatives 1aec, 2a,b, 3aec. Reagents and conditions: (i) Mg/I2, an. Et2O, CH3I, 30e40 ꢀC, 3h, 100%. (ii) Pyridinium chlorochromate, cyclohexane, 70 ꢀC, 3 h,
94%. (iii) NH2OH.HCl, NaHCO3, EtOH, 60 ꢀC, 4 h, 63%. (iv) K2CO3, proper chloroacetylamide, an. DMF, 60 ꢀC, 18 h, (1a,b) or EtONa, 1-(chloroacetyl)piperidin-4-ol, an. DMF, 50e60 ꢀC,
24 h (1c), 41e87%. (v) NaH, an. DMF, 0 ꢀC, 2 h, proper chloropropionylamide, rt, 18 h, 24e56%. (vi) epichlorohydrin, EtONa, DMF, 50e60 ꢀC, 12 h, 53%. (vii) morpholine or 2,6-
dimethylmorpholine, 50e60 ꢀC, 18 h (3a,b) or piperidin-1-ol, an. DMF, 60 ꢀC, 18 h (3c), 34e96%.
Scheme 2. Synthesis of derivatives 4aec and 5aec. Reagents and conditions: (i) full-H2 mode, CH3OH, Pd/C 5%, flow 1 mL/min rt, quantitative yield. (ii) CH3SO3H, N2, 90 ꢀC, 10 min,
56%. (iii) bromocyclopentane, K2CO3, KI, an. DMF, 65 ꢀC, 22 h, 73%. (iv) NH2OH.HCl, NaHCO3, EtOH, 60 ꢀC, 2 h, 59%. (v) K2CO3, proper chloroacetylamide, an. DMF, 50e60 ꢀC, 18 h,
33e100%. (vi) epichlorohydrin, an. DMF, 40e50 ꢀC, 18 h, 50%. (vii) morpholine or 2,6-dimethylmorpholine, 50e60 ꢀC, 18h (5a,b) or piperdin-1-ol, an. DMF, 50e60 ꢀC, 18 h (5c),
41e46%.
cAMP-stabilizing environment, and a solvent-filled pocket (S). All
catechol-based inhibitors known to date are invariably stabilized in
the active site by the same crucial interaction of the P-clamp (Ile502
and Phe538) with the aromatic ring of the catechol shown by the
rolipram complex (PDB code: 1OYN) and by further, mostly water-
mediated, interactions within the M, the Q and the S portions of the
catalytic pocket. In a SAR investigation involving different members
of the GEBR library of rolipram-related PDE4 inhibitors, we have
previously demonstrated that in a selected pool of inhibitors, a
large conformational variability can be observed at the level of the
tail, which, depending on length and flexibility, can take a
protruding, a twisted or an extended conformation [16].
Based on our previous study, we identified the protruding
compounds as the most promising candidates for the development
of the next generation of GEBR compounds. Indeed, they feature a
tail that develops within the S pocket of the catalytic site, a region
where the ligand can interact with the UCR2 of the enzyme, thus
achieving further stabilization and an enhanced specificity for long
PDE4D isoforms.
Within the GEBR library, the subset of ligands featuring a linear
scaffold were found to adopt a protruding conformation and were
therefore selected for subsequent optimization.
5

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
Table 2
proximity. Indeed, in the case of 1b (GEBR-41b), we have shown
that the methyl group interacts with hydrophobic residues in the Q
region and provides further stabilization energy to the ligand in its
protruding conformation inside the catalytic pocket. However, in
the crystal structure, the one-carbon longer tail of 2a (GEBR-42a)
displays two alternative conformations in the two independent
catalytic domain-ligand complexes in the dimer, a protruding and a
twisted conformation, the former resembling GEBR-41b while the
latter featuring van der Waals interactions between the morpholine
moiety and the methyl group in the linker region. Hence, owing to
its enhanced flexibility, 2a can feature a twisted conformation that
reduces the interactions with the capping portion of the regulatory
domain, thus substantially reducing the overall potency of the
ligand.
Inhibitory activity on PDE4 catalytic domain and full-length PDE4D3 enzyme of
compounds 1aec, 2a, 2b, 3aec.
Compounds
IC50
m
Ma PDE4D Cat
IC50 mMa PDE4D3
1a
1b
1c
2a
2b
3a
3b
3c
2.8 0.1
2.5 0.2
Inactive
33.1 4.6
4.6 0.5
6.9 0.7
4.1 0.2
Inactive
0.16 0.01
0.47 0.02
11.2 0.6
3.6 0.3
0.72 0.05
1.1 0.8
a
IC50 values were obtained by plotting dose-response curves against the catalytic
domain only (PDE4D Cat) and against the PDE4D3 isoform. The reported data are the
mean values of three replicates SD (standard deviation).
As previously reported, the reaction between 6-(cyclo-
pentyloxy)-5-methoxindan-1-one and hydroxylamine gave only E-
oxime derivative, and therefore only E-isomers of series 4 and 5.
Thus, it was not possible to verify the entire hypothesis formulated
in the molecules design of constrained derivatives. On the other
hand, by reducing the possibility of movement of the amino ter-
minal part with respect to the catechol portion, the molecule's
entry into the catalytic site is completely prevented and molecules
(4 and 5) were inactive. Our results confirmed that a certain degree
of chain flexibility is necessary in order to facilitate the interaction
of the ligands with the catalytic site.
Despite our chemical modifications, a clear understanding of the
optimal length and flexibility of the tail portion of the ligand would
come only from the successful determination of the crystal struc-
ture of a PDE4D-ligand complex that includes also the portion of
the regulatory domain that caps the entrance of the catalytic pocket
of the enzyme. Such structure would provide the necessary evi-
dence of the interactions that are needed to stabilize the protein
ligand complex and, possibly, to favor selectivity over different
PDE4 isoforms. Further studies aimed at obtaining new inhibitors
that can interact more efficiently both with the catalytic and with
the regulatory domains of long PDE4D isoforms are ongoing.
In particular, in the new ligands reported here we enhanced the
hydrophobicity by adding a methyl group on the first carbon atom
after the aromatic ring of the catechol. In linear compounds, the
hydrogen of the first sp2 carbon atom of the linker points towards
the highly solvated bottom of the catalytic pocket. Given the length
and the flexibility of the tail, we hypothesized that the introduction
of a hydrophobic methyl group in the chain would allow a rotation
of the sp2 carbon to favor an upward orientation of the methyl and
provide further stabilization energy by interacting with Phe196.
This would also allow the rest of the flexible tail to maintain a
protruding conformation and force the following N and O atoms to
lay over the solvation sphere of the metal ions. However, the crystal
structure of the complex between compound 1b (named as GEBR-
41b in the GEBR library) and the PDE4D catalytic domain reveals
that the methyl group actually points towards the bottom of the
catalytic pocket, forming stabilizing van der Waals interactions
with residues Tyr159, Met273, Leu319, Ile336 and Phe340 near the
metal binding pocket (Fig. 6A and B).
Unlike with compound 1b, where the ligand is completely
protruding, the crystal structure of the complex with compound 2b
(GEBR-42a) shows that in one of the two independent crystallo-
graphic monomers the ligand assumes a protruding conformation,
while in the other a twisted conformation. In both monomers, the
methyl group of 2b point toward the region of the Q portion con-
taining Met273, Leu319 and Phe372, where stabilizing van der
Waals interactions take place. However, relative to compound 1b,
the methyl group moves away from the underlying hydrophilic
metal binding region, as shown by Fig. 7. Interestingly, while the
protruding conformation of 2b is very similar to that of 1b, its
twisted conformation features a van der Waals contact between the
morpholine moiety and the methyl group in the linker. Here, the
methyl group is rotated by approximately 90ꢀ compared to the
methyl group in compound 1b and lies almost parallel to the un-
derlying solvated region of the metal binding site.
4. Materials and Methods
4.1. Chemistry
All chemicals were obtained from Sigma-Aldrich s.r.l. (Milan,
Italy). All compounds were tested for purity by TLC (Kieselgel
60F254 DC-Alufolien, E. Merck, Darmstadt, Germany). Product
purification, when necessary, was performed by using Flash Isolera
One Biotage instrument, Silicagel column SNAP ULTRA-HP Spher-
eTM 25 mm.
Melting points are not corrected and were measured with a
Buchi 540 instrument.
IR spectra were recorded with a Perkin-Elmer 390 spectropho-
tometer or a Spectrum Two FT-IR Spectrometer (PerkinElmer, Inc.,
Waltham, MA, USA).
Data collection and statistical analyses are reported in Table 3.
3. Conclusions
1H NMR spectra were recorded on a Varian Gemini 200
(200 MHz) instrument or BRUKER DPX-300 (300 MHz); chemical
As we demonstrated in our previous SAR study, rolipram-related
GEBR-library compounds featuring a flexible tail that protrudes
from the catalytic pocket are capable of interacting with the UCR2
regulatory portion of the enzyme in long PDE4D isoforms. Hence,
they are promising candidates for further drug development
studies aiming at reducing the side effects that are usually associ-
ated with PDE4 inhibition. Here, starting from a GEBR-library
subset of protruding PDE4D inhibitors, we chemically modified
the linker region of the compounds via the introduction of a methyl
group, with the aim to enhance the interaction of the linker portion
of the compounds with the hydrophobic residues that are in its
shifts were referred to as
(TMS) as internal standard; signals were characterized as
d (ppm) relative to tetramethylsilane
s
(singlet), d (doublet), t (triplet), q (quartet), sept (septet), m
(multiplet), or br s (broad signal); J were reported in Hertz. 13C NMR
spectra were recorded on a BRUKER DPX-300 (300 MHz) or a JEOL
JNM ECZ-400S/L1 (400 MHz).
Elemental analyses were within 0.4% of the theoretical values
and were determined with an EA 1110 elemental analyzer (Fison-
Instruments, Milan, Italy). Table of elemental analyses data and
images of all IR, 1H NMR an 13C NMR are available in online Sup-
plementary materials.
6

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
Fig. 5. IC50 curves for the most active compounds (1a, 1b, 2a, 2b, 3a, 3b) relative to the catalytic domain only (A) and against the long PDE4D3 isoform (B). The experimental
conditions are reported in the Materials and Methods section. The reported data are the mean values of three replicates SD (standard deviation).
4.1.1. Synthesis of 1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanol
7
temperature was maintained at 30 ꢀC until magnesium disappears.
The reaction mixture was cooled with an ice bath and the 3-
To a mixture of magnesium turniture (0.67 g, 28 mmol) and I2
(trace) in an. diethyl ether (10 mL) CH3I (3.8 g, 1.67 mL, 28 mmol)
solved in an. diethyl ether (10 mL) was slowly added and the
(cyclopentyloxy)-4-methoxybenzaldehyde 6 (5.0 g, 22 mmol)
solved in an. diethyl ether (10 mL) was slowly added and the
mixture was stirred at 30e40 ꢀC for 3 hs. Then, ice water and
7

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
Fig. 6. A) Crystal structure of the PDE4D catalytic domain in complex with 1b (GEBR-41b) (PDB code: 7AY6), featuring the ligand in a protruding conformation. The side chains of
hydrophobic amino acids that make contacts with the methyl group in the linker region are shown as sticks. B) Structure superposition of the GEBR-41b complex (ligand shown in
purple) with the PDE4D catalytic domain in complex with GEBR-7b (ligand shown in orange) (PDB code: 6F6U). The two ligands, which differ only for the methyl group in the linker
region, are almost completely superimposable (rmsd 0.649 Å). Here, the side chains of the amino acids contacting the methyl group of GEBR-41b are omitted for clarity. The 2Fo-Fc
electron density map of the ligand is shown in the Supp. Inf. (Fig. S1). Figures were created using Pymol [29].
(15 mL), dried (Na2SO4) and concentrated under reduced pressure
to obtain a yellow oil which crystallized as white solid from diethyl
ether. Mp: 127e129 ꢀC Yield: 94% (lett 95%) [21].
4.1.3. Synthesis of (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]
ethanone oxime 9
To a solution of ketone 8 (1.1 g, 4.69 mmol) in 96% EtOH (10 mL)
hydroxylamine hydrochloride (0.7 g, 10 mmol) solved in water
(10 mL), NaHCO3 (0.84 g, 10 mmol) and water (10 mL) were added
and the mixture was stirred at 60 ꢀC for 4 hs. After cooling to room
temperature, the reaction mixture was poured into ice water
(50 mL) to obtain a solid which was filtered, washed with water and
recrystallized from 95% EtOH to obtain a yellowish solid. Yield: 63%.
Mp: 95e97 ꢀC [22]. IR (KBr) cmꢂ1: 3254 (OH). 1H NMR (300 MHz,
CDCl3): d 1.50e2.10 (m, 8H, cyclopent.), 2.27 (s, 3H, CH3), 3.90 (s, 3H,
Fig. 7. Structure superposition of the two monomers of the PDE4D catalytic domain in
complex with 2a (GEBR-42a) (PDB code: 7B9H) and with 1b (GEBR-41b) (PDB code:
7AY6). The ligands in the 2a complex structure bind in two different conformations:
while in one monomer it adopts a protruding conformation (shown in yellow), in the
other (shown in purple) it is twisted toward the bottom of the catalytic pocket and
features van der Waals contacts between the morpholine moiety and the methyl group
in the linker region. As a comparison, the protein structure of the 1b complex is
depicted in light violet and the 1b compound in orange. The 2Fo-Fc electron density
map of the ligands are shown in the Supp. Inf. (Fig. S1).
OCH3), 4.78e4.95 (m, 1H, OCH cyclopent.), 6.15 (brs, 1H, OH dis-
appears with D2O), 6.88 (d, J ¼ 8.0 Hz, 1H, H-5 Ar), 7.14 (dd,
J ¼ 8.0 Hz, 2.0 Hz, 1H, H-6 Ar), 7.28 (d, J ¼ 2.0 Hz, 1H, H-2 Ar). 13
C
NMR (101 MHz, CDCl3)
d 155.55, 147.03, 129.07, 119.00, 112.03,
111.07, 80.05, 77.23, 76.88, 76.06, 56.00, 32.88, 24.22, 12.03. Anal.
Calcd. for C14H19NO3.
4.1.4. General procedure for the synthesis of derivatives 1a and 1b
Into a solution of 9 (0.249 g, 1 mmol) in anhydrous DMF
(2 mL) K2CO3 (0.414 g, 3 mmol) was suspended; then, 4-(chlor-
oacetyl)morpholine or 4-(chloroacetyl)2,6-dimethylmorpholine
(3 mmol) solved in anhydrous DMF (2 mL) was added dropwise
and the mixture was heated at 50e60 ꢀC for 18 hs under nitrogen
atmosphere. After cooling to room temperature, the mixture is
poured into water (100 mL) and extracted with AcOEt (2 ꢁ 20 mL);
the organic phases are washed with brine (3 ꢁ 20 mL), water
(3 ꢁ 20 mL), dried (MgSO4) and concentrated under reduced
pressure to obtain yellowish oils which were purified by flash
chromatography, using diethyl ether as eluent.
saturated NH4Cl solution (15 mL) was added and the aqueous
phases were extracted with diethyl ether (2 ꢁ 20 mL). The com-
bined organic phases were washed with brine (50 mL), dried
(Na2SO4) and concentrated under reduced pressure to obtain a
yellow oil, which crystallizes as white solid from diethyl ether.
Yield: 100%.
4.1.2. Synthesis of 1-[3-(cyclopentyloxy)-4-methoxyphenyl]
ethanone 8
To a suspension of pyridinium chlorochromate (3.2 g, 15 mmol)
in cycloexane (5 mL) a solution of 7 (2.36 g, 10 mmol) solved in
cycloexane (10 mL) is added and the mixture was heated at 70 ꢀC
for 3 hs. After cooling to room temperature, diethyl ether was
repeatedly added (5 ꢁ 10 mL), the gummy black residue was
washed, and the organic yellow solution decanted. The combined
organic phases were then washed with 1 M NaOH (15 mL), brine
4.1.4.1. (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanone O-(2-
morpholin-4-yl-2oxoethyl) oxime 1a. Yellow oil. Yield: 57%. IR
(CHCl3): cmꢂ11673 (CO). 1H NMR (CDCl3):
d 1.56e2.03 (m, 8H, 4CH2
cyclopent.), 2.29 (s, 3H, CH3), 3.38e3.80 (m, 8H, 4CH2 morph.), 3.89
(s, 3H, OCH3), 4.63e4.81 (m, 1H, OCH cyclopent.), 4.87 (s, 2H,
8

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
Table 3
Data collection and refinement statisticsa.
GEBR-41-b
GEBR-42-a
Wavelength (Å)
Resolution range (Å)
Space group
Unit cell (a, b, c) (Å)
Total reflections
Unique reflections
Multiplicity
1.00
1.00
64.84e1.66 (1.69e1.66)
P 21 21 21
64.84, 98.66, 120.24
174845 (15148)
89739 (8130)
5.70 (5.50)
98.50 (90.10)
9.10 (1.20)
17.33
0.108 (1.31)
0.996 (0.600)
89592 (8076)
4517 (432)
0.1828 (0.2653)
0.2153 (0.3038)
6264
54.13e1.50 (1.53e1.50)
P 21 21 21
64.75, 98.63, 119.82
245551 (23667)
122897 (11934)
12.90 (10.10)
99.80 (97.70)
11.50 (1.40)
18.44
0.115 (2.023)
0.999 (0.631)
122638 (11909)
6199 (632)
0.1938 (0.3668)
0.2201 (0.3806)
6331
Completeness (%)
Mean I/s(I)
Wilson B-factor (Å2)
R-merge
CC1/2
Reflections used in refinement
Reflections used for R-free
R-work
R-free
Number of non-hydrogen atoms
Protein
5411
5460
Ligands
124
182
Solvent
729
689
Protein residues
RMS(bonds) (Å)
RMS(angles) (ꢀ)
Ramachandran favored (%)
Ramachandran allowed (%)
Ramachandran outliers (%)
Rotamer outliers (%)
Clashscore
659
0.014
1.44
98.02
1.83
0.15
0.32
4.91
659
0.009
1.11
97.86
1.68
0.46
0.80
5.51
Average B-factor (Å2)
Protein
25.69
24.14
27.94
26.52
Ligands
37.65
37.57
Solvent
35.19
36.61
a
Statistics for the highest-resolution shell are shown in parentheses.
CH2CO), 6.87 (d, J ¼ 8.0 Hz, 1H, H-5 Ar) 7.17 (dd, J ¼ 8.0, 2.0 Hz, 1H,
H-6 Ar) 7.27 (d, J ¼ 2.0 Hz, 1H, H-2 Ar). 13C (101 MHz CDCl3):
was purified by flash chromatography, using as the eluent firstly
dietyl ether, then a mixture of diethyl ether/methanol (8:2) Yellow
oil. Yield: 42%. IR (CHCl3) cmꢂ1: 2949 (OH), 1644 (CO). 1H NMR
d
167.60, 156.21, 151.52, 147.78, 128.55, 119.42, 113.47, 112.69, 111.24,
80.67, 72.71, 66.98, 66.59, 56.09, 45.79, 42.32, 32.83, 24.14, 12.95.
Anal. (C20H28N2O5) C, H, N.
(CDCl3):
d
1.32e1.94 (m, 12H, 4CH2 cyclopent. þ 2CH2pip.), 2.20 (s,
3H, CH3), 3.04e3.23 (m, 2H, CH2N pip.), 3.40e3.55 (m, 2H, CH2N
pip.), 3.78 (s, 3H, OCH3), 3.82e4.01 (m, 1H, CHOH pip.), 4.70e4.84
(m, 3H, OCH cyclopent. þ CH2CO), 6.77 (d, J ¼ 8.6 Hz, 1H, H-5 Ar)
7.06 (dd, J ¼ 8.6, 2.0 Hz, 1H, H-6 Ar) 7.18 (d, J ¼ 2.0 Hz, 1H, H-2 Ar).
4.1.4.2. (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanone O-[2-
(2,6-dimethylmorpholin-4-yl)-2oxoethyl] oxime 1b. Yellow oil.
13C (CDCl3):
d 168.19, 155.35, 151.53, 147.23, 128.89, 121.39, 113.24,
Yield: 87%. IR (CHCl3): cmꢂ1 1671 (CO). 1H NMR (CDCl3):
d 1.13e1.37
111.52, 81.10, 70.61, 67.25, 56.09, 43.24, 34.03, 32.81, 23.45, 16.76.
Anal. (C21H30N2O5) C, H, N.
(m, 6H, 2CH3), 1.55e2.02 (m, 8H, 4CH2 cyclopent.), 2.27 (s, 3H, CH3),
2.33e2.48 and 2.76e2.87 (2 m, 2H, 2CHO morph.), 3.40e3.72 (m,
4H, 2CH2N morph.), 3.88 (s, 3H, OCH3), 4.32e4.64 (m, 1H, OCH
cyclopent.), 4.86 (s, 2H, CH2CO), 6.85 (d, J ¼ 8.0 Hz, 1H, H-5 Ar) 7.18
4.1.6. General procedure for the synthesis of derivatives 2a and 2b
To a solution of oxime 9 (0.373 g, 1.5 mmol) in an. DMF (2 mL),
NaH (in a 50% mineral dispersion, 0.15 g, 3 mmol) was added at 0 ꢀC
and the mixture was stirred at room temperature for 2 hs. Then, 4-
(3-chloropropanoyl)morpholine or 4-(3-chloropropanoyl)-2,6-
dimethylmorpholine (3 mmol) solved in an. DMF (2 mL) was
slowly added at 0 ꢀC and the reaction mixture was stirred at room
temperature for 18 h under nitrogen atmosphere. The mixture was
poured into water (50 mL) and extracted with diethyl ether
(2 ꢁ 20 mL); the organic phases were washed with brine
(3 ꢁ 20 mL), water (3 ꢁ 20 mL), dried (MgSO4) and evaporated
under reduced pressure to obtain a yellow oil which was purified by
flash chromatography, using diethyl ether as eluent.
(dd, J ¼ 8.0, 2.0 Hz, 1H, H-6 Ar) 7.29 (d, J ¼ 2.0 Hz, 1H, H-2 Ar). 13
C
NMR (101 MHz, CDCl3):
d 167.28, 156.21, 151.57, 147.55, 128.50,
119.45, 113.52, 112.75, 111.25, 80.69, 72.85, 72.04, 71.79, 56.17, 51.89,
50.87, 47.37, 40.87, 32.89, 24.19, 18.83, 12.92. Anal. (C22H32N2O5) C,
H, N.
4.1.5. Synthesis of (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]
ethanone O-[2-(4-hydroxypiperidin-1-yl)-2oxoethyl] oxime 1c
Sodium ethoxide was prepared from Na (35 mg, 1.5 mmol) and
absolute EtOH (10 mL). The solvent was evaporated under reduced
pressure and compound 9 (0.373 g, 1.5 mmol) solved in an. DMF
(10 mL) was added to the white solid residue. After stirring for
20 min, 1-(chloroacetyl)piperidin-4-ol (0.532 mg, 3 mmol) solved
in an. DMF (1 mL) was added dropwise and the mixture was stirred
at 50e60 ꢀC for 24 hs under nitrogen atmosphere. After cooling to
room temperature, the mixture was poured into water (50 mL) and
extracted with AcOEt (2 ꢁ 20 mL); the organic phases were washed
with brine (3 ꢁ 20 mL), water (3 ꢁ 20 mL), dried (MgSO4) and
concentrated under reduced pressure to obtain a yellow oil which
4.1.6.1. (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanone O-(3-
morpholin-4-yl-3-oxopropyl) oxime 2a. Yellow oil. Yield: 56%. IR
(film): cm-11630 (CO). 1H NMR (CDCl3):
d 1.53e2.04 (m, 8H, 4CH2
cyclopent.), 2.21 (s, 3H, CH3), 2.81 (t, J ¼ 7.0 Hz, 2H, CH2CO),
3.48e3.59 (m, 4H, 2CH2N morph.), 3.62e3.74 (m, 4H, 2CH2O
morph.), 3.87 (s, 3H, OCH3), 4.50 (t, J ¼ 7.0 Hz, 2H, CH2O), 4.78e4.83
9

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
(m, 1H, OCH cyclopent.), 6.88 (d, J ¼ 10.0 Hz, 1H, H-5 Ar) 7.17 (dd,
J ¼ 10.0, 2.0 Hz, 1H, H-6 Ar) 7.29 (d, J ¼ 2.0 Hz, 1H, H-2 Ar). 13C
Yellow oil. Yield: 96%. IR (CHCl3): cmꢂ1 3445 (OH). 1H NMR (CDCl3):
d
1.10e1.30 (m, 6H, 2CH3 morph.), 1.51e2.08 (m, 8H, cyclopent.),
(101 MHz, CDCl3):
d
170.26, 155.88, 151.74, 147.64, 128.32, 119.56,
2.25 (s, 3H, CH3), 2.35e2.94 (m, 6H, 3CH2N), 3.62e3.85 (m, 2H,
2CH2O morph.), 3.88 (s, 3H, OCH3), 4.10e4.30 (m, 3H,
OCH2 þ CHeOH), 4.79e4.90 (m, 1H, OCH cyclopent.), 6.86 (d,
J ¼ 8.0 Hz, 1H, H-5 Ar), 7.16 (dd, J ¼ 8.0, 2.0 Hz, 1H, H-6 Ar), 7.28 (d,
112.64, 111.28, 80.73, 70.45, 66.81, 56.12, 32.88, 32.84, 24.21, 24.09,
13.02. Anal. (C21H30N2O5) C, H, N.
4.1.6.2. (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanone O-[3-
J ¼ 2.0 Hz, 1H, H-2 Ar). 13C NMR (101 MHz, CDCl3)
d 155.37, 151.40,
(2,6-dimethylmorpholin-4-yl)-3-oxopropyl] oxime 2b. Yellow oil.
147.56, 128.88, 119.21, 112.48, 111.28, 80.62, 75.98, 66.18, 61.26,
60.14, 58.21, 56.19, 32.91, 24.21, 18.97, 12.82. Anal. (C23H36N2O5) C,
H, N.
Yield: 24%. IR (film): cmꢂ1 1640 (CO). 1H NMR (CDCl3):
d 1.09e1.32
(m, 6H, 2CH3 morph.), 1.55e2.02 (m, 8H, 4CH2 cyclopent.), 2.20 (s,
3H, CH3), 2.64e2.90 (m, 6H, 2CH2N morph. þ CH2CO), 3.39e3.64
(m, 2H, 2CHO morph.), 3.87 (s, 3H, OCH3), 4.38e4.58 (m, 2H, CH2O)
4.73e4.83 (m, 1H, OCH cyclopent.), 6.84 (dd, J ¼ 10.0 Hz,1H, H-5 Ar)
7.11 (dd, J ¼ 10.0, 1.8 Hz, 1H, H-6 Ar) 7.28 (dd, J ¼ 1.8 Hz, 1H, H-2 Ar).
4.1.9. Synthesis of (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]
ethanone O-[2-hydroxy-3-(4-hydroxypiperidin-1-yl)propyl] oxime
3c
13C (101 MHz CDCl3):
d
169.72,154.88, 151.38, 147.55, 129.01, 128.44,
To a solution of oxirane 10 (1.26 g, 4.13 mmol) in an. DMF
(10 mL) piperidin-1-ol (0.83 g, 8.26 mmol) solved in an. DMF (2 mL)
was added and the reaction mixture was stirred at 60 ꢀC for 18 hs.
After cooling to room temperature, the reaction mixture was
poured into water (50 mL); the aqueous phase was extracted with
dietyl ether (3 ꢁ 20 mL); the organic phase was washed with water
(2 ꢁ 20 mL), brine (3 ꢁ 20 mL), dried (MgSO4) and concentrated
under reduced pressure to obtain a yellow oil which was purified by
flash chromatography using as eluents firstly dietyl ether, then a
mixture of dietyl ether/methanol (9:1). Finally, the crude oil was
crystallized by adding petroleum ether (p.eb. 4ꢂ01e60 ꢀC). White
solid. Mp: 51e52 ꢀC. Yield: 34%. IR (CHCl3): cm 3429 (OH). 1H
127.43, 119.18, 112.57, 111.28, 80.68, 71.95, 70.36, 56.38, 33.23, 32.91,
24.08, 18.82, 12.78. Anal. (C23H34N2O5) C, H, N.
4.1.7. Synthesis of 1-[3-(cyclopenthyloxy)-4-methoxyphenyl]
ethanone O-(oxiran-2-ylmethyl)-oxime 10
To a sodium ethoxide solution prepared from sodium (0.27 g,
11.3 mmol) and absolute ethanol (20 mL) a solution of oxime 9
(2.82 g, 11.32 mmol) in absolute EtOH (10 mL) was added and the
mixture was stirred at room temperature for 20 min. The solvent
was evaporated under reduced pressure and the residue was sus-
pended in anhydrous DMF (20 mL); then, epichlorohydrin (1.6 mL,
17 mmol) solved in anhydrous DMF (2 mL) was slowly added and
the mixture was heated at 40e50 ꢀC for 12 hs. After cooling to room
temperature, the mixture was poured into water (50 mL) and
extracted with diethyl ether (3 ꢁ 20 mL); the organic phases were
washed with water (3 ꢁ 20 mL), brine (20 mL), dried (MgSO4) and
evaporated under reduced pressure to obtain a yellow oil which
was purified by flash chromatography using DCM as eluent. Yellow
oil. Yield: 53%. IR (KBr): cmꢂ1 1672 (C]N). 1H NMR (CDCl3):
NMR (CDCl3):
d
1.50e2.06 (m, 12H, 4CH2 cyclopent. þ 2CH2 pip.),
2.24 (s, 3H, CH3), 2.47e2.65 (m, 2H, CH2N), 2.75e3.08 (m, 4H,
2CH2N pip.), 3.70e3.79 (m, 1H, OCH pip.), 3.87 (s, 3H, OCH3),
4.10e4.30 (m, 3H, OCH2 þ CHeOH), 4.75e4.90 (m, 1H, OCH cyclo-
pent.), 6.87 (d, J ¼ 8.0 Hz, 1H, H-5 Ar), 7.15 (dd, J ¼ 8.0, 2.0 Hz, 1H, H-
6 Ar), 7.25 (d, J ¼ 2.0 Hz, 1H, H-2 Ar). 13C (101 MHz, CDCl3):
d 155.38,
151.53, 147.23, 128.75, 121.39, 113.26, 111.52, 81.10, 75.39, 67.09,
66.42, 60.80, 56.09, 51.30, 34.69, 32.81, 23.45, 16.78. Anal.
(C22H34N2O5) C, H, N.
d
1.51e2.05 (m, 8H, cyclopent.), 2.25 (s, 3H, CH3), 2.65e2.96 (m, 2H,
CH2O epox.), 3.16e3.45 (m, 1H, CHO epox.), 3.86 (s, 3H OCH3),
4.00e4.45 (m, 2H, OCH2), 4.75e4.90 (m, 1H, OCH cyclopent.), 6.84
(d, J ¼ 8.0 Hz, 1H, H-5 Ar), 7.15 (dd, J ¼ 8.0 Hz, 2.0 Hz, 1H, H ꢂ6 Ar),
7.28 (d, J ¼ 2.0 Hz, 1H, H-2 Ar). Anal. (C17H23NO4) C,H,N.
4.1.10. Synthesis of 3-(4-hyidroxy-3-methoxyphenyl)propionic acid
12
A solution of trans-ferulic acid 11 (0.3 g, 1.54 mmol) in MeOH
(14 mL) was reduced with H-CUBE apparatus at room temperature
in full H2 mode (flow ¼ 1 mL/min) using 5% Pd/C as catalyst. The
obtained solution was concentrated under reduced pressure to
afford the product as white solid. Mp: 88.5e90.0 ꢀC Yield:100% (lit.:
Mp 91e92 ꢀC, Yield 98%) [26].
4.1.8. General procedure for the synthesis of derivatives 3a and 3b
To oxirane 10 (0.60 g, 1.96 mmol) morpholine or 2,6-
dimethylmorpholine (4 mL) was added and the mixture was
heated at 60 ꢀC for 18 h. After cooling to room temperature, the
mixture was poured into water (50 mL) and the aqueous phase was
extracted with dietyl ether (3 ꢁ 20 mL); then, the organic phases
were washed with water (2 ꢁ 20 mL), dried (MgSO4) and concen-
trated under reduced pressure to obtain yellow oils which were
purified by flash chromatography using as the eluents, firstly dietyl
ether, then a mixture of dietyl ether/methanol (9:1).
4.1.11. Synthesis of 6-hydroxy-5-methoxyindan-1-one 13
A solution of 12 (1.2 g, 6.12 mmol) in methanesulfonic acid
(5 mL) was heated to 90 ꢀC under nitrogen atmosphere and the
reaction was stirred and monitored by TLC until acid 12 disappears.
After cooling to room temperature, the reaction mixture was
carefully poured into water (100 mL); the aqueous phase is
extracted with DCM (3 ꢁ 30 mL), the organic phases were washed
with water (3 ꢁ 20 mL), dried (MgSO4) and concentrated under
reduced pressure to obtain a dark solid. Mp: 189e190 ꢀC [lit:
194 ꢀC] Yield: 56% (lit: 65% [30].
4.1.8.1. (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanone O-(2-
hydroxy-3-morpholin-4-ylpropyl) oxime 3a. Yellow oil. Yield: 88%.
IR (CHCl3): cmꢂ1 3452 (OH). 1H NMR (CDCl3):
d 1.52e2.02 (m, 8H, 4
CH2 cyclopent.), 2.24 (s, 3H, CH3), 2.41e2.75 (m, 6H, 3CH2N),
3.66e3.80 (m, 4H, 2CH2O morph.), 3.86 (s, 3H, OCH3), 4.05e4.30
(m, 3H, OCH2 þ CHeOH), 4.75e4.90 (m, 1H, OCH cyclopent.), 6.85
(d, J ¼ 10.0 Hz,1H, H-5 Ar), 7.14 (dd, J ¼ 10.0, 2.0 Hz,1H, H-6 Ar), 7.25
4.1.12. Synthesis of 6-cyclopentyloxy-5-methoxyindan-1-one 14
To a solution of indanone 13 (0.356 g, 2 mmol) in an. DMF
(10 mL) K2CO3 (0.41 g, 3 mmol) and KI (0.01 g, 0.06 mmol) were
added and the suspension was heated at 65 ꢀC. Then cyclopentyl
bromide (0.38 g, 2.57 mmol) was slowly added and the mixture was
stirred at 65 ꢀC for 22 hs. After cooling to room temperature,
toluene (50 mL) was added; the organic phase was washed with
1 M NaOH (2 ꢁ 20 mL) and the combined aqueous phases were
extracted with toluene (20 mL); the organic phases were washed
(d, J ¼ 2.0 Hz, 1H, H-2 Ar). 13C (101 MHz CDCl3):
d 155.02, 151.30,
147.51, 128.99, 119.13, 112.56, 111.32, 80.61, 77.51, 67.00, 66.56,
65.83, 61.29, 56.02, 53.89, 32.75, 24.10, 15.26, 12.69. Anal.
(C21H32N2O5) C, H, N.
4.1.8.2. (1Z)-1-[3-(cyclopentyloxy)-4-methoxyphenyl]ethanone O-[3-
(2,6-dimethylmorpholin-4-yl)-3-hydroxypropyl]
oxime
3b.
10

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
with H2O (3 ꢁ 20 mL), dried (MgSO4) and concentrated under
reduced pressure to obtain a yellow oil which was then used
without further purification. Yield: 73%. IR (CHCl3): cmꢂ11694 (CO).
3H, OCH3), 4.00e4.18 (m, 2H, CH2N pip.), 4.68e4.95 (m, 4H,
CH2CO þ OCH cyclopent. þ CHOH pip.), 6.80 (s, 1H, Ar), 7.30 (s, 1H,
Ar). 13C (101 MHz, CDCl3):
d 169.39, 163.27, 152.17, 144.54, 139.66,
1H NMR (CDCl3):
d
1.512.11 (m, 8H, 4CH2 cyclopent.), 2.67 (t,
131.87, 109.34, 107.69, 81.11, 70.53, 67.25, 56.02, 43.24, 34.03, 32.81,
27.99, 25.09, 23.45. Anal calcd (C22H30N2O5) C, H, N.
J ¼ 5.0 Hz, 2H, CH2), 3.01 (t, J ¼ 5.0, 2H, CH2), 3.93 (s, 3H OCH3),
4.72e4.86 (m, 1H, OCH cyclopent.), 6.87 (s, 1H, Ar), 7.17 (s, 1H, Ar).
Anal. calcd. for (C15H18O3):
4.1.15. Synthesis of (1E)-6-(cyclopentyloxy)-5-methoxyidan-1-one-
O-(oxiran-2ylmethyl) oxime 16
4.1.13. Synthesis of (1E)-6-(cyclopentyloxy)-5-methoxyindan-1-one
oxime 15
To a sodium ethoxide solution prepared from Na (0.14 g,
5 mmol) in absolute EtOH (9.34 mL) a solution of oxime 15 (1.31 g,
5 mmol) in absolute EtOH (10 mL) was added and the mixture was
stirred at room temperature for 20 min. Then, the solvent was
evaporated under reduced pressure, the obtained crude was solved
in an. DMF (10 mL) and epichlorohydrin (0.93 g, 10 mmol) was
added dropwise. The mixture was heated at 40e50 ꢀC for 18 hs.
After cooling to room temperature, the reaction mixture is poured
into water (50 mL) and the aqueous solution is extracted with
diethyl ether (3 ꢁ 20 mL); the organic phase was washed with
water (3 ꢁ 20 mL), brine (20 mL), dried (MgSO4) and concentrated
under reduced pressure to obtain a yellow oil which was purified by
flash chromatography on silica gel using diethyl ether as eluent.
To a solution of indanone 14 (0.40 g, 1.62 mmol) in 95% ethanol
(4.5 mL) hydroxylamine hydrochloride (0.5 g, 6.48 mmol) solved in
water (2 mL) was added. Then, NaHCO3 (0.55 g, 6.48 mmol) and
water (3 mL) were added to small portions and the reaction
mixture was heated at 60 ꢀC for 2 hs. After cooling to room tem-
perature, the mixture was poured into water (50 mL). The solid
obtained was filtered, washed with water and recrystallized froꢂm1
95% ethanol. Ivory solid. Mp: 138e140 ꢀC. Yield: 59%. IR (KBr): cm
1689 (C]N), 1049 (NeO). 1H NMR (CDCl3):
d 1.50e2.13 (m, 8H,
4CH2 cyclopent.), 2.60e2.90 (m, 2H, CH2), 3.60e3.72 (m, 2H, CH2),
3.84 (s, 3H OCH3), 4.69e4.89 (m, 1H, CHO cyclopent.), 6.74 (s, 1H,
Ar), 6.80 (s,1H, Ar). 13C NMR (CDCl3, 300 MHz):
d
24.15, 26.61, 28.35,
Yield: 50%. 1H NMR (CDCl3):
d 1.51e2.10 (m, 8H, 4CH2 cyclopent.),
32.75, 56.05, 80.40, 106.49, 107.81, 127.96, 141.53, 147.60, 152.77,
164.36. Anal. Calcd. for (C15H19NO3).
2.60e2.80 and 2.84e3.06 (2 m, 6H, 2CH2 þ CH2O epox.), 3.20e3.45
(m, 1H, OCH epox.), 3.89 (s, 3H, OCH3), 4.02e4.20 and 4.34e4.44
(2 m, 2H, CH2O), 4.75e4.95 (m, 1H, CHO cyclopent.), 6.80 (s, 1H Ar),
7.12 (s, 1H Ar). Anal calcd (C18H23NO4,) C, H, N.
4.1.14. General procedure for the synthesis of derivatives 4aec
To a solution of 15 (0.261 g, 1 mmol) in anhydrous DMF
(2 mL) K2CO3 (0.414 g, 3 mmol) and 4-(chloroacetyl)morpholine or
4-(chloroacetyl)2,6-dimethylmorpholine or 1-(chloroacetyl)piper-
idin-4-ol (3 mmol) solved in anhydrous DMF (2 mL) were added
dropwise and the mixture was heated at 50e60 ꢀC for 18 hs under
nitrogen atmosphere. After cooling to room temperature, the
mixture was poured into water (100 mL) and extracted with AcOEt
(2 ꢁ 20 mL); the organic phases were washed with brine
(3 ꢁ 20 mL), water (3 ꢁ 20 mL), dried (MgSO4) and concentrated
under reduced pressure to obtain yellow oils which were purified
by flash chromatography, using diethyl ether as eluent.
4.1.16. General procedure for the synthesis of derivatives 5a and 5b
To oxirane 16 (0.37 g, 1.16 mmol) morpholine or 2,6-
dimethylmorpholine (2 mL) was added and the mixture was
heated to 50e60 ꢀC and stirred for 18 hs. After cooling to room
temperature, the reaction mixture is poured into water (50 mL), the
aqueous phase was extracted with diethyl ether (3 ꢁ 20 mL); the
organic phase was washed with water (2 ꢁ 20 mL), dried (MgSO4)
and concentrated under reduced pressure to obtain yellow oils
which were purified by flash chromatography on silica gel, using as
eluents firstly dethyl ether, then a mixture of diethyl ether/meth-
anol (9:1).
4.1.14.1. (1E)-6-(cyclopentyloxy)-5-methoxyindan-1-one O-(2-
morpholin-4-yl-2-oxoethyl)oxime 4a. Ivory solid. Mp: Yield: 44%.
IR (CHCl3): cmꢂ1 1643 (CO), 1603 (C]N). 1H NMR (CDCl3):
4.1.16.1. (1E)-6-(cyclopentyloxy)-5-methoxyindan-1-one O-(2-
hydroxy-3-morpholin-4-ylpropyl)oxime 5a. Yellow solid. Mp:
84e85 ꢀC. Yield: 100%. IR (KBr): cmꢂ1 3444 (OH). 1H NMR (CDCl3):
d
1.53e2.13 (m, 8H, 4CH2 cyclopent.), 2.95e3.12 (m, 4H, 2CH2),
3.57e3.83 (m, 8H, 4CH2 morph.), 3.89 (s, 3H, OCH3), 4.78e4.95 (m,
d 1.51e2.08 (m, 8H, 4CH2, cyclopent.), 2.55e2.70 and 2.83e3.02
3H, OCH cyclopent. þ CH2CO), 6.80 (s, 1H, Ar), 7.17 (s, 1H, Ar). 13
C
(2 m, 10H, 2CH2 þ CH2N þ 2CH2N morph.), 3.70e3.95 (m, 7H,
OCH3 þ 2CH2O morph.), 4.08e4.30 (m, 3H, CHOH þ CH2O),
4.70e4.90 (m, 1H, CHO cyclopent.) 6.77 (s, 1H, Ar), 7.10 (s, 1H, Ar).
(101 MHz, CDCl3):
d 167.76, 164.64, 153.19, 147.63, 141.87, 127.58,
107.82, 106.81, 80.57, 77.24, 77.19, 77.12, 72.72, 66.86, 56.07, 45.78,
42.29, 32.75, 28.41, 27.28, 24.07 Anal calcd (C21H28N2O5) C, H, N.
13C (101 MHz, CDCl3):
d 164.13,152.17,144.54,139.63,131.53,109.35,
107.69, 81.11, 75.37, 66.79, 66.45, 60.69, 56.02, 53.96, 32.81, 28.00,
25.16, 23.45. Anal. Calcd. for (C22H32N2O5) C, H, N.
4.1.14.2. (1E)-6-(cyclopentyloxy)-5-methoxyindan-1-one O-[2-(2,6-
dimethylmorpholin-4-yl)-2-oxoethyl]oxime 4b. Yellow oil. Yield:
46%. IR (CHCl3): cmꢂ1 1641 (CO), 1603 (C]N). 1H NMR (CDCl3):
4.1.16.2. (1E)-6-(cyclopentyloxy)-5-methoxyindan-1-one O-[3-(2,6-
d
1.22 (d, J ¼ 6.2, 6H, 2CH3), 1.51e2.03 (m, 8H, 4CH2 cyclopent.),
dimethylmorpholin-4-yl)-2-hydroxypropyl]oxime 5b. Yellow oil.
2.86e3.08 (m, 4H, 2CH2), 3.40e3.78 (m, 4 H, 2CH2N morph.), 3.88
(s, 3H, OCH3), 3.90e4.15 (m, 1H, CHO morph.), 4.38e4.55 (m, 1H,
CHO morph.), 4.76e4.96 (m, 3H, OCH cyclopent. þ CH2CO), 6.74 (s,
Yield: 62%. IR (film): cm-1 3400 (OH). 1H NMR (CDCl3):
d 1.10e1.30
(m, 6H, 2CH3), 1.52e2.02 (m, 8H, 4CH2 cyclopent.), 2.71e2.89 and
2.90e3.06 (2 m, 10H, 2CH2 þ CH2N þ 2CH2N morph.), 3.85e3.97
(m, 5H, OCH3 þ 2CHO morph.), 4.13e4.23 (m, 3H, CHOH þ CH2O),
4.75e4.90 (m, 1H, OCH cyclopent.), 6.78 (s, 1H, Ar), 7.10 (s, 1H, Ar).
1H, Ar), 7.13 (s, 1H, Ar). 13C (101 MHz, CDCl3):
d 169.44, 163.27,
152.17, 144.54, 139.66, 131.87, 109.34, 107.69, 81.11, 70.73, 69.56,
56.02, 52.44, 32.81, 27.99, 25.09, 23.45, 18.03. Anal calcd
(C23H32N2O5) C, H, N.
13C (101 MHz, CDCl3):
d 164.13, 144.54, 139.63, 131.53, 109.35,
107.69, 81.11, 75.37, 70.25, 66.59, 60.59, 59.78, 56.02, 32.81, 28.00,
25.16, 23.45, 18.20. Anal. Calcd. for (C24H36N2O5) C, H, N.
4.1.14.3. (1E)-6-(cyclopentyloxy)-5-methoxyindan-1-one O-[2-(4-
hydroxypiperidin-1-yl)-2-oxoethyl]oxime 4c. Yellow oil. Yield: 41%.
IR (CHCl3): cmꢂ1 1638 (CO), 1604 (C]N). 1H NMR (CDCl3):
4.1.17. Synthesis of (1E)-6-(cyclopentyloxy)-5-methoxyindan-1-one
O-[2-hydroxy-3-(-4-hydroxypiperidin-1-yl)propyl] oxime 5c
To a solution of 16 (0.29 g, 0.91 mmol) in an. DMF (5 mL)
piperidin-1-ol (0.1 g, 1 mmol) solved in an. DMF (2 mL) was added
d
1.50e2.12 (m, 8H, 4CH2 cyclopent.), 2.81e3.02 (m, 4H, 2CH2),
3.15e3.38 (m, 4H, 2CH2 pip.), 3.63e3.74 (m, 2H, CH2N pip), 3.89 (s,
11

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
and the mixture was heated at 60 ꢀC and stirred for 18 hs. After
cooling to room temperature, the reaction mixture was poured into
water (50 mL), the aqueous phase was extracted with diethyl ether
(3 ꢁ 20 mL); the organic phase was washed with water (2 ꢁ 20 mL),
with brine (3 ꢁ 20 mL), dried (MgSO4) and concentrated under
reduced pressure to obtain a yellow oil which was purified by flash
chromatography on silica gel, using as eluents firstly dethyl ether,
then a mixture of diethyl ether/methanol (9:1). Yield: 33%. IR
the bacmid that was used to obtain high-titer recombinant bacu-
lovirus by PEI MAX (Polysciences Europe GmbH, Hirschberg, Ger-
many) transfection of Sf9 at 0.8 ꢁ 106 cells/mL [31]. Expression of
the protein was done in Sf9 cells infected at 1,5 ꢁ 106 cells/mL for
72h at 27 ꢀC. Cell pellets, harvested by centrifugation, were resus-
pended in 50 mM Hepes pH 7.5, 500 mM NaCl, 10 mM MgCl2, 10%
glycerol, 5 mM imidazole, 10 mg/mL DNaseI, 1 mM TCEP added of
protease inhibitors (Roche, Mannheim, Germany) and homoge-
nized using Emulsiflex C3 cell disruption system (Avestin). Cell
lysate was cleared by centrifugation, after 5 min incubation with
benzonase (Merck Millipore, Darmstadt, Germany), and the su-
pernatant was affinity purified on Ni-NTA resin (Qiagen, Hilden,
Germany). The protein fraction was eluted with 50 mM Hepes pH
7.5, 150 mM NaCl, 10% glycerol, 400 mM imidazole, 1 mM TCEP and
then IEX purified on HiTrap HP Q column (GE Healthcare, Chicago,
Illinois, USA) prior dilution in 100 mM Hepes pH 7.5, 10% glycerol
and 1 mM DTT.
(CHCl3) cmꢂ1: 3522 (OH). 1H NMR (CDCl3):
4CH2 2CH2 pip.), 2.73e3.20
d
1.65e2.00 (m, 12H,
(m, 10H,
cyclopent.þ
2CH2 þ CH2N þ 2CH2N pip.), 3.81 (s, 3H, OCH3), 3.90e4.35 (m, 4H,
CH2O þ CHOH þ OCH pip.), 4.70e4.82 (m, 1H, CHO cyclopent.), 5.33
(br s, 3H, OH þ H2O disappears with D2O), 6.72 (s, 1H, Ar), 7.05 (s,
1H, Ar). 13C NMR (101 MHz, CDCl3)
d 164.13, 152.17, 144.54, 139.63,
131.53, 109.35, 107.69, 81.11, 75.37, 67.09, 66.41, 60.85, 56.02, 51.30,
34.69, 32.81, 28.00, 25.16, 23.45. Anal. Calcd. (C23H34N2O5 H2O) C, H,
N.
4.2. NMR analysis
4.3.3. PDE4D activity and inhibition assays
An NADH-coupled assay was used to detect the levels of AMP
resulting from cAMP hydrolysis and therefore to monitor enzyme
activity [32]. Inhibition assays were carried out in 96 transparent
NMR experiments were performed at 298 K on a Bruker Avance
III 400 MHz spectrometer (Milan, Italy). The NMR experiments
were carried out in 500
mL of CDCl3. All proton and carbon chemical
polystyrene plates (Grainer Bio, Kremsmünster, Austria). A 132.5 mL
shifts were assigned unambiguously using bi-dimensional experi-
ments (COSY, NOESY and HSQC). Edited 1He13C HSQC experiments
were performed to confirm and follow the resonances of carbons.
Phase sensitive 2D-NOESY experiments with gradient pulses in
mixing time were performed with a mixing time of 700 ms in order
to observe homonuclear correlation via dipolar coupling.
volume of a reaction mixture containing 50 mM Tris pH 7.5,
100 mM NaCl, 10 mM MgCl2, 10 mM KCl, 0.3 mM ATP, 0.6 mM PEP,
1.8 U of Pyruvate Kinase, 1.8 U of Lactate Dehydrogenase, 1.6 U
Myokinase, and the enzyme was pre-incubated for 10 min at 25 ꢀC.
Then, cAMP was added to a final concentration of 0.004 mM to start
the hydrolysis reaction. The oxidation of NADH to NADþ was
monitored at 340 nm (NADH ε342 ¼ 6220 Mꢂ1cmꢂ1) using a
€
4.3. Biological studies
Spark10 M plate reader (Tecan, Mannedorf, Switzerland). One unit
(U) is defined as the amount of enzyme required to hydrolyze
4.3.1. PDE4D catalytic domain expression and purification
1 m
mol of cAMP to 50-AMP per minute, while specific activity is
The DNA fragment encoding for the catalytic domain (cat) of
PDE4D (amino acids 244e578) was cloned into a pET3a vector
(Merck Millipore, Darmstadt, Germany) with a C-terminal 6His-tag
and transformed into E. coli BL21(DE3) pLysS cells (Thermo Fisher
Scientific, Waltham MA, USA). Cells were cultured at 37 ꢀC in Luria-
Bertani (LB) medium supplemented with 50 mg/liter ampicillin
until OD600 ¼ 0.6. Protein expression was induced using isopropyl
expressed as U mgꢂ1 of enzyme. Inhibition assays were carried out
following the same procedure and varying the concentration of the
compounds within the range 0.01 mMe300 mM. Compounds were
first dissolved in 100% DMSO at 100 mM concentration and then
diluted to 1% concentration in the assay buffer. IC50 was determined
by plotting the enzyme fractional activity against the logarithm of
compound concentration. Dose-response curve fitting was done
using Sigma Plot (Systat Software) and Eq. (1):
1-thio-b-D-galactopyranoside (IPTG) at a final concentration of
0.5 mM and was carried out overnight at 25 ꢀC. Cells were pelleted,
resuspended in buffer A (20 mM Tris HCl pH 7.5, 150 mM NaCl),
supplemented with 0.5 mM PMSF and DNase (Sigma-Aldrich), and
lysed by sonication. The soluble fraction was first purified by af-
finity chromatography using a Ni-NTA column (Qiagen, Hilden,
Germany) equilibrated with buffer A; elution was done with
400 mM imidazole. Further purification of the sample involved first
a size-exclusion chromatography step using a Sephacryl 100 HR
HiPrep 26/60 column (GE Healthcare, Chicago, Illinois, USA) and
then an ion-exchange chromatography step using a HiPrep Q HP 16/
10 column (GE Healthcare, Chicago, Illinois, USA). The final protein
sample was dialyzed against buffer A and used at 10 mg/ml for
crystallization experiments and at 0.025 mg/ml for inhibition
assays.
-x)
y ¼ min þ ((max - min) / (1 þ 10(logIC
)
(Eq. 1)
50
where y is the fractional activity of the enzyme in the presence of
inhibitor at concentration [I], max is the maximum value of y
observed at [I] ¼ 0, and min is the minimum limiting value of y at
higher inhibitor concentration. All measurements were done in
triplicates.
4.4. Crystallization and X-ray crystallography
Crystals of the PDE4D catalytic domain were grown by vapor
diffusion using the hanging drop method. One
catalytic domain solution at 10 mg/ml concentration was mixed
with 1 L of the crystallization buffer consisting of 10e20% PEG
mL of the PDE4D
m
4.3.2. PDE4D3 expression and purification
3350, 10e100 mM Hepes pH 7.5, 100e200 mM MgCl2. The crystals
thus obtained were soaked for 48h in half-concentration crystalli-
zation buffer solutions containing 1e6 mM inhibitor. Finally, crys-
tals were flash-frozen in 20% ethylene glycol.
Diffraction data were collected at the X06DA-PXIII beamline at
Swiss Light Source, Paul Scherrer Institute, Villigen, Switzerland
using a wavelength of 1 Å. Diffractions images were processed
using XDS and the crystal structures were solved by molecular
replacement using the crystal structure of the PDE4D catalytic
The PDE4D3 protein construct was expressed in baculovirus-
infected insect cells. The codon-optimized PDE4D3 human
sequence was synthesized by GenScript (Piscataway, NJ, USA)
including two point mutations - S54D (PKA phosphomimetic) and
S579A (preventing known inactivating phosphorylation [9]. The
construct, bearing a C-terminal 6His-tag, was inserted in pFastBac
dual vector and transformed into E. coli DH10EMBacY (strain kindly
provided by I. Berger, University of Bristol, Bristol, UK) to generate
12

C. Brullo, F. Rapetti, S. Abbate et al.
European Journal of Medicinal Chemistry 223 (2021) 113638
Goethem, P. D'Ursi, A. Orro, L. Milanesi, S. Guariento, E. Cichero, P. Fossa,
E. Fedele, O. Bruno, New insights into selective PDE4D inhibitors: 3-(Cyclo-
domain (PDB code: 2PW3) as the search probe. All the refinements
were carried out with Phenix [33]. Crystallographic data are re-
ported in Table 3.
pentyloxy)-4-methoxybenzaldehyde
O-(2-(2,6-dimethylmorpholino)-2-
oxoethyl) oxime (GEBR-7b) structural development and promising activities
to restore memory impairment, Eur. J. Med. Chem. 124 (2016) 82e102,
https://doi.org/10.1016/j.ejmech.2016.08.018.
Accession codes
[14] R. Ricciarelli, C. Brullo, J. Prickaerts, O. Arancio, C. Villa, C. Rebosio, E. Calcagno,
M. Balbi, B.T.J. van Hagen, E.K. Argyrousi, M.A. Pronzato, O. Bruno, E. Fedele,
Memory-enhancing effects of GEBR-32a,
promise for the treatment of Alzheimer's disease, Sci. Rep. 7 (2017) 1e14,
https://doi.org/10.1038/srep46320.
a new PDE4D inhibitor holding
The final crystallographic coordinates of the crystal structure
shown here are available as PDB entry 7AY6 and 7B9H.
[15] O. Bruno, E. Fedele, J. Prickaerts, L.A. Parker, E. Canepa, C. Brullo, A. Cavallero,
E. Gardella, A. Balbi, C. Domenicotti, E. Bollen, H.J.M. Gijselaers, T. Vanmierlo,
K. Erb, C.L. Limebeer, F. Argellati, U.M. Marinari, M.A. Pronzato, R. Ricciarelli,
GEBR-7b, a novel PDE4D selective inhibitor that improves memory in rodents
at non-emetic doses, Br. J. Pharm. 164 (2011) 2054e2063, https://doi.org/
10.1111/j.1476-5381.2011.01524.x.
[16] T. Prosdocimi, L. Mollica, S. Donini, M.S. Semrau, A.P. Lucarelli, E. Aiolfi,
A. Cavalli, P. Storici, S. Alfei, C. Brullo, O. Bruno, E. Parisini, Molecular bases of
PDE4D inhibition by memory-enhancing GEBR library compounds,
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
Acknowledgements
Biochemistry
57
(2018)
2876e2888,
https://doi.org/10.1021/
acs.biochem.8b00288.
[17] V. Cavalloro, K. Russo, V. Vasile, L. Pignataro, A. Torretta, S. Donini,
M.S. Semrau, P. Storici, D. Rossi, F. Rapetti, C. Brullo, E. Parisini, O. Bruno,
S. Collina, Insight into GEBR-32a: chiral resolution, absolute configuration and
enantiopreference in PDE4D inhibition, Molecules 25 (2020) 935, https://
doi.org/10.3390/molecules25040935.
The authors wish to thank the X06DA-PXIII beamline personnel
of the Paul Scherrer Institute, Villigen, Switzerland, for help with
data collection.
[18] E.J. Barreiro, A.E. Kümmerle, C.A.M. Fraga, The methylation effect in medicinal
chemistry, Chem. Rev. 111 (2011) 5215e5246, https://doi.org/10.1021/
cr200060g.
Appendix A. Supplementary data
[19] M.J. Ashton, D.C. Cook, G. Fenton, J.A. Karlsson, M.N. Palfreyman, D. Raeburn,
A.J. Ratcliffe, J.E. Souness, S. Thurairatnam, N. Vicker, Selective type IV phos-
phodiesterase inhibitors as antiasthmatic agents. The syntheses and biological
activities of 3-(cyclopentyloxy)-4-methoxybenzamides and analogues, J. Med.
Chem. 37 (1994) 1696e1703, https://doi.org/10.1021/jm00037a021.
[20] A.S. Paraskar, A. Sudalai, Enantioselective synthesis of (-)-cytoxazone and
(þ)-epi-cytoxazone, novel cytokine modulators via Sharpless asymmetric
epoxidation and l-proline catalyzed Mannich reaction, Tetrahedron 62 (2006)
5756e5762, https://doi.org/10.1016/j.tet.2006.03.079.
[21] K. Euikyung, C. Hyung-Ok, J. Sung-Hak, K. Jong Hoon, L. Jae-Mok, S. Byung-
Chul, X.X. Myung, K.R. Chung, Improvement of therapeutic index of phos-
phodiesterase type IV inhibitors as anti-asthmatics, BMCL 13 (2003)
2355e2358, https://doi.org/10.1016/s0960-894x(03)00405-0.
[22] E.J.E. Freyne, G.S.M. Diels, G.J.I. Andres, G.F.J. Fernandez, 1,3-Dihydro-1-(phe-
nylalkyl)-2H-imidazole-2-one Derivatives Having PDE IV and Cytokine Ac-
tivity, PCT Int. Appl., 1996. WO 9631485.
[23] L.J. Lombardo, Preparation of oxime carbamates and oxime carbonates as
bronchodilators and antiinflammatory agents, Eur. Pat. Appl. (1992) 470805.
[24] B.R. Kim, G.H. Sung, J.J. Kim, Y.J. Yoon, A development of rapid, practical and
selective process for preparation of Z-oximes, J. Kor. Chem. Soc. 57 (2013)
295e299, https://doi.org/10.5012/jkcs.2013.57.2.295.
[25] M. Tada, H. Hirano, A. Suzuki, Photochemistry of host-guest complex III. Effect
of guest cation of the photoreactivity of acetophenone oxime derivatives
having crown ether moiety, Chem. Soc. Jpn. 53 (1980) 2304e2308, https://
doi.org/10.1246/bcsj.53.2304.
[26] T. Rafiq, J.L. Sorensen, Synthesis of Actinomycetes natural products JBIR-94,
JBIR-125, and related analogues, Tetrahedron Lett. 56 (2015) 7108e7111,
https://doi.org/10.1016/j.tetlet.2015.11.020.
[27] F.C. Meng, F. Mao, W.J. Shan, F. Qin, L. Huang, X.S. Li, Design, synthesis, and
evaluation of indanone derivatives as acetylcholinesterase inhibitors and
metal-chelating agents, BMCL 22 (2012) 4462e4466, https://doi.org/10.1016/
j.bmcl.2012.04.029.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2021.113638.
References
[1] E.R. Kandel, The molecular biology of memory: cAMP, PKA, CRE, CREB-1,
CREB-2, and CPEB, Mol. Brain 5 (2012) 14, https://doi.org/10.1186/1756-
6606-5-14.
[2] M.D. Houslay, P. Schafer, K.Y. Zhang, Phosphodiesterase-4 as a therapeutic
target, Drug Discov. Today 10 (2005) 1503e1519, https://doi.org/10.1016/
S1359-6446(05)03622-6.
[3] J. Prickaerts, P.R.A. Heckman, A. Blokland, Investigational phosphodiesterase
inhibitors in phase I and phase II clinical trials for Alzheimer's disease, Expet
Opin. Invest. Drugs 26 (2017) 1033e1048, https://doi.org/10.1080/
13543784.2017.1364360.
[4] C. Zhang, Y. Xu, H.T. Zhang, M.E. Gurney, J.M. O'Donnell, Comparison of the
pharmacological profiles of selective PDE4B and PDE4D inhibitors in the
central nervous system, Sci. Rep. 7 (2017) 40115, https://doi.org/10.1038/
srep40115.
[5] M.E. Gurney, E.C. D'Amato, A.B. Burgin, Phosphodiesterase-4 (PDE4) molecular
pharmacology and Alzheimer's disease, Neurotherapeutics 12 (2014) 49e56,
https://doi.org/10.1007/s13311-014-0309-7.
[6] D.C. Lynch, D.A. Dyment, L. Huang, S.M. Nikkel, D. Lacombe, P.M. Campeau,
B. Lee, C.A. Bacino, J.L. Michaud, F.P. Bernier, J.S. Parboosingh, A.M. Innes,
Identification of novel mutations confirms PDE4D as a major gene causing
acrodysostosis, Hum. Mutat. 34 (2013) 97e102, https://doi.org/10.1002/
humu.22222.
[7] C. Silve Acrodysostosis, A new form of pseudohypoparathyroidism? Ann.
Endocrinol. 76 (2015) 110e112, https://doi.org/10.1016/j.ando.2015.03.004.
[8] O. Bruno, R. Ricciarelli, J. Prickaerts, L. Parker, E. Fedele, PDE4D inhibitors: a
potential strategy for the treatment of memory impairment? Neuropharma-
[28] H. Irie, S. Tanaka, Y. Zhang, J. Koyama, T. Taga, K. Machida, Synthesis of
methoxyonchine and use of 1H- and 13C-nuclear magnetic resonance spectra
for structure determination of geometrical isomers of indan-1-one oxime
derivatives, Chem. Pharm. Bull. 36 (1988) 3134e3137, https://doi.org/
10.1248/cpb.36.3134.
cology
85
(2014)
290e292,
https://doi.org/10.1016/
j.neuropharm.2014.05.038.
[9] 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, https://doi.org/10.1038/nbt.1598.
[10] D. Fox, A.B. Burgin, M.E. Gurney, Structural basis for the design of selective
phosphodiesterase 4B inhibitors, Cell. Signal. 26 (2014) 657e663, https://
doi.org/10.1016/j.cellsig.2013.12.003.
[11] 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, BMC (Biomed. Chromatogr.) 23 (2015) 3426e3435, https://
doi.org/10.1016/j.bmc.2015.04.027.
[12] O. Bruno, A. Romussi, A. Spallarossa, C. Brullo, S. Schenone, F. Bondavalli,
N. Vanthuyne, C. Roussel, New selective phosphodiesterase 4D inhibitors
differently acting on long, short, and supershort isoforms, J. Med. Chem. 52
(2009) 6546e6557, https://doi.org/10.1021/jm900977c.
€
[29] The PyMOL Molecular Graphics System, Schrodinger, LLC, Portland, OR.
[30] L. Huang, H. Miao, Y. Sun, F. Meng, X. Li, Discovery of indanone derivatives as
multi-target-directed ligands against Alzheimer's disease, Eur. J. Med. Chem.
87 (2014) 429e443, https://doi.org/10.1016/j.ejmech.2014.09.081.
[31] C. Bieniossek, T. Imasaki, Y. Takagi, I. Berger MultiBac, Expanding the research
toolbox for multiprotein complexes, Trends Biochem. Sci. 37 (2012) 49e57,
https://doi.org/10.1016/j.tibs.2011.10.005.
[32] S.P. Chock, C.Y. Huang, An optimized continuous assay for cAMP phosphodi-
esterase and calmodulin, Anal. Biochem. 138 (1984) 34e43, https://doi.org/
10.1016/0003-2697(84)90765-6.
[33] P.D. Adams, P.V. Afonine, G. Bunkoczi, V.B. Chen, I.W. Davis N. Echols,
J.J. Headd, L.W. Hung, G.J. Kapral, R.W. Grosse-Kunstleve, A.J. McCoy,
N.W. Moriarty, R. Oeffner, R.J. Read, D.C. Richardson, J.S. Richardson,
T.C. Terwilliger, P.H. Zwart, PHENIX: a comprehensive Python-based system
for macromolecular structure solution, Acta Crystallogr. Sect. D Biol. Crys-
tallogr. 66 (2010) 213e221, https://doi.org/10.1107/S0907444909052925.
[13] C. Brullo, R. Ricciarelli, J. Prickaerts, O. Arancio, M. Massa, C. Rotolo,
A. Romussi, C. Rebosio, B. Marengo, M.A. Pronzato, B.T. Van Hagen, N.P. Van
13