F. Xie et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5371–5375
5373
Table 1
Biological activities of synthesized compoundsa
Compound
KIX–KID inhibitionb (IC50
)
CREB-mediated transcription inhibitionb (IC50
)
VP16-CREB-mediated transcription inhibitionb (IC50
)
1
4
6
7
8
9
9a
0.36 0.15
0.41 0.067
7.14 0.57
ꢀ50c
2.09 0.39
9.42 0.47
13.74 3.41
ꢀ50c
6.14 0.88
13.17 4.41
>100
>50
0.17 0.064
13.06 3.69
>50
>50
NDd
>50
>50
NDd
NDd
a
All the stock solutions were prepared in DMSO except 9, which was dissolved in DMF due to its instability in DMSO.
All the IC50 values are shown in lM. The data are presented as mean SD of at least two independent experiments performed in duplicates (KIX–KID inhibition assay) or
b
triplicates (reporter assays). If the IC50 was not reached at the highest concentration tested at 50
l
M or 100
lM, it is presented as >50 or >100.
c
The IC50 is roughly at the highest concentration tested.
d
Not determined.
decreasing its nucleophilicity. Nitrating 18 gave a chromatograph-
ically inseparable mixture of 19a and 19 in a combined 49% yield in
a ratio of 69:31 as determined by HPLC. Final hydrolysis of the
acetyl group in the mixture of 19a and 19 with 20% H2SO4/EtOH
yielded products 9a and 9, which could be separated by careful sil-
ica gel chromatography (Scheme 3).
All the six newly synthesized compounds were evaluated for
their potencies in inhibiting KIX–KID interaction in vitro and
CREB-mediated gene transcription. For those compounds demon-
strating inhibition of CREB’s activity, their effects on VP16-CREB-
mediated gene transcription in HEK 293T cells were also evaluated.
The results are presented in Figure 2 and Table 1. The deoxygeneat-
ed compound 4 displayed comparable activity to compound 1 in
inhibiting KIX–KID interaction in vitro, suggesting the scheme
shown in Figure 3 is not a major pathway for the observed
in vitro inhibition of KIX–KID interaction because thiopyridine
1-oxide is a better leaving group than thiopyridine. On the other
hand, the cellular inhibition of CREB-mediated gene transcription
Figure 3. Potential formation of a Meisenheimer complex between compound 1
and cellular nucleophiles.
and synthesis of compounds 4, 6–9 (Schemes 1–3). The relative
leaving group ability follows the order of 1 > 4 > 6 while compound
7 has reduced capability to form a Meisenheimer-type complex
due to the electron-donating nature of the –NH2 group. Com-
pounds 8 and 9 do not have a leaving group and thus cannot form
the corresponding adduct 5.
Compound 6 was synthesized by condensing NBD-Cl (2) with 4-
chloroaniline (10) in the presence of NaHCO3 (Scheme 2).20,25
Reduction of –NO2 in 6 with Fe/HCl26 gave compound 7 in 58%
yield. In order to synthesize compound 8, we designed the synthe-
sis by coupling a known acid 1627 with aniline 10. Acid 16 was re-
ported to be prepared from commercially available benzofurazan
(11) by a sequence of reactions involving formylation by LDA/
DMF, reduction by NaBH4, concomitant nitration and oxidation
by HNO3/H2SO4.27 However, we were only able to obtain alcohol
14 from this series of reactions. The acid 16 was eventually pre-
pared from alcohol 14 via IBX oxidation followed by Pinnick oxida-
tion.28 The desired amide 8 was finally delivered by coupling acid
16 with aniline 10 facilitated by MsCl.20 The synthesis of com-
pound 9 was not so straightforward. Initial attempts to prepare 9
from Borch reductive amination29 between aldehyde 15 and 10
were unsuccessful. We found that the nitro group in 15 had dele-
terious effects on this reaction because aldehyde 12 could be suc-
cessfully coupled with aniline 10 to give 17 in 67% yield under the
same reaction condition. Direct nitration of 17 only afforded the
undesired product with a nitro group incorporated into the aniline
ring presumably due to its electron-rich nature. So an acetyl group
was introduced on the aniline nitrogen to generate 18 by
by 4 was reduced by about four-fold to an IC50 of 9.42 lM com-
pared to compound 1. These results suggest that the discordance
between in vitro and cellular IC50 of compound 1 is not due to its
charged nature, which may result in reduced cell permeability as
compound 4 is not charged. But its cellular potency is also much
weaker than its in vitro KIX–KID interaction inhibition potency.
Compared to compound 1, the selectivity of deoxygenated
compound 4 was not improved because it inhibited VP16-CREB-
mediated gene transcription with an IC50 of 13.17
4-chloroaniline-substituted compound 6 was significantly less
potent in inhibiting KIX–KID interaction in vitro (IC50 = 7.14 M)
than 1. Its cellular activity (IC50 = 13.74
M) was only ꢀtwo-fold
lM. The
l
l
less potent than its in vitro activity. More importantly, this com-
pound displayed enhanced selectivity as evidenced by the lack of
inhibition of VP16-CREB-mediated transcription activity up to
100 lM, the highest concentration tested. Reduction of the –NO2
group in 6 to compound 7 essentially abolished all the activities
measured (Table 1). The amide compound 8 exhibited two-fold
enhanced activity in inhibiting KIX–KID interaction in vitro
(IC50 = 0.17
lM) compared to 1. However, no cellular activity was
observed (IC50 > 50
l
M). This discrepancy might be due to its poor
cellular permeability or extracellular inactivation. Insertion of one
methylene unit into compound 6 resulted in 9, which showed
reduced in vitro and cellular activity. The regioisomer 9a was inac-
tive in all the assays measured. Collectively, these results suggest
that the electrophilic nature of the nitrobenzofurazan nucleus is
required for potent KIX–KID interaction inhibition and cellular
inhibition of CREB-mediated gene transcription. Furthermore,
modulation of leaving group ability of the substituents at
4-position of nitrobenzofurazan can enhance selectivity.
In summary, 2-[(7-nitrobenzo[c][1,2,5]oxadiazol-4-yl)thio]pyr-
idine 1-oxide (compound 1) was identified as a novel inhibitor of
Scheme 2. Synthesis of compounds 6 and 7.