6
286
H. Su et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6284–6288
Table 1
Table 3
HDACs inhibitory activities of indicated compounds
HDAC1 and HDAC4 inhibitory activities of 1j and 1k
Compd
R
HDACs IC50
(
lM)
HDAC1 (IC50
,
lM)
HDAC4 (IC50, lM)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
Ethyl
Propyl
iso-Propyl
Butyl
iso-Butyl
Amyl
Hexyl
Octyl
Benzyl
2-Phenoxyethyl
2-(p-Tolyloxy)ethyl
3-Phenoxypropyl
3-(p-Tolyloxy)propyl
4-Phenoxybutyl
4-(p-Tolyloxy)butyl
—
0.054
0.052
0.168
1.349
0.494
4.328
2.790
1.587
3.108
0.041
0.101
0.195
1.068
0.057
1.671
0.053
1j
1k
2.3
1.05
>50
>50
SAHA
HDAC1 and HDAC4, for the two enzyme are considered as repre-
sentative of class I and class IIa HDACs respectively. As demon-
strated in Table 2, compounds 1j–o were tested at 5 lM against
Figure 3. Western Blot analysis of p21WAF1/CIP1 expression,
acetylation in U937 cells after treatment with the indicated compounds (at 5
SAHA and MS-275 (5 M) were used for a comparison.
a
-tubulin and H3
M).
human recombinant (hr) HDAC1 and HDAC4 enzymes, in compar-
ison with SAHA. It turns out that all of the compounds 1j–o are
highly selective and potent inhibitors of HDAC1 but not against
HDAC4 compared with pan-HDAC inhibitor SAHA. The structure–
activity relationships are not consistent with the trend of activity
observed against mixtures of HDAC members, confirming selectiv-
ity occur within certain subtype of HDACs. Compounds with longer
side chains (lm, ln and 1o) are more potent against HDAC1 as well
as HDAC4. In addition, the methyl group substitute at phenyl of
side chain gave a little increase (compare 1k with 1j and 1o with
l
l
In addition, class I and pan-HDAC inhibitors usually induce over-
WAF1/CIP1
expression of p21
. Here, the capacity of these N-hydroxyb-
enzamides based derivatives (1j–n) to induce over-expression of
WAF1/CIP1
p21
pound 1l and 1m induced expression of p21
while compounds 1j and 1k compounds were relatively weakly
compare to SAHA). The trend of potency in up-regulation of
was also tested (Fig. 3). The results showed that com-
WAF1/CIP1
moderately,
(
1
n) or significantly improved (compare 1m with 1n) of both the
WAF1/CIP1
p21
of these compounds was quite consistent with their
HDAC1and HDAC4 inhibiting capacity; compound with three
methylene moiety (1o) was the most potent. According to the
IC50 value of compounds 1j and 1k, they displayed moderate activ-
ity to human recombinant (hr) HDAC1, but were almost inactive
against HDAC4 (Table 3). In conclusion, analogs 1j–o are potent
HDAC inhibitors with remarkable HDAC1 selectivity, validating
the structural difference around the rim of the catalytic pockets be-
tween HDAC1 and HDAC4.
HDAC1 inhibitory activity.
We also analyzed the effect of these compounds on granulocytic
differentiation, cell cycle and apoptosis induction. SAHA and MS-
2
75 (5
ation was evaluated by measuring the CD11c expression level upon
0 h of stimulation with 1i–o at 5 M. Amongst the tested com-
lM) were used as reference drugs. Granulocytic differenti-
3
l
pounds 1j, 1k and 1o showed a percent of CD11c positive PI nega-
tive cells higher than SAHA, although none of them being more
potent than MS-275 as cytodifferentiating agent (Fig. 4A). At the
tested conditions, 1j–n were able to induce accumulation of the
cells in the G1 phase, whereas 1o gave a block of the cycle at the
S phase (Fig. 4B). Apoptosis induction was measured with the An-
nexin V/propidium iodide (PI) double staining method. The results
revealed that half of the tested compounds (1j, 1k and 1o) had
higher apoptosis induction potency in the U937 cell line than
MS-275, though the apoptosis induction potency of all of them
was relatively weaker than SAHA in our assay. Taken together,
our findings showed that compounds with higher levels of differ-
entiation ability usually behaved better in apoptosis induction.
Finally, the anti-proliferative activity was evaluated against four
human tumor cell lines with SAHA as controls. As shown in Table 4,
the majority of the compounds displayed excellent anti-prolifera-
tive profiles compared with SAHA, especially in A549 and HEPG2
cells. Compounds 1n and 1o were particularly potent with respect
The effects of compounds 1j–n (5
lM, 24 h) on histone H3 and
a-tubulin acetylation levels, taken as markers of class I HDACs and
HDAC6 inhibition, respectively, were tested in the human leuke-
mia U937 cell line (Fig. 3). SAHA and MS-275 have been taken as
reference compounds.
As seen by Western blot analysis, the five tested compounds
showed moderate levels of acetyl-H3, although all of them being
equally weaker than SAHA. Moreover, most of the tested compounds
(1k, 1l, 1m and 1n) demonstrated strong induction of a-tubulin
acetylation with the exceptions of 1j, which showed lower activity
possibly due to poor cell permeability or differences in the kinetics
of induction. Above all, these data confirmed the capability of such
compounds to inhibit class I HDACs as well as HDAC6.
Table 2
Human recombinant HDAC1 and HDAC4 inhibitory activities of the compounds 1j–o
at 5
lM
Compd
% of residual activity
to growth inhibition of HCT116, with IC50 of 0.68 and 0.58 lM,
respectively, but the others are less potent than SAHA in the same
cell line. The limited difference among the side chains of these
compounds does affect their cytotoxicities against the four cancer
cells, but no outstanding trends were observed here.
In summary, we have designed and synthesized a series of novel 5-
p-tolylthiazol-2-amines HDAC inhibitors. Utilizing the N-hydroxy-
benzamide as a zinc binding group, we maintained the tolylthiazol
as one hydrophobic group and surveyed substitutes with different
HDAC1
HDAC4
1
1
1
1
1
1
j
k
l
m
n
o
6.47
4.16
6.54
89.31
79.16
92.65
64.25
62.74
54.14
0.49
0.45
À0.55
À0.35
1.68
SAHA