In our previous study, we designed and synthesized a novel series of N-hydroxycinnamamide-based HDAC inhibitors (HDACIs), among which the representative compound 14a exhibited promising HDACs inhibition and antitumor activity. In this current study, we report the development of a more potent class of N-hydroxycinnamamide-based HDACIs, using 14a as lead, among which, compound 11r gave IC50 values of 11.8, 498.1, 3.9, 2000.8, 5700.4, 308.2, and 900.4 nM for the inhibition of HDAC1, HDAC2, HDAC3, HDAC8, HDAC4, HDAC6, and HDAC11, exhibiting dual HDAC1/3 selectivity. Compounds 11e, 11r, 11w, and 11y showed excellent growth inhibition in multiple tumor cell lines. In vivo antitumor assay in U937 xenograft model identified compound 11r as a potent, orally active HDACI. To the best of our knowledge, this work constitutes the first report of oral active N-hydroxycinnamamide-based HDACIs with dual HDAC1/3 selectivity.
In our previous study, we designed and synthesized a novel series of N-hydroxycinnamamide-based HDAC inhibitors (HDACIs), among which the representative compound 14a exhibited promising HDACs inhibition and antitumor activity. In this current study, we report the development of a more potent class of N-hydroxycinnamamide-based HDACIs, using 14a as lead, among which, compound 11r gave IC50 values of 11.8, 498.1, 3.9, 2000.8, 5700.4, 308.2, and 900.4 nM for the inhibition of HDAC1, HDAC2, HDAC3, HDAC8, HDAC4, HDAC6, and HDAC11, exhibiting dual HDAC1/3 selectivity. Compounds 11e, 11r, 11w, and 11y showed excellent growth inhibition in multiple tumor cell lines. In vivo antitumor assay in U937 xenograft model identified compound 11r as a potent, orally active HDACI. To the best of our knowledge, this work constitutes the first report of oral active N-hydroxycinnamamide-based HDACIs with dual HDAC1/3 selectivity.
Epigenetics are widely
implicated in tumor initiation and progression
by different modifications of DNA and histones. There is a growing
body of evidence demonstrating the importance of histone modification
in silencing tumor suppressor genes.[1] Among
the various histone modifiers, histone acetyltransferases (HATs) and
histone deacetylases (HDACs) are two reversible enzymes regulating
histone acetylation status, attaching or removing acetyl groups from
the lysine residues in histone tails.[2] HATs
prevent the positive lysine residues from interacting more closely
with the DNA phosphate backbone, resulting in an “open”
chromatin state, whereas the HDACs remove acetyl groups, resulting
in a “closed” configuration, which blocks the access
of the transcription machinery to DNA, suppressing gene expression.[3]To date, 18 mammalianHDAC isoforms have
been identified in humans;
11 of them (Class I, II, IV) are Zn2+-dependent HDAC enzymes
(classical HDACs).[4] Class I HDAC, comprising
HDAC 1, 2, 3, and 8, are homologous to yeastRPD3 (reduced potassium dependency-3) protein; class II members include
HDAC 4, 5, 6, 7, 9, and 10 and are structurally related to yeast Hda
1 (histone deacetylases 1). Class IV HDAC has the sole member, HDAC11.
Class III HDACs, called Sirtuins (SIRT1–7), require NAD+ for their activity.[5]HDAC1
mainly localizes in the nucleus due to lack of the nucleus
export signal. The knockout of HDAC1 in mice caused proliferation
defects and development retardation because of an arrest of cell growth
that is associated with upregulation of the p21 and p27 (cyclin-dependent
Kinase inhibitors).[6] A recent study showed
that HDAC1 largely accumulated in mature areas of differentiated tumor
on the mice with teratoma, indicating that HDAC1 is a possible biomarker
of benign teratoma.[7] HDAC3 is one of the
most frequently upregulated genes in humancancers[8] and is involved in each of the three major targets of cancer
therapy: cell cycle control, differentiation, and apoptosis.[9] There are many direct links between HDAC3 and
various tumor types. In colon cancer cells HCT116 and Caco-2, silencing
of HDAC3 expression resulted in growth inhibition, a decrease in cell
survival, and increased apoptosis through stimulating p21 promoter
activity and expression.[10] In HeLa cells,
the majority of cellular HDAC3 is found to associate with SMRT and
N-CoR complexes, and knockdown of HDAC3 resulted in hyperacetylation
of histone H3 and apoptosis.[9,11] In metastatic breast
cancer cell MDA-MB-231, HDAC3 efficiently inhibited CREB3-enhanced
NF-κB activation. Moreover, high HDAC3 expression is also associated
with gastric cancer,[12] Glioma,[13] renal cancer,[14] liver
cancer,[15] and chronic lymphocytic leukemia.[16] On the basis of the above evidence, developing
HDACIs more specifically against HDAC1/3 may prove to be a worthwhile
goal.In the past 10 years, over 490 clinical trials of more
than 20
HDACIs candidates have been initiated, culminating in the approval
of two antitumor drugs vorinostat (SAHA) and romidepsin (FK228). Recently,
development of class or isoform selective HDACIs has drawn increased
attention. Although selective HDACIs were hypothesized to have fewer
side effects than other pan-HDACIs, their therapeutic advantages have
yet to be confirmed clinically. HDACIs are classified into different
classes depending on their chemical structures, namely, hydroxamates,
benzamides, aliphatic acids, cyclic tetra peptides, electrophillic
ketones, and miscellaneous groups, among which hydroxamates are considered
as the most common HDACIs. Although some class I selective,[17,18] HDAC6 (class IIb)[19] selective, and class
IIa[20] selective hydroxamates inhibitors
have been reported (Figure 1), hydroxamates
are generally thought to have limitation in selectivity against desired
HDAC isoforms due to their very strong chelating ability with zinc
ion.[21] To the best of our knowledge, most
of the subclass I selective inhibitors in research are benzamides
HDACIs, such as MS-275 (30), MGCD0103 (31), CI994 (32), and so on (Figure 1).[22,23] Additionally, recent study demonstrated
aryl substituents in the o-phenylenediamine of benzamide
HDACIs led to dramatic improvements in potency and selectivity for
HDAC1 and 2 versus HDAC3 (Figure 1).[24,25]
Figure 1
Structure
of selective HDACIs.
Structure
of selective HDACIs.Recently, we have embarked on development of N-hydroxycinnamamide-based HDACIs; in our previous study, compound
(S,E)-tert-butyl
(1-(4-(3-(hydroxyamino)-3-oxoprop-1-en-1-yl)-2-methoxyphenoxy)-3-(1H-indol-3-yl)propan-2-yl)carbamate (14a) exhibited
modest HDAC inhibitory.[26] Here in this
article, we report the design, synthesis, and biological evaluation
of a new series of selective hydroxymates HDACIs describing their
structure–activity relationship (SAR) studies, isotype-selectivity,
and anticancer activities. The representative compound 11r shows dual selectivity and high potency toward HDAC1/3. Several
compounds show an excellent effect in proliferation inhibition against
U937, HEL, KG1, HL60, K562, PC-3, A549, MDA-MB-231, HCT116, and MCF-7
cell lines. In vivo antitumor assay in U937 xenograft model showed
compound 11r is a potent, orally active HDACI.Modification
site and general structure of target compounds.
Results and Discussion
Chemistry
We designed and synthesized
a library of
cinnamamide derivatives based on the lead compound 41 (14a) (Figure 2). The library
was designed to fully explore the influence of variation in the methoxy
group of ferulic acid and the cap group. The synthesis of compounds 10a and its derivatives 11a–11x is described in Scheme 1. Synthesis of 3 was started with l-try by methyl ester protection
and Boc-protection. Methyl ester protection of 4-hydroxycinnamic acid 5 gave 6. Reduction of 3 by LiAlH4 led to 4, which was further subjected to Mitsunobu
reaction with 6 and gave 7. Compound 7 was directly converted to the target hydroxamic acid 10a. Intermediate 8, the N-deprotected
product of 7, reacted with different carboxylic acids
or Boc-protected amino acids by TBTU-mediated amide formation to afford 9b–9z. Compounds 9b–9z were converted to hydroxamic acid compounds 11a–11z by NH2OK in dry methanol.
Figure 2
Modification
site and general structure of target compounds.
Reagents and conditions: (a)
CH3COCl, CH3OH, 95%; (b) (Boc)2O,
Et3N, CH2Cl2, 90%; (c) LiAlH4, anhydrous THF, 86%; (d) PPh3, DEAD, anhydrous
THF, 55%; (e) AcOEt/HCl, 80%; (f) carboxylic acids or Boc-protected
amino acids, TBTU, Et3N, anhydrous CH2Cl2 (60–65%); (g) NH2OH, KOH, anhydrous CH3OH, (30–40%).Compound 4 was treated with CH3I and phase
transfer catalyst TBABr in H2O/THF containing KOH to get
the methyl protected 12 (Scheme 2). Subsequent reaction of 12 with 6 through
Mitsunobu reaction in the presence of PPh3 and DEAD afforded
compound 13. Finally, the hydroxamic acid group was achieved
in usual conditions to get compound 14.
Reagents and conditions:
(a)
CH3I, KOH, TBABr, H2O, THF, 50%; (b) PPh3, DEAD, anhydrous THF, 55%; (c) NH2OK, anhydrous
CH3OH, 30%.Scheme 3 shows the preparation of secondary
amine 21a–21c. Intermediate 8 was mixed with benzaldehyde, 4-chlorobenzaldehyde, or 4-fluorobenzaldehyde
in anhydrous CH3OH, respectively, then NaBH4 was added to give 18a–18c. Boc-protected
products 19a–19c were converted to
hydroxamic acids 20a–20c. Subsequent
deprotection gave the end-products 21a–21c.
Reagents and conditions: (a)
benzaldehyde (4-chlorobenzaldehyde or 4-fluorobenzal dehyde), Et3N, NaBH4, anhydrous CH3OH, 85%; (b)
(Boc)2O, Et3N, CH3OH, 65%; (c) NH2OK, anhydrous CH3OH, 30–40%; (d) AcOEt/HCl,
70%.Unlike Scheme 1, Scheme 4 used 3-hydroxycinnamic acid instead
of 4-hydroxycinnamic
acid to give another new series of compounds 27a–27c. The reagents and conditions of Scheme 4 are the same as those in Scheme 1.
HeLa Cell Extract Inhibition of the Target
Compounds
In our previous study we found the lead compound 41 displayed
modest inhibition toward HDACs.[24] At the
beginning of our present research, we made an attempt to find a new
lead compound by modifying the R1 and R2 groups
of 41 (Table 1). We changed the
R2 group of 41 by replacing the methoxy group
with H to get compound 10a (Scheme 1) and changed the R1 group of 10a with a
methyl group to get compound 14 (Scheme 2). We used HeLa cell extract as the enzyme source to efficiently
screen our target compounds. The result in Table 1 showed that compound 10a exhibited slightly
increased HDAC inhibitory activity compared with 41 and 14. This indicated that the methoxy group in R2 and methyl group in R1 were not necessary, so we selected 10a as the lead compound for further optimization and modification.
Table 1
HDAC Inhibition Activity of Compounds 41, 10a, and 13
cpd
R1
R2
IC50 of HeLa extract (nM)a
41
H
CH3O
610.5 ± 20.8
10a
H
H
530.7 ± 18.6
14
CH3
H
600.1 ± 30.7
Assays were performed
in replicate
(n ≥ 2); data are shown as mean ± SD.
Assays were performed
in replicate
(n ≥ 2); data are shown as mean ± SD.In the following work, we designed
an analogue to probe the effect
of the N-substituent by replacing the Boc group of 10a with other functional groups (Scheme 1). First, we synthesized compounds 11a–11o with different substituents including Boc-protected amino
acids, carboxylic acids, sulfo acids, and several anti-inflammatory
acids (Table 2). It was obvious that the compounds
with the structure of acid amide substituents were more potent than
those with sulfamide substituents (Table 2).
For example, the IC50 value of 11g in HeLa
extract was 1703.4 nM, much lower than compound 11h (>5000
nM). The same result was also shown in compounds 11e and 11f. Additionally, compound 11i was more potent
than 11j, which indicated that a short side alkyl chain
might advance HDAC inhibitory activity. Among compounds 11a–11o, the most potent compounds, 11e and 11k, with IC50 values of 15.4 and 20.8
nM, respectively, were far more potent than SAHA (120.8 nM). The common
element they all share is a simple benzamide as their N-substituent, so further modification focused on compound 11e.
Table 2
HDAC Inhibition Activity of Compounds 10a and Its Derivatives 11a–11o
Assays were performed
in replicate
(n ≥ 2); data are shown as mean ± SD.
Assays were performed
in replicate
(n ≥ 2); data are shown as mean ± SD.Next, we synthesized 11p and 11q, which
had different alkyl chain lengths relative to 11e. Activity
data showed 11e with the shortest side chain had the
best activity, which also agreed with the conclusion shown in 11i and 11j, that a short side alkyl chain may
promote their activity. Then we synthesized compounds 11r–11y (Scheme 1) to probe
the effect of substituent in benzene ring of benzamide (Table 3). Surprisingly, the para-substituted compounds 11r, 11w, 11x, and 11y showed a marked increase in the HDAC inhibition potency compared
with 11e, the IC50 values of 11r, 11w, 11x, and 11y were 5.6,
6.7, 20.4, and 4.8 nM, respectively, however, the ortho-substituted,
metha-substituted, and disubstituted compounds 11s, 11t, 11u, and 11v exhibited inferior
activity relative to 11e. The result showed para-substituents
on a benzene ring have an obvious promoting effect on HDAC inhibition.
Table 3
HDAC Inhibition Activity of Compounds 11e and Its Derivatives 11p–11x
Assays were performed
in replicate
(n ≥ 2); data are shown as mean ± SD.
Assays were performed
in replicate
(n ≥ 2); data are shown as mean ± SD.Scheme 4 was designed to evaluate the activity
of 3-hydroxycinnamide-based series. We modified some of the most potent
compounds, 11e, 11r, and 11w, to achieve compounds 27a–27c.
The HDAC inhibition result in Table 4 showed
compounds 27a–27c exhibited poor
activity compared with 11e, 11r, and 11w, which revealed that 4-hydroxycinnamide-based series were
superior to the 3-hydroxycinnamide-based series. Then 21a–21c were designed to improve water solubility and chemical stability
by replacing the amide of 11e, 11r, and 11w with a secondary amine. Compounds 21a–21c showed mild inhibition, and the activity was 15 times lower than 11e, 11r, and 11w. The possible
explanation for why compounds 11r, 21c,
and 27c with similar structure exhibited quite different
activity was described in the molecule docking section.
Table 4
HDAC Inhibition Activity of Compounds 11e, 11r, 11w, 21a–21c, and 27a–27c
Assays were performed in replicate
(n ≥ 2); values are shown as mean ± SD.
Assays were performed in replicate
(n ≥ 2); values are shown as mean ± SD.
In Vitro HDAC Isoform-Selectivity
of the Compounds
In order to explore the HDAC isoform selectivity
profile, we chose
the most potent compounds 11e, 11p, 11r, 11w, and 11y described in the
above discussion, as well as secondary amine analogue 21c and 3-hydroxycinnamide-based analogue 27c to conduct
enzyme inhibitory assays against HDAC1, HDAC2, HDAC3, and HDAC6. Results
presented in Table 5 reveal that compounds 11e, 11r, 11w, and 11y exhibited dual selectivity against HDAC1/3, among which compound 11r exhibited the most obvious dual selectivity, particularly
against HDAC3. IC50 value of 11r against HDAC3
was 3.9 nM, which was ∼80-fold and ∼130-fold lower than
that of HDAC6 and HDAC2, respectively. The positive control, SAHA,
had almost no selectivity toward these HDAC isoforms. To ascertain
the selectivities of our compounds across the broader family of HDAC
isoforms, we next profiled representative 11r against
HDAC8 (class I), HDAC4 (class IIa), and HDAC11 (class IV). Compound 11r also displayed low micromolar activity against HDAC8,
HDAC4, and HDAC11 compared to the low nanomolar activity against HDAC1
and HDAC3 (Table 6). These data establish 11r to be a potent and HDAC1/3 dual selective inhibitor.
Table 5
In Vitro Inhibition of HDACs Isoforms
of Representative Compoundsa
cpd
IC50 of HDAC1 (nM)
IC50 of
HDAC2 (nM)
IC50 of HDAC3 (nM)
IC50 of
HDAC6 (nM)
HDAC2/ HDAC3
isoform selectivity
HDAC6/ HDAC3
isoform selectivity
HDAC2/ HDAC1
isoform selectivity
HDAC6/ HDAC1
isoform selectivity
11e
10.3
535.5
14.1
142.2
37.9
10.1
51.7
13.7
11p
13.2
432.1
143.0
143.7
3.0
1.0
32.6
10.8
11r
11.8
498.1
3.9
308.2
127.7
79.0
42.2
26.1
11w
16.7
457.0
5.5
101.7
83.1
18.5
27.2
6.01
11y
6.0
413.6
3.2
185.6
129.2
58
68.7
30.8
27c
329.9
498.5
403.2
164.4
1.2
0.4
1.5
0.5
21c
355.4
791.2
47.0
293.6
16.8
6.2
2.2
0.8
SAHA
34.6
184.7
90.1
63.0
2.0
0.7
5.3
1.8
Assays were performed in replicate
(n ≥ 2); the SD values are <20% of the
mean.
Table 6
In Vitro
Inhibition of HDACs Isoforms
of 11ra
classes of
HDACs
isoforms
IC50
fold selective
for HDAC3
fold selective
for HDAC1
class II
HDAC1
11.8
HDAC2
498.1
127.7
42.2
HDAC3
3.9
HDAC8
2000.8
513.0
169.6
class IIa
HDAC4
5700.7
1461.7
483.1
class IIb
HDAC6
308.2
79.0
26.1
class IV
HDAC11
900.4
230.9
76.3
Assays were performed in replicate
(n ≥ 2); the SD values are <20% of the
mean.
Assays were performed in replicate
(n ≥ 2); the SD values are <20% of the
mean.Assays were performed in replicate
(n ≥ 2); the SD values are <20% of the
mean.
In Vitro Antiproliferative
Assay
Compounds 11e, 11p, 11r, 11w, and 11y with the most
potent HDACs inhibitory activity were selected
to test their antiproliferative activity against 10 types of solid
or hematological tumor cell lines, which were most frequently used
in evaluating HDACIs (Table 7). Data showed
cell lines HL60, MCF-7, and A549 had lower sensitivity with our tested
compounds than the other cell lines. All the tested compounds exhibited
much more superior antiproliferative potency than SAHA against U937,
K562, HEL, KG1, MDA-MB-231, PC-3, and HCT116 cells. Compounds 11e, 11r, 11w, and 11y, which displayed the highest HDAC enzyme inhibitory potency, also
exhibited the best antiproliferative activity. Among these analogues,
compound 11r exhibited the highest potency, and humanleukemic monocyte lymphomaU937 seemed to be the most sensitive cell
line to our HDACIs. So cell line U937 was chosen for further evaluation.
Furthermore, cytotoxicity of 11r against Human Umbilical
Vein Endothelial Cells (HUVEC) was tested in our lab, and the IC50 value was 61.48 ± 6.17 μM, which revealed our
compounds’ selectivity over nontransformed cells when compared
with tumor cells.
Table 7
In Vitro Antiproliferative Activity
of Representative Compounds
IC50 (μM)a
cpd
U937
K562
HEL
KG1
HL60
MDA-MB-231
PC-3
MCF-7
HCT116
A549
11e
0.33
0.79
0.20
0.39
2.11
0.24
0.33
3.47
0.37
3.39
11p
0.32
0.68
0.27
0.72
1.59
0.41
0.53
2.95
0.57
3.91
11r
0.16
0.51
0.19
0.22
1.69
0.22
0.46
2.68
0.52
2.74
11w
0.18
1.01
0.19
0.24
1.04
0.27
0.51
2.7
0.37
2.96
11y
0.34
0.89
0.16
0.47
1.68
0.15
0.29
2.32
0.22
3.27
21c
0.73
0.94
0.27
0.49
2.17
0.31
0.37
3.25
0.58
3.19
SAHA
1.45
3.24
0.49
1.59
4.26
1.72
3.57
3.78
2.81
3.9
Assays were performed in replicate
(n ≥ 2); the SD values are <20% of the
mean.
Assays were performed in replicate
(n ≥ 2); the SD values are <20% of the
mean.
Western Blot Analysis and
Flow Cytometry Analysis
Although
the HDAC inhibitory activity of our HDACIs was confirmed in cell-free
assays, the results were not as clear in cell based assays. In order
to investigate the HDAC inhibitory activity of our HDACIs in cells
further, we chose compounds that were used in evaluating HDAC isoform-selectivity
to conduct Western blot assay in U937 cell. Intracellular levels of
acetylated histoneH3, acetylated histoneH4 (known substrates for HDAC1,
2, and 3), and acetylated α-tubulin (a known substrate for HDAC6)[27] were compared with SAHA and LBH589 (one of the
most potent pan-HDACIs) as reference compounds (Figure 3). Results show compounds 11e, 11p, 11r, 11w, and 11y can markedly
increase the level of acetylated histoneH3 and acetylated histoneH4,
which were much superior to SAHA, while comparable to LBH589. However,
the levels of acetylated α-tubulin of these compounds, especially 11r, were much lower than LBH589. These data demonstrated
that some of our compounds showed lower potency against HDAC6 than
class I HDACs, which is in agreement with the isoform selectivity
data observed in cell-free assays.
Figure 3
Western blot analysis of acetylated tubulin,
acetylated histone
H3, acetylated histone H4, and pro-caspase3 in U937 cell lines after
24 h treatment with compounds at 1 μM using SAHA (SA) and LBH589
(LBH) as positive control. Hsp90 was used as a loading control.
Western blot analysis of acetylated tubulin,
acetylated histone
H3, acetylated histone H4, and pro-caspase3 in U937 cell lines after
24 h treatment with compounds at 1 μM using SAHA (SA) and LBH589
(LBH) as positive control. Hsp90 was used as a loading control.As our compounds displayed HDAC1/3
dual selectivity and some studies
reported that knockdown of HDAC1 and HDAC3 could increase the percent
of apoptotic cells in carcinoma cells,[11] we also evaluated the level of procaspase3 (inactive form of apoptotic
effector caspase3) in U937 cells (Figure 4).
Cleavage of pro-caspases3 results in the production of an active effector,
caspase3, which can cleave essential structural proteins such as cytokeratins,
nuclear lamins, and also an inhibitor of caspase-activated DNase (iCAD),
which liberates the DNase (CAD) to digest chromosomal DNA and cause
cell death.[28] We can see from the result
that our compounds 11e, 11k, 11r, 11w, and 11y dramatically reduced the
level of procaspase3, which was most likely due to the fact that procaspase3
was cleaved to the active form (caspase3), which could promote apoptosis
of U937 cells. At the same time, we performed flow cytometry analysis
to confirm the effect of these compounds in inducing apoptosis. In
this assay, we chose compounds 11r and 11w as representative compounds and SAHA as the positive control. Results
demonstrated that 11r, 11w, and SAHA can
significantly induce apoptosis in a time-dependent and dose-dependent
manner. At the same concentration (1 μM), 11r and 11w induced 61.76% and 64.25% cell apoptosis, respectively,
which were much higher than that of SAHA (26.83%). Detailed activity
evaluation and apoptosis mechanism studies are currently underway
in our laboratory.
Figure 4
Induction of apoptosis at 12 and 24 h by compounds 11r, 11w, and SAHA in different concentrations
in U937
cells.
Induction of apoptosis at 12 and 24 h by compounds 11r, 11w, and SAHA in different concentrations
in U937
cells.
Stability of Compounds
in Artificial Gastric Juice, Artificial
Intestinal Juice, Rat Liver Homogenate, and Human Plasma
Having demonstrated that 11r had ideal in vitro activity,
we next set out to assess its tumor growth inhibitory effects in subcutaneous
cell lines and primary tumor derived xenograft models. Before we conducted
the in vivo study, we developed HPLC methods to briefly study the
stability of 11r in artificial gastric juice, artificial
intestinal juice, rat liver homogenate, and human plasma. Compound 11r was incubated under each condition at 37 °C for 24
h, extracted, and analyzed by HPLC. We observed that 11r was stable in artificial gastric juice, artificial intestinal juice,
rat liver homogenate, and human plasma (Figure 5).
Figure 5
Stability of compound 11r in artificial gastric juice,
artificial intestinal juice, rat liver homogenate, and human plasma.
Points were achieved after 0, 1, 2, 4, 6, 8, and 24 h, respectively,
and values are shown as mean ± SD.
Stability of compound 11r in artificial gastric juice,
artificial intestinal juice, rat liver homogenate, and human plasma.
Points were achieved after 0, 1, 2, 4, 6, 8, and 24 h, respectively,
and values are shown as mean ± SD.
In Vivo Antitumor Activity Assay
Because of the great
stability of 11r, it was feasible to evaluate its in
vivo activity orally. We established a subcutaneous U937 xenograft
model using SAHA as positive control. U937 cells (5 × 106) were subcutaneously implanted in the right flanks of female
nude mice (BALB/c-nu). When tumor size reached about 100 mm3, mice were randomized to five per group and were treated with compound 11r or SAHA (100 mg/kg/day) by oral gavage for 16 days. Tumor
growth inhibition (TGI) and relative increment ratio (T/C) were calculated
at the end of treatment to reveal the antitumor effects in tumor weight
and tumor volume, respectively. Compound 11r demonstrated
potent in vivo oral antitumor activity with higher TGI value (55.1%)
and lower T/C value (37%) than SAHA (TGI = 32.1%; T/C = 47%). In the
mice group treated by 11r, no significant body weight
loss and no evident toxic signs in liver and spleen were detected
(Figures 6 and 7).
Figure 6
Antitumor
activity comparison of 11r and SAHA against
U937 human tumor xenografts implanted in mice, expressed as mean tumor
volume. Values are shown as mean ± SD (5 mice per group).
Figure 7
Picture of dissected U937 tumor tissues.
Antitumor
activity comparison of 11r and SAHA against
U937humantumor xenografts implanted in mice, expressed as mean tumor
volume. Values are shown as mean ± SD (5 mice per group).Picture of dissected U937tumor tissues.
Docking Results
Figure 8a revealed
the docked conformation of compounds 11r, 21c, and 27c. The docked conformations of 11r and 21c were very similar in the linker and ZBG, while
they were quite different in the aromatic ring and indole ring of
cap group. Moreover, conformation of 27c was different
from 11r and 21c, leading to diverse binding
modes shown in Figure 8c–e. Figure 8c,d showed 11r and 21c had similar docking mode and hydrogen bonds; the greater potency
of 11r might be amenable to the stiffening of the scaffold
introduced by the amide bond, which forces the lipophilic p-chlorophenyl moiety to yield closer van der Waals interactions
with the protein rather than projecting itself toward solvent as in 21c. Furthermore, in cap group, hydrogen bonds formed between 21c and residues Asp92 and Asp93 had a longer bond length
(2.19 and 1.97 Å) than 11r (1.95 and 1.59 Å),
which also might be factors that made 11r more effective
against HDAC3 than 21c. Compared with 11r (Figure 8c), compound 27c (Figure 8e) could form two diverse hydrogen bonds with amide
N–H of Phe200 and imidazole N–H of His172 through carbonyl
oxygen atom and etheric oxygen atom, respectively, but at the same
time, 27c lost two hydrogen bonds with Asp92 in the cap
group and Tyr298 around zinc ion. In addition, the carbonyl oxygen
of 27c had the same distance with 11r to
zinc ion (2.1 Å), but the hydroxyl oxygen of 27c had a longer distance to the zinc ion (4.4 Å) than 11r (1.9 Å), which indicated that 27c could only form
a weak monodentate interaction with zinc, while 11r could
form a strong bidentate interaction with zinc. This might be the reason
why 27c was much less potent than 11r (Figure 8b).
Figure 8
(a) Docked conformational alignment modeling of compounds 11r, 21c, and 27c. For clarity,
HDAC3 (derived by modification of PDB code 4A69 using Tripos SYBYL 8.0) used for docking is not shown. (b) Proposed
docked mode of 27c and 11r for binding zinc.
(c–e) Proposed binding preferences comparison of compounds 11r (c), 21c (d), and 27c (e) at
HDAC3. The zinc ion is shown as a red sphere, and the dashed lines
stand for the hydrogen bonds (atom types: H, white; N, blue; O, red).
(a) Docked conformational alignment modeling of compounds 11r, 21c, and 27c. For clarity,
HDAC3 (derived by modification of PDB code 4A69 using Tripos SYBYL 8.0) used for docking is not shown. (b) Proposed
docked mode of 27c and 11r for binding zinc.
(c–e) Proposed binding preferences comparison of compounds 11r (c), 21c (d), and 27c (e) at
HDAC3. The zinc ion is shown as a red sphere, and the dashed lines
stand for the hydrogen bonds (atom types: H, white; N, blue; O, red).
Conclusions
For
the first time, N-hydroxycinnamamide-based
derivatives have been shown to be potent dual HDAC1/3 selective HDACIs.
The representative compound 11r showed low nanomolar
IC50 values against HDAC1 (11.8 nM) and HDAC3 (3.9 nM),
with micromolar or submicromolar IC50 values against HDAC2,
HDAC8, HDAC4, HDAC6, and HDAC11. In addition, a few of the examples
of the new series in antiproliferative study exhibited high potency
against several solid or hematological cells, even though antiproliferative
activity of the compounds seemed inferior to their HDAC inhibition
activity. For example, antiproliferative activity of 11r was about 10 times higher than SAHA, while its HDAC inhibition activity
was about 100 times higher than SAHA. We supposed 11r had poor transcellular permeability; this speculation was confirmed
by parallel artificial membrane permeation assay (PAMPA)[29] conducted on 11r and SAHA (data
not shown). Future work to improve potency should focus on enhancing
permeability of 11r.Compound 11e, 11r, 11w,
and 11y increased histone H3 and H4 acetylation in the
same level of LBH589 (one of the most potent hydroximates HDACIs in
clinical) in U937 cell line. At the same time, some of the compounds
decreased the level of pro-caspase3, which was consistent with the
result of flow cytometry analysis in inducing apoptosis. Compound 11r was evaluated for its in vivo antitumor study in a U937
xenograft mice model and showed higher efficacy compared to the approved
drug SAHA. More detailed studies of the antitumor profiles of these
promising compounds are underway in our lab. Using 11r as lead, continued studies are underway to search for more promising
HDACIs with improved transcellular permeability and isoform selectivity.
Experimental Section
Materials and Methods
1H NMR spectra were
obtained on a Bruker DRX spectrometer at 300 MHz with TMS as an internal
standard, δ in parts per million, and J in
hertz. High-resolution mass spectrometry was performed by Shandong
Analysis and Test Center in Ji’nan, China. ESI-MS spectra were
recorded on an API 4000 spectrometer. All reactions were monitored
by TLC using 0.25 mm silica gel plates (60GF-254). UV light and ferric
chloride were used to visualize the spots. Silica gel was used for
column chromatography purification. Flash chromatography was accomplished
using the automated CombiFlash Rf system from Teledyne ISCO and was
done using silica gel of 200–300 mesh. Melting points were
determined on an electrothermal melting point apparatus. All tested
compounds are >95% pure by HPLC analysis, performed on an Agilent
1100 HPLC instrument using an ODS HYPERSIL column (5 μm, 4.6
mm × 250 mm) according to one of the following methods. Method
A: compounds 10a, 11a–11x, and 27a–27c were eluted with 35%
acetonitrile/65% water (containing 0.4% formic acid)–65% acetonitrile/35%
water (containing 0.4% formic acid) over 20 min, with detection at
290 nm and a flow rate of 1 mL/min. Method B: compounds 21a-21c were eluted with 30% acetonitrile–70% 0.05
M potassium dihydrogen phosphate/phosphoric acid (PH = 3.0) over 20
min, with detection at 290 nm and a flow rate of 1 mL/min. The temperature
of the column was 25 °C, and quantity of injection was 20 μM.(S)-Methyl 2-amino-3-(1H-indol-3-yl)propanoate
hydrochloride (2), (S)-Methyl 2-((tert-butoxycarbonyl)amino)-3-(1H-indol-3-yl)propanoate(3), and (S)-tert-butyl(1-hydroxy-3-(1H-indol-3-yl)propan-2-yl)carbamate (4) were
synthesized as described previously.[24]
General Procedure for the Preparation of 6 and 23
(E)-Methyl 3-(4-hydroxyphenyl)acrylate. (6)
p-Coumaric acid (16.4g, 100mmol)
was dissolved in 200 mL of MeOH, then acetyl chloride (24g, 300mmol)
was added dropwise at 0 °C; the mixed solution was refluxed at
75 °C for 5 h. The solvent was evaporated under vacuum; the product
was dissolved with EtOAc, washed by saturated NaHCO3 solution
(2 × 100 mL), 1 M aqueous citric acid (2 × 100 mL), and
brine (2 × 100 mL), dried over MgSO4 overnight, and
evaporated under vacuum to obtained product compound 6, a white solid (15.6g, 87.5%). Mp: 140–141 °C. 1H NMR (300 MHz, DMSO-d6) δ
3.69 (s, 3H), 6.43 (d, J = 16.2 Hz, 1H), 6.81 (d, J = 8.7 Hz, 2H), 7.54–7.59 (m, 3H), 10.01 (s, 1H).
ESI-MS m/z: 179.2 [M + H]+.
The solution of compound 8 (0.76g, 2mmol) in CH3OH was added Et3N (0.22g,
2.2mmol) for neutralization,
followed by the addition of methanol (0.21g, 2mmol); the resulting
solution was stirred for 2 h and the solvent evaporated. The residue
was dissolved in anhydrous CH3OH, NaBH4 (0.15g,
4mmol) was added slowly at 0 °C, the mixture was stirred at room
temperature overnight, and after the reaction finished, the reaction
was quenched by adding ice water slowly. CH3OH was evaporated,
and the residue was extracted by EtOAc (3 × 30 mL). The organic
layer was washed with brine (2 × 30 mL) and saturated Na2CO3 (2 × 30 mL) and dried over MgSO4 overnight. The solvent was evaporated under vacuum to get crude
product 11a, a colorless oil (0.79g, 90%). 1H NMR (300 MHz, DMSO-d6) δ 2.95
(d, J = 6.3 Hz, 2H), 3.14–3.18 (m, 1H), 3.70
(s, 3H), 3.87 (s, 2H), 3.94 (d, J = 5.1 Hz, 2H),
6.50 (d, J = 15.9 Hz, 1H), 6.92 (t, J = 7.5 Hz, 1H), 6.95 (d, J = 8.7 Hz, 2H), 7.06 (t, J = 7.4 Hz, 1H), 7.14 (d, J = 2.4 Hz, 1H),
7.16–7.22 (m, 1H), 7.26–7.33 (m, 5H), 7.43 (d, J = 7.8 Hz, 1H), 7.63 (d, J = 15.9 Hz,
1H), 7.65 (d, J = 8.7 Hz, 2H), 8.99 (s, 1H), 10.83
(s, 1H). ESI-MS m/z: 441.5 [M +
H]+.
KOH (28g, 509mmol) and NH2OH·HCl (23.35g, 343mmol)
were dissolved, respectively, in 70 and 120 mL of MeOH to get solution
A and solution B. Next, solution A was added dropwise to solution
B. After filtering the precipitate (KCl), a NH2OK solution
was obtained. Compound 7 (0.45g, 1 mmol) was dissolved
in the NH2OK solution and stirred overnight. After the
reaction was complete, it was evaporated under vacuum. The residue
was acidified with 1 N HCl to a pH 3–4 and then extracted with
EtOAc (3 × 20 mL). The organic layer was washed with brine (2
× 30 mL) and dried over Na2SO4 overnight.
The crude material was purified via flash chromatography to afford
the compound 10a (0.15g, 30% yield). Mp: 151–152
°C. 1H NMR (300 MHz, DMSO-d6) δ 1.37 (s, 9H), 2.79–3.00 (m, 2H), 3.90–4.04
(m, 3H), 6.33 (d, J = 15.6 Hz, 1H), 6.92–7.00
(m, 4H), 7.08 (td, J = 0.9 Hz, J = 7.6 Hz, 1H), 7.13 (d, J = 1.8 Hz, 1H), 7.34 (d, J = 7.8 Hz, 1H), 7.42 (d, J = 15.6 Hz,
1H), 7.48 (d, J = 8.7 Hz, 2H), 7.56 (d, J = 7.8 Hz, 1H), 8.97 (s, 1H), 10.66 (s, 1H), 10.83 (d, J = 1.8 Hz, 1H). HRMS (AP-ESI) m/z calcd for C25H29N3O5 [M + H]+ 452.2180, found 452.2182. Retention time: 18.9
min, eluted with 40% acetonitrile/60% water (containing 0.4% formic
acid).
In vitro
HDACs inhibition assays were conducted as previously described.[22] In brief, 10 μL of enzyme solution (HeLa
nuclear extract, HDAC1, HDAC2, HDAC3. or HDAC6) was mixed with various
concentrations of tested compound (50 μL). The mixture was incubated
at 37 °C for 5 min, then luorogenic substrate Boc-Lys(acetyl)-AMC
(40 μL) was added. After incubation at 37 °C for 30 min,
the mixture was stopped by the addition of 100 μL of developer
containing trypsin and TSA. Twenty minutes later, fluorescence intensity
was measured using a microplate reader at excitation and emission
wavelengths of 390 and 460 nm, respectively. The inhibition ratios
were calculated from the fluorescence intensity readings of tested
wells relative to those of control wells, and the IC50 values
were calculated using a regression analysis of the concentration/inhibition
data.
In Vitro Antiproliferative Assay
All cell lines were
maintained in RPMI1640 medium containing 10% FBS at 37 °C in
a 5% CO2 humidified incubator. Cell proliferation assay
was determined by the MTT method. Briefly, cells were passaged the
day before dosing into a 96-well cell plate, allowed to grow for 12
h and then treated with different concentrations of compound sample
for 48 h. A 0.5% MTT solution was added to each well. After incubation
for another 4 h, formazan formed from MTT was extracted by adding
150 μL of DMSO rocking for 15 min. Absorbance was then determined
using an ELISA reader at 570 nm.
Western Blot Analysis
Briefly, U937 cells with different
treatment were collected and lysed with lysis buffer (20 mM Tris-HCl
[pH 7.5], 1% NP-40, 1 mM sodium vanadate, 1 mM EDTA, 1 mM EGTA, 50
mM NaF, and 1 mM phenylmethyl sulfonylfluoride [PMSF]) for 30 min,
then centrifuged for 15 min at 14 000 rpm at 4 °C, and
the supernatant was the whole-cell extracts. Total protein extracts
(30 μg per lane) were separated by 12% SDSpolyacrylamide gel
electrophoresis and transferred onto PVDF membranes (Cat. IPVH00010,
Millipore). Membrane was blocked with 5% S36 milk in TBS-T (10 mM
Tris [pH 7.4], 150 mM NaCl, and 0.1% Tween 20) for 1 h at room temperature,
then incubated with a 1:1000 or 1:2000 dilution of primary antibody
overnight at 4 °C. Then the membrane was washed three times and
incubated at 1:2000 dilution of antimouse or antirabbit goat-HRP-conjugated
secondary antibodies for 2 h at room temperature. Finally the membrane
was washed another three times and developed by enhanced chemiluminescence
(ECL, Cat. WBKLS0050, Millipore).
Apoptosis Assay
U937 cells (1.5 × 105) were treated with different
concentrations of 11r, 11w, or SAHA for
12 or 24 h. Cells were harvested following
incubation, washed twice in cold PBS, centrifuged, and resuspended
in 1× annexin-binding buffer. Cells were diluted in 1× annexin-binding
buffer to 1 × 106 cells/mL, preparing a sufficient
volume to have 100 μL per assay. Five microliters of Alexa Fluor
488 annexin V and 1 μL of 100 μg/mL PI were added to each
100 μL of cell suspension. Cells were incubated at room temperature
for 15 min. After the incubation period, 150 μL of 1× annexin-binding
buffer was added, mixed gently, and kept on ice. The stained cells
were analyzed by flow cytometry, measuring the fluorescence emission
at 530 and 575 nm (or equivalent) using 488 nm excitation. Alexa Fluor
488 annexin V/Dead Cell Apoptosis Kit with Alexa Fluor 488 annexin
V and PI for Flow Cytometry was used for this assay (Invitrogen).
Stability of Representitive Compound in Artificial Gastric Juice,
Artificial Intestinal Juice, Rat Liver Homogenate, and Human Plasma
Preparation of artificial gastric juice: 160 mL of water and 2
g of pepsin were added to 3.28 mL of diluted hydrochloric acid. After
mixing, the mixture was diluted to 200 mL with water. Preparation
of artificial intestinal juice: 1.36 g of sodium dihydrogen phosphate
was dissolved in 100 mL of water, and the pH was adjusted to 6.8 using
0.4% NaOH. Then 2 g of pancreatin was added, and the mixture was diluted
to 200 mL with water. Preparation of rat liver homogenate: Wistar
rats (Experimental Animal Center, Shandong University, female, 200 ±
20 g) were sacrificed after fasting 12 h. Liver was homogenized in
quadruple phosphate butter (pH 7.4) and centrifuged (4000 rpm), 10
min. The liver homogenate was stored at −20 °C. Human
plasma was obtained from Institution of Clinical Pharmacology, Qilu
Hospital. Artificial gastric juice, artificial intestinal juice, rat
liver homogenate, and human plasma were preincubated at 37 °C
prior to the addition of 50 μL of stock solution of mutual prodrugs
(1 mg/mL in CH3OH). Sample aliquots were taken for processing
and analysis after 24 h incubation at 37 °C. All of the samples
underwent extraction using 600 μL acetonitrile and were filtered
(0.22 μm) after shocking 30 s and centrifugation at 12 000
rpm, 10 min. Chromatographic analysis was achieved by HPLC method
on an Agilent 1100 HPLC instrument using a ODS HYPERSIL column (5
μm, 4.6 mm × 100 mm) according the following method: compound
was eluted with 42% acetonitrile/58% water (containing 0.4% formic
acid) over 20 min, with detection at 290 nm and a flow rate of 1 mL/min.
The temperature of the column was 25 °C, and the quantity of
injection was 20 μM.
In Vivo Antitumor Assay against U937 Xenograft
For
in vivo antitumor efficacy research, 1 × 107 human
histiocytic lymphoma cells (U937) were inoculated subcutaneously in
the right flank of male athymic nude mice (5–6 weeks old, SLAC
LABORATORY ANIMAL, Shanghai). Ten days after injection, tumors were
palpable, and mice were randomized into treatment and control groups
(5 mice per group). The treatment groups received compound 11r (100 mg/kg/d) or SAHA (100 mg/kg/d) by oral administration, and
the blank control group received equal volume of PBS solution containing
40% DMSO. During treatment, subcutaneous tumors were measured with
a vernier caliper every three days, and body weight was monitored
regularly. After treatment, mice were sacrificed and dissected to
weigh the tumor tissues and to examine the internal organ injury.
Tumor growth inhibition (TGI) and relative increment ratio (T/C) were
used as the evaluation indicators to reveal the antitumor effects
in tumor weight and tumor volume, respectively. Data were analyzed
by Student’s two-tailed t test. A P level < 0.05 was considered statistically significant.Tumor
volumes (V)
were estimated using the equation (V = ab2/2, where a and b stand
for the longest and shortest diameter, respectively). T/C was calculated
according to the following formula:where RTV is the relative tumor volume = Vt/V0 (Vt, the tumor volume measured at the end of treatment; V0, the tumor volume measured at the beginning
of treatment).
Molecular Docking Studies
Compounds
were docked into
the active site of HDAC3 (PDB code 4A69) using Tripos SYBYL 8.0. Before the docking
process, the protein structure was treated by deleting water molecules,
adding hydrogen atoms, fixing atom types, and assigning AMBER7 FF99
charges. A 100-step minimization process was performed to further
optimize the protein structure. The molecular structures were generated
with the Sybyl/Sketch module and optimized using Powell’s method
with the Tripos force field with convergence criterion set at 0.05
kcal/(Å mol) and assigned charges with the Gasteiger–Hückel
method. Molecular docking was carried out via the Sybyl/FlexX module.
Other docking parameters were kept to the default values.
Authors: Roland W Bürli; Christopher A Luckhurst; Omar Aziz; Kim L Matthews; Dawn Yates; Kathy A Lyons; Maria Beconi; George McAllister; Perla Breccia; Andrew J Stott; Stephen D Penrose; Michael Wall; Marieke Lamers; Philip Leonard; Ilka Müller; Christine M Richardson; Rebecca Jarvis; Liz Stones; Samantha Hughes; Grant Wishart; Alan F Haughan; Catherine O'Connell; Tania Mead; Hannah McNeil; Julie Vann; John Mangette; Michel Maillard; Vahri Beaumont; Ignacio Munoz-Sanjuan; Celia Dominguez Journal: J Med Chem Date: 2013-12-05 Impact factor: 7.446
Figure 1
Figure 2
Scheme 1
Scheme 2
Scheme 3
Scheme 4
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8