1,2,4-Oxadiazolidin-3,5-diones and 1,3,5-triazin-2,4,6-triones as cytosolic phospholipase A2α inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.02.067

1,2,4-Oxadiazolidin-3,5-dione and 1,3,5-triazin-2,4,6-trione scaffolds were employed as templates to incorporate the pharmacophore requirements of cytosolic phospholipase A₂α substrate mimetics. Inhibitors that are active in both enzyme, and ce

Full text of DOI:10.1016/j.bmcl.2006.02.067

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2978–2981
1,2,4-Oxadiazolidin-3,5-diones and 1,3,5-triazin-2,4,6-triones
as cytosolic phospholipase A2a inhibitors
Ariamala Gopalsamy,a,* Hui Yang,a John W. Ellingboe,a John C. McKew,b Steve Tam,b
Diane Joseph-McCarthy,b Wen Zhang,c Marina Shenc and James D. Clarkc
aChemical and Screening Sciences, Wyeth Research, Pearl River, NY 10965, USA
bChemical and Screening Sciences, Wyeth Research, Cambridge, MA 02140, USA
cInflammation Research, Wyeth Research, Cambridge, MA 02140, USA
Received 10 January 2006; revised 22 February 2006; accepted 24 February 2006
Available online 20 March 2006
Abstract—1,2,4-Oxadiazolidin-3,5-dione and 1,3,5-triazin-2,4,6-trione scaffolds were employed as templates to incorporate the phar-
macophore requirements of cytosolic phospholipase A2a substrate mimetics. Inhibitors that are active in both enzyme, and cell-
based assays were identified from both classes. From the SAR work carried out and modeling efforts around these templates,
the triazinetrione scaffold with an additional substitution point was found to be more favorable.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Leukotrienes and prostaglandins are important media-
tors of inflammation, generated via the 5-lipoxygenase
and cyclooxygenase pathways, respectively. Prostaglan-
dins and leukotrienes are unstable and are not stored
in cells, but are instead synthesized1 from arachidonic
acid in response to stimuli. Arachidonic acid, which is
fed into these two distinct inflammatory pathways, is
released from the sn-2 position of membrane phospho-
lipids by cytosolic phospholipase A2a2 (cPLA2a, a
group IVA phospholipase). When the phospholipid sub-
strate of PLA2 is phosphatidylcholine with an ether link-
age in the sn-1 position, the lysophospholipid produced
is the immediate precursor of platelet-activating factor
(PAF), another potent mediator of inflammation.3
ease states including osteoarthritis, rheumatoid arthritis,
and asthma.4
Recent reports5 from our laboratory disclosed our efforts
in this area that identified indoles 2 as cPLA2a inhibitors.
A substrate-based approach was taken using zafirlukast
16 as a starting point since it was designed to be an arachi-
donic acid mimetic with a template that supports an acid-
ic and a lipophilic group. In continuation of this project,
we were evaluating alternate templates that can be
utilized to package the required pharmacophore. In this
report, we describe the identification of two alternate
scaffolds—1,2,4-oxadiazolidin-3,5-dione 3 and 1,3,5-tria-
zin-2,4,6-trione 4 (Figs. 1 and 2).
The compounds of interest could be readily prepared7 as
described in Schemes 1 and 2 . Benzophenone oxime 6
was reduced to the hydroxyl amine intermediate 7,
which was cyclized with chlorocarbonylisocyanate to
give the scaffold oxadiazolidinedione 8 with the benzhy-
dryl lipophilic group. Mitsunobu reaction with the
required alcohol bearing the appropriate acid region
precursor afforded the aldehyde 10, which was oxidized
to give the final target 11. Triazin-2,4,6-trione analogs
were synthesized starting from benzhydrylamine 12
and the isocyanate of interest to give the urea 14. Cycli-
zation of 14 to triazinetrione 15 followed by alkylation
afforded the aldehyde, which was oxidized to give the
final target acid 17.
Most anti-inflammatory therapies have focused on pre-
venting production of either prostaglandins or leukotri-
enes, but not both of them. It is highly desirable to
identify compounds that inhibit the actions of cPLA2a
for use in treating or preventing inflammatory condi-
tions, particularly where these classes of lipid mediators
might have an additive or synergistic pro-inflammatory
effect. Consequently, the direct inhibition of the activity
of cPLA2a has been suggested as a useful mechanism for
a therapeutic agent with application in number of dis-
Keywords: 1,2,4-Oxadiazolidin-3,5-dione; 1,3,5-Triazin-2,4,6-trione;
Cytosolic phospholipase A2a inhibitors.
*
Corresponding author. E-mail: gopalsa@wyeth.com
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.02.067

A. Gopalsamy et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2978–2981
2979
Figure 1. LTD4 receptor antagonist Zafirlukast (1) and cPLA2a
substrate mimetic inhibitor (2).
Scheme 2. Reagents and conditions: (a) CH2Cl2, rt, 12 h, 90%; (b)
ClCONCO, THF, 0 ꢁC to rt, 3 h, 78%; (c) NaH, THF, 16, rt, 18 h,
67%; (d) sulfamic acid, NaClO2, THF–H2O, rt, 3 h, 81%.
methylene unit as in compound 22 led to a slight loss
in activity, but further extension (compounds 23) was
not desirable. Introduction of unsaturation (compounds
24–25) diminished the activity as well (see Table 1).
Figure 2. Template modification for cPLA2a substrate mimetic 2.
To explore the feasibility of scaffold exchange in 2,
oxadiazolidin-3,5-dione analog 11 was synthesized
retaining the benzhydryl moiety for the lipophilic region
and the benzoic acid portion of the molecule. While this
direct analog was not potent, a small change in the lipo-
philic group by introducing a methylene spacer between
the ring and the benzhydryl group afforded 18 with a
significant boost in enzyme activity. However, analog
19 with a more flexible lipophilic moiety was inactive.
At this stage we decided to maintain the lipophilic
region as diphenylethyl group and systematically vary
the acid region. Moving the acid to a 3-benzoic acid
derivative as in analog 20 was unfavorable. Adding an
additional 3-hydroxyl substitution to the 4-benzoic
acid (analog 21) led to a 4-fold loss in activity. Chain
extension between the scaffold and benzoic acid by one
Table 1. 1,2,4-Oxadiazolidin-3,5-diones
Compd
R1
R2
Glu-Mic
IC50 (lM)8
2
Reference compound
3
O
11
>50
Ph Ph
Ph Ph
COOH
O
O
18
19
20
21
22
23
24
25
6.7
>50
>50
27
COOH
COOH
COOH
Ph Ph
O
O
Ph Ph
Ph Ph
Ph Ph
Ph Ph
Ph Ph
Ph Ph
OH
COOH
O
2
12.4
>50
>50
23
COOH
COOH
O
3
O
O
COOH
Scheme 1. Reagents and conditions: (a) NH2OH, pyridine, ethanol,
60 ꢁC, 3 h, 89%; (b) BH3-Py, MeOH, 10% HCl (pH 2), rt, 18 h, 68%;
(c) ClCONCO, THF, 0 ꢁC to rt, 3 h, 80%; (d) 9, Ph3P, DEAD, THF,
rt, 8 h, 68%; (e) sulfamic acid, NaClO2, THF–H2O, rt, 3 h, 86%.
COOH

2980
A. Gopalsamy et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2978–2981
A similar approach was undertaken to evaluate the sec-
ond scaffold of interest—triazine-2,4,6-trione. However,
this scaffold has the potential to carry three different sub-
stituents. For direct comparison to 2, analogs were pre-
pared retaining the benzhydryl group for the lipophilic
region and the ethoxy-4-benzoic acid for the acid mimetic.
The third substituent was varied from a small ethyl group
to a bulky benzyl group (compounds 26–29). It was inter-
esting to notice that the compounds were active and the
activity range was within a 2-fold variation, indicating
that while the third substituent gives a boost in activity,
it does not show any significant preference for another
bulky group. However, a rigid phenyl group directly at-
tached to the scaffold, 30 was inactive.
2,4,6-triones, chain extension was well tolerated and
the butenyloxy benzoic acid chain (compound 32) was
found to be slightly more potent than the ethoxy benzoic
acid group. As we observed before, with this acid chain,
the size of the third substituent did not play a major role
since compounds 32–35 showed similar potency (see
Table 2).
One of the most potent analogs 35 from this effort was
further characterized in secondary cell-based assays.9
The compound was found to be cell permeable and
was able to inhibit both LTB4 (IC50 = 0.5 lM) and
PGF2-a (IC50 = 0.8 lM) synthesis in a mast cell line
(MC-9 assay). In addition, the compound no longer
inhibited PGF2-a synthesis when exogenous arachidonic
acid was added to bypass cPLA2a indicating that the
compound is also selective for the target in this cell-
based assay. The compound was also active (IC50
TXB2: 5.2 lM) in the ionophore stimulated rat whole
blood assay (COX-1-dependent assay).
Unlike the oxadiazolidin-3,5-dione scaffold in the case
of triazin-2,4,6-trione, installing diphenylethyl group
(compound 31) for the lipophilic region did not improve
the enzyme potency rather rendered the molecule 10-
fold less active. Probably the rigid benzhydryl group is
well accommodated in the binding pocket since this scaf-
fold has a larger six-membered ring. Also, with triazin-
Modeling efforts were undertaken to validate the scaf-
fold tolerance. Towards this end, compound 35 and a
series of triazine analogs were docked into the X-ray
structure of cPLA2a10 via Monte Carlo simulation11.
Figure 3 shows a model of compound 35 bound to
cPLA2a with a partial molecular surface (colored by
element) of the active site region. The acid interacts with
the backbone amide nitrogens of Gly 197 and Gly 198,
which are in an approximate cis conformation thought
to stabilize the developing oxyanion during ester hydro-
lysis, and also makes a weak interaction with Ser 228
Table 2. 1,3,5-Triazin-2,4,6-triones
Compd R1
R2
R3
Glu-Mic
IC50 (lM)8
2
Reference compound
3
˚
(4.0 A to the hydroxyl oxygen), the nucleophile that
attacks the sn-2 ester12. The linker oxygen makes a
hydrogen bond with the backbone amide nitrogen of
Ala 578 closer to the opening of the pocket.
O
O
O
O
O
O
26
Ethyl
7.5
5
Ph Ph
COOH
COOH
COOH
COOH
COOH
COOH
27
28
29
30
31
32
33
34
35
Allyl
Ph Ph
Ph Ph
Ph Ph
Ph Ph
n-Butyl
Benzyl
5.6
3.5
Phenyl >50
Benzyl
Ethyl
30
Ph Ph
Ph Ph
O
O
O
O
1.6
COOH
Allyl
1.3
2.4
1.3
Ph Ph
Ph Ph
Ph Ph
COOH
COOH
COOH
n-Butyl
Benzyl
Figure 3. Interaction of 1,3,5-triazin-2,4,6-trione inhibitor 35 with
cPLA2a protein structure. The figure was generated using PyMOL
(DeLano Scientific, 2002, San Carlos, CA, USA).

A. Gopalsamy et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2978–2981
2981
Reaction completion was monitored by LC–MS and the
solvent was removed in Savant evaporator and com-
pounds were purified by HPLC and the purity was >90%.
LC conditions: HP 1100, 23 ꢁC, 10 lL injected; column:
YMC-ODS-A 4.6 · 50 mm 5l; gradient A: 0.05% TFA/
water, gradient B: 0.05% TFA/acetonitrile; time: 0 and
1 min: 98% A and 2% B: 7 min: 10% A and 90% B; 8 min:
10% A and 90% B; 8.9 min: 98% A and 2% B; post-time
1 min; flow rate 2.5 mL/min; detection: 215 and 254 nm,
DAD. Semi-Prep HPLC: Gilson with Unipoint software;
Column: Phenomenex C18 Luna 21.6 mm · 60 mm, 5 l;
solvent A: water (0.02% TFA buffer); solvent B: acetoni-
trile (0.02% TFA buffer); solvent gradient: time 0: 5% B;
2.5 min: 5% B; 12 min: 95% B; hold 95% B 3 min; flow
rate: 22.5 mL/min; detection: 215 and 254 nm. (b) 1H
NMR data for compound 35: (DMSO) d 7.85 (d, 2H),
7.26–7.36 (m, 15H), 7.25 (s, 1H), 7.0 (d, 2H), 5.88 (br s,
2H), 4.92 (s, 2H), 4.6 (s, 2H), 4.39 (s, 2H).
In conclusion, we have explored the feasibility of
pharmacophore repackaging with two different novel
scaffolds—1,2,4-oxadiazolidin-3,5-dione and 1,3,5-tria-
zin-2,4,6-trione. While 1,2,4-oxadiazolidin-3,5-dione
provided modest inhibitors, 1,3,5-triazin-2,4,6-trione
with an additional substituent to pick up further interac-
tion was more favorable. Good SAR trend for the
enzyme activity was observed for this novel scaffold
along with activity in the cell-based systems.
Acknowledgment
The authors thank discovery analytical chemistry group
at Wyeth Research, Pearl River, NY, for spectral data.
8. For Glu-Mic assay details, see Ref. 5b.
9. (a) Gilliard, J. W.; Ford-Hutchinson, A. W.; Chan, C.;
Charleson, S.; Denis, D.; Foster, A.; Fortin, R.; Legere, S.;
McFarlane, C. S.; Morton, H.; Piechuta, H.; Riendeau,
D.; Rouzer, C. A.; Rokach, J.; Young, R. N.; MacIntyre,
D. E.; Peterson, L.; Bach, T.; Eiermann, G.; Hopple, S.;
Humes, J.; Hupe, D.; Luell, S.; Metzger, J.; Meurer, R.;
Miller, D. K.; Opas, E.; Pachalok, S. Can. J. Physiol.
Pharmacol. 1989, 67, 456; (b) Hagmann, W. Biochem. J.
1994, 299(Pt. 2), 467.
10. Dessen, A.; Tang, J.; Schmidt, H.; Stahl, M.; Clark, J. D.;
Seehra, J.; Somers, W. Cell 1999, 97, 349.
11. McMartin, C.; Bohacek, R. S. J. Comput. Aided Mol. Des.
1997, 11, 333.
References and notes
1. Smith, W. L. Biochem. J. 1989, 259, 315.
2. (a) Clark, J. D.; Schievella, A. R.; Nalefski, E. A.; Lin,
L.-L. J. Lipid Mediators Cell Signal. 1995, 12, 83; (b) Six,
D. A.; Dennis, E. A. Biochim. Biophys. Acta 2000, 1488,
1.
3. Wasserman, S. I. Hosp. Pract. 1988, 15, 49.
4. (a) Sapirstein, A.; Bonventre, J. V. Biochim. Biophys. Acta
2000, 1488, 139; (b) Nagase, T.; Uozumi, N.; Ishii, S.;
Kume, K.; Izumi, T.; Ouchi, Y.; Shimizu, T. Nat.
Immunol. 2000, 1, 42; (c) Lehr, M. Drugs Future 2000,
25, 823.
12. While Arg 200 is proposed to stabilize the binding of the
negatively charged phosphate group of the phospholipid
substrate, Ser 228 acts as the nucleophile and attacks the
sn-2 ester to form the acyl enzyme intermediate. The
oxyanion hole, putatively formed by Gly 197 and Gly 198,
polarizes the sn-2 ester and stabilizes the tetrahedral
intermediate (see Ref. 10). In the model of 35, the acid
group most closely mimics the ester. Models of less potent
analogs, with shorter linkers, showed the acid group
interacting with Arg 200. However, compounds predicted
to bind deeper in the active site cleft showed enhanced
affinity.
5. (a) McKew, J. C.; Lovering, F.; Clark, J. D.; Bemis, J.;
Xiang, Y.; Shen, M.; Zhang, W.; Alvarez, J. C.; Joseph-
McCarthy, D. Bioorg. Med. Chem. Lett. 2003, 13, 4501; (b)
McKew, J. C.; Foley, M. A.; Thakker, P.; Behnke, M. L.;
Lovering, F. E.; Sum, F.-W.; Tam, S.; Wu, K.; Shen, M. W.
H.;Zhang, W.; Gonzalez, M.;Liu, S.;Mahadevan, A.;Sard,
H.; Khor, S. P.; Clark, J. D. J. Med. Chem. 2005, 49, 135.
6. Matassa, V. G.; Maduskuie, T. P.; Shapiro, H. S.; Heep,
B.; Snyder, D. W.; Aharony, D.; Krell, R. D.; Keith, R. A.
J. Med. Chem. 1990, 33, 1781.
7. (a) Parallel solution-phase synthesis was carried out in
8 mL vials in 0.1 mmol scale using an orbital shaker.