2
626
G. Li et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2624–2628
OH
O
Table 1
H
a
O
b
Inhibition of HIF-1 transcriptional activity in HeLa cell-based HRE and CMV dual
luciferase assay and cell growth inhibition
H
OEt
Compound
GN26361
IC50a
(
l
M)
GI50b
(lM)
1
0
14
O
O
2.2 ± 0.2
1.3 ± 0.1
4.2 ± 0.8
1.1 ± 0.3
0.53 ± 0.07
2.4 ± 0.6
1.3 ± 0.2
14.6 ± 2.7
16.4 ± 0.6
13.5 ± 0.6
6.9 ± 0.5
30
5.2 ± 0.8
5.3 ± 5.2
8.7 ± 0.7
7.7 ± 2.3
23.6 ± 0.6
15.6 ± 1.3
B
1
1
1
1
a
b
c
O
O
O
H
OH
O
c
H
B
N
H
d
OH
1e
1f
15
16
1g
1h
HO
OH
a
HeLa cells expressing HRE-dependent firefly luciferase reporter construct (HRE-
O
Luc) and constitutively expressing CMV-driven Renilla luciferase reporter with
SureFECT Transfection Reagent were established with Cignal™ Lenti Reporter
(SABiosciences, Frederick, MD) according to the manufacturer’s instructions. The
consensus sequence of HRE was 5 -TACGTGCT-3 from the erythropoietin gene.
Cells stably expressing the HRE-reporter gene were selected with puromycin. The
cells were incubated for 12 h with or without drugs under the normoxic or hypoxic
condition. After removal of the supernatant, the luciferase assay was performed
O
d
H
N
H
OH
0
0
1
g
Scheme 4. Reagents and conditions: (a) ethyl 4-bromobutyrate, K
overnight, 57%; (b) LiOH, THF–H
overnight, 48%; (d) (i) KHF (4 M), MeOH, rt, 2 h, (ii) HCl (1 N), rt, overnight, 58%.
2
3
CO , DMF, rt,
using a Luciferase Assay System (Promega, Madison, WI) according to the
2
O, rt, overnight, 93%; (c) 13, HATU, DIPEA, DMF, rt,
manufacturer’s instructions. The drug concentration required to inhibit the relative
light units by 50% (IC50) was determined from semi-logarithmic dose–response
plots, and the results represent means ± SD of triplicate samples.
HeLa cells were incubated for 72 h with various concentrations (100 nM–
00 lM) of compounds under the normoxic condition, and viable cells were tested
by the MTT assay. The drug concentration required to inhibit cell growth by 50%
(GI50) was determined from semi-logarithmic dose–response plots, and results
represent means ± SD of triplicate samples.
2
b
2
0 was converted to the corresponding phenol 21 by treated with
BBr . Alkylation of 21 with 7 was carried out using NaH as a base
3
and hydrogenesis of the resulting 22 afforded 3,5-(di-ortho-carbo-
ranyl)phenoxyacetanilide derivative 1h in 54% yield.
1
Biological activity: Synthesized substituted ortho-carboranylphe-
noxyacetanilide derivatives 1a–h were evaluated for their ability to
inhibit hypoxia-induced HIF-1 transcriptional activity using a cell-
based reporter assay in HeLa cells expressing HRE-dependent firefly
luciferase reporter construct (HRE-Luc) and constitutively express-
ing CMV-driven Renilla luciferase reporter. Cell growth inhibition
compared with GN26361 is probably due to the lipophilicity of the
pinacol ester functional group, which would affect the transmem-
brane property. The meta derivative 1b of GN26361 decreased
the inhibition of HIF-1 transcriptional activity with IC50 of
4
.2 ± 0.8 lM. Previously, we demonstrated the effects of sub-
(
GI50) by the synthesized compounds was also determined by MTT
1
stituents at ortho-carborane (R ) in GN26361 on HIF-1 transcrip-
assay. GN26361 was used as a positive control for comparison. The
results are summarized in Table 1. The IC50 of GN26361 toward
tional activity and found that ethyl group was the most suitable
2
0
among other substituted derivatives (H, Me, and i-Bu). In the case
of the meta derivatives, the ethyl substituent (1c) increased the inhi-
HIF-1 transcriptional activity was 2.2 ± 0.2
ester 1a was more potent than GN26361 (IC50 = 1.3 ± 0.1
higher inhibitory activity of 1a toward HIF-1 transcription
l
M, whereas the pinacol
lM). The
bition of HIF-1 transcriptional activity with IC50 of 1.1 ± 0.3
IC50 of compound 1e having an ethylene linker was slightly
decreased to 2.4 ± 0.6 M, whereas that of compound 1g having a
propylene linker dropped to 14.6 ± 2.7 M, indicating that the
lM. The
l
OMe
l
OMe
OMe
length such as methylene and ethylene linkers is appropriate for
the potency of HIF-1 inhibition. The highest inhibition was achieved
a
b
by compound 1d with the IC50 of 0.53 ± 0.07 lM, suggesting that the
Br
Br
TMS
OMe
TMS
rigid conformation by introducing an unsaturated ethyne linker in
the molecule is attributed to the potency of HIF-1 inhibition.
Interestingly, deboronated derivative 1f also inhibited HIF-1
1
7
18
19
OH
H
H
H
H
transcriptional activity with IC50 of 1.3 ± 0.2
ortho-carboranyl) substituent was not effective; the IC50 of com-
pounds 1h dropped to 16.4 ± 0.6 M. Although the cytotoxicity of
lM. However, 3,5-(di-
c
d
e
l
the compounds was excluded from the HIF-1 transcriptional activity
by correcting the measurement of constitutively expressing CMV-
driven Renilla luciferase activity, cell growth inhibition by the com-
pounds (GI50) was related to the inhibition of the hypoxia-induced
HIF-1 transcriptional activity except compound 1b.
We next examined the effects of compounds 1c, 1d, 1f, which
showed significant inhibition of HIF-1 transcriptional activity, and
di-carboranyl compound 1h, on human HSP60 chaperone activity
using the porcine heart malate dehydrogenase (MDH) refolding
2
0
21
O
O
O
O
B
B
H
H
O
O
f
O
H
O
H
N
H
N
H
OBn
OH
2
7,28
assay.
The results are summarized in Figure 2. Inhibition ratio of each com-
pound at 2 M concentration against HSP60 chaperon activity is
indicated by vertical axis. Although epolactaene tert-butyl ester
GN26361 was used as a positive control for comparison.
2
2
1h
l
Scheme 5. Reagents and conditions: (a) ethynyltrimethylsilane, PdCl
CuI, diethylamine, THF, reflux, 5 h, quant.; (b) KOH, MeOH, rt, 6 h, 78%; (c) B10
N,N-dimethylaniline, chlorobenzene, microwave, 130 °C, 37%; (d) BBr , CH Cl
overnight, 100%; (e) 7, NaH, THF, rt, 2 h, 72%; (f) H , Pd/C, MeOH–THF, rt, overnight,
4%.
2
(PPh
3
)
2
, PPh
H
2
, rt,
3
,
,
14
27
(
ETB) exhibited significant inhibition (ꢀ50%) at 6
inhibition was observed at 2 M. Moderate inhibition was observed
in the cases of compounds 1c and 1d, whereas inhibition ratio of
lM, a weak
3
2
l
2
5