Histone deacetylase inhibitors with a primary amide zinc binding group display antitumor activity in xenograft model

Article abstract of DOI:10.1016/j.bmcl.2009.04.011

Histone deacetylase (HDAC) inhibition causes hyperacetylation of histones leading to differentiation, growth arrest and apoptosis of malignant cells, representing a new strategy in cancer therapy. Many of the known HDAC inhibitors (HDACi) that are in clin

Full text of DOI:10.1016/j.bmcl.2009.04.011

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3081–3084
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Histone deacetylase inhibitors with a primary amide zinc binding group
display antitumor activity in xenograft model
a
b
a
a,
Barbara Attenni a, , Jesus M. Ontoria , Jonathan C. Cruz , Michael Rowley , Carsten Schultz-Fademrecht
,
*
Christian Steinkühler a, Philip Jones a
a IRBM/Merck Research Laboratories, Via Pontina Km 30,600, 00040 Pomezia (Rome), Italy
b MRL Boston, 33 Avenue Louis Pasteur, 02115 Boston (MA), USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Histone deacetylase (HDAC) inhibition causes hyperacetylation of histones leading to differentiation,
growth arrest and apoptosis of malignant cells, representing a new strategy in cancer therapy. Many of
the known HDAC inhibitors (HDACi) that are in clinical trials possess a hydroxamic acid, that is a strong
Zn2+ binding group, thereby inhibiting some of the class I and class II isoforms. Herein we describe the
identification of a selective class I HDAC inhibitor bearing a primary carboxamide moiety as zinc binding
group. This HDACi displays good antiproliferative activity against multiple cancer cell lines, and demon-
strates efficacy in a xenograft model comparable to vorinostat.
Received 9 March 2009
Revised 1 April 2009
Accepted 2 April 2009
Available online 9 April 2009
Keywords:
HDAC inhibitors
Primary carboxamide
Zinc binding group
Ó 2009 Elsevier Ltd. All rights reserved.
Histone deacetylases (HDACs) are enzymes involved in
determining the pattern of acetylation of chromatin proteins (his-
tones) in eukaryotic cells.1 The acetylation of histones results in
transcriptionally active chromatin, increasing the accessibility of
nucleosomal DNA; this facilitates gene transcription and leads to
differentiation, growth arrest and apoptosis representing a new
strategy in cancer therapy.1,2
can result in inhibition of other metallo-enzymes or sequestration
of metal ions. Thus there has been a considerable interest in devel-
oping non-hydroxamate HDAC inhibitors.
In the mid 90s researchers at Merck discovered apicidin, an
antiprotozoal agent that inhibits parasite histone deacetylase.7
Apicidin is structurally complex but a targeted screening of the
sample collection, looking for compounds containing the L-Aoda
The HDACs belong to four structurally and functionally different
classes: class I (HDAC 1–3 and 8), class II (HDAC 4–7, 9 and 10)
while class IV (comprising HDAC 11) exhibits properties of both
class I and class II HDACs.3 All the above HDACs are zinc dependent
hydrolases, whereas class III HDACs are structurally distinct and
require the cofactor NAD+ for their deacetylase function.
The vast majority of HDAC inhibitors (HDACi) currently under-
going development, such as vorinostat (1) (ZolinzaÒ, formerly
known as SAHA) approved by the FDA for the treatment of the
cutaneous T-cell lymphoma,4 contain a hydroxamic acid (Fig. 1).
Other hydroxamic acid HDACis include panobinostat (3),5 and beli-
nostat (4).6 This functionality binds to a zinc ion in the active site
displaying little isoform selectivity between class I, II and IV
HDACs.
amino acid, identified a more tractable lead which was further
optimized to 6 (Fig. 2).8
This compound displays an interesting level of antiproliferative
activity in tumor cells and was demonstrated to function in xeno-
graft studies. Derivatization of this scaffold gave rise to an even
more potent compound containing a naphthyl in the 5 position
of imidazole.9 Similarly an ethyl ketone was also shown to improve
O
O
O
H
N
OH
OH
N
H
N
H
O
Me2N
2: Trichostatin A
1: Vorinostat (SAHA)
Hydroxamic acids have some liabilities associated with this
functional group including poor oral absorption and they often dis-
play poor pharmacokinetics as a result of hydrolysis or rapid gluco-
ronidation. The presence of the strong metal chelating group also
O
H
H
OH
N
N
N
H
S
OH
H
N
O O
O
HN
4: Belinostat (PXD101)
* Corresponding author. Tel.: +39 0691093539; fax: +39 0691093654.
E-mail address: barbara_attenni@merck.com (B. Attenni).
3: Panobinostat (LBH-589)
Present address: Lead Discovery Center, 44227 Dortmund, Germany.
Figure 1. Known HDAC inhibitors containing hydroxamic acid.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.04.011

3082
B. Attenni et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3081–3084
Table 2
SAR of the capping group
O
N
O
CH3
Optimisation
N
H
H
N
N
N
H
N
NH
OH
N
OMe
HN
O
O
N
H
OCH3
O
HN
O
RA
O
Compound
Ra
HDAC1 IC50 (nM)a
23
PRO (HeLa) IC50 (nM)a
H3C
N
O
5: Apicidin
6
H
O
IC50 (nM)
HDAC1
PRO (HeLa)
S
14
5800
44
290
27
680
N
O
O
Figure 2. Identification of a lead series from the natural product apicidin.
15
16
17
18
73
>25,000
>25,000
26,400
35,000
N
40
activity. Combining the two pieces gives
7 that displays
N
IC50 = 27 nM against HDAC1 and inhibits proliferation of HeLa cells
with IC50 = 220 nM (Table 1). Knowing that an alkyl ketone is a rel-
atively weak zinc chelator we rationalized that significant binding
affinity was due to the two capping groups: the methyl indolyl and
substituted imidazole. Having such a potent HDACi we were inter-
ested in further exploring other weak zinc chelating groups in the
hope of further refining the isoform selectivity. Apart from amino
benzamides, sulfides and ketones few other HDACi have been
reported.10
O
O
N
S
220
35
O
N
O
O
S
NO2
19
19
6860
Trying to identify a selective class I HDAC inhibitor we changed
the ketone functionality into different possible metal binding moi-
eties (Table 1).
a
Values are means of two or more experiments.
The hydroxamic acid 9 was confirmed to be a potent HDACi
(IC50 = 3 nM) and displayed IC50 = 180 nM in a 72 h anti-prolifera-
tion assay in HeLa cervical cancer cells.8 Another two derivatives
which displayed interesting cellular activity were N-methoxy
amide 10 and the methylamide 13 (HDAC1 IC50 = 34 and 120 nM,
that simple functionality such as a carboxylic acid 8 show promis-
ing levels of anti-proliferation activity in the cell based assay.
Acylsulfonamides are described as lipophilic and more permeable
bioisosters of carboxylic acids, however 11 showed a significant
HeLa IC50 = 0.66 and 1.7 lM, respectively). Interestingly the methyl
amide although showing weak enzyme inhibition displayed low
loss of antiproliferative activity in cells (PRO HeLa IC50 = 8 lM).
The more acidic analogue 12 led to an even further loss of activity
micromolar antiproliferative activity. We were pleased to observe
both in enzyme and in cells (HDAC1 IC50 = 94 nM, PRO HeLa
IC50 = 33.5 lM). Given the interesting cellular activity we started
to profile further compound 8. It suffered from extensive metabo-
lism in rat liver microsomes in the presence of NADPH
Table 1
SAR of zinc binding group
(Clint = 177
lL/min/mg P) while more stability was detected in
human and dog liver microsomes (Clint = 9 and 3
lL/min/mg P,
N
respectively). Moreover, the compound was glucoronidated in
the presence of UDPGA particularly in human microsomes
(Clint = 24 lL/min/mg P). Nevertheless we decided to conduct
N
H
R
HN
O
OCH3
further SAR to try to improve its pharmacokinetics properties
(Table 2).
H3C
N
H
The fragments selected were those which had previously been
demonstrated to improve microsomal stability in the ketone ser-
ies.9 Unfortunately in the carboxylic acid series although maintain-
ing enzyme inhibitory activity, 14–19 lost cellular activity. The
best derivatives, 14, showed only 6 lM anti-proliferation activity,
one order of magnitude weaker than 8.
Considering the promising smaller shift in cells observed for
methyl amide 13 (Table 1), we decided to perform a further SAR
around the amide group (Table 3).
The simple primary amide 20 showed the same level of antipro-
liferative activity of the carboxylic acid 8 (HDAC1 IC50 = 30 nM,
Compound
R
HDAC1 IC50 (nM)a
PRO (HeLa) IC50 (nM)a
200
O
O
7
27
CH3
8
28
3
390
180
OH
O
O
O
O
O
OH
9
N
H
O
10
11
12
34
47
94
120
660
N
H
CH3
PRO HeLa IC50 = 0.44
only twofold on the enzyme (IC50 = 66 nM) was sixfold less active
in cells (IC50 = 2.7 M). Similarly the tertiary dimethylamide 25
displayed HDAC1 IC50 = 170 nM and weak HeLa anti-proliferation
activity (IC50 = 3.8 M). While the introduction of a branched resi-
due such as the isopropyl group present in 22 causes a substantial
loss of activity (PRO HeLa IC50 = 12 M). As predicted the amino-
benzamide 27 showed an anti-proliferative activity in the low
micromolar range (HeLa IC50 = 1.34 M) despite showing only
weak enzyme activity. Micromolar anti-proliferation activity in
lM). The ethyl substitution although losing
O O
S
8000
33,500
1700
CH3
N
H
l
O O
F
S
F
l
N
H
F
l
CH3
13
N
H
l
a
Values are means of two or more experiments.

B. Attenni et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3081–3084
3083
Table 3
SAR of carboxamide
Compound 20 was then evaluated in a HCT116 xenograft study,
it was dosed at 100 mpk for 18 days once a day, and efficacy was
compared to that seen with vorinostat which was dosed at
150 mpk using the same dosing regime. Interestingly the amide
20 displayed similar efficacy to 1 (29% tumor growth inhibition
compared with the vehicle group was observed) with less than
5% body weight loss (Fig 4).
To our knowledge this is the first example of a small molecule
HDACi bearing a primary amide zinc binding motif demonstrating
efficacy in xenograft studies.
N
R
N
H
HN
O
O
OCH3
H3C
N
H
Entry
R
HDAC1 IC50 (nM)a
30
PRO (HeLa) IC50 (nM)a
440
H
N
In parallel, target engagement was measured by assessing the
level of histone acetylation seen in tumor biopsies of HCT116 xeno-
graft mice (4 per group). At efficacious doses of 100 mpk ip an
20
H
H
N
21
22
23
66
480
2700
12,000
6800
CH3
N
H
10.000
N
H
1000
24
25
26
200
170
350
5900
3800
4700
N
H
50
1.000
0.100
HCT116IC
CH3
N
CH3
S
N
N
N
H
27
110
1340
N
H
0.00
5.00
10.00
15.00
20.00
25.00
30.00
NH2
Time (Hour)
a
Values are means of two or more experiments.
Figure 3. Mouse PK (100 mpk, ip).
Table 4
1200
1000
800
600
400
200
0
Activity of compound 20 on HDAC isoforms and various cell linea, IC50 (nM)b
Compound
1
2
3
4
6
8
Vorinostat
20
30
30
82
160
57
250
>10
>10
l
l
M
M
43
120
1700
>5 lM
Cervical HeLa
Colon HCT116
Lung A549
Vorinostat
20
460
440
1000
860
1800
3560
a
Assay condition as described in Ref. 8.
Values are means of two or more experiments.
b
0
3
6
9
12
15
18
Days after treatment
the cell based assay was also detected either with benzyl (23) and
phenyl (24) amides (HeLa IC50 = 6.8 and 5.9 M, respectively).
Figure 4. Antitumor activity of 1 (j) and 20 (N) in a human HCT116 colon cancer
xenograft model in nude mice (12 per group) following ip dosing in vehicle (d)
(error bars represent mean SEM values).
l
Compound 20 was profiled on individual HDAC isoforms, where
it was demonstrated to show fivefold selectivity for HDAC1 over
HDAC2 and nearly 10-fold selective over HDAC3 (IC50 = 30, 160
and 250 nM, respectively). This is in contrast either to vorinostat,
which displayed similar level of activity on HDAC 1, 2 and 3, and
to ketone 7 almost equipotent respect to these 3 HDACs.8 Excellent
selectivity was also seen over HDAC 4 and 8 (Table 4).
Compound 20 was also evaluated on several other cell lines
where it showed comparable activity to vorinostat. With the
encouraging cellular activity it was decided to evaluate 20 in an
in vivo efficacy study, a human colon HCT-116 carcinoma xeno-
graft model. The compound was profiled alongside vorinostat. In
preliminary studies, the maximum tolerated dose (MTD) of both
compounds was determined following intraperitoneal (IP) once
daily dosing for 2 weeks, as defined by no mortality and body
weight loss less than 5%. The amide 20 showed good exposure at
the MTD dose (100 mpk) in mice, and the plasma concentration
remained above IC50 for inhibition of HCT116 proliferation for
nearly 12 h (Fig. 3).
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0
2
4
6
8
10 12 14 16 18 20 22 24
Time of Pretreatment (hrs)
Figure 5. Histone acetylation in HCT116 tumor cells in xenografts after ip dosing of
20 at 100 mpk (j) and vehicle (ꢀ) (error bars represent mean SEM values).

3084
B. Attenni et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3081–3084
O
Br
O
i) Cs2CO3, DMF
N
N
TFA, DCM
quant.
O
O
HO
OH
N
H
N
H
30
NHBoc
O
ii) NH4OAc, toluene
Reflux 80%
+TFA-
NHBoc
O
NH3
O
28
29
EDC, HOBt
DMF, R2CO2H
CO2H
OMe
HOBt, EDC
DIPEA DMF
N
OH
N
H
H3C
N
H
HN
O
O
R2
N
14-19
N
OH
H
N
N
H
R1
N
H
HN
O
O
HN
O
O
OMe
OMe
EDC, HOBt
DMF, NH2R1
H3C
N
H3C
9-13
H
N
8
H
20-27
Scheme 1. Synthesis of compound 11–29.
increase of acetylated histone was seen for more than 8 h with
maximal activity at 4 h post administration (Fig. 5).
References and notes
1. Brittain, D.; Weinmann, H.; Ottow, E. Annu. Rep. Med. Chem. 2007, 42, 337;
Rodriquez, M.; Aquino, M.; Bruno, I.; De Martino, G.; Taddei, M.; Gomez-
Paloma, L. Curr. Med. Chem. 2006, 13, 1119; Jones, P.; Steinkuhler, C. Curr.
Pharm. Des. 2008, 14, 545.
2. Glozak, M. A.; Seto, E. Oncogene 2007, 26, 5420; Minucci, S.; Pelicci, P. G. Nat.
Rev. Cancer 2006, 6, 38.
3. (a) De Ruijter, A. J.; van Gennip, A. H.; Caron, H. N.; Kemp, S.; van Kuilenburg, A.
B. Biochem. J. 2003, 370, 737; (b) Voelter-Mahlknecht, S.; Ho, A. D.; Mahlknecht,
U. Int. J. Mol. Med. 2005, 16, 589.
4. Grant, S.; Easley, C.; Kirkpatrick, P. Nat. Rev. Drug Disc. 2007, 6, 21.
5. Maiso, P.; Carvajal-Vergara, X.; Ocio, E. M.; Lòpez-Pérez, R.; Mateo, G.;
Gutiérrez, N.; Atadja, P.; Pandiella, A.; San Miguel, J. F. Cancer Res. 2006, 66,
5781.
The synthesis of these derivatives starts from the readily
prepared
a-amino acid 28 (Scheme 1) which was first alkylated
with 2-(bromoacetyl)naphthalene and then cyclized to the corre-
sponding imidazole 29 in refluxing toluene in the presence of an
excess of NH4OAc. Double deprotection gave intermediate 30
which was added to a dichloromethane solution of 5-methoxy-2-
methyl-3-indoleacetic acid or the other carboxylic acids, under
standard coupling conditions, to get the free acid 8 or compounds
14–19. Derivatives 9–13 and 20–27 were obtained after the reac-
tion of 8 with the suitable coupling reagents.
6. Plumb, J. A.; Finn, P. W.; Williams, R. J.; Bandara, M. J.; Romero, M. R.; Watkins,
C. J.; La Thangue, N. B.; Brown, R. Mol. Cancer Ther. 2003, 2, 721.
7. Darkin-Rattray, S. J.; Gurnett, A. M.; Myers, R. W.; Dulski, P. M.; Crumley, T. M.;
Allocco, J. J.; Cannova, C.; Meinke, P. T.; Colletti, S. L.; Bednarek, M. A.; Singh, S.
B.; Goetz, M. A.; Dombrowski, A. W.; Polishook, J. D.; Schmatz, D. M. Proc. Natl.
Acad. Sci. U.S.A. 1996, 93, 13143.
8. Jones, P.; Altamura, S.; De Francesco, R.; Gonzalez Paz, O.; Kinzel, O.; Mesiti, G.;
Monteagudo, E.; Pescatore, G.; Rowley, M.; Verdirame, M.; Steinkühler, C. J.
Med. Chem. 2008, 51, 2350.
In conclusion, we have described the first example of a selective
class I HDAC inhibitor with a primary carboxamide moiety as zinc
binding group. Furthermore 20 displays selectivity for HDAC 1 and
demonstrated efficacy in an HCT116 xenograft model comparable
with vorinostat.
Acknowledgements
9. Pescatore, G.; Kinzel, O.; Attenni, B.; Cecchetti, O.; Fiore, F.; Fonsi, M.; Rowley,
M.; Schultz-Fademrecht, C.; Serafini, S.; Steinkühler, C.; Jones, P. Bioorg. Med.
Chem. Lett. 2008, 18, 5528.
The authors thank Ottavia Cecchetti and Sergio Serafini for their
support of this work.
10. Paris, M.; Porcelloni, M.; Binaschi, M.; Fattori, D. J. Med. Chem. 2008, 51, 1505.