In order to introduce a trans double bond to change the linking
mode of the aliphatic carbon chains, we synthesized 2l and 2m
with 3,4-bis (4-methoxyphenyl)-1H-pyrrole (8a) and (E)-3-(4-
methoxyphenyl) propenoyl chloride (21) in the presence of
sodium hydride (Scheme 3).
respectively to give benzyloxycarbonyl protected acid chloride
15. P-methoxyphenylacetic acid was esterified, brominated and
substituted
to
obtain
methyl-2-methoxy-2-(4-
methoxyphenyl)acetate 19, then hydrolyzed by LiOH•H2O to
obtain a carboxylic acid, followed by treatment with oxalyl
chloride to give the methoxy-substituted acid chloride 21. 3,4-
bis(4-methoxyphenyl)-1H-pyrrole 8a was amidated with 15 or 21
in the presence of n-butyllithium in anhydrous THF to give 2a or
2c respectively. By palladium-catalyzed reduction, 2a was
converted to a 7-hydroxy substituted derivative 2b (Scheme 4).
We evaluated the hypoxia-induced HIF-1 inhibitory activity of
neolamellarin A and its derivatives using a cell-based dual-
luciferasereporter assay in Hela cells, which expressed both
HRE-dependent firefly luciferase reporter gene and constitutively
CMV-driven Renilla luciferase reporter gene. Inhibition of cell
growth (GI50) of synthetic compounds against Hela cells under
normoxia was determined by MTT assay. Compound 22
(phenoxyacetanilide), a previously reported HIF-1 inhibitor, was
used as a positive control [29]. All of the results are summarized in
Table 1.
Scheme 3. Synthesis of compound 2l and 2m. Reagents and conditions: (a)
oxalyl chloride, DMF, anhydrous CH2Cl2, r.t., 1h; (b) (i) NaH, r.t., 1 h (ii) 21,
DMF, 35 °C, 2 h, 74%; (c) BBr3, anhydrous CH2Cl2, r.t., 2 h, 63%.
The result showed that compound 22 exhibited the best HIF-1
inhibitory activity (IC50 = 6.7 ± 1.1 μM) among all tested
compounds. Although all synthesized compounds exhibited
weaker HIF-1 inhibitory activity than 22, C-7 hydroxy-
substituted analogue 2b (IC50 =11.9 ± 3.6 μM) showed
comparable potency with neolamellarin A (IC50 =10.8 ± 1.0 μM) ,
and specially, these two compounds were much less cytotoxic
against HeLa cells than compound 22 (GI50 of 1: 59.5 ± 0.5 μM
and 2b: 44.2 ± 1.6 μM vs 22: 9.3 ± 1.7 μM) under normoxia
condition.
Finally, we prepared neolamellarin A derivatives with
hydroxyl or methoxyl substituents at the C-7 position to
investigate the effect of the C-7 substituents on HIF-1 inhibitory
activity. According to our convergent synthesis strategy, the acid
chloride intermediates with different substituents at -C is the
Comparing the IC50 values of compounds 1, 2h and 2k (10.8,
18.8 and 14.7 μM, respectively) with different length of aliphatic
chain suggested that a two-carbon aliphatic carbon chain linker
was more favorable to HIF-1 inhibitory activity than a single or
triple carbon chain. In addition, the introduction of a trans-olefine
on the aliphatic chain resulted in a sharp decrease in inhibitory
activity (2l, IC50 > 50 μM), which may be due to the increased
rigidity of the skeleton.
The substituent on the three benzene rings seemed to produce
a noteworthy influence on HIF-1 inhibitory activity. Comparing
the IC50 values of 2j with 2k (26.8 and 14.7 μM), 2g with 2h
(24.4 and 18.8 μM) and 2l with 2m (>100 and 16.6 μM), it
indicated that compounds with p-hydroxyl group exhibit more
potent HIF-1 inhibitory activity than the corresponding
compounds with p-methoxyl groups, whereas their cytotoxic
effects were also higher than those of methoxyl-substituted
derivatives.
In addition, compound 2b, the 7-hydroxy substituted
derivative, showed higher activity than compound 2i (IC50 = 11.9
vs 22.5 μM), suggesting the benefit of 7-hydroxy to HIF-1
inhibition, which was evidenced by the fact that the methylation
of 7-hydroxy of 2b sharply reduced the inhibitory activity (IC50
of 2c > 50 μM). All the compounds we synthesized had no
significant cytotoxicity at the concentrations they effectively
inhibited HIF-1 activation.
Scheme 4. Synthesis of compound 2b and 2c. Reagents and conditions: (a)
TBAB, CHCl3, NaOH, 45 °C, 2.5 h, 49%; (b) CbzCl, TMEDA, CH2Cl2, 0 °C
0.5 h then r.t. 2h, 37%; (c) Oxalyl chloride, DMF, anhydrous CH2Cl2, r.t., 1 h,
99%. (d) SOCl2, MeOH, 0 °C 0.5 h then r.t. 1 h, 99%; (e) AIBN, NBS, CCl4,
80 °C, 4 h, 99%; (f) MeONa, MeOH, 65 °C, 3 h, 81%; (g) LiOH•H2O,
MeOH/H2O, 30 °C, 1 h, 60%; (h) oxalyl chloride, DMF, anhydrous CH2Cl2,
r.t., 1 h, 99%.. (i) (i) n-BuLi, THF, -78 °C then r.t. 0.5 h (ii) 13 or 19, r.t., 2 h,
2a: 40%, 2c: 56%; (j) H2, Pd/C, MeOH/THF, 25 °C, 24 h, 58%.
In conclusion, neolamellarin A, a novel marine pyrrole
alkaloid derived from sponge, has been shown to inhibit hypoxia-
induced HIF-1 activity. In this work, we have succeeded in
synthesizing neolamellarin A and its derivatives by a convergent
synthetic strategy and evaluated their HIF-1 inhibitory potent
using a cell-based dual-luciferasereporter assay in Hela cells. Of
all the compounds, neolamellarin A (1) and its 7-hydroxy
derivative 2b exhibited significant HIF-1 inhibitory activity and
key step. Under
a
strong alkaline condition, p-
methoxybenzaldehyde 13 reacted with chloroform via
condensation reaction to obtain α-hydroxy acid 14, which was
treated with benzyloxycarbonyl chloride and then oxalyl chloride
4