Diastereoselective synthesis and profiling of bicyclic imidazolidinone derivatives bearing a difluoromethylated catechol unit as potent phosphodiesterase 4 inhibitors

Article abstract of DOI:10.1039/C8OB01039K

Metal-mediated C-H functionalization of cyclic N-oxides was exploited to access a series of new difluoromethylated analogs of imidazolidinone-based PDE4 inhibitor CMPI in a diastereoselective manner. Among the products synthesized, compounds with fine-tuned activity/selectivity profiles compared to both CMPI and the clinically applied apremilast were identified. From these studies, an unusual fused 1,2-oxazinoimidazolidinone heterocyclic system was suggested as a novel scaffold for the design of potent and selective PDE4 inhibitors. Computational studies suggest that the oxygen atom in the imidazolidinone unit can bind to the metal ion center (most likely Mg²⁺). DFT calculations of the relative interaction energies of inhibitors with Mg²⁺ and Zn²⁺ ions were performed on a model of the bimetal active site of PDE4.

Full text of DOI:10.1039/C8OB01039K

View Article Online
View Journal
Organic &
Biomol ec ular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: A. Y. Sukhorukov,
V. Dorokhov, I. Golovanov, S. Ioffe and V. Tartakovsky, Org. Biomol. Chem., 2018, DOI:
1
0.1039/C8OB01039K.
Volume 14 Number
1
7
January 2016 Pages 1–372
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Organic &
Biomolecular
Chemistry
www.rsc.org/obc
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1477-0520
COMMUNICATION
Takeharu Haino et al.
Solvent-induced emission of organogels based on tris(phenylisoxazolyl)
benzene
rsc.li/obc

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 1 of 9
View Article Online
DOI: 10.1039/C8OB01039K
Journal Name
ARTICLE
Diastereoselective Synthesis and Profiling of Bicyclic
Imidazolidinone Derivatives Bearing a Difluoromethylated
Catechol Unit as Potent Phosphodiesterase 4 Inhibitors
Received 00th January 20xx,
Accepted 00th January 20xx
a,b
a
a
a
DOI: 10.1039/x0xx00000x
Valentin S. Dorokhov, Ivan S. Golovanov, Vladimir A. Tartakovsky, Alexey Yu. Sukhorukov* and
Sema L. Ioffe
a
www.rsc.org/
Metal-mediated C-H functionalization of cyclic N-oxides was exploited to access a series of new difluoromethylated
analogs of imidazolidinone-based PDE4 inhibitor CMPI in a diastereoselective manner. Among the products synthesized,
compounds with fine-tuned activity/selectivity profiles as compared to both CMPI and the clinically applied Apremilast
were identified. From these studies, an unusual fused 1,2-oxazinoimidazolidinone heterocyclic system was suggested as a
novel scaffold for the design of potent and selective PDE4 inhibitors. Computational studies suggest that the oxygen atom
2
+
in the imidazolidinone unit can bind to the metal ion center (most likely Mg ). DFT calculations of the relative interaction
2+
2+
energies of inhibitors with Mg and Zn ions were performed on a model of the bimetal active site of PDE4.
1
a,3c,4
PDE4’s in vomiting center of brain.
Some potent PDE4
Introduction
inhibitors, for example Cilomilast, are even more selective
1
c
toward PDE4D as compared to other isotypes. Thus,
nowadays, a lot of research is being done toward the design of
more potent and isotype-selective PDE4 inhibitors, preferably
Inhibitors of cAMP-specific type 4 phosphodiesterase (PDE4)
have gained a considerable interest as potential drugs against
inflammatory diseases, such as COPD, asthma, psoriatic
5
1
administered using the inhalation route. Notably, PDE4B
arthritis, and some others. PDE4 enzyme catalyses the
isotype is considered to be an optimal target for the
development of PDE4 inhibitors with high therapeutic index.
hydrolysis of cyclic adenosine 3’,5’-monophosphate (cAMP) in
immune cells (neutrophils, macrophages, monocytes and T-
cells), thereby regulating the concentration of this important
second messenger. The decrease of intercellular cAMP levels
as a result of PDE4 activity leads to the release of several pro-
inflammatory mediators such as TNF-α, IL-12, IL-2 and LTB4.
Inhibition of cAMP-specific PDE4 activity prevents the pro-
inflammatory cell activation, and also reduces the dysfunction
of structural lung cells, such as relaxation of airway smooth
2
muscle. Recently, PDE4 inhibitors Roflumilast (for treatment
of COPD) and Apremilast (for treatment of psoriatic arthritis)
1
were approved by FDA.
a,b
However, most of PDE4 inhibitors, including those applied in
clinic have dose-limiting gastric adverse effects, such as
3
nausea, diarrhea and weight loss. These side effects are
usually associated with non-selective inhibition of PDE4
isotypes, in particular PDE4D, which dominates over other
a.
N. D. Zelinsky Institute of Organic Chemistry, 119991, Leninsky prospect, 47,
Moscow, Russian Federation. E-mail: sukhorukov@ioc.ac.ru;
Fax: +7 499 1355328; Tel: +7 499 1355329
Higher Chemical College of the Russian Academy of Sciences, D. Mendeleev
b.
University of Chemical Technology of Russia, 125047, Miusskaya sq., 9, Moscow,
Russian Federation
Electronic Supplementary Information (ESI) available: Experimental procedures,
Figure 1. Difluoromethylated analogs of PDE4 inhibitor CMPI
1
13
19
copies of FTIR, NMR H, C, F, COSY, HSQC, NOESY spectra of new compounds,
details of biological assays, molecular docking and DFT calculations. See
DOI: 10.1039/x0xx00000x
Recently, we reported the first asymmetric synthesis and
evaluation of both enantiomers of a highly potent PDE4
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 2 of 9
View Article Online
DOI: 10.1039/C8OB01039K
ARTICLE
Journal Name
6
inhibitor CMPI (Figure 1), which was originally designed by CMPI has a larger heterocyclic core (as compared to
7
GlaxoSmithKline as a strained analog of Ro-20-1724. Here, we Roflumilast and Cilomilast), which can extend to the bimetal
used our synthetic strategy, improved by the recently ion site (Figure 3, (A)). The bicyclic imidazolidinone unit with
developed transition metal-promoted methods for C-H certain stereochemistry can exploit hydrophilic interactions
8
functionalization
difluoromethylated derivatives
imidazolidinone scaffold from CMPI (cycles
moiety of Roflumilast (cathechol unit in Figure 1). In vitro/in Substitution of aryl unit
of
N-oxides,
to
access
new with residues in M pocket (hydrogen bonds with ASP392 and
I
, which combine the bicyclic HIS234), as well as direct metal-ligand coordination as can be
6
a
A
,
B
) and the aryl judged from our previous docking studies.
in CMPI for the difluromethylated
C
C
vivo evaluation of these compounds identified new potent cathechol from Roflumilast was expected to increase the
PDE4 inhibitors with fine-tuned activity/selectivity profiles. potency of the inhibitor. This was supported by docking of
These studies revealed structure-activity relationships hybrid compound 1a into the catalytic site of PDE4B using the
1
1
throughout these series and provided insights into the binding crystal structure 1XMU (AutoDock Vina software was used,
pattern of cyclic urea fragment to the catalytic site of PDE4 see Supporting information for details and protocol).
enzymes.
Compound 1a showed more interactions as compared to CMPI
and ca. 1 kcal/mol higher calculated affinity (Figure 3, (B)).
Results and discussion
Structure design
The binding mode of catechol-based inhibitors is well defined
with numerous co-crystal structures available for the
9
complexes of PDE4B. The active site of PDE4B is comprised of
three subunits, namely Q pocket (containing the glutamine
switch and hydrophobic P-clamp), M pocket (containing a
bimetal ion center) and S pocket (solvent-filled pocket). As can
(
A)
(B)
be seen from the crystal structure of PDE4B with Roflumilast
9
1XMU), the aryl unit is located in the Q pocket, making
(
π
-π
Figure 3. Molecular docking of CMPI and compound 1a into the catalytic site of
PDE4B (AutoDock Vina software). (A) Residues in Q and M pockets of PDE4B
interacting with CMPI and the resulting affinity. (B) Residues in Q and M pockets
of PDE4B interacting with 1a and the resulting affinity. Large-size images can be
found in Supporting information.
stacking and hydrophobic interactions with P-clamp consisting
of PHE446, ILE410, and PHE414 residues (Figure 2, (A)). Oxygen
atoms of cathechol unit are involved in a bidentate hydrogen
bond with the γ-amide group of the GLU443 residue.
Difluoromethyl group is located in Q
TYR233, ASN395, THR407, and TYR403, while the hydrophobic
cyclopropylmethyl group lies across the Q pocket (formed by
residues MET411, MET431, and PHE414). Overall, the less
electron-rich difluromethoxy-substituted cathechol in
1
pocket formed by
2
Roflumilast fits the lipophilic Q pocket better then the
methoxy-substituted aryl in CMPI, Cilomilast and related PDE4
9
inhibitors. Moreover, an important advantage of OCHF
2
group
over the methoxy-group is its ability to serve as a hydrogen
10a
bond donor to the oxygen atom of THR407 residue as well
1
0b,c
as a counterpart in the multipolar interaction
residue (Figure 2, (B)).
with ASN395
Figure 4. Calculated (docking) binding energies of various bicyclic
imidazolidinones I in kcal/mol (AutoDock Vina software).
Ar = 3-(cyclopropylmethoxy)-4-(difluoromethoxy)phenyl.
Since the size and type of heterocyclic ring
B should influence
binding in M pocket, we performed docking of predesigned
structures possessing various bicyclic imidazolidinone cores
selected examples are shown in Figure 4, see Supporting
information for the full study). As can be seen from these
studies, the increase of the size of ring , as well as the
I
(
B
introduction of heteroatoms and unsaturation resulted in the
lowering of the predicted binding energy. However, apart from
a pyrrolidine (1a), an unusual 1,2-oxazine ring (2a) was
identified as a promising scaffold from this modeling. It should
(
A)
(B)
Figure 2. Binding modes of Roflumilast (crystal structure 1XMU). (A) Residues in
Q
pocket of PDE4B interacting with Roflumilast. (B) Close view on the be noted, that no examples of 1,2-oxazine-based PDE4
difluoromethyl group making hydrogen bond with THR407 and multipolar
interaction with ASN395. Large-size images can be found in Supporting inhibitors has been reported in the literature so far. On the
information.
2
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 3 of 9
View Article Online
DOI: 10.1039/C8OB01039K
Journal Name
other hand, our recent studies revealed high potential of this
ARTICLE
1
2
scaffold for the design of anti-inflammatory compounds.
Thus, pyrroloimidazolidinone 1a and
oxazinoimidazolidinone 2a were selected as target structures
Figure 4). In order to gain insights into effects of substituents
in rings and , derivatives of 1a and 2a lacking the
difluoromethyl group in cathechol unit , lacking NH hydrogen
in ring (R H), having sulfur instead of carbonyl oxygen in
ring (X = S) and alkyl groups (R = Alk) in the ring B were also
1,2-
(
A
,
B
C
C
A
≠
1
A
synthesized (vide infra).
Synthetic strategy
The general strategy used to access target molecules
shown in Scheme 1. According to this approach, the precursors
of and are aminomethyl-substituted heterocycles P-1
which can be obtained by reduction of azido derivatives P-2
1 and 2 is
1
2
,
.
Controlled reduction of intermediates P-2 would allow either Scheme 2. Synthesis of nitroalkenes 4a and 4b from protocatechuic aldehyde
to selectively reduce C=N bond with retaining of 1,2-oxazine
ring, or to perform reductive contraction of the 1,2-oxazine to On the next stage, racemic 1,2-oxazine-N-oxides 7a and 7b
1
3
pyrrolidine cycle.
precursors to both pyrroloimidazolidine
oxazinoimidazolidine series of products.
Therefore, azides P-2 can serve as were assembled by the inverse electron demand [4+2]-
1
and 1,2- cycloaddition of nitroalkenes 4a or 4b with ethyl vinyl ether
promoted by tin tetrachloride (eq. (1), Scheme 3). Remarkably,
in contrast to nitrostyrene 4b, the difluoromethyled derivative
2
4a entered cycloaddition with poor conversion. This can be
attributed to its lesser ability to form a reactive complex with
SnCl because of a more electron-deficient character of aryl
4
1
5
group. The unreacted nitroalkene 4a could be recovered and
subsequently used in the cycloaddition. After three repeated
iterations, the desired nitronate 7a was obtained in acceptable
5
9% yield based on the nitroalkene 4a.
Scheme 1. Suggested synthetic strategy to compounds
1 and 2
Similar strategy was used in the synthesis of CMPI
a
6
enantiomers, yet here we attempted to use the advantage of
our recently developed transition metal promoted red-ox С
−
Н
8
activation of cyclic N-oxides of type P-3. Moreover, it was not
evident, whether this methodology is tolerated by electron-
deficient difluoromethyl group.
Diastereoselective synthesis of pyrroloimidazolidinone series
Initially, the synthesis of hitherto unknown nitroalkenes 4a and
4
b
was performed from protocatechuic aldehyde
to the general route shown in Scheme 2. At the first stage, the
more acidic hydroxyl group at the C-4 position in was
difluoromethylated or benzylated using standard protocols,
which gave aldehydes 5a and 5b, respectively. Subsequent
alkylation of the free OH group in with
bromomethylcyclopropane in the presence of K CO furnished
aldehydes . Condensation of the latter with nitroethane
catalyzed by ammonium acetate afforded nitroalkenes 4a and
3 according
3
1
4
Scheme 3. Synthesis of 1,2-oxazine-N-oxides 7a-c
The enantiomerically pure cyclic nitronate 7c, a precursor of
7S,7aR)-enantiomer of 1a, was prepared employing trans-1-
phenyl-2-(vinyloxy)cyclohexane (1S,2R- ) as chiral auxiliary
eq. (2), Scheme 3). Only two of four possible stereoisomers
of nitronate 7c were formed with one being predominant (dr =
5
(
2
3
8
6
6
(
4b.
9
:
1). The major isomer was isolated by column
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 4 of 9
View Article Online
DOI: 10.1039/C8OB01039K
ARTICLE
Journal Name
chromatography with high diastereomeric and enantiomeric novel approach to achieve selective functionalization of
1
α-
position in nitronates based on their transformation into N-
6
purity (de 94%, ee 96%).
silyloxyenamines and subsequent reaction with nucleophiles in
8
the presence transition metal Lewis acids. Following this
OR2
OR2
method, N-silyloxyenamines 9a-c were prepared in almost
quantitative yields by silylation of nitronates 7a-c (Scheme 4).
O
O
However, interaction of enamines
9
with NaN
3 2
/Zn(OTf)
17
8
system produced only complex mixtures of products.
c
Me3SiBr
4
3
6
N
Fortunately, we were able to introduce the nitroxy-group using
N
NEt3, -50°C
R O
3
O
OSiMe3
3
R O
O
O
CH2Cl2, 99%
the recently developed Cr(III)-promoted addition of the nitrate
8a,18
9
a-c
7
a-c
anion to N-silyloxyenamines of type
9
.
Subsequent
NaN3
Zn(OTf)2
DMF, 2 h
nucleophilic substitution of the nitroxy-group for the azide
anion provided azides 11a-c in high yields (Scheme 4).
Nu
LA
Cr(NO3)3
THF, 2 h
N
O
O
On the next stage, the stereocenter at the C-3 atom in azides
SiMe3
3
11 was installed by NaBH CN/AcOH reduction, which produced
OR2
OR2
exclusively 3,4-trans-isomers of corresponding tetrahydro-1,2-
oxazines 12a-c (Scheme 5). Hydrogenation of these 1,2-
O
O
2
oxazines over Raney nickel catalyst in the presence of Boc O
gave aminomethyl-substituted pyrrolidines 13a and 13b via a
NaN3
ONO2
N3
domino-process involving successive reduction/protection of
DMF, 60°C,
5 h, 90%
N
FT-IR
N
3
azido-group, cleavage of the N−O bond, and intramolecular
3
R O
O
-1
R O
O
1632 cm
reductive amination to give a pyrrolidine ring (Scheme 5).
Interestingly, benzyloxy group in 12b was not cleaved under
these conditions. The resulting Boc-protected pyrrolidines 13
1
1a (97%), 11b (91%)
10a (44%), 10b (43%)
10c (53%)
11c (95%)
a R3 = Et, R2 = CHF2; b R3 = Et, R2 = Bn; c R3 = 2-phenyl-cyclohexyl, R2 = CHF2
(
without
pyrroloimidazolidinones rac-1a and 1b upon gentle reflux in
DMSO. Enantiomerically pure product ( )-(7S,7aR)-1a (96% ee)
Having prepared both racemic and enantiomerically enriched was obtained in a same manner from azide 12c (Scheme 5).
nitronates , we then proceeded with azidation of the methyl
group at the C-3 atom (Scheme 4). Recently, we reported a
purification)
were
cyclized
into
target
Scheme 4. Synthesis of azides 11a-c
−
7
Scheme 5. Synthesis of target compounds 1a-e from azides 11a-c
4
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 5 of 9
View Article Online
DOI: 10.1039/C8OB01039K
Journal Name
ARTICLE
The synthesis of free phenol 1c was accomplished by catalytic Nitronate 18b, which is the intermediate in the synthesis of
reductive debenzylation of pyrroloimidazolidinone rac-1b over 2,2-dimethyl-substituted derivative of 2a, was prepared in a
Pd-C (Scheme 5). Unfortunately, we were not able to perform straightforward way by the SnCl
difluoromethylation of phenol 1c to give product 1a using cycloaddition of nitroalkene 4a to 2-methylpropene (Scheme
BrF CCO Et/NaOH as well as the recently developed 7). Interestingly, unlike the reaction with ethyl vinyl ether, no
TMSCF Br/KOH system, likely because of a competitive significant effect of difluoromethyl group on the reactivity and
addition of difluorocarbene to the N H unit. conversion was found in this case (cf. Scheme 3, eq. (1) and
4
-promoted [4+2]-
2
2
1
9
2
−
Racemic pyrroloimidazolidinone 1a was further modified by Scheme 7). This is likely because cationic polymerization of 2-
methylation of carbamide nitrogen to give product 1d and by methylpropene is much slower as compared to vinyl ethers.
transformation of carbamide into thiocarbamide with
Lawesson’s reagent (product 1e) (Scheme 5).
Diastereoselective synthesis of 1,2-oxazinoimidazolidinone
series
Fused 1,2-oxazinoimidazolidine heterocyclic system was not
known in the literature, and no strategies for its assembly have
been developed so far. To synthesize the proposed 1,2-
oxazine-based structure 2a possessing an extra oxygen atom in
Scheme 7. Synthesis of nitronate 18b
the ring
B, 1,2-oxazine-N-oxide 18a unsubstituted at C-6
position is needed (Scheme 6). The synthesis of this
intermediate cannot be achieved by a conventional [4+2]-
Functionalization of the methyl group at the C-3 position in N-
oxides 18 was accomplished in a similar manner to nitronates
cycloaddition used to prepare 1,2-oxazine-N-oxides
), since ethylene does not react with nitroalkenes.
7 (Scheme
7
(Scheme 8), yet the initial introduction of bromide proved to
be more advantageous as compared to the nitrate anion in this
case. Thus, silylation of N-oxides 18a and 18b with TMSBr/Et
followed by treatment of intermediate N-syloxyenamines 19a
3
3
N
8
b
and 19b with CoBr
The latter were reacted with NaN
to give azidomethyl-substituted 1,2-oxazines 21a
yields from nitronates 18. Reduction of 21a with NaBH
2
in THF furnished bromides 20a and 20b
in DMF without purification
in 73-86%
CN in
.
3
,
b
3
acetic acid provided tetrahydro-1,2-oxazine 22a as a mixture
of diastereomers in 6.5 : 1 ratio in favor of desired 3,4-trans-
isomer. Analogous reaction of 21b gave single isomer of the
corresponding reduction product 22b (Scheme 8).
Final stage of the synthesis was the construction of
imidazolidinone ring
reduce azido-group leaving the N
Catalytic hydrogenation of 22a even under very mild
conditions (1 bar H , Ra-Ni) led to the reduction of both
groups. Fortunately, we found, that protection of the NH
group (arising from azide) with Boc resulted in somewhat
decrease of the rate of N O bond hydrogenolysis. Thus, mild
hydrogenation of oxazine 22a over Raney nickel in the
presence of Boc O afforded Boc-protected derivative of amine
Accordingly, nitronate 18a was prepared in a step-wise 23a. The latter was transformed into target bicyclic derivative
manner through the cyclization of -iodo-substituted nitro 2a by deprotection with TFA followed by treatment with
A
. The major challenge here was to
−
O bond intact (Scheme 8).
2
2
-
−
Scheme 6. Synthesis of nitronate 18a
2
δ
compound 17a. The synthesis of the latter was accomplished triphosgene/Et N system.
3
starting from aldehyde 6a, which was transformed into Reduction of the azido-group in the more sterically hindered
unsaturated ester 14a on the first stage by Wittig olefination. intermediate 22b was less problematic. Almost no cleavage of
Michael addition of nitroethane to 14a in the presence of DBU 1,2-oxazine ring was observed even when hydrogenation was
furnished the corresponding nitroester 15a as an equimolar performed at elevated hydrogen pressure. Interaction of the
mixture of diastereomers. Nitroesters 15a were transformed resulting amine 23b with triphosgene/Et N afforded the
3
into iodides 17a by the DIBAL-H reduction of ester and desired product 2b in good overall yield of 51% from 22b
subsequent mesylation of alcohols 16a followed by treatment Hydrogenolysis of N
O bond in 1,2-oxazine 2b at elevated
.
−
with NaI in acetone. Smooth cyclization of isomeric iodides 17a temperature and pressure afforded the open-chain derivative
to nitronate 18a was achieved upon the action of DBU in 24b in high yield (Scheme 8).
CH Cl in nearly quantitative yield.
2
2
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 6 of 9
View Article Online
DOI: 10.1039/C8OB01039K
ARTICLE
Journal Name
Scheme 8. Synthesis of target compounds 2a,b and 24b from cyclic nitronates 18a and 18b
In vitro PDE4B1 inhibition studies
2
OCHF fragment. Almost no activity in case of 1b can be
attributed to the sterical hindrance of benzyl-substituted
catechol unit, which does not fit the lipophilic Q pocket of the
enzyme.
All synthesized compounds were tested for their ability to
inhibit FAM-cAMP hydrolysis catalyzed by recombinant human
2
PDE4B1 using a standard in vitro enzymatic assay. Clinically
0
1
a,b
Analysis of 1,2-oxazinoimidazolone series
2 allows to shed
applied PDE4 inhibitor Apremilast
was used as a reference
some light on the effect of the ring B and the substitution in it
compound in these studies. The results are summarized in
Table 1.
on the PDE4B inhibition activity. Thus, the replacement of the
pyrrolidine ring in 1a by 1,2-oxazine cycle results in about
twofold decrease in IC50 value. Addition of two methyl groups
to the 1,2-oxazine ring (position C-2, product 2b) leads to a
dramatic drop of the activity to IC50 1 μM. Opening of 1,2-
oxazine ring results in almost complete disappearance of the
Table 1. IC50 values for synthesized products and reference compounds (PDE4B1)
Compound
IC50, nM
18
rac-1a
(7S,7aR)-1a
17
inhibition activity (IC50 > 10
ꢀ
demonstrates that the presence of ring
M for product 24b). This
may be important for
rac-1b
> 10000
3200
25
B
rac-1c
rac-1d
retaining high potency (cf. with low potent Ro-20-1724, Table
1, Figure 1).
rac-1e
59
rac-2a
39
Surprisingly, introduction of methyl group at the nitrogen
atom of the imidazolidinone ring in 1a resulted in a slight
decrease of activity only (cf. 1a and 1d in Table 1). Substitution
of carboxamide oxygen atom for sulfur led to a more
noticeable drop of inhibitory activity (cf. 1a and 1e in Table 1).
rac-2b
1000
> 10000
18
rac-24b
Apremilast (reference compound)
6a
CMPI
27
ca. 1
ca. 2000
2
1
Roflumilast
Ro-20-1724
7
These results suggest that the N−H hydrogen bond donor in
the imidazolidinone ring is not essential for binding of
inhibitor. From the docking of 1a it can be assumed that C=O
As can be seen from the data in Table 1, racemic compound
rac-1a exhibits higher potency than rac-CMPI, and comparable
to the activity of clinically used PDE4 inhibitor Apremilast (yet
lower than Roflumilast). Single enantiomer (7S,7aR)-1a has
essentially the same activity as the racemate suggesting that
the eudysmic ratio is ca. 1 (unlike CMPI, which enantiomers
2
+
group binds to the metal ion, most likely Mg (Figure 5, (A)).
In the top scoring docking pose of 1a, carbonyl group points
directly onto the magnesium ion with the relatively close
calculated Mg-O distance of 3.0 Å.
Metalloprotein-ligand interactions are rarely observed for
PDE4 inhibitors in crystal structures. A characteristic example
is the complex of PDE4D with Zardaverine (1XOR, Figure 5,
6
a
differ considerably in the inhibitory activity ).
The presence of difluoromethoxy group in ring
C
is essential
(
B)), in which ligand’s carbonyl group occupies one
for high inhibitory activity as can be seen from the IC50 values
2+
coordination site of Zn instead of a water molecule. Metal-
9
of products 1b and 1c bearing OBn and OH groups instead of
6
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 7 of 9
View Article Online
DOI: 10.1039/C8OB01039K
Journal Name
ARTICLE
ligand binding may play an important role in the inhibition of Binding energies were calculated as ꢁG° of substitution of
PDE4 since it would block the access of cAMP to the catalytic water molecule for ligand either on zinc or magnesium ions in
+
23
domain of the enzyme. We were interested in the energy of the optimized complex [CX-1] (Table 2). As can be seen from
this interaction, since it is not adequately estimated by docking this data, oxygen binding ligand 1a’ almost equally favorable
software with nonpolarized molecular force field methods. interacts with either magnesium or zinc centers, yet sulfur
2
+
Therefore, we performed DFT-calculations of binding of binding ligand 1e’ can efficiently interact only with Zn
2
+
, 1e’ (compounds 1a and 1e (coordination of Mg is unfavourable). Overall, the energy of
Zardaverine and ligands 1a’
without the aryl group) to magnesium and zinc ions using a metal-ligand interactions may significantly contribute to the
o
model of the catalytic site of PDE4 (Table 2).
binding of ligands to the active site of PDE4 enzyme (
ꢁG -3 − -
kcal/mol) and this should be taken into account in the design
8
of new PDE4 inhibitors. Using our theoretic model, the energy
of these interactions can be easily calculated by DFT methods.
PDE4 selectivity studies in vitro
PDE4 isoforms have highly conserved catalytic domain, yet
contain multiple differentially spliced exons and unique N-
terminal regions. Classical PDE4 inhibitors typically do not
exhibit any selectivity across PDE4 isoforms. On the other
hand, different intercellular location of these isoforms allows
for modulation of cAMP signaling in specific cells by an
isoform-selective inhibitor.
(
A)
(B)
ASP392PDE4D
O
H2O
H2O
HIS238PDE4D
H
O
O
OH2
N
H2O Mg
Zn
For the most active compounds 1a, 1d, 1e and 2a the
H2O
O
O
selectivity of inhibition on a panel of 9 isotype PDE4 was
examined in comparison with the reference drug Apremilast
and PDE4 inhibitor CMPI (Figure 6). As can be seen from this
data, apremilast does not exhibit any selectivity across the
PDE4 isotype panel, whereas CMPI is twice less active towards
PDE4C enzyme as compared to other isotypes. On the other
NH
N
ASP275PDE4D
N
H
CX-1]+
HIS234PDE4D
[CX-3]+
[
(C)
(D)
Figure 5. Metalloprotein-ligand interactions for PDE4 inhibitors. (A) Docking of
into the catalytic site of PDE4B (close view on the interactions in M pocket). hand, compounds 1a and, especially, 1,2-oxazine 2a show
1a
2+
(
B) Interaction of Zardaverine with the Zn ion in M pocket of PDE4D (crystal
structure 1XOR). (C) DFT model of PDE4 bimetallic catalytic site. (D) Computed more interesting selectivity profiles. Both compounds are
geometry of complex of 1a’ with the model of PDE4 bimetallic catalytic site
2
(MN15L/Def2TZVP, smd, H
information.
O). Large-size images can be found in Supporting significantly less potent for inhibition of PDE4C (2a is ca. ten-
times more active towards PDE4B1 compared to PDE4C1).
Furthermore, somewhat higher activity is observed for PDE4B
Table 2. Calculated binding energies of ligands to the metal ion in CX-1
isotypes as compared to PDE4D isotypes. This is opposite to
the selectivity profile of well-known PDE4 inhibitor Cilomilast,
1
which is more PDE4D-selective. Overall, as can be judged
c
2
,
4
from the Gini coefficients G0.1
compounds 1a and 2a show
a
much better selectivity profiles than Apremilast and CMPI. It
can be expected that one of enantiomers of 1,2-oxazine 2a
may have even higher activity and selectivity index.
Ligand
Calculated ꢁG° of binding to metal, kcal/mol
2+ b 2+ b
Mg (opt complex )
-
Zn (opt complex )
+
Zardaverine
-7.9 [CX-2]
+
+
1
a’
e’
-3.8 [CX-3]
-3.4 [CX-4]
Another interesting point is the higher activity of 1a and,
especially 2a, towards the long isoform PDE4D7 (possessing a
unique N-terminal region) compared to another long isoform
PDE4D3 and short isoform PDE4D2. Although this requires
special studies, we may speculate that some specific
interactions of the ligand with the residues outside the
catalytic site take place in a specific “capped” conformation of
+
+
1
+1.7 [CX-5]
-3.1 [CX-6]
a
b
Defined as calculated ꢁG° of substitution of water molecule for ligand. For
+
DFT-optimized structures of complexes [CX-n] see Supporting information. All
complexes were characterized by only real vibrational frequencies.
The development of a theoretic model of the bimetallic
catalytic site was performed at MN15L/Def2TZVP (smd, H
level of theory using crystallographic structure 1XOR (Figure 5,
B)) as an initial approximation. In our model, histidines HIS164
2
O)
25
PDE4D7.
For N-methylated compound 1d and cyclic thiourea derivative
(
1
e
the selectivity picture is similar to CMPI. This suggests that
the imidazolidinone ring with an N H hydrogen bond donor
may be important for distinguishing PDE4 isoforms.
and HIS200 were changed to imidazole rings, ASP201 and
ASP318 residues were simplified to formiate ions, the bridge
−
2
2
oxygen atom was treated as hydroxyl group, water molecules
were left on the metal ions (Figure 5, (C)). Geometry of the
+
resulting complex [CX-1] was optimized and vibrational
analysis was performed, which revealed no imaginary
frequencies.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 8 of 9
View Article Online
DOI: 10.1039/C8OB01039K
ARTICLE
Journal Name
Conclusions
In conclusion, we developed
a
general route to
difluoromethylated analogs of potent imidazolidinone-based
PDE4 inhibitor CMPI employing efficient C-H functionalizations
of 1,2-oxazine-N-oxides. A series of CMPI analogs possessing
structural modifications in rings
diastereoselective manner.
A
,
B
and
C were synthesized in
Structure-activity/selectivity
relationships throughout these series were studied and
compounds with fine-tuned activity/selectivity profiles
compared to both CMPI and the clinically applied Apremilast
were identified. From these studies, an unusual bicyclic 1,2-
oxazine-based heterocyclic system was suggested as a novel
scaffold for the design of selective and potent PDE4 inhibitors.
Biological studies in combination with docking and DFT
calculations provided important insights into the binding
pattern of bicyclic urea unit to the M pocket of the catalytic
site in PDE4 enzymes. DFT calculations of the energy of metal-
ligand interaction on a model of PDE4 bimetallic active site
were performed. Calculations suggest that metal-ligand
interactions may play a pivotal role in binding to the active site
of PDE4 enzyme.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
The financial support from Russian Foundation for Basic
Research
(grant
17-33-80172_mol_ev)
is
greatly
acknowledged. Authors thank Dr. Yulia Khoroshutina and Dr.
Roman Novikov for taking 2D NMR spectra, Mr. Kevin Kurtz
and Mr. Kenneth D’Arigo from BPS Bioscience for conducting
standard PDE4 inhibition assays on a contract basis.
2
4
Figure 6. PDE4 inhibition selectivity diagrams. G0.1 – Gini coefficient at 0.1 ꢀM
Notes and references
PDE4B inhibition studies in vivo
1
a) A. Gavaldà and R. S. Roberts, Expert Opin. Ther. Pat. 2013,
For two potent compounds 1a and 2a, a PDE4B cell signaling
pathway assay was performed. In vivo PDE4B inhibition activity
was measured using a cell-based CRE luciferase reporter assay
in comparison with Apremilast. In these experiments cells
were co-transfected with PDE4B and CRE luciferase reporter.
Intracellular levels of cAMP were monitored via expression of
luciferase from the CRE reporter, acting as an intracellular
cAMP sensor.
23, 997-1016; b) M. Mazur, J. Karczewski, M. Lodyga, R. Żaba
and Z. Adamski, J. Dermatol. Treat. 2015, 26, 326-328; c) J.
M. Michalski, G. Golden, J. Ikari and S. I. Rennard, Clin.
Pharmacol. Ther. 2012, 91, 134-142; d) A. M. Mulhall, C. A.
Droege, N. E. Ernst, R. J. Panos and M. A. Zafar, Expert Opin.
Investig. Drugs 2015, 24, 1597-1611; e) C. P. Page and D.
Spina, Curr. Opin. Pharm. 2012, 12, 275-286; f) G. Song, X.
Zhu, J. Li, D. Hu, D. Zhao, Y. Liao, J. Lin, L.-H. Zhang and Z.-N.
Cui, Bioorg. Med. Chem., 2017, 25, 5709-5717; g) S.
Guariento, A. Karawajczyk, J. A. Bull, G. Marchini, M. Bielska,
X. Iwanowa, O. Bruno, P. Fossa and F. Giordanetto, Bioorg.
Med. Chem. Lett., 2017, 27, 24-29; h) J.-S. Zhang, X. Liu, J.
Weng, Y.-Q. Guo, Q.-J. Li, A. Ahmed, G.-H. Tang and S. Yin,
Org. Chem. Front., 2017, 4, 170-177; i) B. S. Chinta and B.
Baire, Org. Biomol. Chem., 2017, 15, 5908-5911.
Table 3. EC50 values for compounds 1a and 2a obtained in a cellular assay
Compound
EC50, ꢀM
1
2
a
a
1.4
> 10
2.8
2
3
a) M. D. Houslay, G. S. Baillie and D. H. Maurice, Circul. Res.
2007, 100, 950-966; b) A. Kodimuthali, S. S. L. Jabaris and M.
Pal, J. Med. Chem. 2008, 51, 5471-5489.
a) B. Beghè, K. F. Rabe and L. M. Fabbri, Am. J. Respir. Crit.
Care Med. 2013, 188, 271-278; b) W. M. Brown, Int. J. Chron.
Obstruct. Pulmon. Dis. 2007, 2, 517-533; c) A. Robichaud, P.
Apremilast
The results of these studies (Table 3) revealed that 1a is twice
more potent than Apremilast, yet the 1,2-oxazine-based
compound 2a exhibited poor activity in this experiment.
8
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Please do not adjust margins
Organic & Biomolecular Chemistry
Page 9 of 9
View Article Online
DOI: 10.1039/C8OB01039K
Journal Name
ARTICLE
B. Stamatiou, S. L. C. Jin, N. Lachance, D. MacDonald, F. 19 L. Li, F. Wang, C. Ni and J. Hu, Angew. Chem. Int. Ed. 2013
,
Laliberté, S. Liu, Z. Huang, M. Conti and C.-C. Chan, J. Clin.
Invest. 2002, 110, 1045-1052.
a) J. A. Cherry and R. L. Davis, J. Comp. Neurol. 1999, 407,
52, 12390–12394.
20 P. H. Schafer, A. Parton, L. Capone, D. Cedzik, H. Brady, J. F.
Evans, H. W. Man, G. W. Muller, D. I. Stirling and R. Chopra,
Cell. Signal., 2014, 26, 2016-2029.
4
5
2
87-301; b) X. Miró, S. Pérez-Torres, P. Puigdomènech, J. M.
Palacios and G. Mengod, Synapse 2002, 45, 259-269; c) F. 21 A. Hatzelmann, E. J. Morcillo, G. Lungarella, S. Adnot, S.
Mori, S. Pérez-Torres, R. De Caro, A. Porzionato, V. Macchi, J.
Beleta, A. Gavaldà, J. M. Palacios and G. Mengod, J. Chem.
Neuroanat. 2010, 40, 36-42.
a) L. Carzaniga, G. Amari, A. Rizzi, C. Capaldi, R. De Fanti, E.
Ghidini, G. Villetti, C. Carnini, N. Moretto, F. Facchinetti, P.
Sanjar, R. Beume, C. Schudt and H. Tenor, Pulm. Pharmacol.
Ther., 2010, 23, 235-256.
22 a) C.-G. Zhan and F. Zheng, J. Am. Chem. Soc. 2001, 123,
2835-2838; b) K. Y. Wong and J. Gao, FEBS J. 2011, 278,
2579-2595.
Caruso, G. Marchini, L. Battipaglia, R. Patacchini, V. Cenacchi, 23 V. Sladek and I. Tvaroška, J. Phys. Chem. B, 2017, 121, 6148-
R. Volta, F. Amadei, A. Pappani, S. Capacchi, V. Bagnacani, M. 6162.
Delcanale, P. Puccini, S. Catinella, M. Civelli and E. Armani, J. 24 P. P. Graczyk, J. Med. Chem., 2007, 50, 5773-5779.
Med. Chem. 2017, 60, 10026-10046; b) R. S. Roberts, S. 25 a) J. P. Day, B. Lindsay, T. Riddell, Z. Jiang, R. W. Allcock, A.
Sevilla, M. Ferrer, J. Taltavull, B. Hernández, V. Segarra, J.
Gràcia, M. D. Lehner, A. Gavaldà, M. Andrés, J. Cabedo, D.
Vilella, P. Eichhorn, E. Calama, C. Carcasona and M.
Miralpeix, J. Med. Chem., 2018, 61, 2472-2489.
a) P. A. Zhmurov, A. Y. Sukhorukov, V. I. Chupakhin, Y. V.
Khomutova, S. L. Ioffe and V. A. Tartakovsky, Org. Biomol.
Chem. 2013, 11, 8082-8091; b) A. Y. Sukhorukov, Y. D. Boyko,
S. L. Ioffe, Y. A. Khomutova, Y. V. Nelyubina and V. A.
Tartakovsky, J. Org. Chem., 2011, 76, 7893-7900.
Abraham, S. Sookup, F. Christian, J. Bogum, E. K. Martin, R. L.
Rae, D. Anthony, G. M. Rosair, D. M. Houslay, E. Huston, G. S.
Baillie, E. Klussmann, M. D. Houslay and D. R. Adams, J. Med.
Chem. 2011, 54, 3331-3347; b) D. Fox, A. B. Burgin and M. E.
Gurney, Cell. Signal. 2014, 26, 657-663.
6
7
8
M. F. Brackeen, J. A. Stafford, D. J. Cowan, P. J. Brown, P. L.
Domanico, P. L. Feldman, D. Rose, A. B. Strickland, J. M. Veal
and M. Verghese, J. Med. Chem. 1995, 38, 4848-4854.
a) Y. A. Naumovich, V. E. Buckland, D. A. Sen'ko, Y. V.
Nelyubina, Y. A. Khoroshutina, A. Y. Sukhorukov and S. L.
Ioffe, Org. Biomol. Chem. 2016, 14, 3963-3974; b) A. Y.
Sukhorukov, M. A. Kapatsyna, T. L. T. Yi, H. R. Park, Y. A.
Naumovich, P. A. Zhmurov, Y. A. Khomutova, S. L. Ioffe and
V. A. Tartakovsky, Eur. J. Org. Chem. 2014, 8148-8159; c) P.
A. Zhmurov, Y. A. Khoroshutina, R. A. Novikov, I. S.
Golovanov, A. Y. Sukhorukov and S. L. Ioffe, Chem. - Eur. J.
2017, 23, 4570-4578.
9
1
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. Schlessinger and K. Y. J.
Zhang, Structure 2004, 12, 2233-2247.
0 (a) J. A. Erickson and J. I. McLoughlin, J. Org. Chem., 1995, 60,
1
626-1631; (b) P. Zhou, J. Zou, F. Tian and Z. Shang, J. Chem.
Inf. Model., 2009, 49, 2344-2355; (c) K. Müller, C. Faeh and F.
Diederich, Science, 2007, 317, 1881-1886.
1 O. Trott and A. J. Olson, J. Comput. Chem., 2010, 31, 455-
1
1
4
61.
2 a) A. C. Nirvanappa, C. D. Mohan, S. Rangappa, H. Ananda, A.
Y. Sukhorukov, M. K. Shanmugam, M. S. Sundaram, S. C.
Nayaka, K. S. Girish, A. Chinnathambi, M. E. Zayed, S. A.
Alharbi, G. Sethi, Basappa and K. S. Rangappa, PLoS One
2016, 11, e0163209; b) A. Y. Sukhorukov, A. C. Nirvanappa, J.
Swamy, S. L. Ioffe, S. Nanjunda Swamy, Basappa and K. S.
Rangappa, Bioorg. Med. Chem. Lett. 2014, 24, 3618-3621.
3 A. Y. Sukhorukov and S. L. Ioffe, Chem. Rev. 2011, 111, 5004-
1
1
5
4 S. Zheng, G. Kaur, H. Wang, M. Li, M. Macnaughtan, X. Yang,
041.
S. Reid, J. Prestegard, B. Wang and H. Ke, J. Med. Chem.,
2008, 51, 7673-7688.
1
1
5 For the mechanism see: S. E. Denmark, B. S. Kesler and Y. C.
Moon, J. Org. Chem., 1992, 57, 4912-4924.
6 The (4S,6S,7S,8R)-configuration of major isomer 7c was
reasonably proposed based on our previous studies on [4+2]-
cycloadditions with vinyl ether 8.
6
b
1
1
1
7 In the H NMR spectra, the characteristic H-6 proton was
absent indicating that the acetal unit at C-6 is cleaved.
8 Slight epimerization of C-6 center was observed in the case
of enantiopure nitrate 10c (d.r. 4.5 : 1), yet the epimer could
be separated on the next stages.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins