Discovery of (2 E)-3-{2-butyl-1-[2-(diethylamino)ethyl]-1H-benzimidazol-5- yl}-N-hydroxyacrylamide (SB939), an orally active histone deacetylase inhibitor with a superior preclinical profile

ARTICLE

pubs.acs.org/jmc

Discovery of (2E)-3-{2-Butyl-1-[2-(diethylamino)ethyl]-1H-

benzimidazol-5-yl}-N-hydroxyacrylamide (SB939), an Orally Active

Histone Deacetylase Inhibitor with a Superior Preclinical Profile

Haishan Wang,*^,†Niefang Yu,^†Dizhong Chen,^†Ken Chi Lik Lee,^†Pek Ling Lye,^†Joyce Wei Wei Chang,^†

Weiping Deng,^†Melvin Chi Yeh Ng,^†Ting Lu,^†Mui Ling Khoo,^†Anders Poulsen,^†Kanda Sangthongpitag,^‡

Xiaofeng Wu,^‡Changyong Hu,^‡Kee Chuan Goh,^‡Xukun Wang,^‡Lijuan Fang,^‡Kay Lin Goh,^‡

Hwee Hoon Khng,^‡Siok Kun Goh,^‡Pauline Yeo,^§Xin Liu,^§Zahid Bonday,^‡Jeanette M. Wood,^‡

Brian W. Dymock,^†Kantharaj Ethirajulu,^§and Eric T. Sun^†

^†Chemistry Discovery, ^‡Biology Discovery, and ^§Pre-Clinical Development, S*BIO Pte Ltd., 1 Science Park Road,

No. 05-09 The Capricorn, Singapore Science Park II, Singapore 117528, Singapore

S Supporting Information

b

ABSTRACT: A series of 3-(1,2-disubstituted-1H-benzimida-

zol-5-yl)-N-hydroxyacrylamides (1) were designed and synthe-

sized as HDAC inhibitors. Extensive SARs have been

established for in vitro potency (HDAC1 enzyme and COLO

205 cellular IC₅₀), liver microsomal stability (t_1/2), cytochrome

P450 inhibitory (3A4 IC₅₀), and clogP, among others. These

parameters were ﬁne-tuned by carefully adjusting the substitu-

ents at positions 1 and 2 of the benzimidazole ring. After

comprehensive in vitro and in vivo proﬁling of the selected compounds, SB939 (3) was identiﬁed as a preclinical development

candidate. 3 is a potent pan-HDAC inhibitor with excellent druglike properties, is highly eﬃcacious in in vivo tumor models (HCT-

116, PC-3, A2780, MV4-11, Ramos), and has high and dose-proportional oral exposures and very good ADME, safety, and

pharmaceutical properties. When orally dosed to tumor-bearing mice, 3 is enriched in tumor tissue which may contribute to its

potent antitumor activity and prolonged duration of action. 3 is currently being tested in phase I and phase II clinical trials.

’ INTRODUCTION

selective knockdown studies on HDACs suggest that the class I

HDACs, particularly HDACs 1 and 3, are essential to the proliferation

and survival of mammalian carcinoma cells.^8,9Class I HDACs, especi-

ally HDACs 1, 2, and 3, are considered as the most relevant targets for

cancer therapy because inhibitors of these enzymes usually show

strong antiproliferative and apoptosis-inducing activity.^9,10Class II

HDACs are more speciﬁcally involved in regulating cell migration

and angiogenesis. Non-histone proteins are deacetylated by classes

II and IV HDACs located predominantly in the cytoplasm⁷as well

as by class I HDACs. For example, HDAC6 deacetylates tubulin and

HSP90, and both HDAC1 and SIRT1 deacetylate p53.¹¹Class III

HDACs (sirtuins) may also play an important role in regulating

tumor onset and/or progression. At present it remains unclear

exactly what role sirtuins may play in oncogenesis and if their

dominant role may be as tumor suppressors or oncogenes.⁷

HDAC inhibitors have been studied for their therapeutic

eﬀects on cancer cells.¹²Suberoylanilide hydroxamic acid (SAHA,

vorinostat) is the ﬁrst HDAC inhibitor approved by the FDA in

2006 for the treatment of cutaneous T-cell lymphoma (CTCL).¹³

In November 2009, FDA also approved romidepsin (FK228) for

treatment of CTCL in patients who have received at least one prior

Genomic DNA is packaged with histone to form chromatin,

which is further condensed to chromosomes.¹The histone proteins

play an important role in the control of gene expression via modiﬁca-

tion through chemical reactions such as acetylation, phosphorylation,

and methylation. The amino termini of histones extend from the

nucleosomal core and are modiﬁed by histone acetyltransferases and

histone deacetylases (HDACs) during the cell cycle.²In most cases,

histone acetylation enhances transcription while histone deacetylation

represses transcription.^3,4Inhibition of HDACs leads to the accumu-

lation of acetylated histones, resulting in a variety of cell type depen-

dent responses such as apoptosis, necrosis, diﬀerentiation, cell sur-

vival, inhibition of proliferation, and cytostasis. HDAC enzymes are

divided into four diﬀerent classes:^5,6class I (HDACs 1ꢀ3, 8), class IIa

(HDACs 4, 5, 7, 9), class IIb (HDACs 6, 10), class III (SIRTs 1ꢀ7),

and class IV (HDAC 11). Classes I, II, and IV are zinc dependent

enzymes. Class III HDACs, the sirtuins, employ NAD^þas a cofactor

for their catalytic activity instead of zinc and are generally not inhibited

by class I and class II inhibitors. Regulation of chromatin structure via

modulation of histone acetylation is generally regarded as the primary

mechanism of action of the class I HDACs. Subsequent eﬀects on cell

proliferation and apoptosis single out these nuclear proteins as key

targets for an anticancer HDAC inhibitor.⁷HDACs1, 2, 3, and8have

been associated with uncontrolled tumor growth. For example,

Received: March 28, 2011

Published: June 02, 2011

r

2011 American Chemical Society

4694

dx.doi.org/10.1021/jm2003552 J. Med. Chem. 2011, 54, 4694–4720

|

Journal of Medicinal Chemistry

ARTICLE

Figure 1. Clinically tested HDAC inhibitors.

systemic therapy.¹⁴These two approved drugs have validated the

therapeutic use of HDAC inhibitors in cancer therapy.

There are a number of HDAC inhibitors that are currently

undergoing clinical trials either as a single agent therapy or in com-

bination with standards of care (SOC) or other targeted therapies for

the treatment of solid and hematologic malignancies (Figure 1). Most

of the compounds are pan-inhibitors, but there are also isoform

selective or class selective HDAC inhibitors in development.^15,16The

morphology of the HDAC inhibitor pharmacophore, as exempliﬁed

by vorinostat (Figure 1), is characterized by three portions: a metal- or

zinc-binding group (ZBG), a hydrophobic group (CAP) for protein

surface recognition or interaction, and a linker to connect both ZBG

and CAP. The common linkers are aliphatic chain (e.g., six-carbon

chain in vorinostat), aromatic ring (e.g., 1,4-phenylene in entinostat),

and vinylꢀaromatic (e.g., styryl in belinostat). The most common

ZBGs are hydroxamic acid and benzamide. The disulﬁde bond of

romidepsin can be reduced in vivo by glutathione (GSH) to form Zn-

binding free thiols and conjugates with GSH. There are other types of

ZBGs such as ketone, thiol, R-acetylmercapto ketone, and so on,¹⁵but

hydroxamic acid remains the most potent ZBG reported for inhibition

of class I HDACs.

Figure 2. Benzimidazole based hydroxamic acid 1 and structures of 2,

2aꢀc, and 3.

series of hydroxamates.¹⁷We recently reported N-hydroxy-1,2-

disubstituted-1H-benzimidazol-5-ylacrylamides (1) (Figure 2)

as novel HDAC inhibitors, and SB639 (2),¹⁸one of the represen-

tative compounds, showed promising pharmacological and

pharmacokinetic properties. In general, compounds with good

in vitro potency from the initial series 1 (e.g., 2, 2aꢀc, Figure 2) are

As part of our ongoing eﬀort to discover novel anticancer

agents, we have designed and synthesized a number of chemical

4695

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Figure 3. Representative R¹side chains for optimization of benzimidazole based hydroxamic acid 1. B8, C1, C4, and C7 are Boc protected R¹groups

and used for syntheses of intermediates for 1.

metabolically unstable in human liver microsomal assays, with the

exception of 2.¹⁸A preliminary metabolism study using rat liver

hepatocytes revealed that 2 was metabolized by oxidation of the R¹

group (pyrrolidinylethyl) and reduction of hydroxamic acid (CON-

HOH) to amide (CONH₂) which is no longer a potent ZBG. The

formation of the amide metabolite was also observed after 1 h of

incubation of 2 with human hepatocytes. Thus, metabolic stability

was identiﬁed as a major issue for this series of compounds. From a

medicinal chemistry perspective, the primary focus was then to

synthesize compounds metabolically stable in both human assay

systems and preclinical species. Herein, we describe the further

optimization and development of series 1 that led to the discovery

of SB939 (3) (Figure 2),^15,16,19currently in multiple phase I and

phase II clinical trials.^20,21

reported earlier¹⁸and extended herein to introduce more com-

plex R¹and R²groups. Representative R¹side chains are listed in

Figure 3, and R²side chains are listed in Figure 4. Four types of R¹

were selected. Type A series have linkers of three carbons

between N¹of the benzimidazole ring and the basic nitrogen

in the R¹group. Type B series have two-carbon linkers,. C series

have a rigid ring system with a two-carbon or three-carbon

equivalent linker, and type D series R¹groups are neutral. R¹side

chains with a free basic NH were further derivatized via reductive

amination, alkylation with alkyl halides, or acylation. For example

(Figure 3), B21ꢀ29, B31ꢀ37, B61ꢀ62, and B71ꢀ77 are derived

from B2, B3, B6, and B7, respectively. Three types of R²were

selected: Type a series are R-branched either cyclic or acyclic.

Type b series are R-unsubstituted with either a methylene linker,

hydrogen, methyl or trifluoromethyl, and type C series are basic.

Chemical synthesis and characterization of key compounds

which were used for SAR determination and in vivo evaluations

are described in the Experimental Section, and those compounds

used for general supportive SAR establishment are described in

’ RESULTS AND DISCUSSION

Chemistry. An efficient synthesis of a wide range of benzimid-

azole based hydroxamic acids 1 (Figure 2) was developed and

4696

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Figure 4. Representative R²side chains for optimization of benzimidazole based hydroxamic acid 1.

Scheme 1^a

^aReagents and conditions: (a) MeOH, H₂SO₄, reﬂux, 95%; (b) Et₃N, dioxane, 80ꢀ100 ꢀC, 53ꢀ98%; (c) SnCl₂2H₂O (5 equiv), AcOHꢀMeOH

3

(1:9), 40 ꢀC, 15ꢀ65%; (d) NH₂OH HCl (10 equiv)/NaOMe (20 equiv)/MeOH, 0 ꢀC to room temp, 10ꢀ90%; (e) conc HCl, HOAc, 70 ꢀC, de-Boc

3

of 9w to aﬀord 9w1. Codes for R¹and R²groups are deﬁned in Figures 3 and 4, respectively.

the Supporting Information (see Table S1 for a full list of the

target compounds synthesized for this report).

Scheme 1 (route A) illustrates the procedure used for preparing

available trans-3-nitro-4-chlorocinnamic acid (4) which was ester-

iﬁed in acidic methanol to give the methyl ester (5) in almost

quantitative yield. Then the chloride ortho to the activating electron-

withdrawing nitro group was displaced by the appropriate amine

compounds of formula 10 (and 3), starting with commercially

4697

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Scheme 2^a

^aReagents and conditions: (a) SnCl₂2H₂O (5 equiv), AcOHꢀMeOH (1:9), 40 ꢀC; (b) coupling reagent (EDCI/HOBt), DIEA, DCM; (c) HOAc,

3

90ꢀ100 ꢀC; (d) NH₂OH HCl (10 equiv)/NaOMe (20 equiv)/MeOH, 0 ꢀC to room temp; (e) HCl, 70 ꢀC, de-Boc of 14d to give 14d1. Codes for R¹

3

and R²groups are deﬁned in Figures 3 and 4, respectively.

Scheme 3^a

^aReagents and conditions: (a) HCHO, MeOH, NaBH(OAc)₃, room temp; (b) NH₂OH HCl (10 equiv)/NaOMe (20 equiv)/MeOH, 0 ꢀC to

3

room temp.

component R¹NH₂(6) in the presence of a base (e.g., triethylamine)

to give a substituted aniline (7) in 53ꢀ98% yield. The key one-pot

reductive cyclization with aldehyde R²CHO (8) forms the benzimi-

dazole ring.²²In this reaction the nitro group of 7 was reduced by a

reducing agent such as tin(II) chloride in acetic acid, and in the

presence of the appropriate aldehyde component 8, cyclization was

achieved to give benzimidazoles (9) in 15ꢀ65% yield. Methyl esters

9/9aꢀab were treated with hydroxylamine (hydroxylamine hydro-

chloride þ excessive sodium methoxide in methanol) and converted

to the target hydroxamates 3/10aꢀab in 10ꢀ90% yield. For most of

the compounds, the reaction mixture of 3/10aꢀab was quenched

with triﬂuoroacetic acid (TFA) or HCl and puriﬁed by preparative

reverse phase HPLC, aﬀording the ﬁnal products 3/10aꢀab as TFA

salts. Alternatively, the pH of the reaction mixture of 3/10aꢀab was

adjusted to around 8, and the freebase of 3/10aꢀab was isolated by

ﬁltration. The pure base was further converted to the hydrochloride

salt, and the stoichiometry of the salt was determined by elemental

analysis. Compounds used for in vivo evaluations were all hydro-

chloride salts. Chemical structures were conﬁrmed by 1D and 2D

NMR, LCꢀMS, and elemental analysis.

was reduced to aniline 11. Then 11 was subsequently reacted

with acid R²CO₂H (12) to form amide(s) 13 under routine

coupling conditions. Normally, 13 was not isolated or puriﬁed but

used directly for the next step of cyclization forming benzimidazole

14 when heated in acetic acid. This method is applicable to most

acids (12) including acetic acid and TFA. Compound 14d1 was

made by removal of the Boc protecting group of 14d. Methyl esters

14aꢀc, 14d1, and 14e were converted to hydroxamic acids 15aꢀe

as described for synthesis of 3/10aꢀab.

R¹side chains of benzimidzoles 9 could be further derivatized

before converting to hydroxamate. Scheme 3 describes the

derivatization of R¹side chains bearing a secondary amine.

Benzimidazole methyl ester amines 9e, 9n, and 9o were alkylated

by reductive amination. The alkylation products 16aꢀc were

subsequently converted to hydroxamates 17aꢀc.

The above-described methods were also applied to benzimid-

azole methyl ester 18, which has a Boc protected cyclic amine at

the R¹position (Scheme 4). 18aꢀf were made from 7l, 7m, and

7n (Scheme 1) by using methods described for 9 and 9aꢀab.

Cleavage of the Boc group from 18aꢀf was achieved under acidic

conditions (HCl, MeOH, reﬂux) to aﬀord secondary amines

19aꢀf, of which 19aꢀe were converted to hydroxamates 20aꢀe.

When aldehyde 8 was not commercially available, a second

synthetic route (route B, Scheme 2) was used. Nitro compound 7

4698

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Scheme 4^a

^aReagents and conditions: (a) MeOH, 1.25 M HCl, reﬂux; (b) NH₂OH HCl (10 equiv)/NaOMe (20 equiv)/MeOH, 0 ꢀC to room temp; (c) HCHO,

3

NaBH(OAc)₃, MeOH, room temp, 1 h.

Secondary amines 19a (R²= n-butyl) and 19f (R²= n-pentyl)

were alkylated by reductive amination to aﬀord the tertiary

amines 21a,b which were subsequently converted to hydroxa-

mates 22a,b after reaction with hydroxylamine.

The above representative synthetic methods and strategies were

used to prepare the majority of target compounds. Additional

synthetic methods and details not mentioned in the Experimental

Section may be found in the Supporting Information.

In Vitro Biological Evaluation and SAR. HDAC isozyme

HDAC1 and human colon cancer cell line COLO 205 were used

as routine enzymatic and cellular antiproliferation screening

tests, respectively. Vorinostat was used as positive control in all

experiments. After the primary tests, compounds passed the

selection criteria (i.e., IC₅₀(HDAC1) e 50 nM and IC₅₀(COLO

205) e 0.050 μM) and those picked for better understanding of

an individual profile (e.g., 3, 10c, 10n, and 10z) were tested

against additional cancer cell lines.

potency is still weak compared to those with basic centers at R¹

(3, 10aꢀh). Thus, a basic center in the R¹group is crucial for

optimal biological activity of this series.

The above SAR can be satisfactorily explained by docking

studies of compounds into a homology model of the enzyme

active site. Docking compound 3 into the HDAC1 homology

model²⁴revealed a key electrostatic interaction between the R¹

basic center and Asp99 (Figure 5). There may be an additional

hydrogen bonding interaction between benzimidazole N³and

His178 via a water molecule (Figure 5b); thus, both basic centers

contribute to the binding of the most potent compounds. For

compounds with longer three-carbon linkers, due to their

increased distance between the R¹basic center and the benz-

imidazole, the side chain is more ﬂoppy and the conformational

entropy increases. In addition, to make similar interactions with

Asp99 the R¹side chain is forced into a higher energy conforma-

tion; hence, the potency is reduced compared to the more

optimal two-carbon linkers. However, 10f has a bulky gem-

dimethyl group in the R¹side chain and a bulky tert-butylmethyl

group at R²which may force it to adopt a more favorable

conformation for eﬀective interaction with Asp99. The increased

bulk of the R¹side chain increases lipophilic interactions with the

protein. When 10f is docked into the HDAC1 homology model,

the R¹side chain has double the number of atomꢀatom van der

Waals contacts with the protein compared to the R¹side chain of

3. These factors may contribute to the better potency of 10f

among the three-carbon linker series.

Preliminary SAR established in early work¹⁸suggested that

there was a preferred distance between the basic nitrogen of the R¹

side chain and the N¹of the benzimidazole ring for benzimidazole

hydroxamates 1 (R²= PhCH₂CH₂ꢀ). Extension of this SAR to

new hydroxamates 3, 10aꢀab, 15aꢀe, 17aꢀc, 20aꢀe, and22a,b, is

described in Schemes 1ꢀ4, respectively, and biological data are

listed in Table 1. With the same R²group, compounds having R¹

with two-carbon linkers are generally on the more potent side than

those of three-carbon linkers, for example, 3 versus 10a, 10c versus

10b, 10e versus 10d. Of the 60 R¹compounds in this series, there

were 18 with three-carbon linkers (Table S2, Supporting In-

formation) and 42 with two-carbon linkers (Table S3, Supporting

Information); they all followed the general trend of two-carbon

linker being preferred. However, with a bulkier R²(i.e., tert-

butylmethyl) this trend was reversed; for example, 10g is less potent

than 10f. 10f appears to be an outlier, with unexpectedly good

enzyme and cellular potency.

Hydroxamates with directly attached cyclic R¹groups (20aꢀe

and 22a,b, Table 2) have linkers of two to four carbons between

the basic nitrogen in the side chain and the core benzimidazole

nitrogen N¹. They may be considered as rigid close comparators

to the ﬂexible acyclic R¹compounds discussed from Table 1.

With R²ﬁxed as n-butyl, cyclic secondary amines 20aꢀc

exhibited comparable HDAC1 inhibitions (e.g., IC₅₀and BEI)

to their more ﬂexible cousins, but the cellular activities of some

compounds were poor possibly because of the strongly basic

nature of some secondary amines which would have made them

less permeable to cells. However, the data are more complex:

there appears to be a correlation between the number of carbons

(related to lipophilicity (log P)) and the cellular potency. For

example, when R²in 20a was changed to n-hexyl, giving 20d, the

clogP increased by 0.88 unit, leading to signiﬁcant improvement

A basic center is important in the R¹group, as evident by

comparing 10h with 10i (Table 1). 10i lost over an order of

magnitude of its enzymatic and cellular activity, and the binding

eﬃciency index (BEI)²³is signiﬁcantly reduced from 21.3 to 17.0

after replacing the R¹basic nitrogen of 10h with a carbon. By

addition of a basic center to the R²group, such as in 15a, the

enzymatic potency can be improved compared to nonbasic 10i,

but the BEI is still very low, and both the enzymatic and cellular

4699

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 1. R¹SARs: Linker between R¹basic Nitrogen and Benzimidazole N¹and the Role of the Basic Center (Nitrogen)

^aValues are expressed as the mean ( standard deviation (SD) of at least two independent duplicate experiments. ^bValues are expressed as the mean ( SD of at

least two independent triplicate experiments. ^cBinding eﬃciency index BEI = pIC₅₀/[MW (kDa)] = ꢀ1000 log[IC₅₀(M)]/MW; see ref 23.

Figure 5. Compound 3 docked into two diﬀerent homology models (built based on the HDLP X-ray structure 1C3R (a) or human HDAC2 X-ray

structure 3MAX (b)). The basic center of the R¹side chain attached to benzimidazole nitrogen N¹has an electrostatic interaction with Asp99, and the

benzimidazole nitrogen N³can also form a hydrogen bond with His178 via a water molecule (b). Key interactions between these basic centers with

Asp99 and His178 contribute to the potency.

in cellular potency. A similar trend can be seen with 20c and 20e:

20e enjoyed a 3.6-fold improvement in cellular activity compared

to 20c. The trend of additional lipophilicity enhancing cellular

potency continues when the secondary amine is methylated. The

4700

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 2. R¹SARs: Directly Attached Cyclic Ammine as R¹

^aValues are expressed as the mean ( standard deviation (SD) of at least two independent duplicate experiments. ^bValues are expressed as the mean ( SD of at

least two independent triplicate experiments. ^cBinding eﬃciency index BEI = pIC₅₀/[MW (kDa)] = ꢀ1000 log[IC₅₀(M)]/MW; see ref 23.

resulting tertiary amine’s cellular potencies are signiﬁcantly

improved: there was 15-fold cellular potency enhancement for

22a versus 20a and 2.3-fold for 22b versus 20c. N-Methylation

does not appear to aﬀect the interaction between the basic center

and HDAC1 Asp99 but does seem to improve cell permeability

which in turn leads to better cellular activity. However, 20aꢀe

and 22a,b are not as potent as their acyclic analogues with similar

molecular weights (e.g., 3, 10e, 10h, Table 1). Moreover, the

acyclic compounds were achiral and more readily available for

further synthetic and biological studies; thus, further optimiza-

tion of the benzimidazole series focused on acyclic two-carbon-

linker R¹groups.

SARs of three representative benzimidazole hydroxamates 1

with R-substituted R²side chains together with R¹= 3-hydrox-

ylpropyl have been reported previously.¹⁸SAR studies were

herein extended to R¹groups with basic centers (Table 3). Both

enzymatic and cellular potencies were improved for compounds

10j and 10k compared to their counterparts bearing R¹=

3-hydroxylpropyl. There are diﬀerences between compounds

bearing linear R²groups, such as 10a compared to those having

R-substituted R²chains (e.g., 10j and 10k) in terms of HDAC1

inhibition, but diﬀerences in cellular potency were negligible. A

bulky cyclohexyl group at R²(10k) signiﬁcantly lowers BEI

compared to R-unsubstituted or less bulky R-substituted R²

counterparts (i.e.,10a and 10j). By introduction of a cyclopropyl

to replace the simple ethylene linker of 2a (Table 3),¹⁸both

enzymatic and cellular potencies of the resulting 15b were

reduced about 3- to 4-fold compared to 2a. The rigidity and

bulkiness of the R²R-substituted side chain might prevent it from

eﬀective interaction with the HDAC binding pocket. Once the

basic R¹was changed to the less bulky dimethylamino, the

resultant 15c had no change in activity. Further reducing the

size of R¹to a simple primary amine such as 15d led to recovery

of HDAC1 potency, but the cellular potency dropped signiﬁ-

cantly probably because of lower permeability of the primary

amine and globally lower lipophilicity of the molecule. Thus, 2a

remains the best of these four compounds. If the phenethyl chain

of 2a was replaced by a simple n-butyl, the resultant 3 is not quite

as potent as 2a in cells, but it is signiﬁcantly better than 15d which

is of similar molecular weight. This SAR suggests that a simple

alkyl (linear or non-R-substituted) R²chain provides suﬃcient

potency with better binding eﬃciency (i.e., BEI of 19.8 for 3, 21.4

for 10e, and 19.0 for 10f, the higher the better) compared to

complex or aromatic ring containing R²chains (i.e., BEI of 18.6

for 2, 17.9 for 2a, 17.5 for 2b, and 18.2 for 2c).

After establishment of the above-mentioned SAR on R¹and

R², the additive or synergistic eﬀect of R¹and R²was then

explored (Table 4). With a ﬁxed R¹group, both the enzymatic

and cellular potency increased along with an increase of length of

the R²chain or clogP. For example, with R¹group ﬁxed as diethyl-

aminoethyl, the calculated clogP values for compounds 10l (R²=

n-propyl), 10c (R²= isopropyl, Table 1), 3 (R²= n-butyl), and

10m (R²= n-hexyl) are 3.17, 3.52, 3.61 and 4.49, respectively.

Both their enzymatic potency (HDAC1, IC₅₀of 155, 92, 77, and

45 nM, respectively) and cellular potency (COLO 205, IC₅₀of

0.98, 0.81, 0.56, and 0.42 μM, respectively) are increased along

with clogP. If one of the ethyl groups was removed from the R¹

side chain of 3, the resultant compound 10n still maintained

good HDAC1 potency, but cellular potency against COLO 205

dropped slightly probably because of a decrease in lipophilicity.

By increase of the length of the R²side chain of 10n, more potent

compounds 10o and 10p were obtained. This trend also held true

4701

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 3. R²SARs: r-Substituted R²Side Chain

^aValues are expressed as the mean ( standard deviation (SD) of at least two independent duplicate experiments. ^bValues are expressed as the mean (

SD of at least two independent triplicate experiments. ^cBinding eﬃciency index BEI = pIC₅₀/[MW (kDa)] = ꢀ1000 log[IC₅₀(M)]/MW; see ref 23.

for both 10e and 10q, which had a secondary amine (isopropyl-

amino) R¹group: both were more potent than their diethylamino

counterparts 3 and 10m. When the diisopropylamino was intro-

duced as the R¹group of 10r, its steric bulk hindered the interaction

between the R¹basic center and HDAC1 Asp99 (Figure 5).

Compound 10r was also less potent in cells than either 10e or

10q. Removing one methylene group from each of the diethylamino

groups of 3 to give the smallest possible tertiary amino R¹, 10h, still

provided good potency. Returning these two carbons back to the R²

side chain of 3 gave n-hexyl 10s which was more potent than 3 in

both enzymatic and cellular assays. Similarly, transferring a methyl

group from R¹to R²of 10h gave 10t with the smallest secondary

amine at R¹but the same molecular weight as 10h and showed

enhanced HDAC1 potency but reduced cellular potency perhaps

because of lower permeability of the more polar secondary amine.

This reduced cellular potency can be reversed by raising clogP

through addition of one more carbon to the R²group of 10t, giving

for example 10u having regained submicromolar cellular potency.

Adding a further two carbons to give 10v with a high clogP of

Broad Antiproliferative Activity. Selected compounds were

tested against a range of tumor cell lines (Table 5) and compared

with vorinostat, panobinostat, and belinostat. The selected

compounds showed broad antiproliferative activity against re-

presentative tumor cells from ovarian (A2780), colon (HCT-

116), and prostate (PC-3), similar to belinostat and more potent

than vorinostat. Panobinostat is clearly a very potent antiproli-

ferative agent, but as will be discussed later, absolute potency is

not the only consideration when selecting the preferred com-

pound for progression into patients.

QSAR of Enzymatic and Cellular Potency. With a significant

body of in vitro enzymatic and cellular data available, QSAR

studies were possible to carry out in order to focus on the most

critical parameters for further lead optimization. Parameters

included were enzyme and cellular IC₅₀and 18 compounds

having R¹= A series (three-carbon linker, Figure 3, Table S2)

and 42 compounds with R¹= B series (two-carbon linker,

Figure 3, excluding B4 subseries, Table S3). Earlier it was

established that the chain length and total number of carbons

in the R¹and R²side chains affect the in vitro potency as

discussed above; hence, the correlation between IC₅₀and clogP

was investigated. The results (Figure 6) showed that lipophi-

licity does play a role for in vitro potency for this class of

compounds: both enzymatic potency pIC₅₀(HDAC1) and

cellular potency pIC₅₀(COLO 205) have weak positive correla-

tions with clogP for both A and B series (excluding B4 series).

Regression lines for the B series are above those for the A series,

for compounds with similar clogP. The B series compounds are

more potent than A series in both HDAC1 and COLO 205

assays. Figure 6 clearly shows that the B series are generally

more potent than the A series. When both series are combined

(n = 60) for linear regression/correlation, the cellular potency

pIC₅₀(COLO 205) still correlates with clogP (p = 0.0056)

(Figure 6b), while the correlation between enzymatic potency

pIC₅₀(HDAC1) is not significant (p = 0.0907).

4.43 resulted in the most potent compound (HDAC1 IC₅₀

=

15 ( 2 nM) in this series. Although in vitro potency can be

achieved in this way, metabolic consequences cannot be

ignored with long alkyl chains. Fortunately there are choices

within this series of R¹/R²variations that permit study of the

pharmacokinetic and metabolic proﬁles of very closely related

compounds, all of which possess potentially acceptable en-

zyme and cellular potency. For example, the isomeric com-

pounds of MW 358 present an excellent opportunity for such a

comparison: compound 10p is most potent among the iso-

meric series 3, 10c, 10s, and 10w. Further studies were done

(see below), but any of these compounds are worthy of further

study. Clearly in vitro potency can be adjusted or tuned by

judicious placement of carbons between R¹and R²side chains

as long as the total number of carbons shared between R¹and

R²falls in the range of 7ꢀ10.

4702

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 4. SARs of R²Length, Lipophilicity (R¹þ R²), and in Vitro Potencies

^aValues are expressed as the mean ( standard deviation (SD) of at least two independent duplicate experiments. ^bValues are expressed as the mean (

SD of at least two independent triplicate experiments. ^cBinding eﬃciency index BEI = pIC₅₀/[MW (kDa)] = ꢀ1000 log[IC₅₀(M)]/MW; see ref 23.

Relationships between enzymatic potency pIC₅₀(HDAC1)

and cellular potency pIC₅₀(COLO 205) were also examined

(Figure 7). pIC₅₀(HDAC1) positively correlates with pIC₅₀-

(COLO 205) for the A series (p < 0.0001, n = 18), B series (p <

0.0001, n = 42), and the combined A and B series (p < 0.0001, n =

72). Correlation with HDAC1 is also true for other tumor cell

lines: 14 compounds were tested in all four cell lines (Table 5 and

Figure S1, Supporting Information), and they all showed sig-

niﬁcant correlations for COLO 205 (p = 0.0009), HCT-116 (p =

0.0019), A2780 (p < 0.0001), and PC-3 (p = 0.0033). These

studies conﬁrm that HDAC1 plays an important role in cell

proliferation and that inhibitors of class I HDACs, especially

HDACs 1, 2, and 3, usually show strong antiproliferative and

apoptosis-inducing activity.^7ꢀ10In summary, both enzymatic

and cellular potencies may be tuned by judicious choice of the

combination of R¹and R²groups. The resultant potencies are in

general predictable; thus, with this series there is a considerable

freedom to optimize physicochemical and pharmacokinetic

parameters without adversely aﬀecting target potency.

In Vitro Pharmacodynamic Biomarker: Quantification of

Inhibition of HDAC in Cells. A hallmark of HDAC inhibition is

the increase in the acetylation level of histones. Acetylation of

4703

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 5. IC₅₀of Selected Compounds

cellular IC₅₀(μM)^b

HCT-116

compd

HDAC1 IC₅₀(nM)^a

A2780

COLO 205

PC-3

2

29 ( 10

42 ( 21

23 ( 13

77 ( 15

42 ( 7

46 ( 7

92 ( 7

77 ( 18

20 ( 7

36 ( 7

28 ( 6

21 ( 4

30 ( 9

53 ( 1

37 ( 1

23 ( 2

32 ( 13

119 ( 37

63 ( 27

6.8 ( 4.4

0.19 ( 0.14

0.11 ( 0.01

0.031 ( 0.001

0.48 ( 0.21

0.38 ( 0.01

0.44 ( 0.04

1.24 ( 0.15

0.76 ( 0.34

0.24 ( 0.06

0.20 ( 0.04

0.23 ( 0.01

0.20 ( 0.02

0.23 ( 0.01

0.37 ( 0.02

0.23 ( 0.02

0.14 ( 0.01

0.21 ( 0.03

1.62 ( 0.47

0.67 ( 0.41

0.035 ( 0.011

0.13 ( 0.05

0.11 ( 0.07

0.087 ( 0.063

0.56 ( 0.08

0.47 ( 0.06

0.57 ( 0.16

0.81 ( 0.34

0.88 ( 0.37

0.19 ( 0.04

0.17 ( 0.03

0.33 ( 0.14

0.31 ( 0.09

0.28 ( 0.10

0.62 ( 0.16

0.15 ( 0.01

0.14 ( 0.04

0.39 ( 0.04

2.12 ( 0.64

0.70 ( 0.39

0.018 ( 0.006

0.19 ( 0.11

0.16 ( 0.10

0.10 ( 0.03

0.48 ( 0.27

0.34 ( 0.06

0.80 ( 0.30

1.30 ( 0.35

1.11 ( 0.27

0.23 ( 0.12

0.21 ( 0.07

0.33 ( 0.02

0.66 ( 0.16

0.26 ( 0.12

0.92 ( 0.31

0.41 ( 0.06

0.27 ( 0.05

0.33 ( 0.00

2.85 ( 1.61

0.60 ( 0.21

0.048 ( 0.028

0.15 ( 0.08

0.18 ( 0.08

0.12 ( 0.04

0.34 ( 0.06

0.23 ( 0.05

0.50 ( 0.15

0.95 ( 0.24

0.38 ( 0.20

0.21 ( 0.06

0.17 ( 0.06

0.25 ( 0.15

0.16 ( 0.01

0.12 ( 0.08

0.38 ( 0.04

0.31 ( 0.04

0.10 ( 0.01

0.53 ( 0.18

1.21 ( 0.85

0.45 ( 0.19

0.024 ( 0.020

2b

2c

3

10e

10f

10c

10n

10p

10q

10s

10x

10y

10z

10aa

10ab

15e

vorinostat

belinostat

panobinostat

^aValues are expressed as the mean ( standard deviation (SD) of at least two independent duplicate experiments. ^bValues are expressed as the mean (

SD of at least two independent triplicate experiments.

Figure 6. Both enzymatic and cellular potencies (pIC₅₀) correlate with lipophilicity clogP. pIC₅₀= ꢀlog[IC₅₀(M)]. (a) pIC₅₀(HDAC1) correlate with

clogP: A series (O, n = 18, Pearson r = 0.5121, p = 0.0298); B series (ꢁ , n = 42, Pearson r = 0.5347, p = 0.0003). (b) pIC₅₀(COLO 205) correlate with

clogP: A series (O, n = 18, Pearson r = 0.6595, p = 0.0029), linear regression equation pIC₅₀(COLO 205) = 0.3389 clogP þ 4.426, R²= 0.4350; B series

(ꢁ, n = 42, Pearson, r = 0.6947, p < 0.0001), linear regression equation pIC₅₀(COLO 205) = 0.3379 clogP þ 5.112, R²= 0.4827.

histones H3, H4, and H2A can be detected by Western im-

munoblotting. To increase the throughput of the analysis, an

enzyme linked immunosorbent assay (ELISA) was developed to

detect and quantify acetylated histone 3 (AcH3) in the protein

lysates obtained from cancer cell lines treated with the HDAC

inhibitors. Key compounds reported herein were tested for their

ability to increase AcH3 level in an in vitro assay in COLO 205 cells

at 10 μM. The level of AcH3 was benchmarked to that of vorinostat,

which was also tested at 10 μM. Twenty-nine compounds generated

a range of 0.8- to 2.5-fold of AcH3 compared to that of vorinostat

(Table S4, Supporting Information). The AcH3 level was found to

correlate with both enzymatic potency pIC₅₀(HDAC1) (p =

0.0141) and cellular potency pIC₅₀(COLO 205) (p = 0.0266)

(Figure 8). The AcH3 level in COLO 205 cells may be modulated

by different doses of HDAC inhibitors. For a small number of

compounds, the effective dose to induce a 50% increase of AcH3

(EC₅₀) in COLO 205 cells was also determined. The cellular

potency pIC₅₀(COLO 205) also positively correlates with pEC₅₀

(Figure S2 and Table S5, Supporting Information). These results

confirmed that these benzimidazole hydroxamates are targeting

histone deacetylases in cells with equal or greater efficiency than

vorinostat in inducing hyperacetylation of H3. A compound’s

inhibitory potency against HDAC targets in cells can be assessed

with measurement of the AcH3 level it induces.

Inhibition of HDAC Isoenzymes. Only HDAC1 was used for

enzymatic primary screening assays. For selected compounds,

profiling against the isozymes was also performed (Table 6).

BIOMOL substrate (KI-104) was used for profiling HDACs

1ꢀ11. Compounds 3 and 10f were confirmed as substrate

competitive inhibitors of HDAC1 (data not shown); therefore,

4704

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

acetyllysine-based substrates (including BIMOL KI-104) and

concluded that most of the enzymatic activity associated with

class IIa HDACs ectopically expressed in mammalian cells is due

to endogenous class I HDACs present in the class IIa protein

complex.²⁸They also identiﬁed Boc-Lys-(ε-triﬂuoroacetyl)-

MCA as an active and selective substrate for class IIa HDACs.²⁸

When triﬂuoroacetyllysine based substrates were used in the

class IIa HDAC assays, hydroxamates such as vorinostat, belino-

stat, and panobinostat showed much reduced potency,^27,29ꢀ31

becoming HDACs 1ꢀ3 and 6 selective inhibitors.³¹

We proﬁled the compounds in Table 6 in 2005. Unfortunately

no inhibitory data were generated for class IIa HDACs using

triﬂuoroacetyllysine as substrate. However, it is reasonable to

assume that these benzimidazoles behave like the hydroxamates

that have been tested with the triﬂuoroacetyllysine substrate(s)

and hence are weak class IIa inhibitors in vitro under the reported

assay conditions (i.e., protein preparation and substrate). Recent

results from studies on class IIa inhibitors and substrates suggest that

inhibition of the catalytic activity of HDAC4 by small molecules is

not a viable approach for cancer therapy,³²and class IIa HDACs may

function as receptors rather than enzymes (deacetylases) of acet-

yllysine in a protein sequence.³¹From analysis of our data it appears

that whether these benzimidazole hydroxamates are potent or weak

class IIa HDAC inhibitors in vitro is not relevant to the in vitro

antiproliferative activity. Correlation analysis of the 11 benzimida-

zoles in Table 6 shows that the cellular potency (pIC₅₀) positively

correlates with the inhibitory constant pK_iof HDAC3 (Figure S3,

Supporting Information, p = 0.0455, 0.0070, 0.0351, and 0.0142 for

A2780, COLO 205, HCT-116, and PC-3, respectively) and has a

similar trend with pK_i(HDAC1). pK_i(HDAC3) also signiﬁcantly

(p = 0.0151) correlates with pK_i(HDAC1). PC-3 cellular potency

(pIC₅₀) correlates with both pK_i(HDAC3) (p = 0.0142, Figure

S3d) and pK_i(HDAC1) (p = 0.0113). These results suggest that

these benzimidazole-based HDAC inhibitors are mainly targeting

class I HDACs, particularly HDAC3 and HDAC1, to achieve anti-

proliferative activity in tumor cells, in line with early reports.^11ꢀ13

Wilson et al. reported that protein expression of HDAC3 was

signiﬁcantly up-regulated in a panel of human colon tumors

(Caco-2, COLO 205, HCT-116, HT-29, etc.) and in small intestinal

adenomas derived from Apc1638^N/þmice, and a signiﬁcant up-

regulation of HDAC1 and HDAC2 protein expression in colon

tumors was observed. Their ﬁndings demonstrate aberrant expres-

sion of HDAC3 and other class I HDACs in colon cancer.³³HDAC3

not only functions primarily in histone deacetylation but also plays a

role in biological processes beyond transcriptional repression.³⁴This

evidence supports the correlations between HDAC1 (or HDAC3)

inhibitory potency and cellular antiproliferative potency established

for benzimidazole HDAC inhibitors.

Figure 7. Cellular activity pIC₅₀(COLO 205) correlates with enzymatic

activity pIC₅₀(HDAC1). A series alone (O, n = 18, p < 0.0001): linear

regression equation pIC₅₀(COLO 205) = 1.065 pIC₅₀(HDAC1) ꢀ 1.529,

R²= 0.7104. B series alone (ꢁ , n = 42, p < 0.0001): linear regression

equation pIC₅₀(COLO 205) = 0.8775 pIC₅₀(HDAC1) ꢀ 0.1304, R²=

0.5963. For combined A and B series (n = 60, p < 0.0001): linear regression

equation pIC₅₀(COLO 205) = 1.007 pIC₅₀(HDAC1) ꢀ 1.093, R²=

0.7530. Linear regression equation with both pIC₅₀(HDAC1) and clogP:

pIC₅₀(COLO 205) = 0.9639 pIC₅₀(HDAC1) þ 0.0991 clogP ꢀ 1.1436.

R²= 0.7877 for combined A and B series (n = 60).

Figure 8. Level of acetylated histone 3 (AcH3) correlates both enzy-

matic potency pIC₅₀(HDAC1) (O, n = 29, Pearson r = 0.4636, p =

0.0113) and cellular potency pIC₅₀(COLO 205) (þ, n = 29, Pearson r =

0.4356, p = 0.0182). “AcH3 (fold)” is the fold of protein concentration

compared with that of vorinostat.

all the hydroxamates in Table 6 were assumed as competitive

inhibitors for all the isozymes and K_ivalues were calculated using

the ChengꢀPrusoff equation (Table 6). The results showed that

these compounds are pan-HDAC inhibitors, broadly similar to

vorinostat, belinostat, panobinostat, and the 2 series (2, 2b, and

2c), with the exception of HDAC6 in most cases. Compound 10x

is an exception among the compounds tested in Table 6, since it

did show good activity against HDAC6. It contains a cis double

bond in the R²side chain which may be crucial to the HDAC6

inhibition. Vorinostat, panobinostat, and belinostat are more

potent against HDAC6 than the majority of other isozymes. All

of them have aromatic rings as the CAP group, so an aromatic

ring or a cis double bond in the CAP group might be important

for HDAC6 inhibition. Indeed, benzimidazole HDAC inhibitors

with phenethyl R²groups (e.g., 2, 2b, and 2c) do show good

HDAC6 inhibitory potency compared to those with alkyl R²

groups (Table 6).

In Vitro ADME Profiling and SAR. One of our major goals was

to improve upon the metabolic stability of HDAC inhibitors for

optimal oral dosing in patient, in particular the phase 1 metabolic

liability often suffered by hydroxamic acids as demonstrated by their

short half-lives (t_1/2) in liver microsomal assays. Indeed, the 2 series

(2 and 2aꢀc) in general had poor microsomal stability, a clear

liability for oral administration.¹⁸In our screening cascade only

compounds that passed predefined potency criteria were evaluated

in human liver microsomal stability assays (HLM). A total of 38

compounds (Table S6, Supporting Information) were identified

with HLM t_1/2data in a wide range, applicable for a QSAR analysis.

In general, only those compounds with HLM t_1/2g 30 min were

tested in mouse liver microsomal assays (MLM), which generally

turned these compounds over faster. We strongly suspected that

Our selectivity proﬁling results for hydroxamates such as

vorinostat, belinostat, and panobinostat are generally in line with

those published and tested with generic acetyllysine derived

substrate(s).^25ꢀ27Lahm et al. described that class IIa HDACs

possess a weak but measurable intrinsic capability of hydrolyzing

4705

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 6. Inhibition of HDAC Isozymes by Selected Compounds

dissociation constant K_i(nM) (mean ( SD) for HDAC Isozymes^a

compd

1

2

3

4

5

6

7

8

9

10

11

2

11 ( 2

15 ( 1

12 ( 2

28 ( 1

26 ( 5

24 ( 1

11 ( 4

20 ( 4

19 ( 2

7.7 ( 0.4

10 ( 1

60 ( 7

2.5 ( 0.2

26 ( 2

36 ( 4

39 ( 7

39 ( 2

27 ( 1

33 ( 1

26 ( 4

18 ( 3

18 ( 2

21 ( 5

8.6 ( 0.4

14 ( 2

42 ( 3

1.9 ( 0.1

22 ( 2

12 ( 1

11 ( 2

11 ( 1

19 ( 1

27 ( 2

20 ( 1

13 ( 1

15 ( 1

20 ( 2

16 ( 1

10 ( 3

36 ( 3

2.3 ( 0.3

19 ( 2

8.5 ( 1.6

10 ( 2

17 ( 2

20 ( 1

NT

95 ( 27

NT

19 ( 1

21 ( 6

24 ( 3

24 ( 1

30 ( 1

11 ( 2

14 ( 1

16 ( 1

11 ( 2

5.7 ( 0.4

4.4 ( 0.1

49 ( 23

1.2 ( 0.1

24 ( 1

19 ( 1

23 ( 4

23 ( 1

23 ( 6

34 ( 2

34 ( 4

28 ( 1

29 ( 1

13 ( 2

18 ( 1

19 ( 2

60 ( 16

31 ( 1

59 ( 2

13 ( 3

16 ( 7

14 ( 3

24 ( 1

18 ( 1

14 ( 2

13 ( 3

14 ( 2

11 ( 1

8.8 ( 1.5

4.2 ( 0.2

31 ( 6

3.5 ( 0.2

27 ( 1

2b

2c

3

16 ( 1

23 ( 4

8.7 ( 1.8

16 ( 1

18 ( 2

27 ( 1

NT

21 ( 1

247 ( 10

225 ( 61

94 ( 5

104 ( 1

77 ( 5

89 ( 4

35 ( 1

41 ( 3

70 ( 5

12 ( 1

42 ( 2

129 ( 44

2.3 ( 0.1

51 ( 1

48 ( 1

78 ( 9

65 ( 3

54 ( 10

57 ( 8

136 ( 7

55 ( 7

94 ( 8

173 ( 49

22 ( 0.5

22 ( 8

10e

11 ( 1

10f

7.9 ( 1.4

6.2 ( 0.6

6.6 ( 0.3

5.3 ( 1.0

2.4 ( 0.1

3.6 ( 0.6

20 ( 6

11 ( 1

10p

6.8 ( 0.5

6.6 ( 0.5

9.2 ( 0.3

3.1 ( 0.1

5.6 ( 0.3

36 ( 10

0.7 ( 0.1

25 ( 1

282 ( 52

348 ( 94

300 ( 5

4.0 ( 2.5

227 ( 1

29 ( 13

0.7 ( 0.4

10 ( 2

10q

10s

10x

10y

vorinostat

panobinostat

belinostat

0.6 ( 0.1

15 ( 1

^aThe K_iwas calculated using the ChengꢀPrusoﬀ equation: K_i= IC₅₀/ {1 þ ([substrate]/K_m)}. IC₅₀values of HDACs 1ꢀ11 for K_icalculation were

measured using Fluor de Lys substrate (KI-104, BIOMOL International, L.P) at pH 7.5, room temp. NT: not tested.

major metabolites were products of mono-N-deethylation and bis-

N-deethylation of the diethylamino group of 3. The amide meta-

bolite formed via reduction of hydroxamic acid was also observed,

but it was in trace amounts. The details of these studies will be

published elsewhere.³⁵

Physicochemical properties were also evaluated: high through-

put kinetic solubility at pH 7 revealed that these compounds are

very soluble and present no concerns for oral or iv dosing for

in vivo studies (Table 7). The log D values were also measured on

selected compounds and were found to be within the desired

range of around 2.0, higher than the more polar vorinostat. Both

the N³of benzimidazole ring (HCl salt pK_a≈ 4.1) and basic

Figure 9. Relationship between liver microsomal half-life (t_1/2) and

center in the R¹side chain (pK_a≈ 7.4) are able to be protonated

clogP. The log[t_1/2(HLM)] correlates with clogP negatively (b, n = 38,

at acidic pH (i.e., pH < 5ꢀ6 for protonation of basic nitrogen at

Pearson r = ꢀ0.4601, p = 0.0037). Linear regression equation is log[t_1/2

R¹side chain, and pH < 2ꢀ3 for protonation of both basic

centers); thus, the resultant salts are very soluble. For example,

the dihydrochloride salt of 3 is very soluble in pure water (>100

mg/mL, pH 2.2), in saline (0.9% NaCl) (>50 mg/mL), in

phosphate buﬀer (>50 mg/mL, pH 4.5), and in 1% methylcellu-

lose/0.1% Tween 80 (>25 mg/mL) for oral dosage formulation.

Permeability and eﬄux of 3, 10f, 10p and 10s were assessed in

(HLM, min)] = ꢀ0.3079 clogP þ 2.848, R²= 0.2117, but log[t_1/2

(MLM)] has no signiﬁcant correlation with clogP (0, n = 13, Pearson

r = ꢀ0.4358, p = 0.1366).

liver microsomal turnover increased with lipophilicity, and after

analysis of the data from 38 compounds, the liver microsomal stabi-

lity data (log t_1/2) was found to be negatively correlated with clogP

(p = 0.0037) (Figure 9). Because of the limited data for MLM

(Table S6, Supporting Information), there was no significant corre-

lation between log t_1/2(MLM) andclogP (p = 0.1366, n= 13), but a

trend was evident. The MLM t_1/2was generally lower than that of

HLM for the same compound or for compounds with the same

clogP (Figure 9), but no significant correlation (p = 0.0633, n = 13)

was observed between log t_1/2(HLM) and logt_1/2(MLM) data.

According to the QSAR equation established (Figure 9), for t_1/2

(HLM) g 30, 60, and 90 min, clogP should be no more than 4.5,

3.5, and 2.9, respectively. These clogP values are for t_1/2(MLM) of

5, 17, and 35 min, respectively. These numbers were used as

guidelines for target compound design.

Caco-2 cells. The rates of transport apparent permeability P_app

-

(AfB) of 10f, 3, and 10s were 19.4 ꢁ 10^ꢀ6, 7.04 ꢁ 10^ꢀ6, and

9.59 ꢁ 10^ꢀ6cm/s, respectively. These rates are higher than that

reported for vorinostat (1.70 ꢁ 10^ꢀ6cm/s).¹³However, 10p

showed lower permeability (0.97 ꢁ 10^ꢀ6cm/s) and had signi-

ﬁcant eﬄux, P_app(AfB)/P_app(BfA) ratio or eﬄux ratio of 10.8,

probably due to the secondary amine. Compound 2 is permeable

but also has signiﬁcant eﬄux (eﬄux ratio of 9.2). Compounds 3,

10e, 10f, and 10s are highly permeable and have no signiﬁcant

eﬄux, an important improvement over the 2 series.

Fluorescence assays against recombinant P450 isoforms such

as 3A4 and 2D6 were used to assess potential for drugꢀdrug

interactions. Potential candidate compounds were further tested in

human liver microsomes with P450 isozyme selective substrates. To

our knowledge, the IC₅₀obtained in the ﬂuorescence assays using

recombinant CYP proteins was about an order of magnitude more

potent than that obtained from the microsomal assay system. These

Compounds selected for in vivo evaluations were also tested in

rat and dog liver microsomal assays (Table 7). In general the

benzimidazoles are relatively stable in human and dog microsomes

but variable in rodent species. Studies on in vitro metabolism of 3 in

mouse, rat, dog, and human liver microsomes revealed that the

4706

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 7. Proﬁling Data of Selected Compounds for in Vivo Studies

liver microsomal stability (t_1/2, min)

Caco-2 permeability (10^ꢀ6cm/s)^d,f

compd

human

mouse

rat

dog

solubility (μM)^a

log D at pH 7.4 ^b,f

clogP ^c

P_app(AfB)/P_app(BfA)

2

3

60

5

4.3

7.8

3.3

41

18

>200

2.1

2.97

3.61

3.46

4.44

3.88

3.81

1.84

3.01/27.7

7.04/19.4

NT

>60

51

27

36

20

1.6

16

>60

48

2.07

NT

2.41

NT

1.04

10e

10f

19.4/14.9

0.97/10.5

9.59/26.0

1.70^e/ꢀ

10p

1.1

10

10s

>60

10

vorinostat

>60

^aKinetic solubility at pH 7 with 2.5% of DMSO. ^bDetermined in 1-octanol/sodium phosphate buﬀer (pH 7.4). ^cclogP (octanol/water) values were

calculated using MOE 2008.10, Chemical Computing Group. ^dSigniﬁcant eﬄux: eﬄux P_app(AfB)/P_app(BfA) > 3.0 and P_app(BfA) > 1.0 ꢁ

10^ꢀ6cm/s. ^eSee ref 13. ^fNT: not tested.

also showed poor bioavailability in rats, for example, 4.3%

(S*BIO in-house data) for belinostat and 8.0% (S*BIO in-house

data) or 10.9%¹³for vorinostat. Rat was hence felt to be an

unsuitable species for pharmacokinetic evaluation of these

hydroxamates. In nude mice, 10f and 3 were sufficiently bioavail-

able: 51% for 10f (Table S9, Supporting Information) and 34%

for 3 (Table 9). In contrast, vorinostat, belinostat, and panobino-

stat showed F = 8.3%, 6.7%, and 4.6%, respectively.³⁶Both 3 and 10f

were also tested in Beagle dogs and demonstrated good bioavail-

ability of 65 ( 16% and 110 ( 54%, respectively. Consistent with

the increasing microsomal half-life from rat to mouse to dog

(Table 7), F also increased as did intravenous terminal half-life t_1/2

Figure 10. Relationship between CYP3A4 inhibitory potency IC₅₀and

(rat 0.89 h, mouse 2.26 h, dog 3.90 h) for compound 3 (Table 9).

clogP. There are a total of 89 compounds in the ﬁgure, 72 of them with

Orally administered 3 was rapidly absorbed with t_max= 0.24, 0.17,

IC₅₀< 20 μM (b) and 17 of them with IC₅₀g 20 μM (ꢁ). The

and 0.75 h for rat, mouse, and dog at doses of 10, 50, and 10 mg/kg,

inhibitory potency IC₅₀correlates with clogP very signiﬁcantly (b, n =

respectively. 3 exhibits an overall better PK profile and oral bioavail-

ability in non-rodent (dog) compared to rodent (mouse and rat)

and was predicted by allometry to have an acceptable human oral PK

profile.³⁷Mouse and dog were hence selected as suitable preclinical

species for further safety evaluation of benzimidazole hydroxamates

such as 3 and 10f .

In Vivo Antitumor Efficacy. Compounds 3, 10f, 10e, 10s, and

10p were selected for in vivo antitumor evaluation in subcuta-

neous xenograft mouse models (Table 10). All these compounds

were stable in HLM assays (Table 7). 3 and 10f also demon-

strated good PK in mice. 10s is potent and relatively stable in the

MLM assay. Compound 10p is potent, but its MLM stability and

permeability are poor compared to the other four, but it was

included for assessing which in vitro parameter is more important

for antitumor efficacy: high in vitro enzymatic and cellular

potency, a good in vitro ADME profile, or more likely, a

combination of both.

The HCT-116 xenograft model was chosen for primary

eﬃcacy screening, as this model had been successfully used for

HDAC inhibitors such as belinostat³⁸and dacinostat (LAQ824)³⁹

in the literature. Vorinostat was used as positive control for most of

the experiments. Compounds were dissolved in a mixture of 0.5% or

1% methylcellulose and 0.1% Tween 80 in water and administered

orally to the tumor-bearing mice. Compounds were dosed up to the

maximum tolerated dose (MTD), which was deﬁned by <20% loss

in mean body weight per treatment group and no more than one

treatment-related death among 7ꢀ10 treated animals. Antitumor

response was assessed by determination of tumor growth inhibition

(TGI) in treated groups compared to the control (vehicle) group.

72, p < 0.0001), linear regression equation log[IC₅₀(3A4, μM)] =

ꢀ0.5779 clogP þ 2.519, R²= 0.3965.

benzimidazole hydroxamates inhibited mainly recombinant CYP3A4

in the ﬂuorescence assay. The relation between IC₅₀(3A4) and clogP

is clear (Figure 10): There were 72 compounds with IC₅₀(3A4) e

20 μM and 17 compounds with IC₅₀(3A4) > 20 μM (Tables S7 and

S8, Supporting Information). The logarithm IC₅₀(3A4) was corre-

lated with clogP signiﬁcantly, p < 0.0001. According to the QSAR

equation established in Figure 10, clogP should be no more than 4.36

and higher than 2.63 for IC₅₀(3A4) between1 and 10 μM. Consider-

ing that the IC₅₀obtained in the microsomal system is about an order

of magnitude lower than the recombinant system, it is suﬃcient to

ensure that the clogP is below 4.36 to achieve an IC₅₀(3A4) above

10 μM in the physiologically more relevant microsomal assay.

Actually, 87% (77/89) of the compounds has clogP < 4.36, and

84% (64/77) of these 77 compounds has IC₅₀(3A4) g 1 μM.

Compounds generally have no signiﬁcant CYP liabilities; results from

both the ﬂuorescence assay and microsomal probe substrate assay for

three representative compounds (3, 10f, 10u) are shown in Table 8.

Compound 3 has IC₅₀= 5.78 μM for CYP2C19, and the plasma

concentration achieved in human is much lower than this value,

indicating no potential to cause any drugꢀdrug interaction.

Pharmacokinetics (PK) in Preclinical Species. PK studies

were first performed in Wistar rats, but the bioavailbilities (F)

were poor for 10f (4.0%), 10e (9.0%), 10p (1.4%), and 3 (3.1%),

probably related to the poor metabolic stabilities for 3, 10p, and

10e in rat liver microsomes (Table 7). Reference compounds

4707

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 8. Inhibition of CYP Isozymes

IC₅₀(μM)

pooled HLM with CYP isozyme probe substrate

recombinant isozymes^a

compd

3A4

2D6

1A

2C9

2C19

2D6

3A4

3

2.73 ( 0.84

2.28 ( 0.76

0.33 ( 0.07

>10

>25

10.4

5.78

8.10

10.3

>25

13.9

>25

10f

10s

3.17 ( 0.29

^aValues are expressed as the mean and standard deviation (SD) of at least two independent duplicates.

Table 9. Pharmacokinetic Parameters of 3 in Mouse, Rat, and Dog

female nude mouse^a

po

Wistar rat^b

Beagle dog^c

parameter

iv

po

iv

po

dose (mg/kg (nmol/kg))^d

C_max(μM)

10 (23)

50 (116)

6.10

2 (4.6)

10 (23)

0.139

0.24

2 (4.6)

10 (23)

3.56

t_max(h

0.17

0.75

t_1/2(h)

2.26

9.2

2.44

0.89

4.45

1.66

1.15

2.02

3.90

1.45

4.24

3.22

4.12

CL/F ((L/h)/kg)

V_SS(L/kg)

27.2

153

2.29

3.46

2.52

AUC_0ꢀ¥(μM h)

F (%)

4.27

34

0.181

3.1

10.49

65

3

^aThree female BALB/c nude mice (ARC, West Australia), 16.5ꢀ19.4 g, 10ꢀ12 weeks of age, per time point per route. ^bThree male Wistar rats each for

iv (230ꢀ250 g, 6ꢀ8 weeks of age) and po (330ꢀ370 g, 8 weeks of age). ^cSix male non-naive Beagle dogs, age 1ꢀ2 years, 12ꢀ16 kg for iv (n = 3) and po

(n = 3). ^dCompound was tested as the dihydrochloride salt and was dissolved in saline for iv injection and in 0.5% methylcellulose/0.1% Tween 80 in

water for oral administration.

The ﬁve compounds in Table 7 were assessed in three separate

experiments with vorinostat as the positive control. In the ﬁrst

experiment (Table 10), the MTD for vorinostat was established as

300 mg/kg, but 200 mg/kg was shown to be the maximum absorbed

or eﬀective dose; MTD for 10e was 200 mg/kg. The MTD for 10p

was above 200 mg/kg. Both 10f and 10e demonstrated signiﬁcant

antitumor activities at their respective MTDs and at half MTD dose

levels with good tolerability for the entire 21-day treatment period.

Compound 10p did not show signiﬁcant antitumor activity until the

high dose of 200 mg/kg and even then only 41% TGI. This com-

pound is metabolically unstable and of low permeability, even though

it is the most potent compound among the tested compounds. Its

poor ADME most likely signiﬁcantly reduced its exposure below

eﬀective levels in vivo. A higher dose of 10p may have been as

eﬀective as 10e. Both the MTD dose groups of 10f and 10e were

signiﬁcantly better (p < 0.001) than the vorinostat 200 mg/kg group,

but there was no statistically signiﬁcant diﬀerence between the latter

and other treatment groups except for the 50 and 100 mg/kg groups

of 10p. In the second experiment, 10s was tested together with

vorinostat and belinostat, which also has a cinnamyl linker to a

hydroxamic acid. However, the similarities with the compounds

described herein appear limited: belinostat was weakly active up to

200 mg/kg which was the maximum absorbed dose in our hands. 10s

showed signiﬁcant toxicity at 200 mg/kg; hence this compound was

only dosed qd ꢁ 7, showing TGI = 53% (p < 0.05) on day 22. 10s

was also active at its daily dosing MTD (100 mg/kg) with TGI =

50% (p < 0.05) but not at 50 mg/kg. Vorinostat performed well in

this model with 73% TGI, but it was not statically better than the 10s

100 mg/kg group. In a further experiment, 3was tested together with

vorinostat (Table 10). Compound 3 was clearly toxic at the highest

dose tested (200 mg/kg); however, at the MTD dose of 100 mg/kg

and at 50 mg/kg, it demonstrated very signiﬁcant antitumor eﬀects

on day 21 with TGI = 90% (p < 0.001) and 66% (p < 0.001), respec-

tively, with acceptable body weight loss. Vorinostat was also active

compared to the vehicle group with TGI = 41% (p < 0.05). The 100

mg/kg group of 3 was very signiﬁcantly (p < 0.01) active compared

to vorinostat when assessed by Dunnett’s multiple comparison test

using a vorinostat group as control. In a fourth experiment, both 3

and panobinostat were tested at their MTD doses, and they

demonstrated signiﬁcant eﬃcacy compared to the vehicle group

with TGI of 98% (p < 0.0001) and 61% (p < 0.01), respectively. The

tumor growth inhibition of the 3 group was found to be signiﬁcantly

better than that of the panobinostat group (p < 0.001). In the above

four experiments, compounds 3, 10f, and 10e demonstrated very

signiﬁcant (p < 0.001) antitumor activities at their respective MTD

dose levels. At half MTD dose level, their respective mean tumor

volumes were still signiﬁcantly (p < 0.01) diﬀerent from that of the

vehicle (or nontreatment) group. Furthermore, the toxicity in terms

of body weight loss is reversible. For example, the mean body weights

of 3 of treated groups increased by 6.5% in the ﬁrst 4 days after the

last dose and fully recovered in 17 days for the MTD dose group. The

major adverse eﬀects observed for 3 in repeated oral dose toxicity

studies in mice and dogs are dose-dependent mild myelosuppression

and gastrointestinal eﬀects that are consistent with eﬀects reported

for other HDAC inhibitors. However, in clinical trials 3 appears to be

well tolerated with an overall favorable safety proﬁle compared to

other orally dosed HDAC inhibitors.²¹

3 and 10f emerged from this work as the top two candidates

for further evaluations. 3 was selected as the preferred compound

because of its attractive chemistry (e.g., chemical synthesis and

4708

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

Table 10. Antitumor Activities of Selected Compounds in HCT-116 Xenograft Models

entry

1

compd^a

dose (mg/kg)^b

% TGI (day)^c

statistical signiﬁcance^d

max body weight loss (%) (at day)

survivors on day 22

10f

25

30 (22)

55 (22)

85 (22)

19 (22)

60 (22)

90 (22)

ꢀ9 (22)

ꢀ7 (22)

41 (22)

57 (22)

54 (22)

17(22)

ns

10/10

9/10^e

10/10

8/10^f

9/9

10f

50

//

///

ns

5.6 (5)

10f

100

50

13.1 (22)

10e

100

200

50

//

///

ns

5 (8)

10e

7.9 (19)

10p

100

200

300

50

ns

10p

ns

vorinostat

10s

//

/

4.6 (19)

5.5 (19)

1.9 (5)

4.9 (5)

18 (5)

2

ns

10s

100

200

50

50 (22)

53 (22)

73 (22)

6 (22)

//

///

ns

9/9

7/9^g

10s

vorinostat

belinostat

3

8.9 (8)

9/9

100

200

50

26 (22)

29 (22)

66 (21)

90 (21)

toxic(21)

41 (21)

98 (18)

61 (18)

ns

9/9

ns

3.9 (5)

9/9

3

4

///

6.4 (17)

14.6 (19)

25.5 (6)

1.9 (19)

19.9 (20)

17.3 (12)

9/9

3

100

200

100

50

9/9

0/9^h

3

vorinostat

3

/

///

//

9/9

10/10

panobinostat

^aCompounds 3, 10e, 10f, 10p, and 10s were dosed as the dihydrochloride salt, and vorinostat was dosed as parent. Vehicle was 1% methylcellulose/0.1%

Tween 80 in water for entries 1 and 2, and was 0.5% methylcellulose/0.1% Tween 80 in water for entries 3 and 4. ^bDose schedule: po, qd ꢁ 21. Entry 1:

female athymic nude mice (nu/nu, Harlan), 9 weeks of age. Entry 2: 6 weeks of age. Entries 3 and 4: female BALB/c nude mice (ARC, West Australia),

10ꢀ12 weeks of age. ^cThe percent tumor growth inhibition (% TGI) = [(C_tꢀ T_t)/(C_tꢀ C₁)] ꢁ 100. C₁and C_tare the mean tumor volumes for control

group (vehicle) on day 1 and day t, respectively. T_tis the mean tumor volume for treatment group on day t. ^dOne-way ANOVA followed by Dunnett's

multiple comparison test was used to determine the statistical signiﬁcance of (log transformed) tumor volume between a treatment group and the

control (vehicle) group: ns = not signiﬁcant, (/) p < 0.05, (//) p < 0.01, and (///) p < 0.001. Entry 1: There were 2/10 very slow growing tumors in the

vehicle group, which eliminated any statistical diﬀerence between this group and any of the treated groups; thus, the nontreated group was used as the

control group. There was no signiﬁcant diﬀerence in mean tumor volume between the nontreatment and vehicle group (p = 0.94, two-tailed unpaired

Student’s t-test). ^eOne treatment-related death (TRD) occurred on day 3. ^fOne TRD on day 13 and one non-treatment-related death due to accident

on day 5. ^gOne TRD on day 5 and another on day 8, only dosed qd ꢁ7. ^hFour TRDs on day 6 and ﬁve TRDs on day 7.

stability), optimal balance of potency, in vitro ADME properties,

good oral PK in mouse, dog and predicted human exposures, and

strong eﬃcacy in mouse xenograft models.

in mice was mainly due to its low bioavailability. Panobinostat

(LBH589), an analogue of dacinostat, was administered to mice

via either iv or ip injection in recent reports.⁴⁰The 15 mg/kg iv

group of panobinostat was much more eﬃcacious than all the ip

groups (15, 20, 30 mg/kg) in a NSCLC xenograft model. These

data suggest that panobinostat has inadequate oral bioavailability

in mice. In conclusion, the in vitro enzyme and cellular potency

of 3 is improved over vorinostat, close to belinostat but inferior to

panobinostat. We have demonstrated that 3 has a superior oral

PK proﬁle to the agents currently in clinical trials with a higher

C_maxand AUC than vorinostat, belinostat, and panobinostat at

the same oral dose level,³⁶leading to superior eﬃcacy in the

HCT-116 mouse xenograft model.

Compound 3 exhibited broad antiproliferative activity in

tumor cell lines in vitro (Table 5), prompting us to explore its

potential in other solid and hematological subcutaneous tumor

models. Results from representative experiments for various xeno-

graft models are summarized in Table 11. 3 has antiproliferative

IC₅₀values of 0.10, 0.14, 0.34, and 0.48 μM for MV4-11 (acute

monocytic leukemia), Ramos (lymphoma), PC-3 (prostate), and

A2780 (ovarian) cancer cell lines, respectively. Consistent with its in

The goal for our HDAC inhibitor project was to develop a

drug with a superior once daily oral dosage regimen; hence,

compound 3 was compared to the leading clinical compounds

available at the time, such as vorinostat, panobinostat, and

belinostat in the same HCT-116 xenograft model experiment

(Figure 11). Following continuous daily dosing the tumor

volumes of the 3 group were very signiﬁcantly diﬀerent from

the vehicle group on days 15 and 12 (p < 0.001). Vorinostat

reduced tumor growth moderately by days 15 and 12 (p < 0.05);

however, in this experiment, neither panobinostat nor belinostat

showed signiﬁcant diﬀerences from the vehicle group at their

MTDs or maximum absorbed/eﬃcacious dose levels. The mean

tumor size of the 3 group was also signiﬁcantly reduced com-

pared to those of panobinostat, belinostat, and vorinostat groups

on days 12 and 15. In an early report, belinostat was tested in the

HCT-116 model with ip injection (40 mg/kg, qd ꢁ 7, in 10%

DMSO),³⁸and the TGI% was estimated to be 50% on day 10.

Taken together, the low eﬃcacy of orally administered belinostat

4709

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

vitro potency against cell lines of these tumor types, 3 was more

eﬀective against hematopoietic cancers than solid tumors. The

MV4-11 leukemia model was the most sensitive model tested, with

signiﬁcant antitumor activity observed at doses as low as 25 mg/kg

(TGI = 61%, p < 0.001) and 50 mg/kg (TGI = 111%, p < 0.001).

There were 6/10 complete tumor regressions (CR) and 2/10 partial

tumor regressions (PR) of tumors observed at 50 mg/kg. After

discontinuation of dosing on day 21, only 1/6 of the tumors had

begun to regrow in the CR animals by day 28. The 50 mg/kg group

was also very signiﬁcantly more eﬃcacious than the 25 mg/kg group

(p < 0.01). 3 was also tested in a Ramos xenograft model. After 14

days of daily oral dosing, the tumor volumes in the 50 mg/kg group

were signiﬁcantly diﬀerent from the vehicle group with TGI =

75% (p < 0.05). There were 3/10 CRs and 1/10 PRs in the 3

treated group. At the well-tolerated dose of 50 mg/kg, 3 was also

tested in the PC-3 xenograft model together with the other HDAC

inhibitors, and it demonstrated signiﬁcant antitumor activity with

TGI = 66% (p < 0.05). In an ovarian cancer (A2780) xenograft

model, the original dosing scheme was qd ꢁ 21. Unfortunately the

tumor grew very quickly and multiple animals in the vehicle group

reached their predeﬁned end-point (tumor size of 1500 mm³);

thus, mice were euthanized on day 8 (1/10), day 10 (6/10), and

day 12 (1/10). Hence, only tumor volume data up to day 9 were

used to assess eﬃcacy (TGI). Tumor volumes of both 50 and 100

mg/kg groups were very signiﬁcantly (p < 0.001) diﬀerent from

that of the vehicle group with TGI of 44% and 63%, respectively.

The vorinostat group was not signiﬁcantly diﬀerent from the

vehicle group with TGI of 39%. Both the 3 groups were signiﬁ-

cantly diﬀerent from the vorinostat group with p < 0.01 and p <

0.001 for 50 and 100 mg/kg groups, respectively.

In all these in vivo models, 3 demonstrated dose-dependent

antitumor eﬀects (statistically signiﬁcant reductions in tumor

growth compared with vehicle control) and was well tolerated

(e.g., at 100 mg/kg, qd ꢁ 21). The toxicity in terms of body

weight loss is reversible.

Pharmacokinetic/Pharmacodynamic (PK/PD) Relation-

ship of 3. In a preliminary study, limited tissue distribution

was assessed in nude mice bearing established HCT-116 tumors

following a single oral administration of 50 and 100 mg/kg of 3. As

shown in Figure 12, 3 preferentially distributed to tumor tissues;

the tumor concentration (μmol/kg)/plasma concentration (μM)

ratio reached 10 at about 5 h. The area under curve (AUC) of tu-

mor tissue was 1.75- and 2.37-fold of that of plasma for 50 and

100 mg/kg dose, respectively (see Table S9, Supporting Information,

for tissue distribution data). Clearly, 3 is enriched in tumor tissues.

C_maxconcentrations were significantly higher than both HDAC1 and

cell proliferation IC₅₀values. Tumor concentration levels of 3

were above the HDAC1 IC₅₀level for the 50 and 100 mg/kg dose

levels for 10 and 16 h, respectively, and the corresponding times

above IC₅₀of HCT-116 cells were 5.1 and 7.1 h, respectively.

Considering protein binding (91.8 ( 0.3% in human, 83.5 ( 1.8%

in dog, 93.0 ( 4.4% in rat, and 93.9 ( 1.5% in mouse plasmas as

tested at 1000 ng/mL), the theoretical free drug concentration is

still higher than the HDAC1 IC₅₀. Perhaps a more meaningful

measure of sufficient exposure is given by AUC/IC₅₀(equivalent

Figure 11. Antitumor activities of 3, vorinostat, panobinostat, and

belinostat in HCT-116 xenograft model. Tumor volume is expressed as

the mean ( standard error of the mean. 3 was dosed as the dihydrochloride

salt. Vorinostat and belinostat were dosed as parent. Panobinostat was

dosed as lactate, and vehicle is 0.5% methylcellulose/0.1% Tween 80 in

water. Female BALB/c nude mice (ARC, West Australia), 10ꢀ12 weeks of

age, seven mice per group, were dosed po, qd ꢁ 21. Tumor necrosis

occurred among groups of vehicle, belinostat, and vorinostat after day 15;

thus, only TGI on day 15 and before were used for eﬃcacy assessment. See

footnote of Table 10 for TGI calculation and statistical analysis.

Table 11. Eﬃcacy of Orally Administered 3 in Various Xenograft Models

tumor

A2780

compd^a

schedule^b

dose (mg/kg)

% TGI (day)^c

statistical signiﬁcance^d

max body weight loss (day) (%)

survivors (day)

vehicle

qd ꢁ 21

qd ꢁ 14

qd ꢁ 21

1.5 (5)

9.3 (13)

20.0 (20)

8.9 (13)

0.5 (17)

7.7 (17)

0

9/10 (9)^e

vorinostat

200

100

50

39 (9)

63 (9)

44 (9)

ns

9/10 (9)^f

10/10 (9)^g

10/10 (9)^h

8/8 (22)

3

///

3

PC-3

vehicle

3

50

66 (22)

75 (15)

/

8/8 (22)

Ramos

MV4-11

vehicle

10/10 (15)

10/10 (15)ⁱ

10/10 (21)

10/10 (21)^j

10/10 (21)

3

3.7 (12)

0

vehicle

3

50

25

114 (21)

61 (21)

///

7.1 (19)

3.2 (10, 12)

^aVehicle was 0.5% methylcellulose/0.1% Tween 80 in water. ^bFemale BALB/c nude mice (ARC, West Australia), 10ꢀ12 weeks of age, 5 ꢁ 10⁶cells (A2780,

MV4-11), PC-3 (1 mm³fragment of PC-3 tumor), Ramos (7 ꢁ 10⁶cells). ^cThe percent tumor growth inhibition (% TGI) = [(C_tꢀ T_t)/(C_tꢀ C₁)] ꢁ 100. C₁

and C_tare the mean tumor volumes for control group (vehicle) on day 1 and day t, respectively. T_tis the mean tumor volume for treatment group on day t. ^dOne-

way ANOVA followed by Dunnett's multiple comparison test was used to determine the statistical signiﬁcance of (log transformed) tumor volume between a

treatment group and the control (vehicle) group: ns = not signiﬁcant, (/) p < 0.05, (//) p < 0.01, and (///) p < 0.001. ^eOne animal was euthanized because the

tumor size reached the end-point on day 8, and 1/10 survived on day 21 because of spontaneous regression after day 13. ^fVorinostat group, one non-treatment-

related death on day 9, and 1/10 survived on day 21 because of CR (tumor disappeared after day 11). ^g5/10 survived on day 21. ^h2/10 survived on

day 21. ⁱInduced 3/10 CR and 1/10 PR. ^jInduced 6/10 CR and 2/10 PR.

4710

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

PK, and PD (AcH3) are closely related for 3. These ﬁndings have

translated successfully to the assessment of drug exposure and

response in clinical trials.^20,21

Compound 3 Selected as Clinical Candidate. On the basis

of the above pharmacology and ADME data, 3 was selected for

preclinical development. In vitro safety pharmacology was assessed

by testing 3 against 26 different receptors and 33 different non-

HDAC enzymes, which were selected based on their physiological

function. At 10 μM, 3 does not interfere with enzymes of other

families including Zn^2þ-dependent enzymes and has no interactions

with G-protein-coupled receptors, monoamine transporters, or ion

channels. In particular, 3 did not inhibit ³H-dofetilide binding to the

hERG potassium channel or zinc-dependent enzymes such as

MMP-3 and MMP-9 at concentrations up to 10 μM. After

toxicology studies in mice and dogs, the first dose in human was

determined. 3 entered phase I trials in 2007 and is currently in

multiple phase I and phase II clinical trials for treatment of patients

with solid tumors and hematologic malignancies. The phase I data

from treatment of patients with advanced solid malignancies showed

that 3 had a manageable toxicity and favorable PK profile (dose-

proportional PK).²¹The relative AcH3 values measured in PBMCs

increased dose dependently from 10 to 60 mg and in proportion to 3

plasma levels.²¹The ongoing clinical phase II program is designed to

take full advantage of the superior PK/PD properties of 3.

Figure 12. Plasma and tumor tissue distribution of 3 in HCT-116

tumor bearing mice. Female BALB/c nude mice, 10ꢀ12 weeks of age,

were implanted subcutaneously in the right ﬂank with 5 ꢁ 10⁶cells of

HCT-116 human colon carcinoma. When the tumor size reached

approximately 100 mm³, these mice were orally administered with a

single dose of 3 at 50 or 100 mg/kg. 3 was dissolved in 0.5%

methylcellulose/0.1% Tween80 in water. There were seven time points

(predose, 0.17, 0.5, 1, 2, 6, and 24 h) for 50 mg/kg group and eight time

points (predose, 0.17, 0.5, 1, 3, 6, 16, and 24 h) for 100 mg/kg group.

Two mice were scariﬁed at each time point. Blood and tumor tissue

samples were collected and analyzed for reach mouse. Only data points

with concentrations (in μM for plasma or in μmol/kg for tumor tissue)

above the quantiﬁcation limit are shown.

’ CONCLUSIONS

SARs for a series of 3-(1,2-disubstituted-1H-benzimidazol-5-yl)-

N-hydroxyacrylamides have been established for the enzyme

HDAC1, colon cancer cell line COLO 205, liver microsomal

stability (t_1/2), and CYP P450 (3A4) inhibition. These parameters

were tuned by adjusting substituents at positions 1 and 2 of the

benzimidazole ring (i.e., the R¹and R²groups) to balance potency,

ADME, and lipophilicity. In addition, stability due to phase 1 meta-

bolism was signiﬁcantly improved, together with improved perme-

ability, resulting in a new generation of preclinical development

candidates with demonstrably good PK proﬁles in both rodent

(mouse) and non-rodent (dog) species and signiﬁcant antitumor

eﬃcacies in a variety of tumor models. As an oral HDAC inhibitor, 3

demonstrates a superior preclinical proﬁle compared to the leading

clinical compounds available at the time, vorinostat, belinostat, and

panobinostat in our proﬁling studies. 3was selected as the preclinical

development compound with the most favorable overall balance of

in vitro and in vivo properties.

In summary, 3 is a potent pan-HDAC inhibitor with excellent

druglike properties (log D_pH7.4= 2.1, solubility at pH 5 of >10

mg/mL and of >100 mg/mL in water for the hydrochloride salt),

is highly eﬀective in in vivo tumor models, has high and dose-

proportional oral bioavailability, and has very good ADME,

safety, and pharmaceutical properties. 3 has a prolonged duration

of action and is enriched in tumor tissue which may contribute to

its potent antitumor activity. 3 is currently being tested in phase I

and phase II clinical trials in both hematological and solid tumor

patients, and preliminary data show that the superior preclinical

proﬁle is translated to the clinic.

exposure time at IC₅₀level): HCT-116 cells have 14 and 31 h in

tumor for 50 and 100 mg/kg dose levels, respectively, and 8.2 and

13 h in plasma, respectively. Concentrations in plasma are more

easily measured compared to tumor tissue; thus, plasma AUC

(μM h)/IC₅₀(μM) g 8.2 h can be used to predict the antitumor

3

efficacy. For sensitive cell lines like MV4-11 (IC₅₀= 0.10 μM), the

exposure time at the IC₅₀level for the 50 mg/kg dose group was

39 h, which is significantly longer than 8.2 h, hence explaining the

excellent efficacy achieved at 50 mg/kg and even at 25 mg/kg dose

levels in the MV4-11 xenograft model. With consideration of

protein binding, the free drug exposures (about 1/10 of all drug

exposure) are still high, especially with respect to the sensitivity of 3

toward MV4-11. HCT-116 cells are less sensitive, but 3 appears to

reach sufficient levels over the first hours of exposure to significantly

inhibit growth. This suggests that tumor growth can be sufficiently

arrested if 3 can achieve IC₅₀or above levels for about 8 h. Even

when unbound drug levels are considered, plasma levels are still

above IC₅₀for 1 and 2 h at 50 and 100 mg/kg in mice, respectively.

In summary, the antitumor efficacy of 3 in a variety of tumor

models is explained by its excellent PK profile and preferred tissue

distribution coupled with its enzymatic and cellular potency.

As expected for an HDAC inhibitor, 3 increased the level of

both acylated histone 3 and R-tubulin and up-regulated cyclin-

dependent kinase inhibitor p21 protein expression in HCT-116

cells in a dose-dependent manner.¹⁹The relationship of PK

(drug concentration) and PD (AcH3 level) was also studied in

HCT-116 tumor bearing mice after a single oral dose of 3 (50,

100, or 125 mg/kg) or vorinostat (200 mg/kg). AcH3 was

strongly induced in a dose dependent manner for 3 and the

induction of AcH3 was more sustained (up to 16 h for 3 versus up

to 8 h for vorinostat).¹⁹In conclusion, in vivo antitumor eﬃcacy,

’ EXPERIMENTAL SECTION

Flash column chromatography was conducted using silica gel 60

(Merck KGaA, 0.040ꢀ0.063 mm, 230ꢀ400 mesh ASTM). Reverse-

phase preparative high performance liquid chromatography (RPHPLC) was

conducted on a Phenomenex column (Luna, 5 μm, C18 100A, 150 mm ꢁ

21.2 mm) with adjustable solvent gradients, usually 5ꢀ95% of acetonitrile in

4711

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

water þ 0.1% TFA in 18 min at ﬂow rate of 20 mL/min, and was used for

routine puriﬁcation. The preliminary purity and identity of all compounds

were assessed after puriﬁcation by tandem HPLCꢀmass spectrometry

(LCꢀMS) experiments on a Waters Micromass ZQ mass spectrometer in

electrospray ionization (ESI) positive mode after separation on a Waters

2795 separations module. The HPLC separations were performed on a

Phenomenex column (Luna, 5 μm, C18 100A, 50 mm ꢁ 2.00 mm) with a

ﬂow rate of 0.8 mL/min and a 4 min gradient of 5ꢀ95% or 30ꢀ95% of

acetonitrile in water þ 0.05% TFA, using a Waters 2996 photodiode array

detector. Purity and identity were assessed on the integrated UV chromato-

grams (220ꢀ400 nm) and the mass spectra. The ﬁnal purity was determined

using a Waters 2695 separations module on a Waters Xterra RP18, 3.5 μm,

4.6 mm ꢁ 20 mm IS column with a ﬂow rate of 2.0 mL/min, gradient

5ꢀ95% of acetonitrile in water with 0.1% TFA over 5 min, and a Waters 2996

photodiode array detector. A longer column (Phenomenex Gemini, 5 μm,

C18, 110A, 4.6 mm ꢁ 150 mm) together with a ﬂow rate of 1.0ꢀ1.2 mL/

min and a 15 min gradient of 5ꢀ95% acetonitrile in water þ 0.1% TFA was

also used for purity check. Purity was >95% for all compounds reported for

biological data except those indicated in their respective synthesis.

The solution was evaporated, and to the residue were added dichloro-

methane (DCM) and aqueous Na₂CO₃. The DCM (ꢁ3) extracts were

concentrated, and to the residue was added EtOAcꢀhexanes. The resulting

red solid was filtered to afford the title compound 7f (0.672 g, 84%).

LCꢀMS m/z 294.2 ([M þ H]^þ). ¹H NMR (CDCl₃þ CD₃OD) δ 8.21

(d, J = 2.1 Hz, 1H), 7.56 (dd, J = 9.0, 2.1 Hz, 1H), 7.48 (d, J = 16.0 Hz, 1H),

6.81 (d, J = 9.0 Hz, 1H), 6.20 (d, J = 15.9 Hz, 1H), 3.70 (s, 3H), 3.34 (t, J =

6.5 Hz, 2H), 2.56 (t, J = 6.4 Hz, 2H), 2.23 (s, 6H); ¹³C NMR (CDCl₃þ

CD₃OD) δ 167.3, 145.4, 142.6, 134.0, 131.1, 127.1, 121.3, 114.8, 114.0,

56.7, 51.1, 44.6, 40.1.

(E)-3-[4-(3-Dimethylamino-2,2-dimethylpropylamino)-3-

nitrophenyl]acrylic Acid Methyl Ester (7a). 7a was prepared

according to procedure A but using 3-dimethylamino-2,2-dimethylpro-

pylamine as starting material and was obtained as a red solid (35 g, 95%).

LCꢀMS m/z 336.1 ([M þ H]^þ). ¹H NMR (CDCl₃) δ 9.73 (br s or t,

1H, ꢀNHꢀ), 8.33 (d, J = 2.0 Hz, 1H), 7.60 (dd, J = 8.9, 2.0 Hz, 1H),

7.59 (d, J = 16.1 Hz, 1H), 6.88 (d, J = 9.1 Hz, 1H), 6.28 (d, J = 15.9 Hz,

1H), 3.80 (s, 3H), 3.21 (d, J = 4.6 Hz, 2H), 2.36 (s, 2H), 2.34 (s, 6H),

1.04 (s, 6H).

1

All the 1D and 2D NMR experiments for H (400.13 MHz), ¹³C

(E)-3-[4-(2-Diethylaminoethylamino)-3-nitrophenyl]acrylic

Acid Methyl Ester (7b). 7b was prepared according to procedure A but

using 2-diethylaminoethylamine as starting material and was obtained as a

(100.61 MHz), ¹⁵N (40.55 MHz), and ¹⁹F (376.47 MHz) nuclei were

1

performed on a Bruker AVANCE-400 digital NMR spectrometer. ¹Hꢀ H,

13

1

15

¹Hꢀ C, and Hꢀ N 2D experiments (COSY, HMQC, HSQC, and

HMBC) were run with Z-gradient selection. NMR spectra are reported in

ppm with reference to an internal tetramethylsilane standard (0.00 ppm for

¹H and ¹³C) or solvent peak(s) of CDCl₃(7.26 and 77.1 ppm) or CD₃OD

(3.31 and 49.0 ppm), or DMSO-d₆(2.50 and 39.5 ppm). Other NMR

solvents were used as needed. When peak multiplicities are reported, the

following abbreviations are used: s = singlet, d = doublet, t = triplet, q =

quartet, m = multiplet, br = broadened, dd = doublet of doublets, dt =

doublet of triplets, bs = broadened singlet. Coupling constants, when given,

are reported in hertz.

1

yellow solid (29 g, 82%). LCꢀMS m/z 322.1 ([M þ H]^þ). H NMR

(CDCl₃) δ 8.72 (t-like, J = 4.2 Hz, 1H, ꢀNHꢀ), 8.32 (d, J = 2.1 Hz, 1H),

7.61 (dd, J = 9.3, 2.1 Hz, 1H), 7.58 (d, J = 15.9 Hz, 1H), 6.85 (d, J = 9.0 Hz,

1H), 6.29 (d, J = 15.9 Hz, 1H), 3.79 (s, 3H), 3.35 (td, J = 5.4, 4.7 Hz, 2H,

ꢀNHCH₂CH₂ꢀ), 2.77 (t, J = 6.1 Hz, 2H), 2.59 (q, J = 7.1 Hz, 4H), 1.07

(t, J = 7.1 Hz, 6H); ¹³C NMR (CDCl₃) δ 167.5, 146.1, 143.0, 134.3, 131.7,

127.7, 121.6, 115.3, 114.8, 51.6 (OCH₃), 50.7 (ꢀNHCH₂CH₂ꢀ), 46.8

(-NCH₂CH₃), 40.9 (ꢀNHCH₂CH₂ꢀ), 12.0 (ꢀNCH₂CH₃). HRMS

(ESI) m/z: [M þ H]^þcalcd for C₁₆H₂₄N₃O₄, 322.1761; found 322.1767.

(E)-3-[4-(2-Ethylaminoethylamino)-3-nitrophenyl]acrylic

Acid Methyl Ester (7c). 7c was prepared according to procedure A

but using 2-ethylaminoethylamine as starting material and was obtained

as a red solid (36 g, 56%). LCꢀMS m/z 294.1 ([M þ H]^þ). ¹H NMR

(DMSO-d₆) δ 8.49 (t, J = 6.1 Hz, 1H, ꢀNHꢀ), 8.35 (d, J = 2.0 Hz, 1H),

7.96 (dd, J = 9.1, 1.9 Hz, 1H), 7.62 (d, J = 16.0 Hz, 1H), 7.20 (d, J = 9.1 Hz,

1H), 6.52 (d, J = 16.0 Hz, 1H), 3.75 (td, J = 6.5, 6.2 Hz, 2H), 3.70 (s, 3H),

3.08 (t, J = 6.5 Hz, 2H), 2.93 (q, J = 7.2 Hz, 4H), 1.17 (t, J = 7.2 Hz, 6H).

(E)-3-[4-(2-Isopropylaminoethylamino)-3-nitrophenyl]acry-

lic Acid Methyl Ester (7d). 7d was prepared according to procedure

A but using 2-isopropylaminoethylamine as starting material and was

obtained as a red solid (27 g, 57%). LCꢀMS m/z 308.1 ([M þ H]^þ). ¹H

NMR (DMSO-d₆) δ 8.58 (t, J = 5.6 Hz, 1H, ArꢀNHꢀ), 8.33 (d, J = 2.0

Hz, 1H), 7.94 (dd, J = 9.1, 1.9 Hz, 1H), 7.60 (d, J = 16.0 Hz, 1H), 7.14 (d,

J = 9.2 Hz, 1H), 6.49 (d, J = 16.0 Hz, 1H), 3.70 (s, 3H), 3.56 (masked by

water peak, identified by COSY, 2H), 3.01 (septet, J = 6.4 Hz, 1H), 2.94

(t, J = 6.2 Hz, 2H), 1.10 (d, J = 6.4 Hz, 6H).

For TFA salts, the TFA content (w/w) and freebase content (w/w) of

1

TFA salts were determined by ¹⁹F and H NMR using an internal

reference standard (e.g., 2,2,2-triﬂuoroethylamine hydrochloride), and

the molar ratio of TFA over freebase was also calculated. For those

molecules containing a CF₃group, the CF₃group can also serve as an

internal reference standard for TFA quantiﬁcation, and the results of the

TFA/freebase ratio obtained from this method are in line with the above

method using a dedicated internal reference standard. Elemental ana-

lyses of CHN were performed on a Perkin-Elmer 2400 CHN/CHNS

elemental analyzer. HRMS results were obtained from a Bruker micrO-

TOF-Q II instrument with direct injection.

Preparation of trans-4-Chloro-3-nitrocinnamic Acid Methyl

Ester (5). To a suspension of trans-4-chloro-3-nitrocinnamic acid (4,

10.00 g, 43.94 mmol) in MeOH (200ꢀ250 mL) was added concentrated

H₂SO₄(3 mL). The reaction mixture was allowed to stir at 80ꢀ85 ꢀC for

4ꢀ6 h. The progress of the reaction was monitored by HPLC or LCꢀMS.

When the acid (4) was completely converted to the corresponding ester (5),

the reaction flask was cooled in an ice bath. The suspension was filtered, and

the residue was washed with cold methanol several times and dried. Ester 5

was obtained as a yellowish solid (10.09 g, 95% yield). LCꢀMS m/z210 and

212 (very weak signal, [M þ Hꢀ MeOH]^þ). ¹HNMR(CDCl₃) δ8.01 (d,

J = 2.0 Hz, 1H, H-2), 7.64 (dd, J= 8.5, 2.1 Hz, 1H, H-6), 7.64 (d, J = 16.0 Hz,

1H, ꢀCHdCHꢀCO₂Me), 7.58 (d, J = 8.4 Hz, 1H, H-5), 6.50 (d, J = 16.1

Hz, 1H, ꢀCHdCHꢀCO₂Me), 3.83 (s, 3H, OCH₃);¹³CNMR(CDCl₃) δ

166.4 (CO₂Me), 148.3 (C-3), 140.8 (ꢀCHdCHꢀCO₂Me), 134.5 (C-1),

132.5 (C-5), 131.8 (C-6), 128.3 (C-4), 124.5 (C-2), 121.4 (ꢀCHdCHꢀ

CO₂Me), 52.1 (OCH₃).

(E)-3-[4-(2-Diisopropylaminoethylamino)-3-nitrophenyl]-

acrylic Acid Methyl Ester (7e). 7e was prepared according to

procedure A but using 2-diisopropylaminoethylamine as starting ma-

terial and was obtained as a yellow solid (12 g, 79%). LCꢀMS m/z 350

([M þ H]^þ). ¹H NMR (CDCl₃) δ 8.76 (t-like, J = 4.3 Hz, 1H), 8.32 (d,

J = 2.0 Hz, 1H), 7.61 (dd, J = 8.3, 2.7 Hz, 1H), 7.58 (d, J = 15.8 Hz, 1H),

6.85 (d, J = 9.0 Hz, 1H), 6.29 (d, J = 15.9 Hz, 1H), 3.79 (s, 3H), 3.31 (td,

J = 5.3, 6.1 Hz, 2H, ArꢀNHꢀ), 3.08 (septet, J = 6.6 Hz, 2H,

(Me₂CH)₂N), 2.84 (t, J = 6.2 Hz, 2H), 1.07 (d, J = 6.6 Hz, 12H).

(E)-3-[4-(2-Methylaminoethylamino)-3-nitrophenyl]acrylic

Acid Methyl Ester (7g). 7g was prepared according to procedure A but

using 2-methylaminoethylamine as starting material and was obtained as

an orange solid (2.1 g, 100%). LCꢀMS m/z 280 ([M þ H]^þ). ¹H NMR

(CDCl₃) δ 8.54 (t-like, J = 4.2 Hz, 1H), 8.33 (d, J = 2.1 Hz, 1H), 7.63 (dd,

J = 9.0, 2.2 Hz, 1H), 7.59 (d, J = 16.0 Hz, 1H), 6.90 (d, J = 9.0 Hz, 1H),

General Procedure A for Synthesis of Aniline Compound 7.

(E)-3-[4-(2-Dimethylamino-ethylamino)-3-nitrophenyl]acrylic

Acid Methyl Ester (7f). A mixture of 5 (0.658 g, 2.72 mmol), N,N-

dimethylethylenediamine (0.90 mL, 8.20 mmol), and triethylamine

(1.2 mL, 8.6 mmol) in 1,4-dioxane (20 mL) was heated at 80 ꢀC for 5 h.

4712

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

6.31 (d, J = 15.9 Hz, 1H), 3.80 (s, 3H), 3.45 (td, J = 5.8, 5.6 Hz, 2H), 2.96

(t, J = 6.2 Hz, 2H), 2.50 (s, 3H).

0.513 mmol of 7b according to procedure B) was mixed with hydro-

xylamine hydrochloride (0.331 g, 4.76 mmol) and MeOH (3 mL) with

stirring at room temperature. The reaction mixture was cooled over dry

ice, followed by the addition of sodium methoxide (25% in MeOH,

2.4 mL, 10.5 mmol). The mixture was then allowed to warm slowly to

room temperature. The reaction was monitored by LCꢀMS and was

completed in ∼30 min. The reaction mixture was slowly acidified with 6

N hydrochloric acid to pH ≈ 7, and water was added to achieve a clear

solution, which was subsequently purified by RPHPLC to afford

compound 3 as the bis-TFA salt, light pinkish solid (0.061 g, 20% in

two steps from 7b). There was negative reaction in the silver nitrate test.

LCꢀMS m/z 359.2 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 7.94 (d, J = 8.6

Hz, 1H), 7.85 (s, 1H), 7.76 (d, J = 8.5 Hz, 1H), 7.50 (d, J = 15.7 Hz, 1H),

6.49 (d, J = 15.7 Hz, 1H), 4.96 (overlapped with DHO, identified by

COSY, 2H), 3.69 (t-like, J = 7.6 Hz, 2H), 3.44 (q, J = 7.6 Hz, 4H), 3.26 (t,

J = 7.9 Hz, 2H), 1.94 (quintet, J = 7.5 Hz, 2H), 1.57 (sextet, J = 7.6 Hz,

2H), 1.40 (t, J = 7.2 Hz, 6H), 1.05 (t, J = 7.3 Hz, 3H); ¹³C NMR

(CD₃OD) δ 165.5, 157.7, 140.0, 134.8, 134.0, 133.8, 126.5, 119.9, 115.1,

113.6, 50.2, 48.7 (2C), 40.5, 29.4, 26.6, 23.3, 13.9, 8.9 (2C) (TFA peaks

at 163.4. 163.0, 162.7, 162.3; 122.3, 119.5, 116.6).

(E)-3-[4-(2-Aminoethylamino)-3-nitrophenyl]acrylic Acid

Methyl Ester (7h). In method 1, 7h was prepared according to

procedure A but using 1,2-ethylenediamine as starting material and was

obtained as an orange solid (2.8 g, 98%). LCꢀMS m/z 266 ([M þ

H]^þ). In method 2, to a 150 mL flask were added 7i (0.855 g, 2.34

mmol), MeOH (10 mL), and 6 N HCl (2 mL). The mixture was heated

to reflux (65ꢀ70 ꢀC) for 3 h (monitored by HPLC and LCꢀMS). The

solution was evaporated to dryness. Then EtOAc/DCM was added and

the solution was evaporated to dryness. 7h was obtained as a red or

orange solid (0.7069 g, 100% as the mono-HCl salt).

(E)-3-[4-(2-tert-Butoxycarbonylaminoethylamino)-3-nitro-

phenyl]acrylic Acid Methyl Ester (7i). 7i was prepared according to

procedure A but using N-Boc-ethylenediamine as starting material and

was obtained as a bright yellow solid (0.193 g, 53%). LCꢀMS m/z 366.1

([M þ H]^þ). ¹H NMR (CDCl₃) δ 8.41 (br t like, 1H, ꢀNHAr), 8.31 (d,

J = 1.8 Hz, 1H), 7.63 (dd, J = 9.0, 1.7 Hz, 1H), 7.57 (d, J = 16.0 Hz, 1H),

6.98 (d, J = 8.9 Hz, 1H), 6.30 (d, J = 15.9 Hz, 1H), 4.91 (br t like, 1H,

BocNHꢀ), 3.80 (s, 3H), 3.52 (dt, J = 5.8, 5.5 Hz, 2H), 3.45 (dt, J = 6.0, 5.9

Hz, 2H), 1.45 (s, 9H); ¹³C NMR (CDCl₃) δ 166.9, 155.7, 145.8, 142.3,

134.1, 131.5, 127.1, 121.8, 115.4, 113.9, 79.5, 51.2, 42.7, 39.1, 27.9.

General Procedures for Syntheses of Benzimidazoles 9

and 9aꢀab. Procedure B. Preparation of (E)-3-[1-(2-Ethyl-

aminoethyl)-2-hexyl-1H-benzimidazol-5-yl]acrylic Acid

Methyl Ester (9p). To a stirred solution of 7c (8.174 g, 27.87 mmol)

and heptaldehyde (4.85 g, 42.47 mmol) in AcOH and MeOH (1:9 v/v,

Procedure D for Freebase and Salt Formation. Prepara-

tion of (E)-3-[1-(2-Ethylaminoethyl)-2-hexyl-1H-benzimida-

zol-5-yl]-N-hydroxyacrylamide (10p). To a solution of ester 9p

(4.428 g, 12.39 mmol) and NH₂OH HCl (8.66 g, 124.7 mmol) in dry

3

MeOH (50 mL), which was stirred and cooled in a dry iceꢀacetone

bath, was added NaOMe solution in MeOH (25%, 55 mL, 240 mmol).

The reaction mixture was then stirred at room temperature. The progress of

reaction was monitored by LCꢀMS (usually reaction completed within

30ꢀ90min) andquenchedbyadding6NHCl(40mL).Tothemixturewas

added Milli-Q deionized water. The mixture was adjusted pH ≈ 8 with 1 N

NaOH and was evaporated to remove the organic solvent. The resultant

residue was washed with Milli-Q water (ꢁ3) and redissolved in MeOHꢀ

DCM. The solution was filtered and diluted with Milli-Q water. The suspen-

sion was evaporated to remove the organic solvent, and the resultant residue

was washed with Milli-Q water (ꢁ2). The freebase of 10p was obtained

(HPLC purity at 254 nm, 98%). It was recrystallized from MeOHꢀethyl

acetate to afford a white or pale yellow solid. The freebase was dissolved in

MeOH and excess 6 N HCl (final pH < 2), and the clear solution was

evaporated to dryness and then diluted with MeOH and coevaporated with

PhMe (ꢁ1) and EtOAc (ꢁ2). The solid was recrystallized from

MeOHꢀEtOAc to give a white or pale yellow solid (3.298 g, 61.7%).

HPLC purity at 254 nm: 98.4ꢀ99.6%. LCꢀMS m/z 359.2 ([M þ H]^þ).

¹H NMR (CD₃OD) δ 9.33 (residual NH), 8.03 (d, J = 8.3 Hz, 1H), 7.77 (s,

1H),7.73(d,J=8.2Hz, 1H), 7.16(d,J=15.7 Hz, 1H), 6.34 (d, J=15.7 Hz),

1H, 4.88 (overlapped with DHO, identified by COSY, 2H), 3.63 (br t like,

2H), 3.32 (d, J = 7.9 Hz, 2H), 3.15 (q, J = 7.1 Hz, 2H), 1.94 (quintet, J = 7.1

Hz, 2H), 1.53 (quintet, J = 6.7 Hz, 2H), 1.42ꢀ1.31 (m, 4H), 1.33 (t, J = 7.1

Hz, 3H), 0.88 (t, J = 7.0 Hz, 3H); ¹³C NMR (CD₃OD) δ 163.4, 155.8,

138.1, 133.0, 132.0, 130.3, 125.1, 117.4, 112.8, 112.5, 44.5, 43.2, 41.1, 30.5,

28.0, 25.3, 25.2, 21.6, 12.4, 9.6. Anal. (C₂₀H₃₀N₄O₂2.05HCl H₂O) C, H,

300 mL) was added SnCl₂2H₂O (31.45 g, 139.4 mmol) in portions.

3

The resulting mixture was heated to 40 ꢀC with stirring. The progress of

the reaction was monitored by LCꢀMS. When the reaction was completed,

solvent was removed under reduced pressure below 40 ꢀC. The resultant

residue was diluted with EtOAc (50 mL) and then basified (pH > 10) with

saturated aqueous Na₂CO₃and extracted with DCM (ꢁ3). The organic

extracts were combined, dried (Na₂SO₄), filtered, and evaporated to dryness.

The resulting oily residue was purified by flash column chromatography

(silica, 0ꢀ10% of MeOH in DCM). The title compound 9p was obtained as

a yellow solid (4.445 g, 44.6%). LCꢀMS m/z 358.3 ([M þ H]^þ). ¹H NMR

(CDCl₃) δ 7.88 (d, J = 1.2 Hz, 1H), 7.83 (d, J = 16.0 Hz, 1H), 7.43 (dd, J =

8.4, 1.4 Hz, 1H), 7.33 (d, J = 8.4 Hz, 1H), 6.43 (d, J = 15.9 Hz, 1H), 4.22 (t, J

= 6.6 Hz, 1H), 3.80 (s, 3H), 3.01 (t, J = 6.6 Hz, 2H), 2.89 (t, J = 7.9 Hz, 2H),

2.65 (q, J = 7.1 Hz, 2H), 1.91 (quintet, J = 7.8 Hz, 2H), 1.46 (m, 2H),

1.40ꢀ1.30 (m, 4H), 1.07 (t, J = 7.1 Hz, 3H), 0.90 (t, J = 7.0 Hz, 3H); ¹³

C

NMR (CDCl₃) δ167.3, 156.5, 145.5, 142.6, 136.3, 128.0, 121.7, 119.1, 115.1,

109.1, 51.0, 48.2, 43.7 (C ꢁ 2), 31.0, 28.7, 27.1 (C ꢁ 2), 22.0, 14.7, 13.5.

(E)-3-[2-Butyl-1-(2-diethylaminoethyl)-1H-benzimidazol-

5-yl]acrylic Acid Methyl Ester (9). 9 was prepared and purified

according to procedure B with yields of 38ꢀ73% as a yellow solid.

LCꢀMS m/z 358.2 ([M þ H]^þ). ¹H NMR (CDCl₃) δ 7.87 (d, J = 1.0

Hz, 1H), 7.83 (d, J = 15.9 Hz, 1H), 7.43 (dd, J = 8.4, 1.5 Hz, 1H), 7.28 (d,

J = 8.2 Hz, 1H), 6.43 (d, J = 15.9 Hz, 1H), 4.15 (t, J = 7.0 Hz, 2H), 3.81

(s, 3H), 2.90 (t, J = 7.0 Hz, 2H), 2.74 (t, J = 7.0 Hz, 2H), 2.55 (q, J = 7.1

Hz, 4H), 1.90 (quintet, J = 7.8 Hz, 2H), 1.49 (sextet, J = 7.6 Hz, 2H),

0.99 (t, J = 7.3 Hz, 3H), 0.96 (t, J = 7.2 Hz, 6H); ¹³C NMR (CDCl₃) δ

167.8, 157.0, 146.1 (CH), 143.1, 136.8, 128.4, 122.1 (CH), 119.6 (CH),

115.6 (CH), 109.5 (CH), 52.4, 51.5, 47.8, 43.2, 29.7, 27.3, 22.7, 13.8,

3

Cl. N: calcd, 12.42; found, 12.92.

(E)-3-[2-Butyl-1-(2-diethylaminoethyl)-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide Dihydrochloride Salt (3). The free-

base of 3 was prepared according to procedure D. The hydroxamic acid

15

moiety was identified by ¹Hꢀ N HSQC (DMSO-d₆) with δ_N= 169.0

1

15

12.0. The nitrogens in ester 9 were also indentified by ¹Hꢀ N HMBC

ppm (CONHOH). Other nitrogens in 3 were identified by Hꢀ N

HMBC (DMSO-d₆) with δ_Nof 241.4 ppm for N³of the benzimidazole

ring, 152.3 ppm for N¹, and 41.3 ppm for the diethylamino group

(reference to nitromethane δ_N= 380.0 ppm in CDCl₃). The dihy-

drochloride salt of 3 was prepared according to procedure D as white or

off-white solid or powder in ∼60% yield from 9 in two steps. LCꢀMS

m/z 359.2 ([M þ H]^þ). ¹H NMR (DMSO-d₆) δ 11.79 (brs, 1H, NH or

OH), 10.92 (very br s, 1H), 8.18 (d, J = 8.6 Hz, 1H), 7.97 (s, 1H), 7.79

(d, J = 8.6 Hz, 1H), 7.64 (d, J = 15.8 Hz, 1H), 6.65 (d, J = 15.8 Hz, 1H),

with δ_Nof 239.0 ppm for N³of the benzimidazole ring, 151.8 ppm for

N¹, and 42.6 ppm for the diethylamino group (reference to nitro-

methane δ_N= 380.0 ppm in CDCl₃). HRMS (ESI) m/z [M þ H]^þ

calcd for C₂₁H₃₂N₃O₂, 358.2489; found, 358.2492.

General Procedures for Synthesis of Hydroxamates 3, 10,

15, 17, 20, and 22. Procedure C for Small Scales (TFA Salt).

(E)-3-[2-Butyl-1-(2-diethylaminoethyl)-1H-benzimidazol-5-

yl]-N-hydroxyacrylamide (3). Ester 9 (crude 0.177 g, made from

4713

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

5.01 (t-like, J = 7.7 Hz, 2H), 3.48 (m, 2H), 3.30ꢀ3.19 (m, 6H), 1.87

(quintet, J = 7.8 Hz, 2H), 1.47 (sextet, J = 7.5 Hz, 2H), 1.29 (t, J = 7.2 Hz,

6H), 0.97 (t, J = 7.3 Hz, 3H); ¹³C NMR (DMSO-d₆) δ 162.3, 156.0,

137.3 (CH), 132.8, 132.3, 132.0 (br, identified by HMBC), 124.7 (CH),

120.2 (CH), 113.1 (2 ꢁ CH), 48.2, 46.3, 39.0, 28.1, 25.0, 21.7, 13.6, 8.3.

Anal. (C₂₀H₃₀N₄O₂2HCl 0.265H₂O) C, H, N, Cl. Water content =

6H), 0.96 (t, J = 7.3 Hz, 3H); ¹³C NMR (DMSO-d₆) δ 162.3, 156.1, 137.3,

132.9, 132.3, 131.5, 124.8, 120.3, 113.1, 112.8, 50.0, 42.1, 41.0, 28.2, 25.1,

21.7, 18.4, 13.5. Anal. (C₁₉H₂₈N₄O₂2.2HCl 2.5H₂O) C, H, N, Cl.

3

(E)-3-[1-(3-Dimethylamino-2,2-dimethylpropyl)-2-(2,2-di-

methylpropyl)-1H-benzimidazol-5-yl]-N-hydroxyacrylamide

Hydrochloride Salt (10f). 10f was prepared according to procedures

B and D but using 7a as starting material and was obtained as the hydro-

chloride salt, pinkish solid (3.65 g, 19% in three steps). LCꢀMS m/z

387.2 ([M þ H]^þ). ¹H NMR(CD₃OD) δ 8.26 (d, J = 10.9 Hz, 1H), 7.99

(s, 1H), 7.89 (d, J = 8.8 Hz, 1H), 7.75 (d, J = 15.8 Hz, 1H), 6.68 (d, J =

15.8 Hz, 1H), 4.79 (s, 2H), 3.57 (s, 2H), 3.05 (s, 6H), 1.23 (s, 6H), 1.15

(s, 9H); ¹³C NMR (CD₃OD) δ 165.5, 155.3, 140.0, 135.6, 135.0, 132.0,

126.4, 120.7, 116.6, 114.5, 68.6, 53.6, 40.1, 39.7, 35.8, 27.9, 23.9. Anal.

(C₂₂H₃₄N₄O₂2.4HCl 2.3H₂O) C, H, N, Cl.

3

1.09% (Karl Fisher method). HRMS (ESI) m/z [M þ H]^þcalcd for

C₂₀H₃₁N₄O₂, 359.2442; found, 359.2449.

The following compounds (10aꢀo and 10qꢀab) were prepared

according to procedures AꢀD.

(E)-3-[2-Butyl-1-(3-dimethylamino-2,2-dimethylpropyl)-

1H-benzimidazol-5-yl]-N-hydroxyacrylamide (10a). 10a was

prepared according to procedures B and C but using 7a as starting

material and was obtained as the bis-TFA salt (74 mg, 25% in two steps).

LCꢀMS m/z 373.2 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 8.05 (d, J = 8.8

Hz, 1H), 7.90 (s, 1H), 7.78 (d, J = 8.7 Hz, 1H), 7.61 (d, J = 15.8 Hz, 1H),

6.59 (d, J = 15.7 Hz, 1H), 4.61 (s, 2H), 3.49 (s, 2H), 3.30 (overlapped

with CD₂HOD, 2H), 3.07 (s, 6H), 1.96 (quintet, J = 7.2 Hz, 2H), 1.55

(sextet, J = 7.6 Hz 2H), 1.27 (s, 6H), 1.04 (t, J = 7.3 Hz, 3H); ¹³C NMR

(CD₃OD) δ 165.5 (br), 158.2, 139.8, 135.3, 135.1, 132.4, 126.4, 120.6

(br), 115.6, 114.3, 68.7, 53.5, 47.8 (Me ꢁ 2), 39.5, 29.9, 27.2, 23.6

(Me ꢁ 2), 23.3, 13.9 (TFA peaks at 162.0, 161.7, 119.0, 116.1).

(E)-3-[1-(3-Dimethylamino-2,2-dimethylpropyl)-2-isobutyl-

1H-benzimidazol-5-yl]-N-hydroxyacrylamide (10b). 10b was

prepared according to procedures B and C but using 7a as starting material

and was obtained as the bis-TFA salt (29 mg, 15% in two steps). LCꢀMS

m/z 373.14 ([Mþ H]^þ). ¹H NMR (DMSO-d₆) δ10.80 (br s, 0.5H), 9.47

(br s, 1H), 7.92 (d, J = 7.2 Hz, 1H), 7.89 (s, 1H), 7.64 (d, J = 7.4 Hz, 1H),

7.62 (d, J = 15.5 Hz, 1H), 6.54 (d, J = 15.8 Hz, 1H), 4.39 (s, 2H), 3.33 (s,

2H), 2.97 (d, J = 7.26 Hz, 2H), 2.92 (s, 6H, Me₂N), 2.35 (septet, J = 6.8 Hz,

3

(E)-3-[1-(2-Diethylaminoethyl)-2-(2,2-dimethylpropyl)-1H-

benzimidazol-5-yl]-N-hydroxyacrylamide (10g). 10g was prepared

according to procedures B and C but using 7b as starting material and was

obtained as the bis-TFA salt (50 mg, 28% in two steps). LCꢀMS m/z 373.1

([Mþ H]^þ). ¹HNMR(CD₃OD) δ7.95 (m, 2H), 7.86 (d, J=8.8Hz, 1H),

7.72 (d, J= 15.8 Hz, 1H), 6.62 (d, J= 15.8 Hz, 1H), 5.03 (m, 2H), 3.64 (t, J=

8.1 Hz, 2H), 3.42 (q, J = 7.3 Hz, 4H), 3.20 (s, 2H), 1.40 (t, J = 7.3 Hz, 6H),

1.16 (s, 9H); ¹³CNMR(CD₃OD) δ165.7, 155.3, 140.2, 135.1, 134.6, 133.9,

126.5, 120.3, 115.5, 113.8, 50.2, 41.1, 39.4, 34.9, 29.8, 9.0.

(E)-3-[2-Butyl-1-(2-dimethylaminoethyl)-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10h). 10h was prepared according to

procedures B and C but using 7f as starting material and was obtained as

the bis-TFA salt. LCꢀMS m/z 331.1 ([M þ H]^þ). ¹H NMR(DMSO-d₆)

δ 10.68 (br s, 2H), 7.97 (s, 1H), 7.93 (d, J = 8.6 Hz, 1H), 7.71 (d, J = 8.6

Hz, 1H), 7.62 (d, J = 15.7 Hz, 1H), 6.57 (d, J = 15.7 Hz, 1H), 4.74 (t, J =

7.6 Hz, 2H), 3.54 (t, J = 7.6 Hz, 2H), 3.09 (t, J = 7.72 Hz, 2H), 1.83

(quintet, J = 7.4 Hz, 2H), 1.46 (sextet, J = 7.5 Hz, 2H), 0.97 (t, J = 7.3 Hz,

3H); ¹³C NMR (DMSO-d₆) δ 162.6, 156.1, 138.0, 133.3, 131.7, 123.4,

119.2, 115.0, 112.1, 53.4, 42.6, 28.2, 25.2, 21.7, 13.6.

(E)-3-[2-Butyl-1-(3-methylbutyl)-1H-benzimidazol-5-yl]-N-

hydroxyacrylamide (10i). 10i was prepared according to procedures

B and C but using 7k as starting material and was obtained as the TFA salt

(10 mg, 3.3% in two steps). LCꢀMS m/z 330.1 ([M þ H]^þ). ¹H NMR

(DMSO-d₆) δ 10.79(brs, 1H), 7.91(s, 1H), 7.84(d, J = 8.6Hz, 1H), 7.71

(d, J = 8.6 Hz, 1H), 7.65 (d, J = 15.8 Hz, 1H), 6.55 (d, J = 15.8 Hz, 1H),

4.37 (t, J = 7.8 Hz, 2H), 3.10 (t, J = 7.7 Hz, 2H), 1.82 (quintet, J = 7.7 Hz,

2H), 1.72 (septet or m, J = 6.5 Hz, 1H, Me₂CH), 1.65 (q or dt-like, J = 8.0

Hz, 2H), 1.45 (sextet, J = 7.5 Hz, 2H), 0.98 (d, J = 6.4 Hz, 6H, Me₂CH),

0.96 (t, J = 7.3 Hz, 3H).

(E)-3-[1-(3-Dimethylamino-2,2-dimethylpropyl)-2-isopro-

pyl-1H-benzimidazol-5-yl]-N-hydroxyacrylamide (10j). 10j

was prepared according to procedures B and C but using 7a as starting

material and was obtained as the bis-TFA salt. LCꢀMS m/z 359.2 ([M þ

H]^þ). ¹H NMR (DMSO-d₆) δ 10.80 (br s, 1H), 9.71 (br s, 1H), 7.95 (d,

J = 8.7Hz, 1H), 7.90 (s, 1H), 7.66 (d, J = 8.7Hz, 1H), 7.63 (d, J = 15.8Hz,

1H), 6.55 (d, J = 15.8 Hz, 1H), 4.44 (s, 2H), 3.58 (septet, J = 6.8 Hz, 1H,

Me₂CH), 3.36 (s, 2H), 2.92 (s, 6H, Me₂N), 1.40 (d, J = 6.4 Hz, 6H,

Me₂CH), 1.05 (s, 6H, Me₂C).

1H, Me₂CHꢀ), 1.09 (s, 6H, Me₂C), 0.97 (d, J = 6.6 Hz, 6H, Me₂CH); ¹³

C

NMR (DMSO-d₆) δ 162.7, 156.1, 138.2, 135.2, 131.1, 122.7, 118.9, 118.1,

115.3, 113.5, 66.7, 51.1, 46.7, 38.1, 34.6, 27.6, 22.8, 22.0 (TFA peaks at

158.5, 158.1).

(E)-3-[1-(2-Diethylaminoethyl)-2-isobutyl-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10c). 10c was prepared according to

procedures B and C but using 7b as starting material and was obtained as

the bis-TFA salt (17 mg, 10% in two steps). LCꢀMS m/z 359.1 ([M þ

H]^þ). ¹H NMR (DMSO-d₆) δ 10.81 (s, 0.5H), 10.13 (s, 1H), 7.90 (s, 1H),

7.81 (d, J = 8.5 Hz, 1H), 7.66 (d, J = 8.6 Hz, 1H), 7.61 (d, J = 15.8 Hz, 1H),

6.53 (d, J = 15.8 Hz, 1H), 4.72 (t, J = 7.8 Hz, 2H), 3.47 (t, overlapping with

solvent, 2H), 3.30 (m, 4H), 2.93 (d, J = 7.2 Hz, 2H), 2.27 (m, J = 6.7 Hz,

1H), 1.25 (t, J = 7.2 Hz, 6H), 1.03 (d, J = 6.6 Hz, 6H); ¹³C NMR (DMSO-

d₆) δ 162.7, 155.2, 138.4, 133.9, 131.0, 123.0, 118.6, 116.0, 111.6, 48.8, 46.8,

34.1, 27.1, 22.2, 8.5 (TFA peaks at 158.5, 158.1).

(E)-3-[2-Butyl-1-(3-isopropylaminopropyl)-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10d). 10d was prepared according to

procedures B and C but using 7j as starting material and was obtained as

the bis-TFA salt (60 mg, 28% in two steps). LCꢀMS m/z 359.2 ([M þ

H]^þ). ¹H NMR (DMSO-d₆) δ 8.80 (br s, 2H), 8.00 (d, J = 8.7 Hz, 1H),

7.98 (s, 1H), 7.79 (d, J = 8.6 Hz, 1H), 7.65 (d, J = 15.8 Hz, 1H), 6.63 (d, J =

15.8 Hz, 1H), 4.53 (t, J = 7.5 Hz, 2H), 3.29 (septet, J = 6.2 Hz, 1H), 3.16 (t,

J = 7.8 Hz, 1H), 3.11 (m, 2H), 2.15 (quintet, J = 7.1 Hz, 2H), 1.83 (quintet,

J = 7.7 Hz, 2H), 1.44 (sextet, J= 7.6 Hz, 2H), 1.23 (d, J= 6.5 Hz, 6H), 0.95 (t,

J = 7.3 Hz, 3H).

(E)-3-[2-Butyl-1-(2-isopropylaminoethyl)-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10e). 10e was prepared according to

procedures B and D but using 7d as starting material and was obtained as a

off-white hydrochloride salt (7 g, 22% in three steps). LCꢀMS m/z 345.2

([M þ H]^þ). ¹H NMR (DMSO-d₆) δ 10.37 (br s, 0.33H), 9.92 (br s, 2H,

ꢀNH₂^þꢀCHMe₂), 8.16 (d, J = 8.6 Hz, 1H), 7.97 (s, 1H), 7.78 (d, J = 8.6

Hz, 1H), 7.63 (d, J = 15.8 Hz, 1H), 6.67 (d, J = 15.8 Hz, 1H), 4.93 (t, J = 6.4

Hz, 2H), 3.36 (t, J= 5.4 Hz, 2H), 3.31(m, 1H), 3.30 (t, J= 7.8 Hz, 2H), 1.86

(quintet, J = 7.6 Hz, 2H), 1.47 (sextet, J = 7.5 Hz, 2H), 1.27 (d, J = 6.5 Hz,

(E)-3-[2-Cyclohexyl-1-(3-dimethylamino-2,2-dimethylpro-

pyl)-1H-benzimidazol-5-yl]-N-hydroxyacrylamide (10k). 10k

was prepared according to procedures B and C but using 7a as starting

material and was obtained as the bis-TFA salt. LCꢀMS m/z 399.1

1

([M þ H]^þ). H NMR (CD₃OD) δ 8.06 (d, J = 8.8 Hz, 1H), 7.91

(s, 1H), 7.80 (d, J = 8.8 Hz, 1H), 7.68 (d, J = 15.8 Hz, 1H), 6.61 (d, J =

15.8 Hz, 1H), 4.64 (s, 2H), 3.50 (s, 2H), 3.39 (m, 1H), 3.07 (s, 6H), 2.12

(d, J = 12.0 Hz, 2H), 1.96 (d, J = 13.0 Hz, 2H), 1.87 (d, J = 12.8 Hz, 1H),

1.79 (q-like, J = 10.7 Hz, 2H), 1.58 (q-like, J = 12.9 Hz, 2H), 1.47 (t-like,

J = 12.6 Hz, 1H), 1.26 (s, 6H); ¹³C NMR (CD₃OD) δ 165.7, 161.3,

140.1, 135.4, 134.8, 134.0, 126.1, 120.3, 115.5, 114.9, 68.7, 53.1, 47.9,

39.2, 37.0, 32.4, 26.5, 26.3, 23.6 (TFA peaks at 119.6, 116.7).

4714

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

(E)-3-[1-(2-Diethylaminoethyl)-2-propyl-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10l). 10l was prepared according to

procedures B and C but using 7b as starting material and was obtained as

the bis-TFA salt (31 mg, 8.2% in two steps). LCꢀMS m/z 345.1 ([M þ

H]^þ). ¹H NMR (CD₃OD) δ 8.15 (d, J = 8.7 Hz, 1H), 7.94 (s, 1H), 7.89

(d, J = 8.3 Hz, 1H), 7.68 (d, J = 15.8 Hz, 1H), 6.63 (d, J = 15.8 Hz, 1H),

5.08 (t, J = 7.7 Hz, 2H), 3.70 (t, J = 7.7 Hz, 2H), 3.44 (m, 4H), 3.35 (t, J =

6.3 Hz, 2H), 2.03 (sextet, J = 7.9 Hz, 2H), 1.44 (t, J = 7.2 Hz, 6H), 1.20 (t,

J = 7.3 Hz, 3H); ¹³C NMR (CD₃OD) δ 165.5, 157.4, 139.8, 135.5, 133.5,

132.3, 126.8, 120.7, 114.5, 114.3, 50.0, 40.8, 28.5, 21.0, 13.9, 9.1.

(E)-3-[1-(2-Diethylaminoethyl)-2-hexyl-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10m). 10m was prepared according

to procedures B and C but using 7b as starting material and was obtained

as the bis-TFA salt. LCꢀMS m/z 387.16 ([M þ H]^þ). ¹H NMR

(DMSO-d₆) δ 10.60 (br s, 2H), 7.94 (s, 1H), 7.93 (d, J = 8.8 Hz, 1H),

7.73 (d, J = 8.8 Hz, 1H), 7.63 (d, J = 15.8 Hz, 1H), 6.59 (d, J = 15.8 Hz,

1H), 4.81 (t, J = 7.7 Hz, 2H), 3.51 (t, J = 7.7 Hz, 2H), 3.31 (m, 4H), 3.12

(t, J = 7.7 Hz, 2H), 1.85 (quintet, J = 7.6 Hz, 2H), 1.44 (m, 2H),

1.37ꢀ1.30 (m, 4H), 1.26 (t, J = 7.2 Hz, 6H), 0.88 (t, J = 7.0 Hz, 3H); ¹³C

NMR (DMSO-d₆) δ 162.6, 156.2, 137.9, 134.7, 133.1, 131.9, 123.7,

119.4, 114.7, 112.2, 48.6, 46.7, 38.6, 30.8, 28.2, 26.1, 25.4, 21.9, 13.8, 8.5

(TFA peaks at 159.1, 158.8, 158.4, 158.1, 117.8).

(E)-3-[2-Butyl-1-(2-ethylaminoethyl)-1H-benzimidazol-5-yl]-

N-hydroxyacrylamide (10n). 10n was prepared according to proce-

dures B and D but using 7c as starting material and was obtained as the

hydrochloride salt, light pinkish solid (0.98 g, 24% in three steps). LCꢀMS

m/z 331.2 ([M þ H]^þ). ¹H NMR (DMSO-d₆) δ 10.94 (br s, 0.8H), 9.85

(br s, 2H, ꢀNH₂^þꢀEt), 8.13 (d, J = 8.6 Hz, 1H), 7.97 (s, 1H), 7.78 (d, J =

8.5 Hz, 1H), 7.64 (d, J = 15.8 Hz, 1H), 6.65 (d, J = 15.8 Hz, 1H), 4.88 (t, J =

5.9 Hz 2H), 3.37 (m, 2H), 3.29 (t, J = 7.7 Hz, 2H), 2.96 (m, 2H), 1.86

(quintet, J = 7.6 Hz,, 2H), 1.47 (sextet, J = 7.5 Hz, 2H), 1.21 (t, J = 7.2 Hz,

3H), 0.95 (t, J = 7.0 Hz, 3H, ꢀ(CH₂)₃CH₃); ¹³C NMR (DMSO-d₆) δ

162.4, 156.1, 137.4, 132.8, 132.7, 131.9 (br), 124.6, 120.3, 113.2, 112.9, 44.5,

42.2, 41.1, 28.2, 25.2, 21.7, 13.6, 10.8. HRMS (ESI) m/z[Mþ H]^þcalcd for

C₁₈H₂₇N₄O₂, 331.2129; found, 331.2131.

(E)-3-[1-(2-Dimethylaminoethyl)-2-hexyl-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10s). 10s was prepared according to

procedures B and D but using 7f as starting material and was obtained as the

hydrochloride salt, pinkish solid (3.09 g, 26% in three steps). LCꢀMS m/z

359.2 ([M þ H]^þ). ¹HNMR(CD₃OD) δ8.09 (d, J= 8.7 Hz, 1H), 7.90 (s,

1H), 7.87 (d, J= 8.8 Hz, 1H), 7.63 (d, J= 15.8 Hz, 1H), 6.58 (d, J = 15.8 Hz,

1H), 4.98 (t, J = 7.8 Hz, 2H), 3.71 (t, J = 7.8 Hz, 2H), 3.34 (t, J = 10.0 Hz,

2H), 3.07 (s, 6H), 1.98 (quintet, J = 7.8 Hz, 2H), 1.59 (quintet, J = 7.5 Hz,

2H), 1.50ꢀ1.34 (m, 4H), 0.95 (t, J = 7.1 Hz, 3H); ¹³C NMR (CD₃OD) δ

165.5, 157.7, 140.0, 135.5, 133.7, 132.5, 126.8, 120.6, 114.6, 114.2, 54.9,

44.0, 40.9, 32.5, 30.0, 27.4, 26.9, 23.5, 14.3. Anal. (C₂₀H₃₀N₄O₂1.9HCl

3

H₂O) C, H, N, Cl.

(E)-N-Hydroxy-3-[1-(2-methylaminoethyl)-2-pentyl-1H-ben-

zimidazol-5-yl]acrylamide (10t). 10t was prepared according to

procedures B and C but using 7g as starting material and was obtained as

the bis-TFA salt (10 mg, 1.8% in two steps). LCꢀMS m/z 331.2 ([M þ

H]^þ). ¹H NMR (DMSO-d₆) δ 9.13 (2H, s, ꢀNH₂^þꢀCH₃), 7.94 (d, J =

8.1 Hz, 1H), 7.93 (s, 1H), 7.73 (d, J = 8.8 Hz, 1H), 7.64 (d, J = 15.8 Hz,

1H), 6.58 (d, J = 15.8 Hz, 1H), 4.71 (t, J = 6.3 Hz, 2H), 3.38 (t-like, 2H),

3.13 (t, J = 7.8 Hz, 2H), 2.61 (t, J = 4.6 Hz, 3H, ꢀNH₂^þꢀCH₃), 1.84

(sextet, J = 7.7 Hz, 2H), 1.48ꢀ1.35 (m, 4H), 0.92 (t, J = 7.1 Hz, 3H).

(E)-3-[2-Hexyl-1-(2-methylaminoethyl)-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10u). 10u was prepared according to

procedures B and C but using 7g as starting material and was obtained as

the bis-TFA salt (10 mg, yield 3.4% in two steps). LCꢀMS m/z 345.2

([M þ H]^þ). ¹H NMR (CD₃OD) δ 7.87 (s, 2H), 7.8 (d, J = 8.6 Hz,

1H), 7.60 (d, J = 15.7 Hz, 1H), 6.53 (d, J = 15.7 Hz, 1H), 4.81

(overlapped with solvent, 2H), 3.59 (m, 2H), 3.18 (t, J = 7.8 Hz, 2H),

2.81 (s, 3H), 1.95 (m, 2H), 1.56 (m, 2H), 1.49ꢀ1.35 (m, 4H), 0.95 (t,

J = 7.1 Hz, 3H); ¹³C NMR (CD₃OD) δ 140.5, 134.7, 126.2, 119.6, 115.7,

113.2, 42.2, 34.2, 32.5, 30.0, 27.6, 27.1, 23.5, 14.3.

(E)-N-Hydroxy-3-[1-(2-methylaminoethyl)-2-octyl-1H-ben-

zimidazol-5-yl]acrylamide (10v). 10v was prepared according to

procedures B and C but using 7f as starting material and was obtained as

the bis-TFA salt. LCꢀMS m/z 373.2 [(M þ H)^þ]. ¹H NMR (CD₃OD)

δ 7.91 (d, J = 8.4 Hz, 1H), 7.80 (s, 1H), 7.76 (d, J = 8.6 Hz, 1H), 7.38 (d,

J = 15.7 Hz, 1H), 6.42 (d, J = 15.7 Hz, 1H), 4.84 (overlapped with

solvent, 2H), 3.64 (t-like, 2H), 3.24 (t, J = 7.8 Hz, 2H), 2.81 (s, 3H), 1.96

(E)-3-[1-(2-Ethylaminoethyl)-2-pentyl-1H-benzimidazol-5-

yl]-N-hydroxyacrylamide (10o). 10o was prepared according to

procedures B and C but using 7b as starting material and was obtained as

the bis-TFA salt (30 mg, 5.3% in two steps). LCꢀMS m/z 345.2 ([M þ

H]^þ). ¹H NMR (CD₃OD) δ 7.87 (d, J = 8.6 Hz, 1H), 7.81 (s, 1H), 7.75

(d, J = 8.6 Hz, 1H), 7.45 (d, J = 15.7 Hz, 1H), 6.44 (d, J = 15.7 Hz, 1H),

4.79 (t, J = 6.5 Hz, 2H), 3.62 (t, J = 6.4 Hz, 2H), 3.21 (t, J = 7.2 Hz, 2H),

3.18 (q, J = 7.2 Hz, 2H), 1.95 (quintet, J = 7.6 Hz, 2H), 1.55ꢀ1.43 (m,

(m, 2H), 1.54 (m, 2H), 1.46ꢀ1.32 (m, 8H), 0.91 (t, J = 6.6 Hz, 3H); ¹³

C

NMR (CD₃OD) δ 165.7, 157.8, 140.4, 134.5, 126.3, 119.2, 115.6, 113.6,

48.3, 42.5, 34.3, 32.9, 30.33, 30.27, 30.23, 27.5, 27.0, 23.7, 14.4 (TFA

peaks at 163.0, 162.7).

(E)-N-Hydroxy-3-[1-(2-aminoethyl)-2-octyl-1H-benzimi-

dazol-5-yl]acrylamide (10w). Boc protected 9w was prepared

according to procedure B using 7i as starting material. Then it was

deprotected (concentrated HCl in HOAc, 70 ꢀC) to afford 9w1, which

was further converted to the bis-TFA salt of 10w according to procedure

C. LCꢀMS m/z 359.2 [(M þ H)^þ]. ¹H NMR (CD₃OD) δ 7.90 (d, J =

8.5 Hz, 1H), 7.85 (s, 1H), 7.78 (d, J = 8.6 Hz), 1H, 7.49 (d, J = 15.6 Hz,

1H), 6.49 (d, J = 15.7 Hz, 1H), 4.78 (t, J = 6.3 Hz, 2H), 3.56 (t, J = 6.3 Hz,

2H), 3.23 (t, J = 8.3 Hz, 2H), 1.95 (m, 2H), 1.55 (m, 2H), 1.47ꢀ1.30 (m,

8H), 0.91 (t, J = 6.9 Hz, 3H); ¹³C NMR (CD₃OD) δ 165.7, 157.8, 140.3,

134.6, 134.5, 126.4, 119.6, 115.5, 113.4, 43.2, 39.1, 32.9, 30.34, 30.28,

30.23, 27.6, 27.0, 23.7, 14.4 (TFA peaks at 163.0, 162.6).

(E)-3-[1-(2-Ethylaminoethyl)-2-((Z)-hex-3-enyl)-1H-benzi-

midazol-5-yl]-N-hydroxyacrylamide (10x). 10x was prepared

according to procedures B and D but using 7c as starting material and

was obtained as the hydrochloride salt (82 mg, 1.8% in three steps).

LCꢀMS m/z 357.1 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 8.08 (d, J =

8.0 Hz, 1H), 7.87 (s, 1H), 7.83 (d, J = 8.1 Hz, 1H), 7.35 (d, J = 15.7 Hz,

1H), 6.47 (d, J = 15.7 Hz, 1H), 5.57 (dt, J = 10.7, 7.1 Hz, 1H), 5.55 (dt, J =

10.7, 7.1 Hz, 1H), 4.90 (masked peaks, 2H), 3.66 (t-like, 2H), 3.42 (t, J =

7.4 Hz, 2H), 3.20 (q-like,J = 7.0 Hz, 2H), 2.75 (q-like, J = 7.0 Hz, 2H), 2.05

(quintet, J = 7.2 Hz, 2H), 1.38 (t, J = 6.9 Hz, 3H), 0.89 (t, J = 7.5 Hz, 3H);

13

4H), 1.34 (t, J = 7.3 Hz, 3H), 0.98 (t, J = 7.1 Hz, 3H); C NMR

(CD₃OD) δ 165.7, 157.8, 140.5, 135.1, 134.7, 134.3, 126.2, 119.2, 115.8,

113.3, 46.3, 45.0, 42.4, 32.5, 27.3, 27.0, 23.3, 14.2, 11.5.

(E)-3-[2-Hexyl-1-(2-isopropylaminoethyl)-1H-benzimidazol-

5-yl]-N-hydroxyacrylamide (10q). 10q was prepared according to

procedures B and C but using 7d as starting material and was obtained as the

bis-TFA salt (11.2 mg, 2.5% in two steps). LCꢀMS m/z 373.1 ([M þ

H]^þ). ¹H NMR (CD₃OD) δ 7.85 (m, 2H), 7.74 (s, 1H), 7.57 (br d, 1H),

6.52 (br d, 1H), 4.76 (d, J = 6.6 Hz, 2H), 3.58 (d, J = 6.6 Hz, 2H), 3.48 (m,

J = 6.5 Hz, 1H), 3.18 (t, J = 5.9 Hz, 2H), 1.94 (m, 2H), 1.55 (m, 2H),

1.47ꢀ1.37 (m, 4H), 1.37 (d, J = 6.5 Hz, 6H), 0.84 (d, J = 7.0 Hz, 3H).

(E)-3-[2-Butyl-1-(2-diisopropylaminoethyl)-1H-benzimida-

zol-5-yl]-N-hydroxyacrylamide (10r). 10r was prepared according

to procedures B and C but using 7e as starting material and was obtained

as the bis-TFA salt (30 mg, 9.8% in two steps). LCꢀMS m/z 387.1

([M þ H]^þ). ¹H NMR (DMSO-d₆) δ 10.73 (s, 1H), 9.08 (s, 1H), 7.84

(s, 1H), 7.62ꢀ7.57 (m, 2H), 7.60 (d, J = 15.7 Hz, 1H), 6.47 (d, J =

15.7 Hz, 2H), 4.61 (br t-like, 2H), 3.80 (m, 2H), 3.40 (masked by

solvent peak, 2H), 2.96 (t, J = 7.2 Hz, 2H), 1.85 (quintet, J = 7.5 Hz, 2H),

1.47 (sextet, J = 7.4 Hz, 2H), 1.37 (d, J = 6.0 Hz, 6H), 1.34 (d, J = 6.2 Hz,

6H), 0.97 (t, J = 7.4 Hz, 3H).

4715

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

¹³C NMR (CD₃OD) δ 165.4, 157.1, 140.0, 136.2, 135.2, 133.9, 132.6,

127.0, 125.8, 119.8, 114.9, 114.4, 46.4, 45.1, 42.9, 27.1, 25.3, 21.5, 14.4, 11.6.

(E)-N-Hydroxy-3-[1-(2-isopropylaminoethyl)-2-pentyl-1H-

benzimidazol-5-yl]acrylamide (10y). 10y was prepared according

to procedures B and D but using 7d as starting material and was obtained

as the hydrochloride salt, light pinkish solid (1.4 g, 20% in three steps).

LCꢀMS m/z 359.2 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 9.27 (residual

proton after exchange, 0.16 H), 8.09 (d, J = 8.4 Hz, 1H), 7.85 (s, 1H),

7.80 (d, J = 8.2 Hz, 1H), 7.34 (d, J = 15.7 Hz, 1H), 6.46 (d, J = 15.8 Hz,

1H), 4.93 (overlapped with solvent peak, 2H), 3.65 (t-like, 2H), 3.51

(septet, J = 6.3 Hz, 1H), 3.36 (t, J = 7.8 Hz, 2H), 1.99 (quintet, J = 7.2 Hz,

2H), 1.56 (m, 2H), 1.49 (m, 2H), 1.42 (t, J = 6.4 Hz, 6H), 0.98 (t, J = 7.2

Hz, 3H); ¹³C NMR (CD₃OD) δ 165.3, 157.7, 140.0, 135.0, 133.9, 132.5,

127.0, 119.7, 114.8, 114.3, 53.3, 44.0, 42.8, 32.4, 27.1, 27.0, 23.3, 19.2, 14.2.

(E)-3-[1-(3-Dimethylamino-2,2-dimethylpropyl)-2-((Z)-hex-

3-enyl)-1H-benzimidazol-5-yl]-N-hydroxyacrylamide (10z).

10z was prepared according to procedures B and C but using 7a as

starting material and was obtained as the bis-TFA salt. LCꢀMS m/z

399.1 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 8.05 (d, J = 8.7 Hz, 1H), 7.93

(s, 1H), 7.79 (d, J = 8.63 Hz, 1H), 7.64 (d, J = 15.8 Hz, 1H), 6.51 (d, J =

15.8 Hz, 1H), 5.53ꢀ5.40 (m, 2H), 4.61 (s, 2H), 3.48 (s, 2H), 3.37 (t, J =

7.3 Hz, 2H), 3.06 (s, 6H), 2.74 (q, J = 6.8 Hz, 2H), 2.00 (quintet, J = 7.3

Hz, 2H), 1.26 (s, 6H), 0.84 (t, J = 7.5 Hz, 3H); ¹³C NMR (CD₃OD) δ

165.7, 157.9, 140.2, 135.8, 134.549, 134.458, 126.1, 126.0, 120.0, 115.2,

115.1, 68.7, 53.3, 47.9, 39.6, 27.6, 25.9, 23.7, 21.4, 14.4 (TFA peaks at

163.1, 162.8, 119.6, 116.7).

(E)-3-[1-(3-Dimethylamino-2,2-dimethylpropyl)-2-(2,4,4-

trimethylpentyl)-1H-benzimidazol-5-yl]-N-hydroxyacryla-

mide (10aa). 10aa was prepared according to procedures B and C but

using 7a as starting material and was obtained as the bis-TFA salt.

LCꢀMS m/z 429.2 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 8.19 (d, J = 8.8

Hz, 1H), 8.08 (s, 1H), 7.90 (d, J = 8.8 Hz, 1H), 7.76 (d, J = 15.7 Hz, 1H),

6.75 (d, J = 15.8 Hz, 1H), 4.79 (s, 2H), 3.61 (s, 2H), 3.41 (dd, J = 10.1,

6.5 Hz, 1H), 3.32 (dd, J = 15.7, 9.0 Hz, 1H), 3.18 (s, 6H, Me₂N), 2.52 (br

s, 1H), 1.50ꢀ1.45 (m, 2H), 1.36 (d, J = 3.8 Hz, 6H, Me₂C), 1.12 (d, J =

5.5 Hz, 3H, MeCH), 1.02 (s, 9H, Me₃C); ¹³C NMR (CD₃OD) δ 165.6,

157.4, 139.9, 135.2, 135.1, 132.9, 126.4, 120.6, 115.7, 114.6, 68.6, 53.3,

51.4, 47.9, 39.7, 36.3, 31.9, 31.3, 30.2, 23.8, 22.3 (TFA peaks at 163.4,

163.0, 162.7, 162.3, 122.4, 119.5, 116.6, and 113.7).

acid (145 mg, 0.896 mmol) in DCM (7 mL) was added a coupling cocktail

solution containing EDCI (0.288 g, 1.50 mmol), HOBt hydrate (0.230 g,

1.50 mmol), DIEA (0.26 mL, 1.50 mmol), and DCM (7 mL). After the

mixture was stirred for 0.5 h, aniline 11b (0.200 g, 0.686 mmol) was added

and stirred until the reaction was completed. The usual workup afforded

crude amide 13b whichwasaddedtoaceticacid(5mL) andheatedat90ꢀC

overnight. When the reaction was completed, the mixture was concentrated

to dryness, diluted with DCM, washed with NaHCO₃, and dried over

Na₂SO₄. The mixture was filtered and concentrated, and the resulting crude

was purified by flash chromatography (silica, 50ꢀ100% EtOAc in hexanes

with 1% Et₃N) to afford methyl ester 14b (120 mg, 42%) which was

converted to hydroxamic acid 15b as bis-TFA salt (30 mg, 16%) according

to procedure C. LCꢀMS m/z 419.1 ([M þ H]^þ). ¹H NMR (CD₃OD) δ

7.89 (d, J = 8.7 Hz, 1H), 7.83 (s, 1H), 7.75 (d, J = 8.8 Hz, 1H), 7.54 (d, J =

15.8 Hz, 1H), 7.40ꢀ7.27 (m, 5H), 6.49 (d, J = 15.8 Hz, 1H), 4.97

(overlapped with solvent peak, 2H), 3.71 (m, 1H), 3.63 (m, 1H), 3.49 (m,

1H), 3.22 (m, 4H), 2.90 (m, 1H), 2.84 (m, 1H), 2.16 (m, 1H), 2.05 (m,

1H), 1.18 (t, J = 7.2 Hz, 6H). ¹³C NMR (CD₃OD) δ 165.9, 157.4, 140.8,

140.4, 137.1, 135.5, 133.7, 130.0, 128.3, 127.1, 125.7, 118.9, 116.4, 112.4,

50.4, 48.7, 39.9, 24.5, 19.5, 17.9 (CH₂), 8.8 (TFA peaks at 163.2, 162.9,

162.5, and 162.2).

(E)-3-[2-(2-Diethylaminoethyl)-1-(3-methylbutyl)-1H-ben-

zimidazol-5-yl]-N-hydroxyacrylamide (15a). 15a was prepared

according to the procedures described for compound 15b but using

diamine 11k and 3-diethylaminopropionic acid hydrochloride as starting

materials and was obtained as the bis-TFA salt (30 mg, 6.9% from 11k in

three steps). LCꢀMS m/z 373.4 ([M þ H]^þ). ¹H NMR (DMSO-d₆) δ

9.87 (br s, 1H), 7.85 (s, 1H), 7.67 (d, J = 8.5 Hz, 1H), 7.60 (d, J = 15.3

Hz, 1H), 7.57 (d, J = 6.7 Hz, 1H), 6.49 (d, J = 15.8 Hz, 1H), 4.28 (t, J =

7.6 Hz, 2H), 3.68 (t, J = 7.5 Hz, 2H), 3.43 (t, J = 7.6 Hz, 2H), 3.28 (q, J =

7.3 Hz, 4H), 1.69ꢀ1.61 (m, 3H), 1.28 (t, J = 7.2 Hz, 6H), 0.98 (d, J = 6.4

Hz, 6H); ¹³C NMR (DMSO-d₆) δ 162.9, 151.9, 139.8, 138.9, 135.1,

129.7, 122.1, 117.7, 117.4, 111.1, 48.0, 46.6, 41.9, 37.9, 25.4, 22.3, 21.2,

8.5 (TFA peaks at 158.5, 158.1, 117.4, and 114.5).

(E)-3-[1-(2-Dimethylaminoethyl)-2-(2-phenylcyclopropyl)-

1H-benzimidazol-5-yl]-N-hydroxyacrylamide (15c). 15c was

prepared according to the procedure described for compound 15b but

using diamine 11c (37% from 7f) as starting material and was obtained

as bis-TFA salt (25 mg, 5.3% in three steps from 11c). LCꢀMS m/z

391.1 ([M þ H]^þ). ¹H NMR (DMSO-d₆) δ 10.80 (br s, 1H), 10.21 (br

s, 1H), 7.84 (s, 1H), 7.77 (d, J = 7.5 Hz, 1H), 7.615 (d, J = 10.7 Hz, 1H),

7.609 (d, J = 15.2 Hz, 1H), 7.37ꢀ7.31 (m, 4H), 7.26 (m, 1H), 6.52 (d,

J = 15.8 Hz, 1H), 4.77 (m, 2H), 3.55ꢀ3.40 (m, 2H), 2.81 (s, 6H),

2.85ꢀ2.72 (m, 2H), 1.89 (m, 1H), 1.76 (m, 1H).

(E)-3-[1-(2-Ethylaminoethyl)-2-(2,4,4-trimethylpentyl)-1H-

benzimidazol-5-yl]-N-hydroxyacrylamide (10ab). 10ab was

prepared according to procedures B and C but using 7c as starting material

and was obtained as the bis-TFA salt. LCꢀMS m/z 387.15 ([M þ H]^þ).

¹H NMR (CD₃OD) δ 7.96 (d, J = 8.6 Hz, 1H), 7.79 (s, 1H), 7.78ꢀ7.75 (d,

J = 8.7 Hz, 1H), 7.23 (d, J = 15.7 Hz, 1H), 6.37 (d, J = 15.7 Hz, 1H), 4.92

(masked peaks, 2H), 3.70 (m, 2H), 3.36ꢀ3.28 (masked peak, 1H),

3.26ꢀ3.14 (m, 3H), 2.31 (m, 1H), 1.44ꢀ1.27 (m, 2H), 1.35 (t, J = 7.2

Hz, 3H), 1.07 (d, J = 6.6 Hz, 3H, MeCH), 0.92 (s, 9H, Me₃C); ¹³C NMR

(CD₃OD) δ 165.6, 156.9, 140.6, 134.9, 134.5, 134.2, 126.2, 118.7, 116.0,

113.7, 51.6, 46.5, 45.0, 42.7, 35.8, 31.9, 30.8, 30.2, 22.6, 11.4 (TFA peaks at

163.4 and 163.0).

(E)-3-[1-(2-Aminoethyl)-2-(2-phenylcyclopropyl)-1H-benz-

imidazol-5-yl]-N-hydroxylacrylamide (15d). 15d was prepared

according to the procedure described for compound 15b but using

diamine 11i (78% from 7i) as starting material. Amide 13d (126.0 mg,

75% from 7i) was cyclized to 14d, and then the Boc group was removed

with HCl to give 14d1 (104 mg, 91% as hydrochloride salt), which was

further converted to 15d as the bis-TFA salt (14 mg, 10%) according to

procedure C. LCꢀMS m/z 363.1 ([M þ H]^þ). ¹H NMR (CD₃OD) δ

7.82 (s, 1H), 7.75 (d, J = 8.6 Hz, 1H), 7.70 (d, J = 8.5 Hz, 1H), 7.60 (d, J =

15.8 Hz, 2H), 7.37ꢀ7.30 (m, 4H), 7.28ꢀ7.23 (m, 1H), 4.76 (m, 2H),

3.45 (t, J = 6.6 Hz, 2H), 2.89 (m, 1H), 2.67 (m, 1H), 2.03 (m, 1H), 1.95

(m, 1H); ¹³C NMR (CD₃OD) δ 165.9, 157.7, 140.8, 140.2, 137.2, 135.7,

133.6, 129.8 (CH ꢁ 2), 128.1, 127.2 (CH ꢁ 2), 125.6, 118.8, 116.2,

112.2, 42.7, 39.3, 29.3, 19.0, 18.5 (TFA peaks at 162.6, 162.3)

(E)-3-[2-Bicyclo[2.2.1]hept-2-ylmethyl-1-(2-diethylamino-

ethyl)-1H-benzimidazol-5-yl]-N-hydroxyacrylamide (15e).

15e was prepared according to a similar procedure used for 15b and was

obtained as the bis-TFA salt (30 mg, 7.2% in three steps from 11b).

LCꢀMS m/z 411.2 ([M þ H])^þ. ¹H NMR (DMSO-d₆) δ 10.36 (br s,

1H), 7.93 (s, 1H), 7.87 (d, J = 8.6 Hz, 1H), 7.71 (d, J = 8.5 Hz, 1H), 7.63

Representative Procedures for Compounds 11 (Proce-

dure E). Preparation of (E)-3-[3-Amino-4-(2-diethylami-

noethylamino)phenyl]acrylic Acid Methyl Ester (11b). To

a prestirred solution of 7b (3.21 g, 10.0 mmol) in MeOH/HOAc

(3:1, 50 mL), SnCl₂2H₂O was added (9.45 g, 41.9 mmol). The result-

3

ing solution was heated at 45 ꢀC overnight. Then the solvent was removed

under vacuum. The residue was basified and extracted with DCM and

purified by flash chromatography (silica, 0ꢀ10% MeOH in DCM) to afford

11b (2.30 g, 79%). LCꢀMS m/z 292.2 ([M þ H]^þ).

Representative Procedures for Synthesis of Compounds

15 (Procedure F). (E)-3-[1-(2-Diethylaminoethyl)-2-(2-phenyl-

cyclopropyl)-1H-benzimidazol-5-yl]-N-hydroxyacrylamide

(15b). To a prestirred solution of trans-2-phenylcyclopropanecarboxylic

4716

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

(d, J = 15.8 Hz, 1H), 6.56 (d, J = 15.8 Hz, 1H), 4.77 (t, J = 7.8 Hz, 2H),

3.49 (t, J = 7.6 Hz, 2H), 3.31 (m, 4H), 3.06 (dd, J = 16.0 and 7.9 Hz, 1H),

3.06 (dd, J = 16.0 and 7.6 Hz, 1H), 2.25 (m, 1H), 2.15ꢀ2.05 (m, 2H),

1.57ꢀ1.42 (m, 4H), 1.25 (t, J = 7.2 Hz, 6H), 1.27ꢀ1.10 (m, 4H).

(E)-3-{2-Butyl-1-[2-(isopropylmethylamino)ethyl]-1H-benz-

imidazol-5-yl}-N-hydroxyacrylamide (17c). A solution of 3-[2-

butyl-1-(2-isopropylaminoethyl)-1H-benzimidazol-5-yl]acrylic acid methyl

ester 9e (170 mg, 0.495 mmol), formaldehyde (37%, 0.10 mL, 1.34 mmol),

NaBH(OAc)₃(159 mg, 0.75 mmol), and HOAc (36 mg, 0.60 mmol) in

MeOH was stirred at room temperature until depletion of 9e. After base

workup, 16c (170 mg, 96% from 9e) was converted to 17c as the bis-TFA

salt (20 mg, 7.2%) according to procedure C. LCꢀMS m/z 359.2 ([M þ

H]^þ). ¹H NMR (CD₃OD) δ7.97 (d, J = 8.7 Hz, 1H), 7.89 (s, 1H), 7.82 (d,

J = 8.6 Hz, 1H), 7.63 (d, J = 15.8 Hz, 1H), 6.56 (d, J = 15.8 Hz, 1H), 4.93 (t,

J = 8.1 Hz, 2H), 3.80 (septet, J = 6.6 Hz, 1H), 3.64 (t, J = 7.9 Hz, 2H), 3.26

(t, J = 7.9 Hz, 2H), 2.99 (s, 3H), 1.94 (quintet, J = 6.3 Hz, 2H), 1.58 (septet,

J = 7.6 Hz, 2H), 1.39 (d, J = 6.6 Hz, 6H), 1.06 (t, J = 7.3 Hz, 3H). ¹³C NMR

(CD₃OD) δ 165.6, 157.7, 140.2, 134.9, 134.1, 126.5, 120.1, 115.2, 113.6,

59.9, 40.7, 35.6 (NꢀMe), 29.5, 26.6, 23.4, 16.6 (identified by HSQC,

isopropyl Me ꢁ 2), 14.0 (TFA peaks at 162.8, 162.4).

3.48ꢀ3.30 (m, 2H), 3.30ꢀ3.27 (m, 2H), 2.84 (m, 2H), 2.30 (d, J = 13.0

Hz, 2H), 1.89 (quintet, J = 7.5 Hz, 2H), 1.56 (sextet, J = 7.5 Hz, 2H),

1.06 (t, J = 7.2 Hz, 3H).

(E)-3-(2-Butyl-1-(piperidin-3-yl)-1H-benzimidazol-5-yl)-N-

hydroxyacrylamide (20c). 20c was prepared according to the

procedures described for compound 20a but using 7m as starting

material and was obtained as the bis-TFA salt (25 mg, 8.8% in three

steps from 7m). LCꢀMS m/z 343.2 ([M þ H]^þ). ¹H NMR (CD₃OD)

δ 8.12 (d, J = 8.8 Hz, 1H), 7.92 (s, 1H), 7.76 (d, J = 9.2 Hz, 1H), 7.70 (d,

J = 15.8 Hz, 1H), 6.60 (d, J = 15.8 Hz, 1H), 5.03 (tt, J = 12.3, 4.0 Hz, 1H),

3.84 (t, J = 12.0 Hz, 1H), 3.71 (dd, J = 11.9, 3.1 Hz, 1H), 3.56 (d, J = 12.5

Hz, 1H), 3.35ꢀ3.25 (m, 1H), 3.23 (m, 2H), 2.68 (m, 1H), 2.26 (m, 2H),

2.09 (m, 1H), 1.89 (quintet, J = 7.4 Hz, 2H), 1.56 (sextet, J = 7.6 Hz,

2H), 1.06 (t, J = 7.2 Hz, 3H).

(E)-3-(2-Hexyl-1-(pyrrolidin-3-yl)-1H-benzimidazol-5-yl)-

N-hydroxyacrylamide (20d). 20d was prepared according to the

procedures described for compound 20a but using 7l as starting material

and was obtained as the bis-TFA salt (85 mg, 58% from 19d). LCꢀMS

m/z 357.1 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 7.99 (d, J = 8.7 Hz, 1H),

7.94 (s, 1H), 7.76 (d, J = 8.7 Hz, 1H), 7.67 (d, J = 15.8 Hz, 1H), 6.59 (d,

J = 15.8 Hz, 1H), 5.67 (quintet, J = 9.2 Hz, 1H), 4.00ꢀ3.85 (m, 3H),

3.57 (td, J = 11.4, 3.5 Hz, 1H), 3.26 (d, J = 7.9 Hz, 2H), 2.82 (m, 1H),

2.70 (m, 1H), 1.91 (quintet, J = 7.4 Hz, 2H), 1.54 (m, 2H), 1.46ꢀ1.34

(m, 4H), 0.94 (t, J = 7.1 Hz, 3H).

(E)-3-(2-Hexyl-1-(piperidin-3-yl)-1H-benzimidazol-5-yl)-

N-hydroxyacrylamide (20e). 20e was prepared according to the

procedures described for compound 20a but using 7m as starting

material and was obtained as the bis-TFA salt (28 mg, 9.3% in three

steps from 7m). LCꢀMS m/z 371.4 ([M þ H]^þ). ¹H NMR (CD₃OD)

δ 8.12 (d, J = 8.8 Hz, 1H), 7.92 (s, 1H), 7.76 (d, J = 9.2 Hz, 1H), 7.68 (d,

J = 15.8 Hz, 1H), 6.60 (d, J = 15.8 Hz, 1H), 5.05 (tt-like, J = 12.2, 4.0 Hz,

1H), 3.85 (t, J = 12.0 Hz, 1H), 3.72 (dd, J = 11.9, 3.0 Hz, 1H), 3.56 (d, J =

12.1 Hz, 1H), 3.32 (overlapped with solvent peak, 1H), 3.24 (m, 2H),

2.68 (m, 1H), 2.26 (m, 2H), 2.10 (m, 1H), 1.90 (quintet, J = 7.4 Hz, 2H),

1.54 (m, 2H), 1.48ꢀ1.32 (m, 4H), 0.94 (t, J = 7.2 Hz, 3H).

(E)-3-{2-Butyl-1-[2-(ethylmethylamino)ethyl]-1H-benz-

imidazol-5-yl}-N-hydroxyacrylamide (17a). 17a was prepared

according to the procedure used for 17c but using 9p as starting material

and was obtained as the bis-TFA salt (20 mg, 20% in two steps).

LCꢀMS m/z 345.2 ([M þ H]^þ). ¹H NMR (DMSO-d₆) δ 10.59 (br s,

1H), 7.96 (d, J = 8.4 Hz, 1H), 7.95 (s, 1H), 7.75 (d, J = 9.2 Hz, 1H), 7.63

(d, J = 15.8 Hz, 1H), 6.60 (d, J = 15.8 Hz, 1H), 4.80 (t-like, 2H), 3.54 (m,

2H), 3.28 (br, 2H), 3.14 (t, J = 7.8 Hz, 2H), 2.93 (s, 3H), 1.83 (quintet,

J = 7.7 Hz, 2H), 1.46 (sextet, J = 7.2 Hz, 2H), 1.25 (t, J = 7.2 Hz, 3H),

0.96 (t, J = 7.3 Hz, 3H).

(E)-3-{1-[2-(Ethylmethylamino)ethyl]-2-pentyl-1H-benz-

imidazol-5-yl}-N-hydroxyacrylamide (17b). 17b was prepared

according to the procedure used for 17c but using 9q as starting material

and was obtained as the bis-TFA salt (20 mg, 5.3% in two steps).

LCꢀMS m/z 359.2 ([M þ H]^þ). ¹H NMR (DMSO-d₆) δ 10.46 (br s,

1H), 7.94 (s, 1H), 7.92 (d, J = 9.1 Hz, 1H), 7.74 (d, J = 8.4 Hz, 1H), 7.63

(d, J = 15.8 Hz, 1H), 6.58 (d, J = 15.8 Hz, 1H), 4.77 (br, 2H), 3.52 (m,

2H), 3.27 (br, 2H), 3.11 (t, J = 7.8 Hz, 2H), 2.93 (s, 3H), 1.85 (quintet,

J = 7.3 Hz, 2H), 1.44ꢀ1.34 (m, 4H), 1.25 (t, J = 7.2 Hz, 3H), 0.91 (t, J =

7.1 Hz, 3H).

(E)-3-(2-Butyl-1-(pyrrolidin-3-yl)-1H-benzimidazol-5-yl)-N-

hydroxyacrylamide (20a). 3-[2-Butyl-5-(2-methoxycarbonylvinyl)ben-

zimidazol-1-yl]pyrrolidine-1-carboxylic acid tert-butyl ester (18a) was pre-

pared according to procedure C using 7l as starting material and was obtained

as a pale-yellow solid (3.8 g, 44%). LCꢀMS m/z 428.16 ([M þ H]^þ). 18a

(70mg, 0.16 mmol) wasaddedto1.25MHClinMeOH(4mL) andheated

at 80 ꢀC for 4 h, then evaporated to dryness under reduced pressure

to give compound 19a as HCl salt, which is pure enough for next step

without any purification. LCꢀMS m/z 328.1 ([M þ H]^þ). The crude 19a

(∼0.16 mmol) was converted to hydroxamic acid 20a as the bis-TFA salt

(25 mg, 27% from 19a in two steps) according to procedure C. LCꢀMS

m/z 329.1 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 7.97 (d, J = 8.7 Hz, 1H),

7.94 (s, 1H), 7.75 (d, J = 8.0 Hz, 1H), 7.68 (d, J = 16.0 Hz, 1H), 6.59

(d, J = 16.0 Hz, 1H), 5.65 (m, 1H), 4.00ꢀ3.80 (m, 3H), 3.57 (m,

1H), 3.27ꢀ3.20 (overlapped with solvent peak, 2H), 2.81 (m, 1H),

2.69 (m, 1H), 1.89 (quintet, J = 7.7 Hz, 2H), 1.56 (sextet, J = 7.6 Hz,

2H), 0.95 (t, J = 7.2 Hz, 3H).

(E)-3-(2-Butyl-1-(piperidin-4-yl)-1H-benzimidazol-5-yl)-N-

hydroxyacrylamide (20b). Compound 20b was prepared according

to the procedure described for compound 20a but using 7n as starting

material and was obtained as the bis-TFA salt (12 mg, 8.5% from 7n).

LCꢀMS m/z 343.2 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 8.07 (d, J = 8.8

Hz, 1H), 7.92 (1H, s), 7.76 (1H, d, J = 8.0 Hz), 7.70 (1H, d, J = 15.8 Hz),

6.60 (1H, d, J = 15.8 Hz), 5.05 (1H, m), 3.70 (d, J = 12.5 Hz, 2H),

(E)-3-[2-Butyl-1-(1-methylpyrrolidin-3-yl)-1H-benzimid-

azol-5-yl]-N-hydroxyacrylamide (22a). Tertiary amine 21a was

prepared by reductive amination of 19a with formaldehyde and NaBH-

(OAc)₃according to similar procedures used for 17c and then converted

to hydroxamic acid according to procedure C to obtain 22a as the bis-

TFA salt (25 mg, 22% in two steps). LCꢀMS m/z 343.2 ([M þ H]^þ).

¹H NMR (CD₃OD) δ 8.05 (d, J = 8.8 Hz, 1H), 7.94 (s, 1H), 7.76 (d,

J = 8.6 Hz, 1H), 7.65 (d, J = 15.8 Hz, 1H), 6.58 (d, J = 15.8 Hz, 1H), 5.79

(quintet, J = 9.1 Hz, 1H), 4.20ꢀ3.8 (m, 3H), 3.66 (br s, 1H), 3.26 (t, J =

7.9 Hz, 2H), 3.15 (s, 3H), 2.89 (m, 1H), 2.79 (m, 1H), 1.89 (quintet, J =

7.8 Hz, 2H), 1.56 (sextet, J = 7.6 Hz, 2H), 1.05 (t, J = 7.4 Hz, 3H).

(E)-N-Hydroxy-3-[1-(1-methylpiperidin-3-yl)-2-pentyl-1H-

benzimidazol-5-yl]acrylamide (22b). Tertiary amine 21b (144.5

mg, 53% in two steps from 18f) was prepared according to similar

procedures used for 17c and then converted to hydroxamic acid

according to procedure C to obtain 22b as the bis-TFA salt. LCꢀMS

m/z 371.2 ([M þ H]^þ). ¹H NMR (CD₃OD) δ 8.18 (d, J = 7.9 Hz, 1H),

7.92 (s, 1H), 7.77 (d, J = 8.1 Hz, 1H), 7.61 (d, J = 15.7 Hz, 1H), 6.58 (d,

J = 15.7 Hz, 1H), 5.21 (m, 1H), 4.00ꢀ3.80 (m, 2H), 3.75ꢀ3.60 (m, 1H),

3.43ꢀ3.24 (masked peaks, 3H), 3.03 (s, 3H), 2.64 (m, 1H), 2.36ꢀ2.14

(m, 3H), 1.92 (m, 2H), 1.60ꢀ1.49 (m, 4H), 0.96 (t, J = 7.1 Hz, 3H); ¹³C

NMR (CD₃OD) δ165.6, 157.6, 139.9, 134.6, 134.1, 132.5, 126.3, 120.4,

115.5, 115.2, 54.9, 54.4, 53.3, 44.1, 32.4, 27.5, 27.3, 26.8, 23.2, 23.1, 14.2

(TFA peaks at 163.0, 162.7, 119.4, and 116.5).

HDAC Enzyme Assay. The recombinant HDAC enzymes, includ-

ing HDACs 1ꢀ11, were produced by the Protein Biochemistry Group in

S*BIO Pte Ltd. (see supplements of ref 19 for details). All HDACs are

purified full-length proteins except HDAC5 (191ꢀ1122). The assay

protocols have been reported in our early publications.¹⁸The assay was

4717

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

performed in a 96-well or 384-well format using the BIOMOL fluor-

escent-based HDAC activity assay (BIOMOL International, L.P). Fluor

de Lys substrate (KI-104) was used for HDAC1 IC₅₀and HDACs 1ꢀ11

isoform dissociation constant K_idetermination. The reaction mixture

was composed of assay buffer, containing 25 mM Tris, pH 7.5, 137 mM

NaCl, 2.7 mM KCl, 1 mM MgCl₂, 1 mg/mL BSA, tested compounds, an

appropriate concentration of HDAC1, and 250 μM Fluor de Lys generic

substrate. The fluorescence was detected at the excitation wavelength of

360 nm and the emission wavelength of 460 nm using Tecan Ultra

microplate detection system (Tecan Group Ltd., Switzerland). The

analytical software, Prism 4.0 (GraphPad Software, Inc.), was used to

generate IC₅₀values from the data. IC₅₀values for HDACs 2ꢀ11 were

obtained by using analogous protocols but with appropriate adjustment

of protein and substrate concentrations (20 μM substrate for HDAC6,

100 or 250 μM for HDACs 2ꢀ5, 8ꢀ11). The K_iwas calculated using the

ChengꢀPrusoff equation: K_i= IC₅₀/{1 þ ([substrate]/K_m)}.

sample was determined using the SoftMax Pro software in SpectraMax. The

content of AcH3 in a sample was normalized by dividing the AcH3 concen-

tration by the protein concentration: AcH3 relative to vorinostat (fold) =

AcH3 content (HDAC inhibitor)/AcH3 content (vorinostat).

EC₅₀Determination. The assays were performed according to the

above protocol except that COLO 205 cells were treated with HDAC

inhibitors at different doses (in triplicate, nine doses, 4-fold serial

dilutions starting from 100 μM). The absorbance at 450 nm for each

sample was normalized by the protein concentration, and the dose

response curve was plotted (OD₄₅₀vs compound concentration) using

XL-fit (ID Business Solution, Emeryville, CA) to determine EC₅₀values

of HDAC inhibitors.

Homology Modeling. HDAC1 homology model was built from

HDLP as reported in our earlier report.¹⁸A similar model was also built

from a recent published HDAC2 structure PDB entry 3MAX²⁴with 90%

sequence identity to HDAC1. The X-ray structure was prepared using

the protein preparation wizard in Maestro 9.1 as recommended by

Schrodinger (http://www.schrodinger.com/). The homology model

was build using Prime 2.2 with standard settings. Crystal water was

removed from the template. However, water molecule 457 was added to

the homology model, as this was in a position to hydrogen-bond with N³

of the benzimidazole HDAC inhibitors and His178. The identity of

binding site residues of HDAC1 and HDAC2 is 100%.

In Vitro ADME Studies. Solubility and log D were measured using

our published methods.⁴¹Microsomal stability studies and Caco-2

permeability assays were performed according to published methods.⁴²

Primary screen for inhibition of CYP3A4 and 2D6 was carried by using

BD Biosciences’ high throughput inhibitor screening kits (catalogue no.

459100 and no. 459200). For CYP3A4 inhibition, to a 96-well plate,

each 200 μL well reaction mixture containing 1 pmol (5 nM) of P450

CYP3A4, 8.1 μM NADP^þ, 0.41 mM glucose 6-phosphate, 0.4 U/mL

glucose 6-phosphate dehydrogenase, 0.41 mM magnesium chloride,

50 μM BFC in 25 mM potassium phosphate (pH 7.4), and test com-

pound (eight doses, 3-fold serial dilution starting from 20 to 0.00915

μM) or positive control ketoconazole (eight doses, 3-fold serial dilution

starting from 5 to 0.0023 μM) was incubated at 37 ꢀC for 30 min. The

reaction was stopped by the addition of acetonitrile (75 μL), and the

metabolite HFC was detected (403 nm excitation and 535 nm emission)

and quantified. Data were analyzed using Prism 4.0 (GraphPad Software,

Inc.). IC₅₀values for inhibition of CYP1A, CYP3A4, CYP2D6, CYP2C9,

and CYP2C19 using HLM system with isozyme specific probe sub-

strates and inhibitors were determined by Cyprotex (U.K.).

Cell-Based Proliferation Assay for Determination of IC₅₀.

Human colon cancer cell lines were obtained from the American Type

Culture Collection (ATCC; VA, U.S.) or the European Collection of

Cell Culture (ECACC; Wilshire, U.K.). They were cultivated according

to the supplier’s instructions. After subconfluent growth, cells were seeded

in a 96-well plate at log growth phase. Before treatment, the plates were

incubated at 37 ꢀC, 5% CO₂for 24 h (adherent cells) or 2 h (suspension

cells). Treatment with compounds was carried out in triplicate wells for

96 h. Proliferation of adherent cells was monitored using Cyquant cell

proliferation assay (Invitrogen Pte Ltd., Singapore), while proliferation of

suspension cells was monitored using CellTiter96 Aqueous One solution

cell proliferation assay (Promega Pte Ltd., Singapore). Dose response

curves were plotted to determine the IC₅₀values using XL-fit (ID Business

Solution Ltd., U.S.). IC₅₀is the cconcentration needed for inhibition of 50%

cell proliferation of tumor cells.

Histone H3 Acetylation Assay: ELISA Approach and EC₅₀

Determination. COLO 205 cells was cultivated in a 96-well plate at

1 ꢁ 10⁵cells/well for 24 h. COLO 205 cells were subsequently treated

with HDAC inhibitors (10 μM), reference (vorinostat, 10 μM), and

negative control (1% DMSO) in triplicate. After treatment for 24 h, cells

were lysed according to the instructions from Sigma mammalian cell

lysis kit (Sigma no. MCL-1) and the protein concentration was

determined. The ELISA plate (Immulon 2HB plate, Bio Laboratories

Pte Ltd., Singapore) was coated with 4 μg/mL mouse monoclonal

antibody against H3 (Upsate no. 05-499, Upstate Pte Ltd., Singapore) in

PBS, pH 7.4, incubated at 37 ꢀC for 1 h and then at 4 ꢀC overnight.

Subsequently the plate was washed with washing buffer [PBS containing

0.05% (v/v) Tween-20] and blocked with SuperBlock solution (Pierce

no. 37515, Pierce Pte Ltd., Singapore) at 37 ꢀC for 1 h. The SuperBlock

solution was removed, and the plate was washed with washing buffer.

The AcH3 peptide standards (24 doses, 2-fold serial dilutions starting

from 200 μg/mL, Upsate no. 12-360, Upstate Pte Ltd., Singapore) and

the protein lysates from COLO 205 cells treated with HDAC inhibitors

were applied to the plate (final volume, 50 μL/well) in triplicate and

incubated for 1 h at 37 ꢀC. After removal of the samples, the plate was

washed with washing buffer. The plate was incubated with secondary

antibody [100 μL, 0.5 μg/mL rabbit polyclonal antibody against AcH3

(Lys9/14) (Upsate no. 06-599, Upstate Pte Ltd., Singapore)] at 37 ꢀC

for 1 h. Subsequently the plate was washed with washing buffer and

incubated with a detection antibody [100 μL, 1:5000 donkey anti-rabbit

IgG antibody coupled to HRPO (Pierce no. 31458, Pierce Pte Ltd.,

Singapore)] that was applied at 37 ꢀC for 30 min. The plate was washed

with washing buffer, then incubated with substrate [100 μL, Turbo TMB

substrate solution (Pierce no. 34022, Pierce Pte Ltd., Singapore)] at room

temperature for 30 min. The reaction was stopped using 1 M H₂SO₄. The

absorbance was measured at 450 nm on a SpectraMax absorbance microplate

reader (Molecular Devices Corporation, Sunnyvale, CA). The standard

curve was drawn, and the concentration of AcH3 [(Lys9/14), μg/mL] in a

Pharmacokinetic Analysis. The analysis followed protocols

similar to those published for 2⁴²and 3.¹⁹

Human Tumor Xenograft Studies. These studies followed

protocols similar to those previously published.^18,19,41

’ ASSOCIATED CONTENT

S

Supporting Information. Figures S1ꢀ3 showing addi-

b

tional correlations with potency; Tables S1ꢀ10 listing target

compounds, IC₅₀, clogP, acetylation data, microsomal data,

pharmacokinetic parameters, and tissue distribution; Table S11

listing elemental analysis results and HPLC purity data for key

and reference compounds; synthesis and analytical data of

additional compounds for establishing SARs (10acꢀax, 15fꢀv,

and 17dꢀx). This material is available free of charge via the

Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION

Corresponding Author

*Phone:þ65-68275019. Fax: þ65-68275005. E-mail: haishan_wang@

sbio.com.

4718

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

’ ACKNOWLEDGMENT

(11) Glozak, M. A.; Sengupta, N.; Zhang, X.; Seto, E. Acetylation and

deacetylation of non-histone proteins. Gene 2005, 363, 15–23.

(12) Butler, L. M.; Agus, D. B.; Scher, H. I.; Higgins, B.; Rose, A.;

Cordon-Cardo, C.; Thaler, H. T.; Rifkind, R. A.; Marks, P. A.; Richon,

V. M. Suberoylanilide hydroxamic acid, an inhibitor of histone deace-

tylase, suppresses the growth of prostate cancer cells in vitro and in vivo.

Cancer Res. 2000, 60, 5165–5170.

(13) FDA approval and new drug application (NDA) documents for

Zolinza (SAHA, vorinostat): http://www.accessdata.fda.gov/drugsatf-

da_docs/nda/2006/021991s000_ZolinzaTOC.cfm.

The authors thank Evelyn Goh and Lee Sun New for

performing ADME analytical experiments, Tony Ng for support

of in vivo experiments, Wai Chung Ong for carrying out HDAC

assays, Miah Kiat Goh and Chee Pang Ng for preparation of

HDAC proteins, Chang Kai Soh for making the freebase of 3,

Noah Tu for HPLC puriﬁcation system support, Bee Kheng Ng

for preparation of compound assay solutions, Drs. Michael

Entzeroth and Binhui Ni for their early inputs and administrative

work in biology, and Dr. Veronica Novotny-Diermayr for critical

reading of this manuscript.

(14) FK228: http://www.fda.gov/NewsEvents/Newsroom/Press-

Announcements/2009/ucm189629.htm.

(15) Wang, H.; Dymock, B. W. New patented histone deacetylase

inhibitors. Expert Opin. Ther. Pat. 2009, 19, 1727–1757.

(16) Paris, M.; Porcelloni, M.; Binaschi, M.; Fattori, D. Histone

deacetylase inhibitors: from bench to clinic. J. Med. Chem. 2008,

51, 1505–1529. Correction in J. Med. Chem. 2008, 51, 3330.

(17) S*BIO patent applications on HDAC inhibitors.(a) Chen, D.;

Deng, W.; Sangthongpitag, K.; Song, H.; Sun, E. T.; Yu, N.; Zou, Y.

Benzimidazole Derivates: Preparation and Pharmaceutical Applications.

PCT Int. Appl. WO2005028447, March 31, 2005. (b) Lim, Z.-Y.; Wang, H.;

Zhou, Y. Acylurea and Sulfonylurea Connected Hydroxamates. PCT Int.

Appl. WO2005040101, May 6, 2005. (c) Stunkel, W.; Wang, H.; Yin, Z.

Biaryl Linked Hydroxamates: Preparation and Pharmaceutical Applications.

PCT Int. Appl. WO2005040161, May 6, 2005. (d) Deng, W.; Lye, P. L.;

Lim, Y. H. Benzothiophene Derivatives: Preparation and Pharmaceutical

Applications. PCT Int. Appl. WO2006101454, September 28, 2006. (e)

Lee, K. C. L.; Sun, E. T. Imidazo[1,2-a]pyridine Derivatives: Preparation

and Pharmaceutical Applications. PCT Int. Appl. WO2006101455,

September 28, 2006. (f) Deng, W.; Chen, D.; Zhou, Y. Bicyclic Heterocycles

Hydroxamate Compounds Useful as Histone Deacetylase (HDAC) In-

hibitors. PCT Int. Appl. WO2006101456, September 28, 2006. (g) Chen,

D.; Deng, W.; Lee, K. C. L.; Lye, P. L.; Sun, E. T.; Wang, H.; Yu, N.

Heterocyclic Compounds. PCT Int. Appl. WO2007030080, March 15,

2007. (h) Lee, K. C. L.; Sun, E. T.; Wang, H. Imidazo[1,2-a]pyridine

Hydroxamate Compounds That Are Inhibitors of Histone Deacetylase.

PCT Int. Appl. WO2008036046, March 27, 2008.

(18) Wang, H.; Yu, N.; Song, H.; Chen, D.; Zou, Y.; Deng, W.; Lye,

P. L.; Chang, J.; Ng, M.; Blanchard, S.; Sun, E. T.; Sangthongpitag, K.;

Wang, X.; Goh, K. C.; Wu, X.; Khng, H. H.; Fang, L.; Goh, S. K.; Ong,

W. C.; Bonday, Z.; Stunkel, W.; Poulsen, A.; Entzeroth, M. N-Hydroxy-

1,2-disubstituted-1H-benzimidazol-5-yl acrylamides as novel histone

deacetylase inhibitors: design, synthesis, SAR studies, and in vivo

antitumor activity. Bioorg. Med. Chem. Lett. 2009, 19, 1403–1408.

(19) Novotny-Diermayr, V.; Sangthongpitag, K.; Hu, C. Y.; Wu,

X.; Sausgruber, N.; Yeo, P.; Greicius, G.; Pettersson, S.; Liang, A. L.;

Loh, Y. K.; Bonday, Z.; Goh, K. C.; Hentze, H.; Hart, S.; Wang, H.;

Ethirajulu, K.; Wood, J. M. SB939, a novel potent and orally active

histone deacetylase inhibitor with high tumor exposure and eﬃcacy

in mouse models of colorectal cancer. Mol. Cancer Ther. 2010, 93,

642–652.

’ ABBREVIATIONS USED

A2780, human ovary carcinoma cell line; AcH3, acetyl histone

H3; ADME, absorption, distribution, metabolism, and excretion;

AUC, area under curve; BEI, binding eﬃciency index; BFC, 7-benzyl-

oxytriﬂuoromethylcoumarin; Boc, tert-butoxycarbonyl; Cl, clearance;

COLO 205, human colorectal adenocarcinoma; CR, complete tumor

regression, no measurable tumor (less than 3 mm ꢁ 3 mm) for

three consecutive measurements; CYP, cytochrome P450 super

family; DCM, dichloromethane; DIEA, N,N-diisopropylethylamine;

ELISA, enzyme-linked immunosorbent assay; EDCI, N-(3-dimethyl-

aminopropyl)-N⁰-ethylcarbodiimide hydrochloride; EtOAc, ethyl

acetate; F, percent oral bioavailability; HCT-116, human colorectal

adenocarcinoma cell line; HDAC, histone deacetylase; HFC, 7-hy-

droxytriﬂuoromethylcoumarin; HLM, human liver microsomes;

HOBt, 1-hydroxybenzotriazole; MCA, 4-methylcoumarin-7-amide;

MeOH, methanol; MLM, mouse liver microsomes; MV4-11, human

acute monocytic leukemia cell line; PC-3, human prostate cancer

cell line; PR, partial tumor regression, reduction to <50% of initial

tumor volume for three consecutive measurements; PyBOP,

benzotriazole-1-yloxytripyrrolidinophosphonium hexaﬂuoropho-

sphate; Ramos, human Burkitt’s lymphoma cell line; qd, quaque

die, once a day; SAHA, suberoylanilide hydroxamic acid; SAR,

structureꢀactivity relationship; TBME, tert-butyl methyl ether;

TFA, triﬂuoroacetic acid; TGI, tumor growth inhibition; TRD,

treatment related death; ZBG, zinc binding group

’ REFERENCES

(1) Amato, I. Pulling genes’ strings. Chem. Eng. News 2006, 84 (29),

13–20.

(2) Grunstein, M. Histone acetylation in chromatin structure and

transcription. Nature 1997, 389, 349–352.

(3) Struhl, K. Histone acetylation and transcriptional regulatory

mechanisms. Genes Dev. 1998, 12, 599–606.

(20) Razak, A. R.; Hotte, S. J.; Siu, L. L.; Chen, E. X; Hirte, H. W.;

Powers, J.; Walsh, W.; Stayner, L. A.; Laughlin, A.; Novotny-Diermayr, V.;

Zhu, J.; Eisenhauer, E. A. Phase I clinical, pharmacokinetic and pharmaco-

dynamic study of SB939, an oral histone deacetylase (HDAC) inhibitor, in

patients with advanced solid tumours. Br. J. Cancer 2011, 104, 756–762.

(21) Yong, W. P.; Goh, B. C.; Soo, R. A.; Toh, H. C.; Ethirajulu, K.;

Wood, J.; Novotny-Diermayr, V.; Lee, S. C.; Yeo, W. L.; Chan, D.; Lim,

D.; Seah, E.; Lim, R.; Zhu, J. Phase I and pharmacodynamic study of an

orally administered novel inhibitor of histone deacetylases, SB939, in

patients with refractory solid malignancies. Ann. Oncol. [Online early

access]. DOI: 10.1093/annonc/mdq784. Published Online: Mar 8,

2011.

(4) Wade, P. A. Transcriptional control at regulatory checkpoints by

histone deacetylases: molecular connections between cancer and chro-

matin. Hum. Mol. Genet. 2001, 10, 693–698.

(5) de Ruijter, A. J. M.; van Gennip, A. H.; Caron, H. N. P.; Kemp, S.;

van Kuilenburg, A. B. P. Histone deacetylases (HDACs): characteriza-

tion of the classical HDAC family. Biochem. J. 2003, 370, 737–749.

(6) Michan, S.; Sinclair, D. Sirtuins in mammals: insights into their

biological function. Biochem. J. 2007, 404, 1–13.

(7) Ellis, L.; Atadja, P. W.; Johnstone, R. W. Epigenetics in cancer:

targeting chromatin modiﬁcations. Mol. Cancer Ther. 2009, 8, 1409–1420.

(8) Curtin, M.; Glaser, K. Histone deacetylase inhibitors: the Abbott

experience. Curr. Med. Chem. 2003, 10, 2373–2392.

(22) Wu, Z.; Rea, P.; Wickham, G. “One-pot” nitro reductionꢀ

cyclisation solid phase route to benzimidazoles. Tetrahedron Lett. 2000,

41, 9871–9874.

(23) Abad-Zapatero, C. Ligand eﬃciency indices for eﬀective drug

discovery. Expert Opin. Drug Discovery 2007, 2, 469–488.

(9) Glaser, K. B.; Li, J.; Staver, M. J.; Wei, R.-Q.; Albert, D. H.;

Davidsen, S. K. Role of class I and class II histone deacetylases in carcinoma

cells using siRNA. Biochem. Biophys. Res. Commun. 2003, 310, 529–536.

(10) Witt, O.; Deubzer, H. E.; Milde, T.; Oehme, I. HDAC family:

What are the cancer relevant targets? Cancer Lett. 2009, 277, 8–21.

4719

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Journal of Medicinal Chemistry

ARTICLE

(24) For homology modeling, two human HDAC1 models were

built from HDLP X-ray structure 1C3R (see ref 18) and HDAC2

structure 3MAX (see Bressi et al.). Bressi, J. C.; Jennings, A. J.; Skene, R.;

Wu, Y.; Melkus, R.; De Jong, R.; O’Connell, S.; Grimshaw, C. E.; Navre,

M.; Gangloﬀ, A. R. Exploration of the HDAC2 foot pocket: synthesis

and SAR of substituted N-(2-aminophenyl)benzamides. Bioorg. Med.

Chem. Lett. 2010, 20, 3142–3145.

(25) Khan, N.; Jeﬀers, M.; Kumar, S.; Hackett, C.; Boldog, F.;

Khramtsov, N.; Qian, X.; Mills, E.; Berghs, S. C.; Carey, N.; Finn, P. W.;

Collins, L. S.; Tumber, A.; Ritchie, J. W.; Jensen, P. B.; Lichenstein, H. S.;

Sehested, M. Determination of the class and isoform selectivity of small-

molecule histone deacetylase inhibitors. Biochem. J. 2008, 409, 581–589.

(26) Arts, J.; King, P.; Mari€en, A.; Floren, W.; Beli€en, A.; Janssen, L.;

Pilatte, I.; Roux, B.; Decrane, L.; Gilissen, R.; Hickson, I.; Vreys, V.; Cox,

E.; Bol, K.; Talloen, W.; Goris, I.; Andries, L.; Du Jardin, M.; Janicot, M.;

Page, M.; van Emelen, K.; Angibaud, P. JNJ-26481585, a novel “second

generation” oral histone deacetylase inhibitor, shows broad-spectrum

preclinical antitumoral activity. Clin. Cancer Res. 2009, 15, 6841–6851.

(27) Hanessian, S.; Auzzas, L.; Larsson, A.; Zhang, J.; Giannini, G.;

Gallo, G.; Ciacci, A.; Cabri, W. Vorinostat-like molecules as structural,

stereochemical, and pharmacological tools. ACS Med. Chem. Lett. 2010,

1, 70–74.

PK/PD Attributes of SB939, Potent Orally Active HDAC Inhibitor.

Presented at the 18th EORTC-NCI-AACR Symposium on Molecular

Targets and Cancer Therapeutics, Prague, Czech Republic, November

7ꢀ10, 2006.

(38) Plumb, J. A.; Finn, P. W.; Williams, R. J.; Bandara, M. J.;

Romero, M. R.; Watkins, C. J.; La Tangue, N. B.; Brown, R. Pharma-

codynamic response and inhibition of growth of human tumor xeno-

grafts by the novel histone deacetylase inhibitor PXD101. Mol. Cancer

Ther. 2003, 2, 721–728.

(39) Remiszewski, S. W.; Sambucetti, L. C.; Bair, K. W.; Bontempo,

J.; Cesarz, D.; Chandramouli, N.; Chen, R.; Cheung, M.; Cornell-Kennon,

S.; Dean, K.; Diamantidis, G.; France, D.; Green, M. A.; Howell, K. L.; Kashi,

R.; Kwon, P.; Lassota, P.; Martin, M. S.; Mou, Y.; Perez, L. B.; Sharma, S.;

Smith, T.; Sorensen, E.; Taplin, F.; Trogani, N.; Versace, R.; Walker, H.;

Weltchek-Engler, S.; Wood, A.; Wu, A.; Atadja, P. N-Hydroxy-3-phenyl-2-

propenamides as novel inhibitors of human histone deacetylase with in vivo

antitumor activity: discovery of (2E)-N-hydroxy-3-[4-[[(2-hydroxyethyl)-

[2-(1H-indol-3-yl)ethyl]amino]methyl]phenyl]-2-propenamide (NVP-

LAQ824). J. Med. Chem. 2003, 46, 4609–4624.

(40) Atadja, P.; Shao, W.; Wang, Y.; Growney, J.; Feng, Y.; Yao, Y.-

M.; Wallace, A.; Crisanti, C.; Fawell1, S.; Albelda, S. Eﬃcacy of

Panobinostat (LBH589) in Lung Cancer: Potent Anticancer Activity

in Both in Vitro and in Vivo Tumor Models. Presented at the 20th

EORTC-NCI-AACR Symposium on Molecular Targets and Cancer

Therapeutics, Geneva, Switzerland, Oct 21ꢀ24, 2008; Abstract No. 151.

(41) Wang, H.; Lim, Z.-Y.; Zhou, Y.; Ng, M.; Lu, T.; Lee, K.;

Sangthongpitag, K.; Goh, K. C.; Wang, X.; Wu, X.; Khng, H. H.; Goh,

S. K.; Ong, W. C.; Bonday, Z.; Sun, E. T. Acylurea connected straight

chain hydroxamates as novel histone deacetylase inhibitors: synthesis,

SAR, and in vivo antitumor activity. Bioorg. Med. Chem. Lett. 2010,

20, 3314–3321.

(28) Lahm, A.; Paolini, C.; Pallaoro, M.; Nardi, M. C.; Jones, P.;

ꢀ

Neddermann, P.; Sambucini, S.; Bottomley, M. J.; Lo Surdo, P.; Carfi, A.;

Koch, U.; De Francesco, R.; Steink€uhler, C.; Gallinari, P. Unraveling the

hidden catalytic activity of vertebrate class IIa histone deacetylases. Proc.

Natl. Acad. Sci. U.S.A. 2007, 104, 17335–17340.

(29) Jones, P.; Altamura, S.; De Francesco, R.; Paz, O. G.; Kinzel, O.;

Mesiti, G.; Monteagudo, E.; Pescatore, G.; Rowley, M.; Verdirame, M.;

Steink€uhler, C. A novel series of potent and selective ketone histone

deacetylase inhibitors with antitumor activity in vivo. J. Med. Chem.

2008, 51, 2350–2353.

(30) Tessier, P.; Smil, D. V.; Wahhab, A.; Leit, S.; Rahil, J.; Li, Z.;

Dꢀeziel, R.; Besterman, J. M. Diphenylmethylene hydroxamic acids as

selective class IIa histone deacetylase inhibitors. Bioorg. Med. Chem. Lett.

2009, 19, 5684–5688.

(42) Venkatesh, P. R.; Goh, E.; Zeng, P.; New, L. S.; Xin, L.; Pasha,

M. K.; Sangthongpitag, K.; Yeo, P.; Kantharaj, E. In vitro phase I

cytochrome P450 metabolism, permeability and pharmacokinetics of

SB639, a novel histone deacetylase inhibitor in preclinical species. Biol.

Pharm. Bull. 2007, 30, 1021–1024.

(31) Bradner, J. E.; West, N.; Grachan, M. L.; Greenberg, E. F.;

Haggarty, S. J.; Warnow, T.; Mazitschek, R. Chemical phylogenetics of

histone deacetylases. Nat. Chem. Biol. 2010, 6, 238–243.

(32) Ontoria, J. M.; Altamura, S.; Di Marco, A.; Ferrigno, F.; Laufer,

R.; Muraglia, E.; Palumbi, M. C.; Rowley, M.; Scarpelli, R.; Schultz-

Fademrecht, C.; Seraﬁni, S.; Steink€uhler, C; Jones, P. Identiﬁcation of

novel, selective, and stable inhibitors of class II histone deacetylases.

Validation studies of the inhibition of the enzymatic activity of HDAC4

by small molecules as a novel approach for cancer therapy. J. Med. Chem.

2009, 52, 6782–6789.

(33) Wilson, A. J.; Byun, D. S.; Popova, N.; Murray, L. B.; L’Italien,

K.; Sowa, Y.; Arango, D.; Velcich, A.; Augenlicht, L. H.; Mariadason,

J. M. Histone deacetylase 3 (HDAC3) and other class I HDACs regulate

colon cell maturation and p21 expression and are deregulated in human

colon cancer. J. Biol. Chem. 2006, 281, 13548–13558.

(34) Gupta, P.; Ho, P.-C.; Ha, S. G.; Lin, Y.-W.; Wei, L.-N. HDAC3

as a molecular chaperone for shuttling phosphorylated TR2 to PML: a

novel deacetylase activity-independent function of HDAC3. PLoS One

2009, 4, No. e4363. DOI: 10.1371/journal.pone.0004363.

(35) Jayaraman, R.; Wang, H.; Sangthongpitag, K,; Yeo, P.; Hu, C.;

Wu, X.; Liu, X.; Reddy, V. P.; Goh, E.; New, L. S.; Pasha, M. K.;

Ethirajulu, K. Preclinical metabolism and disposition of SB939, an orally

active histone deacetylase (HDAC) inhibitor, and prediction of human

pharmacokinetics. Submitted to Drug Metab. Dispos.

(36) Yeo, P.; Liu, X.; Goh, E.; New, L. S.; Zeng, P.; Wu, X.;

Venkatesh, P.; Kantharaj, E. Development and validation of high-

performance liquid chromatographyꢀtandem mass spectrometry assay

for 6-(3-benzoyl-ureido)-hexanoic acid hydroxyamide, a novel HDAC

inhibitor, in mouse plasma for pharmacokinetic studies. Biomed. Chro-

matogr. 2007, 21, 184–189.

(37) Sangthongpitag, K.; Wang, H.; Yeo, P.; Liu, X.; Goh, E.; New,

L. S.; Zeng, P.; Venkatesh, P., Wu, X.; Hu, C.; Ethirajulu, K. ADME and

4720

dx.doi.org/10.1021/jm2003552 |J. Med. Chem. 2011, 54, 4694–4720

Article Doi

Discovery of (2 E)-3-{2-butyl-1-[2-(diethylamino)ethyl]-1H-benzimidazol-5- yl}-N-hydroxyacrylamide (SB939), an orally active histone deacetylase inhibitor with a superior preclinical profile

DOI: 10.1021/jm2003552

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm2003552

Full text of DOI:10.1021/jm2003552

Products guided by the article

R&D Labs maybe for 1333893-53-0

Relevant to this article

Hot Product