PDE2 inhibition by the PI3 kinase inhibitor LY294002 and analogues

Article abstract of DOI:10.1016/j.bmcl.2004.03.043

Synthetic 2-morpholinochromones, including the known PI3-kinase inhibitor LY294002, have been evaluated in vitro as inhibitors of isolated human platelet phosphodiesterases. Inhibition of the cAMP-phosphodiesterases, PDE2 and PDE3 by LY294002 is reported

Full text of DOI:10.1016/j.bmcl.2004.03.043

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2847–2851
PDE2 inhibition by the PI3 kinase inhibitor LY294002 and
analogues
a
Belinda M. Abbott and Philip E. Thompson
b,
*
a
Department of Medicine, Monash University, Box Hill Hospital, Box Hill 3128, Australia
Kinacia Pty. Ltd, Level 5, Clive Ward Centre, Box Hill Hospital, Box Hill 3128, Australia
b
Received 30 January 2004; accepted 16 March 2004
Abstract—Synthetic 2-morpholinochromones, including the known PI3-kinase inhibitor LY294002, have been evaluated in vitro as
inhibitors of isolated human platelet phosphodiesterases. Inhibition of the cAMP-phosphodiesterases, PDE2and PDE3 by
LY294002 is reported for the first time. Preliminary screening across a range of 2-morpholinochromones has revealed structural
features for optimised PDE2inhibition.
Ó 2004 Elsevier Ltd. All rights reserved.
Cyclic nucleotide phosphodiesterases (PDEs) are
important regulators of cellular functions, catalysing the
hydrolysis and deactivation of the second messengers
cyclic adenosine monophosphate (cAMP) and cyclic
cAMP–PDE3 inhibitors (e.g., cilostazol and milrinone)
and cAMP–PDE4 inhibitors (e.g., rolipram). While
further progress in these areas is fuelled by the phar-
macological characterisation of these inhibitors, the
understanding of those isoforms for which selective
inhibitors are not known has tended to languish. For
example, relatively little is known about PDE2, which is
widely expressed in human tissues and hydrolyses both
cAMP and cGMP. This activity may be significant as
PDE2has contrasting enzyme kinetics to other PDEs
1
guanosine monophosphate (cGMP). The regulation of
cyclic nucleotide levels by PDEs is very complex with
most cells utilising multiple PDE isoforms to effect
hydrolysis. The activities of the isoforms themselves are
subject to regulation by varied expression levels, sub-
cellular compartmentalisation, activation and inhibi-
tion. Activity is further subject to cross-isoform
communication, with cGMP and cAMP competing for
catalytic and allosteric sites on specific isoforms, which
determines the mode and extent of nucleotide hydroly-
sis. In this way, PDEs are active contributors to cell
signalling pathways, regulating intracellular distribution
of cyclic nucleotides and therefore downstream activa-
tion of specific protein kinases.
3
and is also allosterically activated by cGMP. Thus,
PDE2may play a key role in regulating cyclic nucleotide
levels under conditions where other PDEs are ineffec-
tive.
Progress in the pharmacological study of PDE2has
been hampered by the lack of a selective inhibitor. Only
EHNA (erythro-9-(2-hydroxy-3-nonyl)adenine) (Fig. 1)
has been described as exhibiting PDE2-isozyme
The understanding of phosphodiesterases has been
accelerated in recent years by the advances of molecular
cloning and the concomitant development of isoform
selective inhibitors, most famously by the cGMP–
PDE5 selective inhibitor, sildenafil, but also with
COOH
2
F
NH
2
N
N
N
N
H
O
Keywords: Phosphodiesterase; Inhibitor; PDE2; LY294002; 2-Mor-
pholinochromone.
HO
S
O
*
Corresponding author at present address: Department of Medicinal
Chemistry, Victorian College of Pharmacy, Monash University, Park-
ville 3052, Australia. Tel.: +61-3-9903-9672; fax: +61-3-9903-9582;
e-mail: phil.thompson@vcp.monash.edu.au
EHNA
Sulindac sulfone
Exisulind)
(
Figure 1. Structures of known inhibitors of PDE2.
0
960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.03.043

2848
B. M. Abbott, P. E. Thompson / Bioorg. Med. Chem. Lett. 14 (2004) 2847–2851
O
O
O
O
covering diseases such as cancer, scleroderma and cog-
6
nitive disorders. However, the broader selectivity or
otherwise of these analogues has not yet been described.
N
O
N
O
O
Our own interest in PDE2began with our structure–
function studies of the phosphatidylinositol 3-kinase
7
(PI3K) inhibitor, LY294002 (1) (Fig. 2). We were aware
that other synthetic 2-morpholinochromones had been
described as inhibitors of platelet cAMP phosphodies-
terase, in particular PDE3 as reported for U84569 (Fig.
1
(LY294002)
U84569
8
Figure 2. Structure of the 2-morpholinochromones LY294002 1 and
U84569.
2), so we prepared a library of analogues with a view to
studying the specificities of 2-morpholinochromones for
PI3K and PDE3. As the procedure for assaying PDE3
activity demanded separation of PDE2and PDE3 from
4
selectivity with moderate potency. However, EHNA is
9
also a potent inhibitor of adenosine deaminase, an
activity, which compromises its use in ascribing physi-
ological roles to PDE2. A new contributor to our
understanding of PDE2function is the antitumour
agent sulindac sulfone (exisulind) (Fig. 1), which
appears to act by inhibition of both PDE5 and PDE2.
A number of sulindac sulfone and EHNA analogues
have been the subject of recent patent applications
platelet lysate, we were also able to evaluate these
compounds as potential inhibitors of PDE2. The sur-
prising results of that study are presented here.
The library of 2-morpholinochromones synthesised is
summarised in Table 1 and the general synthetic routes
used to prepare them are outlined in Scheme 1. The
synthesis of 2-morpholinochromone analogues was
5
Table 1. Molecular structures and PDE2inhibition of the 2- morpholinochromones
O
R6
R7
O
N
R8
O
Compd
R
6
R
7
R
8
Ref.
PDE2inhibition (% basal)
a
at 50 lM
1
2
3
4
5
6
7
8
9
H
H
Ph
10
10
54 ꢀ 3
43 ꢀ 3
43 ꢀ 2
47 ꢀ 5
47 ꢀ 2
62 ꢀ 7
75 ꢀ 3
60 ꢀ 6
47 ꢀ 3
75 ꢀ 2
78 ꢀ 2
87 ꢀ 2
67 ꢀ 6
69 ꢀ 4
ꢀ 6
H
H
(4-F)Ph
(2-CF )Ph
)Ph
O)Ph
H
H
3
10
H
H
(2-CH
(2-CH
3
10
10
H
H
3
H
H
(2-Cl)Ph
(4-PhO)Ph
Ph
10
10
H
H
Me
Et
H
10
10
H
Ph
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
OMe
OMe
OMe
OMe
OH
OMe
H
H
10
13
H
(2-Cl)Ph
(3-Cl)Ph
(4-Cl)Ph
Ph
H
13
13
H
H
13
H
H
1238
10,122
10,1219
1240
1235
1293
1243
11,1238
11,1235
11
H
OMe
OH
0
ꢀ 8
H
H
ꢀ 11
H
H
OCH
OCH
2
Ph
(3-pyridyl)
ꢀ 4
H
H
2
ꢀ 4
H
H
CBCPh
ꢀ 1
H
H
(CH
Me
Me
Me
Me
Me
H
2
)
2
Ph
ꢀ 7
H
OCH
OCH
2
Ph
(3-pyridyl)
ꢀ 4
H
2
ꢀ 2
H
(CH
Ph
2
)
2
(4-Me-1-piperazinyl)
ꢁ2 ꢀ 6
31ꢀ 14
ꢀ 1
H
12
H
CBCPh
1244
1212
122
13
H
OCH
OCH
Ph
2
Ph
(3-pyridyl)
ꢀ 3
H
2
H
2
ꢀ 17
H
H
9 ꢀ 7
92 ꢀ 1
H
CBCPh
H
13
a
All values are mean ꢀ SE (n P 3).

B. M. Abbott, P. E. Thompson / Bioorg. Med. Chem. Lett. 14 (2004) 2847–2851
2849
O
O
R
X
R
X
OMe
OH
OH
a
b
O
O
R
N
O
OH
X
X = OTf, Br
c
R = H, Me, Et, OMe
O
R
X
O
N
O
Method A
Method C
O
O
O
O
R
Method B
R
O
N
O
N
O
Ar
R
R'
O
O
N
O
Ph
2 2
Scheme 1. Reagents: (a) LiHMDS, 4-morpholinecarbonyl chloride, THF; (b) LDA, 4-acetylmorpholine, THF; (c) triflic anhydride, CH Cl .
1
0–13
performed as described previously. In brief, a range
of bromo or triflate substituted salicylacetamides were
fractions was determined by assays under optimal con-
ditions for the two isozymes with measurement of the
15
prepared by condensation of appropriately substituted
0
AMP product quantified by reverse phase HPLC. The
activity of each isoform was confirmed by the use of
selective isoform inhibitors, namely EHNA for PDE2
2
-hydroxyacetophenones with 4-morpholinecarbonyl
chloride. Alternatively, salicylate esters were condensed
with 4-acetylmorpholine. Cyclodehydration yields the
corresponding 2-morpholinochromones. Both the tri-
flate and bromo precursors are amenable to Suzuki
16
and milrinone for PDE3. The results are shown in
Figures 3 and 4. LY294002 was shown to inhibit the
phosphodiesterase activities of both isozymes in a dose
dependent manner, but was a significantly more potent
inhibitor of PDE2, with an IC50 of 40 lM compared
with an IC50 of 100 lM for PDE3. Under these condi-
tions EHNA was found to have an IC50 of 5 lM versus
10
cross-coupling reactions (Method A) yielding the bi-
14
aryl products (1–13, 25 and 29). Demethylation of (10)
using AlCl /NaI gave the hydroxy compound (14). The
key triflate intermediates used for Suzuki couplings were
also derivatised by the Sonagashira reaction (Method
B) to yield the alkyne analogues (20, 26 and 30).
Hydrogenation of (20) gave the phenethyl substituted
compound (21). Alternatively, the triflates were readily
hydrolysed to the corresponding phenols and alkylated
3
12
12
100%
7
5
2
5%
0%
5%
(
Method C), to yield benzyloxy compounds (18, 22, 27)
12
or 3-pyridylmethoxy compounds (19, 23, 28). The
intermediate 7-methyl-8-hydroxy-2-morpholinochro-
mone was also treated with dibromoethane to give a
bromoethoxy adduct, which in turn was treated with N-
methylpiperazine to give the potent antiplatelet com-
11
pound (24).
0
%
In the first instance, our interest was in measuring
inhibition of phosphodiesterase activity by LY294002.
The isozymes were purified from human platelet lysate
by anion exchange chromatography according to the
0
.1
1
10
100
1000
Concentration (µM)
Figure 3. Inhibition of PDE2activity by EHNA ðjÞ and LY294002
ðNÞ.
9
general procedure of Dickinson et al. Activity of pooled

2850
B. M. Abbott, P. E. Thompson / Bioorg. Med. Chem. Lett. 14 (2004) 2847–2851
1
00%
125%
7
5%
0%
5%
100%
75%
50%
5
2
25%
0
%
0
.01
0.1
1
10
100
1000
Concentration (µM)
0%
0.1
1
10
100
1000
Figure 4. Inhibition of PDE3 activity by milrinone ðjÞ and LY294002
ðNÞ.
Concentration (µM)
Figure 5. Inhibition of PDE2activity compounds 12 ( ) and 20 ðꢂÞ.
PDE2while milrinone had an IC 50 of 0.6 lM versus
PDE3, consistent with the reported literature values.
16
This is the first report of phosphodiesterase inhibition
by LY294002.
the effects it might be expected to have on the cGMP-
inhibited phosphodiesterase (PDE3). In the event,
however, we found that LY294002 was a poor inhibitor
of PDE3 but a moderate inhibitor of PDE2. The IC50 of
approximately 40 lM would not generally be enough to
compromise the existing literature relating to LY294002
but in some experimental settings LY294002 has been
used at high concentrations, in which case caution may
be required in interpreting results.
Having detected this level of potency against PDE2by
LY294002, we decided to explore the activity of the 2-
morpholinochromone structural class more thoroughly.
We screened the compounds of our analogue library at a
single concentration of 50 lM. The results of this screen
are shown in Table 1. A number of compounds showed
strong inhibition of PDE2activity. Of the LY 29 4002
analogues, which were substituted on the pendant aryl
ring, the bulky para-phenoxy substituent (7) showed
There is some precedent for the overlap of kinase and
phosphodiesterase activity beyond 2-morpholinochro-
mones. Natural product flavonoids, such as quercetin,
and purines, such as theophylline and caffeine, have
7
5% inhibition of PDE2activity while the introduction
of ortho-substituents on the pendant aryl ring (com-
pounds 3–6) was generally benign. Improved levels of
inhibition were generally observed across the range of
LY294002 derivatives bearing substituents at the 6-po-
sition, with the 6-methoxy-8-(3-chloro)phenyl analogue
20–22
been reported to inhibit both PI3Ks and PDEs.
The
planarity of the purine and chromone ring structures
would appear to be a fundamental feature of this
overlapping activity. As a great deal of effort is being put
in to determining the ‘kinase selectivity’ of many new
(
12) effectively abolishing activity at 50 lM. Both the 8-
18;23
and 7-phenylethynyl substituted compounds (20 and 30,
respectively), had the same marked effect of near com-
plete inhibition. The remaining 6-, 7-, 8- and 8-methyl-7-
substituted compounds showed inhibitory potency
inferior to that of LY294002.
inhibitors,
it would seem prudent to include other
nucleoside binding targets such as phosphodiesterases
and purine receptors in such screens. Indeed, as in this
case, kinase-inhibitor libraries might be an excellent
resource for discovering new ligands for phosphodiester-
ases.
The results from the preliminary screen prompted us to
give attention to two of the more potent compounds.
The 6-methoxy-8-(3-chloro)phenyl (12) and 8-phenyle-
thynyl (20) compounds were both found to inhibit
PDE2in a dose dependent manner (Fig. 5). Compound
The inhibitory activity of LY294002 and the analogues
shown in this study may prove to be an excellent lead for
the development of PDE2selective ligands. The results
of the screen show three distinct points at which elab-
oration of the 2-morpholinochromone structure might
improve upon inhibitor potency. Firstly, inclusion of a
phenethynyl substituent yielded the two most potent
inhibitors of PDE2of this study, with compound 20
confirmed to be equipotent with EHNA. Furthermore,
addition of a 6-methoxy substituent to the 2-morpholi-
nochromone backbone of LY294002 was found to
enhance activity against PDE2, as was substitution on
the pendant aryl ring of LY294002. The inclusions of
para-phenoxy and meta-chloro substituents on the
pendant aryl ring were of particular interest. Our pre-
liminary studies suggest these modifications are benign
or deleterious to PDE3 and PI3K potency, and studies
to determine the broader selectivity of these compounds
are underway.
2
0 was found to be equipotent with EHNA in this assay,
with an IC50 of 6 lM, while for compound 12 an IC50 of
lM was determined.
9
While LY294002 has been widely used as one of the
tools to delineate the role that phosphatidylinositol
3
-kinases play in the regulation of intracellular signalling
pathways, other properties have been reported for this
compound in more recent times. These include inhibi-
tion across the broad PI3K-like kinase family, including
the serine/threonine kinases DNA-dependent protein
17
kinase (DNA-PK), as well as other kinases such as
creatine kinase 2(CK 2) . Most recently, LY294002 has
been reported to be a potent antagonist of the oestrogen
18
19
receptor. Our own concerns with LY294002 related to

B. M. Abbott, P. E. Thompson / Bioorg. Med. Chem. Lett. 14 (2004) 2847–2851
2851
These studies reveal a new structural motif for inhibitors
of PDE2. We have shown that inhibition of PDE2 can
be improved by relatively simple substitutions and that
these compounds display some discrimination between
PDE2and PDE3. Realising the potential of these
ligands rests with optimisation of phosphodiesterase
activity at the expense of PI3K and other potential
activities. Such a goal may yet be difficult to achieve, but
the addition of a new small molecule template will
greatly assist our understanding of PDE2-inhibitor
interactions and lead the way to the development of
therapeutic opportunities across a range of diseases.
12. Gammill, R. B.; Judge, T. M.; Morris, J. WO Patent
06921, 1990.
0
3. All previously unreported compounds were characterised
1
1
1
by NMR and high resolution ESI-MS (>90% purity and
þ
correct (MH) was seen).
4. Horie, T.; Kobayashi, T.; Kawamura, Y.; Yoshida, I.;
Tominaga, H.; Yamashita, K. Bull. Chem. Soc. Jpn. 1995,
6
8, 2033.
5. Assay conditions: A typical PDE2assay consisted of 70 lL
of HEPES buffer to which 0.5 lL DMSO or inhibitor, 1 lL
1 mM cGMP (final concentration 10 lM) and 20 lL PDE2
were added followed by mixing in an incubator at 37 °C
for 90 s. The enzymatic reaction was initiated by the
addition of 10 lL 300 lM cAMP (final concentration
3
0 lM) and the reaction terminated after 5 min by placing
Acknowledgements
the reaction vessel in an oven at 100 °C for 3 min. Analysis
was carried out using reverse phase HPLC on an Agilent
1100 HPLC system, in an adaptation of the method of
Ogata, S.; Suzuki, A.; Kawase, T. Jpn. J. Oral. Biol. 1995,
The financial support of the Australian Research
Council (Australia Postgraduate Award) and Monash
University (Postgraduate Publications Award) is
acknowledged.
3
7, 1. A 50 lL aliquot of the assay reaction was injected
onto an Waters Xterra RP C18 column equilibrated at 1%
acetonitrile in 0.2M (NH )H PO aq at 0.3 mL/min flow
4
2
4
rate. The elution gradient program was (a) 1% acetonitrile
isocratic for 3 min, (b) 1–10% acetonitrile linear gradient
over 5 min, (c) 10% acetonitrile isocratic for a further
2min and (d) return to the starting conditions over 1 min
and re-equilibration for 10 min. Peak detection was carried
out at 254 nm and compared to AMP and cAMP
standards. Phosphodiesterase activity was determined as
the peak area of AMP (mAU) measurement as a percent-
age of the DMSO control (where the control without any
inhibitor present indicates maximal AMP production
under the assay conditions). Unless otherwise stated, all
assays were completed at least in triplicate. Data is
presented as the mean ꢀ the SE. IC is defined as the
References and notes
1
. Manganiello, V. C.; Degerman, E. Thromb. Haemost.
999, 82, 407; Beavo, J. A.; Brunton, L. L. Nature Rev.
Mol. Cell Biol. 2002, 3, 710.
. Perry, M. J.; Higgs, G. A. Curr. Opin. Chem. Biol. 1998, 2,
4
. Mehats, C.; Anderson, C. B.; Filopanti, M.; Jin, S.-L. C.;
Conti, M. Trends Endocrin. Metabol. 2002, 13, 2 9.
. Podzuweit, T.; Nennstiel, P.; M u€ ller, A. Cell. Signal. 1995,
1
2
3
4
5
72.
7, 733.
. Thompson, W. J.; Piazza, G. A.; Li, H.; Liu, L.; Fetter, J.;
Zhu, B.; Speri, G.; Ahnen, D.; Pamukcu, R. Cancer Res.
5
0
concentration of inhibitor where 50% of the AMP
production has been inhibited as compared to the DMSO
control.
2
000, 60, 3338.
6
. Niew o€ hner, U.; Schauss, D., Hendrix, M.; K o€ nig, G.;
B o€ ss, F.-G.; van der Staay, F.-J.; Schreiber, R.; Schlem-
mer, K.-H.; Grosser, R. WO Patent 050078, 2002;
Whitehead, C. M.; Earle, K. A.; Alila, H. W.; Thompson,
W. J. WO Patent 017999, 2003.
. Vlahos, C. J.; Matter, W. F.; Hui, K. Y.; Brown, R. F.
J. Biol. Chem. 1994, 269, 5241.
. Benjamin, C. W.; Lin, A. H.; Morris, J.; Wishka, D. G.;
Gorman, R. R. J. Pharmacol. Exp. Ther. 1993, 265, 457.
. Dickinson, N.; Jang, E. K.; Haslam, R. J. Biochem.
J. 1997, 323, 371.
16. Sudo, T.; Tachibana, K.; Toga, K.; Tochozawa, S.; Inoue,
Y.; Kimura, Y.; Hiroyoshi, H. Biochem. Pharmacol. 2000,
59, 347.
17. Izzard, R. A.; Jackson, S. P.; Smith, G. C. M. Cancer Res.
1999, 59, 2581.
18. Davies, S. P.; Reddy, H.; Caivano, M.; Cohen, P.
Biochem. J. 2000, 351, 91.
19. Pasapera Limon, A. M.; Herrera-Munoz, J.; Gutierrez-
Sagal, R.; Ulloa-Aguirre, A. Mol. Cell. Endocrinol. 2003,
200, 199.
7
8
9
1
1
0. Abbott, B.; Thompson, P. Aust. J. Chem. 2003, 56, 1099–
106.
20. Kuppusamy, U. R.; Das, N. P. Biochem. Pharmacol. 1992,
44, 1307.
21. Matter, W. F.; Brown, R. F.; Vlahos, C. J. Biochem.
Biophys. Res. Commun. 1992, 186, 624.
22. Foukas, L. C.; Daniele, N.; Ktori, C.; Anderson, K. E.;
Jensen, J.; Shepherd, P. R. J. Biol. Chem. 2002, 277, 37124.
23. Bain, J.; McLauchlan, H.; Elliot, M.; Cohen, P. Biochem.
J. 2003, 371, 199.
1
1. Morris, J.; Wishka, D. G.; Lin, A. H.; Humphrey, W. R.;
Wiltse, A. L.; Gammill, R. B.; Judge, T. M.; Bisaha, S. N.;
Olds, N. L.; Jacob, C. S.; Bergh, C. L.; Cudahy, M. M.;
Williams, D. J.; Nishizawa, E. E.; Thomas, E. W.;
Gorman, R. R.; Benjamin, C. W.; Shebuski, R. J.
J. Med. Chem. 1993, 36, 2026.