The discovery of 6-amino nicotinamides as potent and selective histone deacetylase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.08.023

This communication highlights the development of a nicotinamide series of histone deacetylase inhibitors within the benzamide structural class. Extensive exploration around the nicotinamide core led to the discovery of a class I selective HDAC inhibitor that possesses excellent intrinsic and cell-based potency, acceptable ancillary pharmacology, favorable pharmacokinetics, sustained pharmacodynamics in vitro, and achieves in vivo efficacy in an HCT116 xenograft model.

Full text of DOI:10.1016/j.bmcl.2007.08.023

Bioorganic & Medicinal Chemistry Letters 17 (2007) 5300–5309
The discovery of 6-amino nicotinamides as potent and selective
histone deacetylase inhibitors
Christopher L. Hamblett,a,* Joey L. Methot,a Dawn M. Mampreian,a David L. Sloman,a
Matthew G. Stanton,a Astrid M. Kral,b Judith C. Fleming,b Jonathan C. Cruz,c
Melissa Chenard,c Nicole Ozerova,b Anna M. Hitz,b Hongmei Wang,b
Sujal V. Deshmukh,d Naim Nazef,d Andreas Harsch,d Bethany Hughes,e
William K. Dahlberg,e Alex A. Szewczak,e Richard E. Middleton,e
Ralph T. Mosley,f J. Paul Secristb and Thomas A. Millera
aDepartment of Drug Design and Optimization—Medicinal Chemistry, Merck Research Laboratories,
33 Avenue Louis Pasteur, Boston, MA 02115, USA
bDepartment of Cancer Biology and Therapeutics, Merck Research Laboratories, 33 Avenue Louis Pasteur, Boston, MA 02115, USA
cDepartment of Pharmacology, Merck Research Laboratories, 33 Avenue Louis Pasteur, Boston, MA 02115, USA
dDepartment of Drug Design and Optimization—Drug Metabolism and Pharmacokinetics, Merck Research Laboratories,
33 Avenue Louis Pasteur, Boston, MA 02115, USA
eDepartment of Drug Design and Optimization—Automated Lead Optimization, Merck Research Laboratories,
33 Avenue Louis Pasteur, Boston, MA 02115, USA
fDepartment of Medicinal Chemistry, Merck Research Laboratories, 126 Lincoln Avenue, PO Box 2000, Rahway, NJ 07065, USA
Received 13 July 2007; revised 8 August 2007; accepted 13 August 2007
Available online 16 August 2007
Abstract—This communication highlights the development of a nicotinamide series of histone deacetylase inhibitors within the
benzamide structural class. Extensive exploration around the nicotinamide core led to the discovery of a class I selective HDAC
inhibitor that possesses excellent intrinsic and cell-based potency, acceptable ancillary pharmacology, favorable pharmacokinetics,
sustained pharmacodynamics in vitro, and achieves in vivo efficacy in an HCT116 xenograft model.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Epigenetic defects alter levels of gene expression without
structurally modifying DNA sequence and have been
linked to the progression of a number of diseases, such
as cancer.1 In contrast to DNA mutations, epigenetic
defects that lead to disease are potentially reversible by
pharmacologically targeting the enzymes responsible
for the epigenetic modifications.2
are currently in development as therapeutic agents and
have been shown to regulate a variety of cellular re-
sponses including proliferation, differentiation, and
apoptosis.4 In general, inhibiting HDAC activity results
in the accumulation of acetylated histones and subse-
quent activation of gene transcription.5 Histone hyper-
acetylation induced by HDAC inhibitors has been
shown to correlate with the anti-tumor effects of these
inhibitors.6 These observations suggest that HDAC
inhibition provides an opportunity to reverse epigenetic
defects that lead to disease and represents a promising
new strategy for the treatment of cancer.
Histone acetylation is an epigenetic modification that is
controlled by histone acetyltransferases (HATs) and his-
tone deacetylases (HDACs).3 Several HDAC inhibitors
Class I and II histone deacetylases catalyze the removal
of the acetyl group from the amino-terminal lysine resi-
dues of histones within nucleosomes.7 While the biolog-
ical functions of the many HDAC subtypes are still
Keywords: Epigenetics; Histone acetylation; Histone deacetylases;
Nicotinamides; Benzamides.
*
Corresponding author. Tel.: +1 617 992 2053; fax: +1 617 992
2043; e-mail: christopher_hamblett@merck.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.08.023

C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
5301
being defined, there is compelling evidence that class I
HDACs regulate cell proliferation and therefore are via-
ble targets for cancer therapeutics.8
It was quickly determined that the 6-position of the pyr-
idine ring could be easily functionalized and that deriv-
atization with cyclic amines, such as N-substituted
piperazines, led to a significant enhancement in HDAC
inhibitory activity over analogs derived from corre-
sponding acyclic amino derivatives. For example, acy-
clic dimethyl amine derivative 2 exhibited an HDAC
inhibitory activity of 963 nM compared to piperazine
3 which was 168 nM (Table 1). In general, amino nico-
tinamide substitution with six-membered cyclic amines
gave enhanced HDAC enzymatic activity compared to
larger or smaller heterocyclic ring systems.
Vorinostat (suberoylanilide hydroxamic acid, SAHA,
Zolinza.TM), a hydroxamate-based HDAC inhibitor,
inhibits a subset of both class I and II HDAC enzymes
and has recently been approved for marketing in the
United States for the treatment of cutaneous manifesta-
tions of cutaneous T-cell lymphoma (Fig. 1).9 More spe-
cifically, vorinostat inhibits HDACs 1, 2, 3, and 6 at the
low nanomolar level (HDAC1 IC50 < 0.1 lM) while
inhibiting HDAC8 and HDAC11 at the micromolar
level.
All nicotinamide analogs were prepared in excellent
overall yields by either a linear 3- or 4-step sequence
using one of two general methods: nicotinamides pre-
pared by method A commenced with the acid chloride
displacement of commercially available 6-chloronicoti-
noyl chloride with t-butyl (2-aminophenyl)carbamate
414 to give the Boc-protected chloronicotinamide 5 in
84% yield. The sequence was completed by pyridyl chlo-
ride displacement in the presence of excess (3.0 equiv)
functionalized (i) or unfunctionalized (ii) cyclic amine
at 85 ꢁC in DMSO. The piperazine was then further
functionalized (ii) by acetylation, sulfonylation, or car-
bamoylation to give intermediate 6, followed by TFA
removal of the Boc group (i and ii) to afford the desired
nicotinamide 7 (Scheme 1).
A number of benzamide HDAC inhibitors, such as MS-
275 (HDAC1 IC50 > 1.0 lM),10 CI-994 (HDAC1
IC50 > 100 lM),11
and
MGCD-0103
(HDAC1
IC50 P 0.1 lM),12 generally characterized as less potent
HDAC inhibitors with greater class I selectivity com-
pared to hydroxamate-based inhibitors, are under clini-
cal investigation to further evaluate if they offer
advantages to vorinostat with regard to efficacy and tol-
erability (Fig. 2).
Our medicinal chemistry efforts were initiated to dis-
cover an HDAC inhibitor with improved class I selectiv-
ity, greater tolerability, and improved efficacy compared
to current inhibitors in development. Initial SAR fo-
cused on structural diversification around the pyridine
core of nicotinamide lead 1, which exhibits an HDAC
inhibitory activity of 12.4 lM (Fig. 3).13
Nicotinamides prepared by method B began with the
pyridyl chloride displacement of commercially available
methyl 6-chloronicotinate with excess cyclic amine
(3.0 equiv) to give intermediates 8, followed by hydroly-
sis of the methyl ester with LiOH to give intermediates 9
and EDCI coupling in the presence of excess phenylen-
ediamine to give the desired nicotinamide 7 by this alter-
native strategy (Scheme 2).
Zollinza TM (vorinostat)
O
H
N
N
H
OH
O
Having determined that 6-piperazinyl nicotinamides
gave optimal HDAC inhibitory activity, we focused
our efforts initially on modifications to the piperazine
ring system by preparing a diverse set of analogs 7a–
7m (Table 2). For 6-piperazinyl nicotinamides lacking
N-substitution where R3 = H, no general trends emerged
for enzymatic potency despite the structural and stereo-
chemical modifications made on the piperazine ring
Figure 1. Vorinostat, a first in class hydroxamate-based HDAC
inhibitor approved for the treatment of cutaneous T-cell lymphoma.
O
O
N
O
N
H
N
H
H
N
NH2
O
NH2
N
H
O
CI-994
MS-275
O
N
Table 1. Acyclic versus cyclic comparison at the 6-position
N
H
H
N
N
NH2
O
N
N
MGCD-0103
H
NH2
R
N
Figure 2. Representative benzamides under clinical evaluation.
Compound
R
HDAC1 IC50 (nM)a
N
Me
O
N
Ph
O
O
N
N
2
963
Me
O
O
H
N
NH2
N
Ph
3
168
1
HDAC1 IC50 = 12.4 μM
a Values are reported as an average for n = 2 or greater.
Figure 3. N-(2-Aminophenyl)nicotinamide (1) as a benzamide lead.

5302
C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
O
functionalized after
O
H2N
chloride displacement
4
NHBoc
N
H
Cl
or
HN
O
R6
CH2Cl2, RT
Cl
N
5
ii.
Cl
N
(a)
R5
HN
O
(84%)
NH
R1
(3.0 eq.)
i.
R6
R3 R2
DMSO, 85 o
R5
NH
R1
C
(3.0 eq.)
N
R4
(b) installation of
R4 via acetylation,
sulphonylation,or
carbamoylation
R3 R2
DMSO, 85 o
C
functionalized before
chloride displacement
O
R6
N
H
O
R5
R4
HN
O
R6
N
H
N
N
TFA/CH2Cl2, RT
R5
NH2
N
O
6
R1
N
N
R3 R2
N
R3 R2
7
R4
R1
Scheme 1. Method A: general synthesis of N-(2-aminophenyl)-6-piperazinyl nicotinamides 7.
R5
R4
O
NH
O
R6
OMe
N
R4
R1
R5
R4
8
OMe
R3 R2
DMSO, 85 o
N
N
C
N
Cl
N
R1
R3 R2
LiOH/THF/H2O
O
NH2
O
R6
OH
NH2
R5
R6
N
H
(2.0 eq.)
N
N
R5
R4
NH2
9
N
R3 R2
N
N
R4
R1
EDCI, HOBt, DMF
7
N
R1
R3 R2
Scheme 2. Method B: alternative synthesis of N-(2-aminophenyl)-6-piperazinyl nicotinamides 7.
system (7a–7f). The enzymatic activities for analogs 7a–
7f ranged from 301 nM (7c) to 821 nM (7f). The signif-
icance of the size and position of the substituents on the
piperazine ring became clear when R3 was further elab-
orated to a number of carbamate derivatives. Analogs
with smaller R1 and R2 substituents, such as (S)-Me
(7g, 7i) or (R)-Me (7j), on the external 2-position of
the piperazine were generally more potent than inhibi-
tors with larger substituents, such as (S)-i-Pr (7m) and
(S)-Bn (7h) when R3 = Cbz or Boc. Analogs that con-
tained disubstituted piperazines, such as dimethyl deriv-
atives 7k and 7l with substituents either in the internal or
external positions of the piperazine ring, were typically
less potent than derivatives with mono-methyl substitu-
tion at the same positions when R3 = Cbz. The two most
potent inhibitors to emerge from the SAR were nicotina-
mides 7i with an HDAC1 IC50 of 73 nM and 7j with an
HDAC1 IC50 of 40 nM.
The 6-amino nicotinamides were further evaluated in
our cell-based HCT116 colon carcinoma growth inhibi-
tion15 assay, which is a direct measure of in vitro anti-
proliferative effects, as well as an indirect measure of cell
permeability. Several trends emerged from the cell-based
results (Table 3). Inhibitors lacking N-substitution not
only had poor intrinsic potency, but generally had poor
cell-based potency as well as exemplified by 7a with a
GI50 of 1690 nM. Inhibitors with larger R1/R2 pipera-
zine substituents, such as 7m with R1 = (S)-i-Pr, had a
growth inhibition of 3484 nM, and inhibitors with

C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
Table 2. HDAC1 activity for representative 6-piperazinyl nicotinamides
5303
O
R5
N
H
R4
R3
NH2
N
N
7
N
R2 R1
Compound
Method
R1
R2
R3
R4
R5
HDACl IC50 (nM)a
7a
7b
7c
7d
7e
7f
Ai
Ai
Ai
Ai
Ai
Ai
B
(S)-i-Me
(S)-i-Pr
(S)-Bn
(S)-Me
Me
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
435
304
301
582
529
821
242
352
73
H
H
H
H
H
H
(R)-Me
H
H
Me
Me
H
H
H
H
Me
H
H
7g
7h
7i
Boc
Boc
Cbz
Cbz
Cbz
Cbz
Cbz
H
B
Bn
H
H
Aii
Aii
Aii
Aii
Aii
(S)-Me
H
Me
H
H
7j
(R)-Me
Me
H
H
40
7k
7l
H
(R)-Me
169
255
326
(S)-Me
(S)-i-Pr
7m
H
H
a Values are reported as an average for n = 2 or greater.
Table 3. Enzymatic versus cell-based potency for representative nicotinamides
O
N
H
7
NH2
R
N
Compound
Method
Ai
R
HDAC1 IC50 (nM)a
HCT116-72h GI50 (nM)a
Shift in potency
3.9
N
HN
7a
435
1690
Me
N
O
N
7g
7i
B
242
73
1100
81
4.5
1.1
O
Me
N
Aii
Ph
O
N
O
Me
N
7j
Aii
Aii
40
495
871
12.4
5.1
Ph
Ph
O
O
N
N
O
O
Me
N
7k
169
Me
N
71
Aii
255
326
3317
3484
13.0
10.7
Ph
Ph
O
O
N
N
O
O
Me
N
7m
Aii
N
7n
7o
Aii
Aii
77
416
5.4
Ph
N
O
N
O
S
N
149
2435
16.3
O
a Values are reported as an average for n = 2 or greater.

5304
C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
Table 4. In vitro profile for representative urea analogs
than 400 nM (7u), and a significant shift from intrinsic
to cell-based potency was again observed.
O
N
Given the remarkable intrinsic and cell-based potencies
associated with 7i, we made further alterations around
the piperazine scaffold. We prepared a number of
bridged 2,5-diazabicyclo[2.2.1]heptane analogs 10a–10l
and 3,8-diazabicyclo[3.2.1]octane analogs 11a and 11b
(Tables 5 and 6).
H
NH2
N
N
H
N
N
7
R
O
Me
Compound Methoda
R
HDAC1
HCT116-72h
IC50 (nM)b GI50 (nM)b
7p
7q
7r
Aii
Aii
Aii
Ph
Bn
75
39
47
725
916
892
While a number of the bridged analogs showed compa-
rable or improved intrinsic potencies to 7i as exemplified
by 10c, 10f, 10g, 10k, 10l, and 11b none of the bridged
analogs showed an improved cell-based potency com-
pared to 7i with GI50s ranging from 2262 nM (10l) to
229 nM (10f).
PMB
Ph
7s
Aii
101
1925
Me
Ph
7t
Aii
Aii
47
47
902
433
Me
We also prepared a wide variety of bicyclic amino nico-
tinamides 12a–12e including tetrahydroimidazo[1,2-a]
pyrazine analogs, 2,5-diazabicyclo[4.2.0]octane derivatives,
7u
Ph
a Method Ai: the ureas were prepared at room temperature using the
functionalized piperazine nicotinamide, the corresponding isocyanate
RAC@N@O (2.0 equiv) in DMF for up to 12 h.
Table 5. In vitro characterization of representative 2,5-diazabicy-
clo[2.2.1]heptane nicotinamides
b Values are reported as an average for n = 2 or greater.
O
n
N
disubstituted piperazines, such as dimethyl derivatives
7k and 7l with growth inhibitions of 871 and 3317 nM,
respectively, generally had poor cell-based potencies
and significant shifts from intrinsic potencies. N-sulfon-
amide derivatives also showed a dramatic shift from
intrinsic potency to cell-based potency as demonstrated
by dimethyl piperazine 7o with a greater than 16-fold
shift from HDAC inhibitory activity (149 nM) to
HCT116 growth inhibition (2435 nM). Compounds that
exhibited the greatest cellular potency contained small
alkyl (R1, R2 = Me) mono-substituted piperazines cou-
pled with CBz carbamates (R3). The significance of chi-
rality of the substituent in the distal 2-position (R1/R2)
of the piperazine is most worthy of note. While 7j with
R2 = (R)-Me possessed greater intrinsic potency
(HDAC1 IC50 = 40 nM), 7i with R1 = (S)-methyl and
an intrinsic potency of 73 nM gave a greater than 6-fold
enhancement in cell-based potency (GI50 = 81 nM) over
7j (GI50 = 495 nM).16 While the mechanism for the
remarkable 12-fold shift from HDAC enzymatic activity
to cell-based potency for 7j and other select 6-amino nic-
otinamides that exhibited this shift in potency is unclear,
it may be attributable to an off-target cellular transport
effect. The cell-based anti-proliferative effects of several
analogs were further evaluated in vitro in HeLa cervical
cancer cells. Nicotinamide 7i displayed a remarkable
cell-based potency with a GI50 of 28 nM.
R
H
N
NH2
N
N
10
Compound Method
R
n
HDAC1 HCT116-72h
IC50
(nM)a
GI50
(nM)a
10a
10b
10c
10d
Ai
B
H
1
1
1
1
307
189
68
—
Boc
CBz
Bn
282
284
1273
Aii
Ai
235
Ph
10e
10f
10g
Aii
Ai
Ai
1
1
1
167
39
1480
229
O
Cl
F
54
477
10i
10j
10k
Ai
B
Aii
H
2
2
2
405
669
67
—
7246
1510
Boc
CBz
N
10l
Aii
2
91
2262
O
O
a Values are reported as an average for n = 2 or greater.
Table 6. In vitro characterization for a number of 3,8-diazabicy-
clo[3.2.1]octane nicotinamides
O
N
Having identified that the (2S)-methyl piperazine moiety
coupled with the Cbz carbamate functionality leads to
an exceptional in vitro profile, we further evaluated a
number of urea analogs 7p–7u to determine if we could
improve upon the potency and HCT116 growth inhibi-
tion of 7i (Table 4). While a number of urea analogs
showed comparable or slightly improved HDAC1 enzy-
matic potency compared to 7i as exemplified by analogs
7p–7u ranging in enzymatic activities from 39 nM (7q) to
101 nM (7s), none of the analogs achieved a GI50 of less
H
N
NH2
R
N
N
11
Compound
Method
R
HDAC1
IC50 (nM)a
HCT116-72h
GI50 (nM)a
11a
11b
B
Boc
CBz
167
82
2581
1034
Ai
a Values are reported as an average for n = 2 or greater.

C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
5305
and 1,2,3,4-tetrahydroquinoxalines (Table 7). The most
notable of the bicyclic analogs was 12e with enzymatic
activity of 92 nM and HCT116 growth inhibition of
609 nM. Having extensively derivatized the piperazine
moiety of the nicotinamide scaffold, we determined that
7i was a unique 6-amino nicotinamide HDAC inhibitor
with excellent intrinsic potency and remarkable cell-
based potency worthy of further evaluation in vivo.
to determine what therapeutic advantages class I selec-
tivity offers over less selective HDAC inhibitors.
A homology model of hHDAC1 was built using the
crystal structures of human HDAC8 co-crystallized with
trichostatin A (PDB entry 1T64)18 and HDLP co-crys-
tallized with SAHA (PDB entry 1C3S)19 as templates
(Fig. 4).20 Superposition with the hHDAC8 crystal
structure permitted a direct copy of the Zn and select
waters into the hHDAC1 homology model. Side-chain
torsions of the residues proximate to the Zn were ad-
justed to reflect those observed in hHDAC8. A set of
150 energy minimized conformers were docked into
the hHDAC1 homology model via rigid body alignment
such that the aniline NH2 group and nicotinamide car-
bonyl approximate the position of the hydroxamate hy-
droxyl group and hydroxamate carbonyl, respectively,
from trichostatin A as observed in the hHDAC8 crystal
structure.21 The remainder of each structure was al-
lowed to access the solvent exposed surface via the
Because class I and class II HDACs have very different
biological and functional roles, a significant effort has
been invested in designing inhibitors that exhibit an im-
proved class I selectivity over inhibiting class II
HDACs.17 The potency and subtype selectivity of 7i
were measured in vitro by monitoring the deacetylation
of a synthetic acetyl-lysine bearing peptide by subtype-
specific HDAC complexes purified from mammalian
cells overexpressing recombinant enzyme.13 Nicotin-
amide 7i showed modest selectivity for HDAC1 over
HDAC2, a 13-fold selectivity for HDAC1 over HDAC3,
was essentially inactive against HDAC6, and a greater
than 225-fold selectivity for HDAC1 over HDAC8 (Ta-
ble 8). Further in vivo evaluation of 7i will be necessary
˚
ꢀ11 A channel which provides access to the catalytic
Zn. A set of 25 of these conformers were selected for fur-
ther study based on their ability to uniquely explore var-
ious structural features of the homology model without
encountering extreme steric clashes with the enzyme
model atoms. Each was energy minimized within the
context of the hHDAC1 homology model. Residues fall-
Table 7. In vitro characterization of representative bicyclic nicotina-
mides
˚
ing within 10 A of each conformer were held rigid while
˚
side-chains of residues within 5 A of the compound were
energy minimized in conjunction with the inhibitor.
O
N
H
12
The model shown is the conformer which had the most
stabilizing interactions with the enzyme while encoun-
tering the least amount of strain relative to its original,
ex-enzyme energy minimized state (Fig. 5). Given the
vagaries of homology modeling and the approximations
made to complete these studies, several other conform-
ers nearly isoenergetic with that presented were retained
to provide a more realistic picture of the conformational
space which could be sampled by inhibitor 7i.22
NH2
R
N
N
Compound Method R
HDAC1
IC50 (nM)a GI50 (nM)a
HCT116-72h
N
12a
12b
Ai
Ai
588
139
723
N
N
1118
Ph
N
In our proposed model, the inhibitor’s nicotinamide car-
bonyl oxygen makes a direct interaction with the Zn
N
˚
2.4 A away (Fig. 5). In the case of the template struc-
Ph
Ph
N
12c
12d
Ai
Ai
222
140
866
tures HDLP/SAHA and HDAC8/trichostatin, the dis-
tance between the hydroxamate carbonyl and Zn is 2.7
˚
and 2.2 A, respectively. The aniline NH2 forms a tetra-
hedral pucker such that the protons form hydrogen
bonds with His-140 and His-141 while the nitrogen lone
˚
pair interacts with the Zn 2.4 A away. Relative to the
templates HDLP/SAHA and HDAC8/trichostatin, the
distance between the hydroxamate oxygen and Zn is
˚
1.8 and 2.2 A, respectively.
N
N
N
O
O
N
N
3360
O
Ph
12e
Bb
92
609
O
a Values are reported as an average for n = 2 or greater.
b Method B modification: the pyridyl chloride displacement in this case
required using 12 mol % Pd[P(t-Bu)3]2 and K3PO4 (3 equiv) in PhMe
at 85 ꢁC over the course of 12 h.
Further removed from the catalytic Zn are two more
notable changes which occur during the minimization
Table 8. HDAC In vitro selectivity profile for 7i
HDAC1 IP (nM)a
HDAC2 IP (nM)a
HDAC3 IP (nM)a
939
HDAC6 IP (nM)a
>50,000
HDAC8 IP (nM)a
16,720
73
272
a Values are reported as an average for n = 2 or greater.

5306
C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
Figure 4. Sequence alignment constructed for HDAC1 homology model.
the benzyl carbamate to more closely follow the con-
tours of the protein surface. The piperazine is proximate
to the Leu-271 side-chain and the methyl substituent
descends axially toward Phe-150. The benzyl group ap-
proaches Pro-29. In some of the conformers which are
nearly isoenergetic with that presented, the piperazine
and pyridine are equatorially joined and the carbamate
is axially positioned off the piperazine nitrogen. The
selectivity observed for HDAC1 over HDAC8 can best
be explained in a structural context given that Trp-141
of HDAC8 (Leu-139 for HDAC1) partially occludes
the pocket into which the aniline ring of compound 7i
is docked in the HDAC1 homology model.
Having identified 7i as our lead nicotinamide, we evalu-
ated its pharmacokinetic parameters in 2 species, Spra-
gue–Dawley rat and beagle dog (Table 9). Intravenous
dosing in rat (2 mg/kg) led to a moderate clearance
and low equilibrium volume of distribution, while oral
dosing (4 mg/kg) gave acceptable exposure, oral bio-
availability of 32%, and a terminal plasma elimination
half-life of 2.2 h. Intravenous dosing in dog (0.9 mg/
kg) gave a low clearance as well as a low equilibrium
volume of distribution, while oral dosing (1.8 mg/kg)
led to high exposure, oral bioavailability of 73%, and
a terminal plasma elimination half-life of 5.3 h.
Figure 5. A proposed binding mode for benzamide inhibitor 7i
(purple) in context of the hHDAC1 homology model. The coordina-
tion (green and orange dashed lines) between Zn, inhibitor 7i, and key
amino acid residues (white), as well as proposed hydrogen-bonding
(yellow dashed lines) are shown.
process. One of these is that the phenolic hydroxyl of
Tyr-303 is directed toward the carboxylate of Asp-264
rather than to the nicotinamide carbonyl unlike the ob-
served interaction between the Tyr-OH and hydroxa-
mate carbonyl in the current HDAC/HDLP
crystallographic complexes. The other change noted is
that the phenyl group of Phe-150, located in the channel,
rotates about its axis during minimization such that it is
orthogonal to the starting position. This more closely
approximates the analogous Phe torsion observed in
the HDLP structure rather than that seen in the
hHDAC8 structure. The rotation appears to result from
the proximity of the nicotinamide amide N–H with
which it appears to be forming an edge-to-face p-stack-
ing interaction.
Continuous treatment of HCT116 colon carcinoma cells
for 24 h with 7i resulted in a concentration-dependent
hyperacetylation of histone H2B (H2BK5) with an
EC50 of 2.7 lM, and was accompanied by caspase-3,7
activation consistent with an induction of apoptosis
and cell death. It is important to note that histone acet-
ylation at lysine-5 of histone H2B (H2BK5) induced by
benzamide 7i in HCT116 cells was observed within 1 h
of treatment and was sustained for more than 8 h after
the drug was washed from the cells. Treatment of
HCT116 cells with vorinostat for 1 h also resulted in sig-
nificant H2BK5 acetylation, however this effect was not
sustained and diminished rapidly following removal of
inhibitor.23
The nicotinamide pyridine and linked piperazine of the
inhibitor are modeled to extend through the channel
to the solvent exposed region of the enzyme. In the con-
former depicted, the pyridine joins the piperazine
’’chair’’ conformation in an axial fashion which permits
The overall ancillary pharmacology profile of 7i was
evaluated to establish if there was any significant off-tar-
get activity or other metabolic liabilities associated with
the compound. Nicotinamide 7i was inactive in our

C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
Table 9. Pharmacokinetics profile for 7i
5307
Species
n
IV Dose (mg/kg)
Cl (mL/min/kg)
Vdss (L/kg)
p o Dose (mg/kg)
AUC(N) (lM h kg/mg)
F (%)
t1/2 (h)
Rat
Dog
3
3
2.0
0.9
52
2.1
2.1
4.0
1.8
0.23
3.7
32
73
2.2
5.3
7.3
Effect of 7i on HCT-116 tumor growth
O
500
450
400
350
300
250
200
150
100
50
Vehicle
N
H
'100 mg/kg
'150 mg/kg
NH2
N
N
O
N
O
Me
** : P < 0.01
** : P < 0.01
0
0
3
6
9
12
15
18
21
24
Days after beginning of treatment
Figure 6. Compound 7i dosed IP achieves in vivo efficacy in HCT116 tumor bearing mice compared to vehicle (10% DMSO/45% PEG-400/45%
water).
hERG binding assay with a ki > 30 lM, which does not
predict a liability for 7i to cause drug-induced QT pro-
longation. The CYP inhibition potential of 7i was mea-
In summary, we have discovered a series of 6-amino nic-
otinamides as HDAC inhibitors that possess excellent
enzymatic inhibitory activity and cell-based potency.
An extensive SAR campaign led to the discovery of
6-amino nicotinamide 7i, a class I selective inhibitor that
possesses remarkable anti-proliferative effects in two cell
cultures, displays sustained histone acetylation in vitro,
and achieves in vivo efficacy in HCT116 tumor bearing
mice. The selectivity profile and cell-based potency for
7i are most notable for an HDAC inhibitor in the benz-
amide structural class and further in vivo evaluation of
7i will determine if these characteristics lead to greater
tolerability and efficacy compared to other HDAC
inhibitors under clinical investigation. In general, fur-
ther clinical evaluation of class I selective HDAC inhib-
itors is needed to determine the therapeutic advantages
that greater selectivity might offer with regard to im-
proved tolerability and efficacy compared to less selec-
tive HDAC inhibitors already under clinical
investigation.
sured against
a
number of human isoforms.
Nicotinamide 7i was inactive against CYP3A4 and
CYP2D6 with an IC50 > 50 lM, and showed marginal
activity against CYP2C9 with an IC50 of 24.5 lM and
therefore is not a reversible inhibitor of these major iso-
forms. In a broad-based biochemical assay screen to
check for off-target activities, 7i exhibited minor activity
against human serine/threonine kinase MAPK3 with an
IC50 = 4.52 lM, human dopamine transporter (DAT)
with an IC50 = 3.12 lM, and human serotonin trans-
porter (SERT) with an IC50 = 1.68 lM.
In a 21-day efficacy study in HCT116 tumor bearing
mice, once daily intraperitoneal (IP) administration of
7i at both 100 mg/kg and 150 mg/kg was well tolerated
and inhibited tumor growth by 49% and 53%, respec-
tively, compared with vehicle (Fig. 6). At 7 h post-treat-
ment with 7i, a mean tumor concentration of 6.0 lM
was observed at the 150 mg/kg dose and a mean plasma
concentration of 1.8 lM was observed, while at 26 h, a
mean tumor concentration of 1.2 lM was observed
and a mean plasma concentration of 0.31 lM was
observed.
References and notes
1. (a) Feinberg, A. P.; Vogelstein, B. Nature 1983, 301, 89;
(b) Holliday, R. Science 1987, 238, 163; (c) Jones, P. A.;
Baylin, S. B. Nat. Rev. Genet. 2002, 3, 415.
2. Kouzarides, T. EMBO J. 2000, 19, 1176.
The pharmacokinetics profile and sustained pharmaco-
dynamics of 7i, coupled with its ancillary pharmacology
and established in vivo efficacy, support further investi-
gation of this compound as a potential cancer therapeu-
tic and compare favorably to HDAC inhibitors
currently in development.24
3. Grozinger, C. M.; Schreiber, S. L. Chem. Biol. 2002, 9, 3.
4. Urnov, F. D.; Wolffe, A. Emerg. Ther. Targets 2000, 4,
665.
5. (a) Ng, H. H.; Bird, A. Trends Biochem. Sci. 2000, 25, 121;
(b) Strahl, B. D.; Allis, C. D. Nature 2000, 403, 41;
(c) Jenuwein, T.; Allis, C. D. Science 2001, 293, 1074.

5308
C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
6. (a) Monneret, C. Eur. J. Med. Chem. 2005, 40, 1; (b) Kelly,
W. K.; O’ Connor, O. A.; Marks, P. A. Expert Opin.
Invest. Drugs 2002, 11, 1695.
7. (a) Miller, T. A.; Witter, D. J.; Belvedere, S. J. Med.
Chem. 2003, 46, 5097; (b) Miller, T. A. Expert Opin. Ther.
Pat. 2004, 14, 791.
HDAC8 substrate KI-178), incubated 15 (HDAC8) or 60
(HDACs 1, 2, 3, and 6) minutes at 37 ꢁC, before adding
50 lL of the appropriate development solution. The
development solution for HDACs 1, 2, 3, and 6 was
167·-diluted 20· Developer Concentrate (BIOMOL: KI-
105) plus 10 lM SAHA. For HDAC8, the development
solution was 100·-diluted 5· Developer Concentrate
(BIOMOL: KI-176) plus 10 lM SAHA. The assay was
read in a VictorV plate reader (Perkin-Elmer, Wellesley,
MA) at Ex 360 nm/Em 460 nm. The final substrate
concentration was 30 lM and final HDAC concentrations
in the reaction were 1 nM for HDACs 1, 3, and 6), 2 nM
for HDAC2, and 15 nM for HDAC8. Carboxy-terminal
FLAG-tagged human HDACs 1, 2, 3 (co-expressed with
the domain of SMRT), and 6 were overexpressed in
mammalian cells and affinity purified using an anti-Flag
antibody matrix, eluted from the matrix with 100 lg/mL
of a competing FLAG peptide in 20 mM Tris–Cl, pH 8.0,
150 mM NaCl, 10% glycerol, and protease inhibitor
cocktail (Roche cat. # 1836153).
8. (a) Richon, V. M.; Sandhoff, T. W.; Rifkind, R. A.;
Marks, P. A. Proc. Natl. Acad. Sci. U.S.A. 2000, 97,
10014; (b) Marks, P. A.; Richon, V. M.; Breslow, R.;
Rifkind, R. A. Curr. Opin. Oncol. 2001, 13, 477; (c) Marks,
P. A.; Rifkind, R. A.; Richon, V. M.; Breslow, R. Clin.
Cancer Res. 2001, 7, 759; (d) Marks, P. A.; Rifkind, R. A.;
Richon, V. M.; Breslow, R.; Miller, T.; Kelly, W. K. Nat.
Rev. Cancer 2001, 1, 194; (e) Meinke, P. T.; Liberator, P.
Curr. Med. Chem. 2001, 8, 211; (f) Marks, P. A.; Miller,
T.; Richon, V. M. Curr. Opin. Pharmacol. 2003, 3, 344; (g)
Curtin, M.; Glaser, K. Curr. Med. Chem. 2003, 10, 2373.
9. (a) Marks, P. A.; Breslow, R. Nat. Biotechnol. 2007, 25,
84; (b) Kelly, W. K.; O’Connor, O. A.; Marks, P. A.
Expert Opin. Invest. Drugs 2002, 11, 1695; (c) Jenuwein,
T.; Allis, C. D. Science 2001, 293, 1074.
10. (a) Suzuki, T.; Ando, T.; Tsuchiya, K.; Fukazawa, N.;
Saito, A.; Mariko, Y.; Yamashita, T.; Nakanishi, O.
J. Med. Chem. 1999, 42, 3001; (b) Saito, A.; Yamashita,
T.; Mariko, Y.; Nosaka, Y.; Tsuchiya, K.; Ando, T.;
Suzuki, T.; Tsuruo, T.; Nakanishi, O. Proc. Natl. Acad.
Sci. 1999, 96, 4592; (c) Jaboin, J.; Wild, J.; Hamidi, H.;
Khanna, C.; Kim, C. J.; Robey, R.; Bates, S. E.; Thiele, C.
J. Cancer Res. 2002, 62, 6108.
11. (a) El-Beltagi, H. M.; Martens, A. C. M.; Lelieveld, P.;
Haroun, E. A.; Hagenbeek, A. Cancer Res. 1993, 53, 3008;
(b) Loprevite, M.; Tiseo, M.; Grossi, F.; Scolaro, T.;
Semino, C.; Pandolfi, A.; Favoni, R.; Ardizzoni, A. Oncol.
Res. 2005, 15, 39; (c) Foster, B. J.; Jones, L.; Wiegand, R.;
Lorusso, P. M.; Corbett, T. H. Invest. New Drugs 1997, 15,
187.
12. (a) Delorme, D.; Woo, S. H.; Vaisburg, A.; Moradel, O.;
Leit, S.; Raeppel, S.; Frechette, S.; Bouchain, G. PCT Int.
Appl. 2003, WO 2003024448; (b) Siu, L. L; Carducci, M.;
Pearce, L.; Maclean, M.; Sullivan, R.; Li, Z.; Kalita, A.;
Chen, E. X.; Pili, R.; Longstreth, J.; Martell, R. E.; Reid,
G. K. Abstracts of Posters, C77: Phase I Study of Isotype
Selective Histone Deactylase (HDAC) Inhibitor
MGCD0103 Given as Three-Times Weekly Oral Dose in
Patients (pts) with Advanced Solid Tumors, AACR-NCI-
EORTC International Conference on Molecular Targets
and Cancer Therapeutics: Discovery, Biology, and Clin-
ical Applications, Philadelphia, Pennsylvania, November
14–18, 2005.
13. The HDAC Fluorescent HDAC1 and HDAC8 Fluor-de-
Lys Activity Assays from BioMol Research Laboratories
(Plymouth Meeting, PA) provided the basis for our
HDAC activity assays. The assay buffer for the HDAC1
and HDAC2 assays consisted of 20 mM Hepes, pH 8.0,
137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 0.1 mg/mL
BSA; for the HDAC3 assay it was 20 mM Hepes, pH 8.0,
137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, and 0.25 mg/
ml BSA; for the HDAC6 assay, it was 20 mM Hepes, pH
7.4, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, and
0.1 mg/mL BSA; and for the HDAC8 assay, it was 20 mM
Hepes, pH 8.0, 100 mM NaCl, 20 mM KCl, 1 mM MgCl2,
and 0.1 mg/mL BSA. HDAC enzymatic activities were
determined by the following procedure: 3· serial dilutions
of a 10 mM solution of inhibitor were performed in
DMSO followed by a 20· dilution into Assay buffer. 20 ll
HDAC was preincubated with 5 lL diluted compound at
RT for 10 min. The reaction was initiated by the addition
of 25 lL of the appropriate substrate (HDACs 1, 2, 3, and
6: fluor-de-Lys substrate KI-104; HDAC8: fluor-de-Lys
14. Seto, C. T.; Mathias, J. P.; Whitesides, G. M. J. Am.
Chem. Soc. 1993, 115, 1321.
15. HDAC cell-based proliferation results were determined by
the following procedure: HCT116 cells were plated in 96-
well plates at density of 1000 cells/well. The next day, cells
were treated with either 0.2% DMSO or increasing
concentrations of 7i dissolved in DMSO (final concentra-
tion DMSO 0.2%). After 72 h incubation at 37 ꢁC with 5%
CO2, viable cells were quantitated using Vialight Plus
(Cambrex) according to manufacturer’s instructions.
16. Experimental procedure for the synthesis of 7i: a mixture of
the Boc-protected chloronicotinamide (3.0 g, 8.6 mmol)
and benzyl-(2S)-2-methylpiperazine-1-carboxylate (6.0 g,
25.8 mmol) in PhMe (5 mL) was heated at 85 ꢁC for 12 h.
The reaction mixture was diluted with EtOAc (100 mL),
and washed with satd aq NaHCO3 (1 · 25 mL) and brine
(1 · 25 mL). The crude oil was purified by reverse phase
chromatography (25–100% MeCN/H2O with 0.05% TFA)
and formation of the desired Boc-protected piperazinyl
nicotinamide was confirmed by LC/MS (ESI+): calcd
[M+H]+ 546.3, exp. 546.3. The Boc-protected piperazinyl
nicotinamide was treated with TFA (4 mL) in CH2Cl2
(8 mL) and after 20 min of stirring at room temperature,
the reaction mixture was concentrated and purified by
reverse phase chromatography (15–75% MeCN/H2O with
0.05% TFA). The appropriate fractions were combined,
diluted with EtOAc (150 mL) and washed with satd aq
NaHCO3 (1 · 50 mL) and brine (1 · 50 mL). The organic
layer was dried over Na2SO4, filtered, and concentrated to
give the desired nicotinamide 7i: 1H NMR (600 MHz,
DMSO-d6) d 9.42 (s, 1H), 8.69 (s, 1H), 8.05 (d, J = 8.8 Hz,
1H), 7.34–7.29 (m, 5H), 7.10 (d, J = 7.3 Hz, 1H), 6.92 (t,
J = 7.3 Hz, 1H), 6.85 (d, J = 9.1 Hz, 1H), 6.73 (d,
J = 7.9 Hz, 1H), 6.55 (m, 1H), 5.11–4.85 (m, 2H), 4.83 (br
s, 2H), 4.26-4.21 (m, 3H), 3.85 (dd, J = 5.0 Hz, 3.5 Hz, 1H),
3.34–3.22 (m, 2H), 3.20–3.00 (m, 1H), 1.07 (d, J = 6.5 Hz,
3H); MS (ESI+): calcd [M+H]+ 446.2, exp. 446.2.
17. (a) Heltweg, B.; Dequiedt, F.; Marshall, B. L.; Brauch, C.;
Yoshida, M.; Nishino, N.; Verdin, E.; Jung, M. J. Med.
Chem. 2004, 47, 5235; (b) Curtin, M. L. Expert Opin. Ther.
Pat. 2002, 12, 1375.
18. HDAC8 crystal structure (1T64) Somoza, J. R.; Skene, R.
J.; Katz, B. A.; Mol, C.; Ho, J. D.; Jennings, A. J.; Luong,
C.; Arvai, A.; Buggy, J. J.; Chi, E.; Tang, J.; Sang, B.-C.;
Verner, E.; Wynands, R.; Leahy, E. M.; Dougan, D. R.;
Snell, G.; Navre, M.; Knuth, M. W.; Swanson, R. V.;
McRee, DE.; Tari, L. W. Structure 2004, 12, 1325.
19. HDAC8 crystal structure (1T64) (a) Somoza, J. R.; Skene,
R. J.; Katz, B. A.; Mol, C.; Ho, J. D.; Jennings, A. J.;

C. L. Hamblett et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5300–5309
5309
Luong, C.; Arvai, A.; Buggy, J. J.; Chi, E.; Tang, J.; Sang,
B.-C.; Verner, E.; Wynands, R.; Leahy, E. M.; Dougan,
D. R.; Snell, G.; Navre, M.; Knuth, M. W.; Swanson, R.
V.; McRee, DE.; Tari, L. W. Structure 2004, 12, 1325;
HDLP crystal structure (1C3S); (b) Finnin, M. S.;
Donigian, J. R.; Cohen, A.; Richon, V. M.; Rifkind, R.
A.; Marks, P. A.; Breslow, R.; Pavletich, N. P. Nature
1999, 401, 188.
Wiley: New York, 1988; (b) Halgren, T. A. J. Comp.
Chem. 1999, 20, 720.
22. During the course of energy minimization within the
context of the enzyme model, only slight differences in the
presentation of the co-minimizing residues (those within
˚
5 A of the inhibitor) were noted between the starting and
final enzyme structure. The catalytic Zn and chelating
residues were virtually unchanged during the minimization
process.
20. An alignment of the two crystal structures and the amino
acid sequence for hHDAC1 (retrieved from SwissProt,
accession ID: Q13547) was accomplished using the
sequence/structure alignment tools in MOE (Chemical
Computing Group, Montreal, Canada). This alignment
was then used in Quanta/Modeler5 (Accelrys, San Diego,
CA, USA) to generate a set of homology models. That
with the least number of violations was further energy
minimized using CHARMm 200 steps steepest descents
without the catalytic Zn or any waters present: (a)
23. Acetylation assays were performed using the following
procedures: HCT116 cells were plated in 6-well plates at a
density of 1 · 106 cells per well. The next day, cells were
incubated at 37 ꢁC, 5% CO2 for 24 hours with either 0.2%
DMSO or increasing concentrations of 7i dissolved in
DMSO (final DMSO concentration = 0.2%). Whole cell
lysates were made by disrupting cell membranes with 2%
deoxycholate followed by sonication. Using these HCT116
lysates as antigen, H2BK5 acetylation was quantified
using an indirect ELISA and normalized to total H2B
levels obtained using a second indirect ELISA. Primary
antibodies used were rabbit anti-acetylated H2BK5 (Cell
Signaling) and sheep anti-histone H2B (Abcam). Goat
anti-rabbit IgG HRP (Bio-Rad) and rabbit anti-sheep-
HRP (Jackson ImmunoResearch) were used as secondary
antibodies, respectively. TMB substrate (Pierce) was used
for detection with absorbance read at 450 nm.
´
Sanchez, R.; Sali, A. Comparative protein structure
modeling. In Protein Structure Prediction: Methods and
Protocols; Webster, D. M., Ed.; Humana Press: New
Jersey, 2000; Vol. 143, pp. 97–129; (b) Brooks, B. R.;
Bruccoleri, R. E.; Olafson, B. D.; States, D. J.; Swamina-
than, S.; Karplus, M. J. Comp. Chem. 1983, 4, 187; (c)
Wang, D.-F.; Helquist, P.; Wiech, N. L.; Wiest, O. J. Med.
Chem. 2005, 48, 6936.
21. The 150 conformers were generated and energy minimized
using MMFFs with a distance-dependent dielectric of 2r:
(a) Crippen, C. M.; Havel, T. F. Distance Geometry and
Molecular Conformation. In Chemometrics Research
Studies Series; Bawden, D., Ed.; Research Studies Press,
24. (a) Yoo, C. B.; Jones, P. A. Nat. Rev. Drug Disc. 2006, 5,
37; (b) Bolden, J. E.; Peart, M. J.; Johnstone, R. W. Nat.
Rev. Drug. Disc. 2006, 5, 769; (c) Moradei, O.; Maroun, C.
R.; Paquin, I.; Vaisburg, A. Curr. Med. Chem.: Anti-
Cancer Agents 2005, 5, 529.