574
H.-L. Jang et al. / Bioorg. Med. Chem. Lett. 24 (2014) 571–575
1g (cycloheptane, p-(trifluoromethyl)benzenesulfonate), 1h (cyclo-
octane, methanesulfonate), 1j (cyclooctane, n-propanesulfonate),
and 1m (cyclooctane, p-toluenesulfonate) were near the inhibitory
effect expressed by the reference compound, NS398.
Among all the target compounds, compound 1g possessing
fused cycloheptane ring and p-(trifluoromethyl)benzenesulfonate
moieties showed the highest inhibitory effects on NO and PGE2
productions. It showed higher efficacy and potency than the refer-
ence NO production inhibitor, L-NIL. It inhibited NO production by
#
40
35
30
25
20
15
10
5
IC 50:85.76 3.58 µM
**
***
65.76% at 5
lM concentration, while L-NIL inhibited it by 50% at
20 M. It also inhibited the PGE2 production in RAW 264.7 macro-
phages by 91.48%, which is very near to the inhibition effect pro-
duced by NS398.
l
0
LPS (10ng/ml)
-
-
+
+
1
+
5
+
+
+
-
STX-64 (µM)
-
10
50
100
100
Based on NO and PGE2 production inhibition% data, compounds
1g, 1j, 1m, and 1n with the most promising results were selected
for further screening of their potencies. Their IC50 values were cal-
culated and are summarized in Table 3. Compounds 1g and 1j were
more potent than L-NIL as NO production inhibitor. Compound
1g possessing fused cycloheptane ring showed higher potency
than the corresponding analog with fused cyclooctanone ring,
compound 1n. Among the four tested compounds, compound 1g
demonstrated the highest potency. It is 5.24 times more potent
than L-NIL. Regarding PGE2 inhibitory potency, the four com-
pounds 1g, 1j, 1m, and 1n showed higher potency than NS398.
Compound 1j possessing fused cyclooctanone and n-propanesulfo-
nate moieties was the most potent, and it was 1.74-fold more po-
tent than NS398. We further investigated whether the inhibitory
20
15
10
5
#
IC50:81.04 0.92 nM
***
***
***
***
ꢀꢀꢀ
ꢀꢀꢀ
ꢀꢀꢀ
ꢀꢀꢀ
+
***
+
0
LPS (10 ng/ml)
-
+
-
+
+
1
+
-
effects of compound 1g (0.1, 1 or 5 lM) on PGE2 production are re-
STX-64 (nM)
-
0.1
10
100
1000 1000
lated to COX-2 enzyme activity and protein expression. We ob-
served that compound 1g significantly inhibited COX-2 enzyme
activity (Fig. 4A). Because iNOS expression induced by LPS produce
NO, the protein expression of iNOS was determined. As shown in
Figure 4B, COX-2 protein levels were found to be markedly up-reg-
ulated in response to LPS, and compound 1g dose-dependently
inhibited LPS-induced COX-2 protein expression. Next, we investi-
gated whether inhibitory effects of compound 1g on LPS-induced
NO production are related to iNOS enzyme activity or iNOS expres-
sion. We found that 1g exhibited no effect on the iNOS enzyme
activity (Fig. 4C). However, 1g suppressed the LPS-induced iNOS
expression (Fig. 4B). Thus, we suggest that the inhibition of PGE2
and NO production by 1g are due to its inhibition of the COX-2 en-
zyme activity and protein expressions of COX-2 and iNOS. Thus,
the marked inhibitory effect of 1g on PGE2 production can be
caused by simultaneously suppressed COX-2 enzyme activity and
protein synthesis. However, 1g did not affect iNOS enzyme activity.
Although the present study does not clarify the mechanism
responsible for the inhibition of iNOS and COX-2 expression by
1g, we suggest that 1g could interferes the LPS-induced activation
Figure 5. Inhibitory effects of STX-64 on LPS-induced NO and PGE2 production in
RAW 264.7 macrophages. Following pretreatment with STX-64 (1, 5, 10, 50 or
100 lM) for 1 h, cells were treated with LPS (10 ng/mL) for 24 h. Levels of NO in
culture media were quantified using Griess reaction assay. Following pretreatment
with STX-64 (0.1, 1, 10, 100 or 1000 nM) for 1 h, cells were treated with LPS (10 ng/
mL) for 24 h. Levels of PGE2 in culture media were quantified using EIAs. Data are
presented as the means SDs of three independent experiments. #p <0.05 versus
the control cells; ⁄⁄p <0.01, ⁄⁄⁄p <0.001 versus LPS-stimulated cells; statistical
significances were compared using ANOVA and Dunnett’s post hoc test.
Lipinski’s rule of five.19 This approach has been widely used as a fil-
ter for substances that would likely be further developed in drug
design programs. In addition, we calculated the total polar surface
area (TPSA) since it is another key property that has been linked to
drug bioavailability. Thus, passively absorbed molecules with a
TPSA >140 are thought to suffer from low oral bioavailability.20
Molecules violating more than one of these rules may have prob-
lems with bioavailability. Predictions of ADME properties for the
studied compounds are summarized in Table 3. The results showed
that all the four tested compounds and the reference compound,
NS398, comply with these rules. Theoretically, compounds 1g, 1j,
1m, and 1n should present good passive oral absorption and differ-
ences in their bioactivity cannot be attributed to this property.
In conclusion, we synthesized a series of tricyclic coumarin sul-
fonates in order to explore the relationship between their struc-
tures and their antiinflammatory activities. The compounds
showed varying cytotoxicities and inhibitory activities on the
LPS-induced productions of NO and PGE2. The biological screening
of this series of coumarins led to discovery of a new potential
lead compound 1g. This compound with fused cycloheptane and
p-(trifluoromethyl)benzenesulfonate moieties exhibited the high-
est inhibitory activities on both NO and PGE2 productions in LPS-
induced RAW 264.7 macrophages with low cytotoxicity. It showed
superior potencies to the reference compounds, L-NIL and NS398,
respectively. In silico ADME prediction showed that compound
1g complies with Lipinski’s rule of five and it can be passively ab-
sorbed orally. Compound 1g can be utilized as a promising lead
of TAK1/IKK-a/b/NF-jB through TLR4.
Since STX-64 (Irosustat, BN83495), the first steroid sulfatase
(STS) inhibitor to enter diverse clinical trials for patients with ad-
vanced hormone-dependent cancer is closely related to the struc-
tures of tricyclic coumarin sulfonates in this manuscript, we
investigated whether STX-64 has inhibitory effects on the LPS-in-
duced NO and PGE2 production in RAW 264.7 macrophages. As
shown in Figure 5A and B, STX-64 inhibited LPS-induced NO and
PGE2 production in a dose-dependent manner. However, STX-64
more potently inhibited LPS-induced PGE2 production (IC50
:
81.04 nM) than NO production (IC50: 85.76 M). These results sug-
l
gest that STX-64 showed similar inhibitory effects of four most ac-
tive target compounds (Table 3) via a selective PGE2 inhibition in
LPS-stimulated macrophages.
The bioavailability of compounds 1g, 1j, 1m, and 1n with prom-
ising biological results was assessed using ADME (absorption, dis-
tribution, metabolism, and excretion) prediction methods. In
particular, we calculated the compliance of compounds to the