A. S. Kalgutkar et al. / Bioorg. Med. Chem. Lett. 12 (2002) 521–524
523
cated the dual COX and 5-lipoxygenase inhibitory
properties of hydroxamate derivatives of meclofenamic
acid such as 12.11 Furthermore, 12 demonstrated non-
ulcerogenic, antiinflammatory properties in animal
models. Our results suggest that the nonulcerogenic,
antiinflammatory properties of 12 are due to the selec-
tive inhibition of COX-2. The potency of hydroxamate
derivatives 12–15 as 5-lipoxygenase inhibitors remains
to be determined.
Simple aromatic amide derivatives of meclofenamic acid
including 4-fluorophenyl-, 4-methoxyphenyl-, and 3-
pyridyl amides also were included in the study, but these
compounds did not display significant inhibition of the
COX isozymes when tested at 30 mM. Examples of ter-
tiary amide derivatives such as the piperidinyl- and the
morpholino amide derivatives were also evaluated for
selective COX-2 inhibitory properties. Although the
piperidinyl amide displayed some COX-2 selective inhi-
bition [IC50 (COX-2)=1.0 mM; IC50 (COX-1)=20 mM],
the morpholino amide derivatives was inactive against
the COX isozymes at the concentration range tested (20
mM). Additional examples are needed to fully char-
acterize the SAR around tertiary amide derivatives of
meclofenamic acid.
SAR analysis on the selective COX-2 inhibitory profile
of meclofenamic acid-amino acid conjugates 18–24 was
also undertaken (Table 3). Esterification of the carboxyl-
ate moiety in the amino acid portion of these conjugates
resulted in an overall increase in COX-2 potency and
selectivity, whereas conjugates with a free carboxylic
acid group12 were less potent and selective as COX-2
inhibitors. Comparison of the COX-2 selectivity ratios
for the phenoxyethyl analogue 10 and the aminoacid
conjugates 18 and 19 suggests some SAR trends for
selective COX-2 inhibition by meclofenamic acid
amides. For instance, bulky alkyloxy (ethyl vs methyl)
(19) or aryloxy (10) substituents seem to be detrimental
for COX-1 inhibition. Whether the COX-2 selectivity
ratio can be further increased by incorporating func-
tionalities in the aryl ring in 10 remains to be deter-
mined.
The ability of meclofenamic acid amides to inhibit
COX-2 in intact cells was assayed in RAW264.7 mac-
rophages in which COX-2 activity was induced by
pathologic stimuli. The macrophages were exposed to
lipopolysaccharide and g-interferon to induce COX-2
and then treated with several concentrations of the
phenoxyethyl amide and the 4-nitrobenzylhydroxamate
derivatives 10 and 15, respectively. The IC50 values for
inhibition of prostaglandin D2 (PGD2) by 10 and 15
were 1.8 and 0.6 mM, respectively. The results from
these studies indicate that meclofenamic acid amides, in
addition to inhibiting purified COX-2, are also good
inhibitors of COX-2 activity in intact inflammatory cells.
Table 3. IC50 values for COX-1 and -2 inhibition by amino acid
conjugates of meclofenamic acid
The present study extends our carboxylate derivatiza-
tion strategy to include meclofenamic acid as a repre-
sentative example from the fenamate class of NSAIDs.
Unlike the indomethacin series, wherein both esters and
amides were potent and selective COX-2 inhibitors,
preliminary SAR studies on neutral meclofenamic acid
analogues suggest that only the amide derivatives are
capable of COX-2-selective inhibition. This may be due
to subtle differences in the binding of the ester and amide
derivatives of meclofenamic acid in the COX-2 active site.
Compd
18
R
IC50 (mM)a
Selectivityb
COX-2
COX-1
1.2
0.07
0.2
0.4
0.8
6.0
2.7
8.0
17
20
Site-directed mutagenesis studies have also provided
useful insights into the molecular basis for COX-2 inhi-
bition by meclofenamic acid amides. As observed with
19
4.0
0.3
the indomethacin ester and amide derivatives,6,7
a
representative meclofenamic acid amide (10) inhibited
the Arg120Ala murine COX-2 mutant with potency
comparable to wild-type enzyme [IC50 (wt)=0.4 mM;
IC50 (Arg120Ala)=0.55 mM]. These results are con-
sistent with our prediction that ester and amide deriva-
tives of NSAIDs do not require the positively charged
Arg120 for selective COX-2 inhibition.6 Both meclo-
fenamic acid and amide 10 were unable to inhibit the
Tyr355Phe COX-2 mutant, suggesting that H-bonding
of the carboxylate or the amide linkage with the
hydroxyl group in Tyr355 is critical for COX-2 inhibi-
tion. Regions of the COX-2 protein that account for the
selectivity of celecoxib (or rofecoxib),13 proved unim-
portant for selective COX-2 inhibition by 10, which was
able to inhibit the Val523Ile mutant at comparable
potency to wild-type enzyme. Since Tyr355, which is
important for COX-2 inhibition by meclofenamic acid
amides, is conserved within the two COX isozymes,
20
0.7
3.2
5.5
3.2
>4.0
21
2.6
22
33
23
6.0
24
>33
aValues are average from two separate experiments.
bIC50 COX-1/IC50 COX-2.