Discovery of a potent class I selective ketone histone deacetylase inhibitor with antitumor activity in vivo and optimized pharmacokinetic properties

Article abstract of DOI:10.1021/jm9004303

The optimization of a potent, class I selective ketone HDAC inhibitor is shown. It possesses optimized pharmacokinetic properties in preclinical species, has a clean off-target profile, and is negative in a microbial mutagenicity (Ames) test. In a mouse xenograft model it shows efficacy comparable to that of vorinostat at a 10-fold reduced dose.

Full text of DOI:10.1021/jm9004303

J. Med. Chem. 2009, 52, 3453–3456
3453
Discovery of a Potent Class I Selective
Ketone Histone Deacetylase Inhibitor with
Antitumor Activity in Vivo and Optimized
Pharmacokinetic Properties
Olaf Kinzel,* Laura Llauger-Bufi, Giovanna Pescatore,
Michael Rowley, Carsten Schultz-Fademrecht,
Edith Monteagudo, Massimiliano Fonsi,
Odalys Gonzalez Paz, Fabrizio Fiore, Christian Steinku¨hler,
and Philip Jones
IRBM/Merck Research Laboratories, Via Pontina km 30, 600,
00040 Pomezia, Italy
ReceiVed April 3, 2009
Figure 1. Clinically advanced HDAC inhibitors.
Abstract: The optimization of a potent, class I selective ketone HDAC
inhibitor is shown. It possesses optimized pharmacokinetic properties
in preclinical species, has a clean off-target profile, and is negative in
a microbial mutagenicity (Ames) test. In a mouse xenograft model it
shows efficacy comparable to that of vorinostat at a 10-fold reduced
dose.
into evidence the ability of HDACi to induce cell death of
transformed cells by multiple pathways.5
While the precise mechanism of how HDACi exert their
anticancer effect remains to be fully elucidated, this new class
of chemotherapeutics has demonstrated potent anticancer activi-
ties in preclinical and clinical studies.6 Vorinostat (1, Figure
1), belonging to the class of hydroxamic acid inhibitors, has
been approved by the FDA for the treatment of advanced
cutaneous T-cell lymphoma.7,8 A number of other compounds
are in phase II clinical trials for the treatment of hematological
and solid tumors, e.g., 3 (MS-275)9 and 6 (MGCD-0103),10 both
belonging to the class of aminobenzamides, 7 (romidepsin,
FK228),11 a natural product, and 2 (PXD-101)12 and 5 (LBH-
589),13 both hydroxamic acids (Figure 1).6
In the nucleus of eukaryotic cells DNA is tightly packed
around nucleosomes composed of histone proteins whose
N-terminal tails are subject to a number of modifications like
acetylation, methylation, phosphorylation, and others.1 Histone
deacetylases (HDACsa) and histone acetyl transferases (HATs)
are the two counteracting key enzyme families involved in the
regulation of the acetylation status of histone tails (and other
non-histone target proteins). The HDAC mediated deacetylation
of lysine residues enforces ionic interactions with the negatively
charged DNA backbone. This leads to a more condensed form
of chromatin that is inaccessible to the transcription machinery.
On the contrary, the action of HATs leads to a more relaxed
form of chromatin, favoring transcription.2
There are four classes of histone deacetylases. Classes I, II,
and IV enzymes possess a zinc ion in their active site and consist
of 11 isoforms, whereas class III deacetylases (often referred
to as sirtuins) are NAD+ dependent enzymes. Many of the class
I enzymes (HDACs 1, 2, 3, 8) are preferentially located in the
nucleus where they are associated in larger complexes that
specifically mediate their catalytic activity. Class II enzymes
(HDACs 4, 5, 6, 7, 9 10) may localize in the cytosol, and some
of these proteins were shown to shuttle between the nucleus
and the cytoplasm in response to external stimuli. HDAC 11 is
the only member of class IV.3,4
The search for a second generation inhibitor remains an
important activity in the field of HDACi, with the ultimate goal
being an improved therapeutic window in man. Many hydrox-
amic acids are poorly selective and show activity against classes
I and II enzyme isoforms. Aminobenzamides like 3 and 6 show
improved selectivity for HDACs 1, 2, 3, and recently biaryl
analogues like 4 (Figure 1) have been reported to have further
increased selectivity for HDACs 1 and 2.14-17 For the class I
isoforms the correlation between inhibition and antitumor effect
is well established. Therefore, a class I selective inhibitor would
be desirable. Compound 1 has antiproliferative activity against
cancer cells with an IC50 in the low micromolar range and suffers
from a relatively short half-life in man (high plasma Cmax/Ctrough
ratio).18 A more potent compound with a longer half-life and a
smaller plasma Cmax/Ctrough ratio when dosed orally would be
desirable to improve the overall therapeutic window.
In malignant cells HDACs are overexpressed, mutated, or
aberrantly recruited by oncoproteins, causing hypoacetylation
of histones leading to transcriptional repression. Upon the action
of HDAC inhibitors (HDACi), the balance in the acetylation
status is restored and mostly genes involved in apoptosis, cell
cycle checkpoints (G1/S, G2/M), and cellular differentiation are
activated. Moreover, a growing number of non-histone target
proteins important in the regulation of apoptosis (p53), mitosis
(R-tubulin), and cellular migration have been identified, putting
Previously we have described how our efforts to identify
second generation HDACi started from apicidin (8, Figure 2).19
This cyclic tetrapeptide contains an octanyl ketone chain, which
is believed to chelate the zinc ion present in the active site. A
novel series of potent low molecular weight HDACi containing
the L-Aoda amino acid present in 8 was identified and
subsequently transformed into a related series of imidazole
containing compounds like 9, which displays activity in vivo
albeit with poor PK properties.14,20 Optimization of this new
class of compounds has resulted in structures like 10 with
improved potency. Although 10 displayed a good PK profile in
dogs, it suffered from a high clearance in rats. The related ethyl
ketone analogue 11 has an 80-fold reduced activity on HDAC
* To whom correspondence should be addressed. Phone: (+39)
0691093334. Fax: (+39) 0691093654. E-mail: olaf_kinzel@merck.com.
a Abbreviations: HAT, histone acetyl transferases; HDAC, histone
deacetylases; CYP, cytochrome P450; PK, pharmacokinetic.
10.1021/jm9004303 CCC: $40.75
2009 American Chemical Society
Published on Web 05/14/2009

3454 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 11
Letters
Table 1. Activity on HDAC 1, Antiproliferative Activity against HeLa
Cells, Intrinsic Metabolic Stability, in Vivo Rat Clearance
Figure 2. Key transitions steps from 8 to imidazole HDACi 9 and
10/11.
6 compared to 10 and is slightly more active against HeLa cells
in the proliferation assay (Table 1).21
Herein we describe the successful further optimization of
these inhibitors culminating in the identification of a potent and
HDAC class I selective alkyl ketone HDAC inhibitor with
optimized pharmacokinetic properties.
In vitro and in vivo investigations of the metabolism of 10
identified the naphthyl group as the metabolic soft spot in rats.22
This finding triggered an intense investigation of the aryl
substituent on the imidazole with the objective to identify
moieties that would stabilize our inhibitors. On the basis of our
previous results, the nature of the pendent basic amino group
also contributes to the overall intrinsic stability of the inhibitors.
We therefore set out for a focused optimization of both the aryl
substituents (R1) and the capping group (R2) starting from lead
11.
The synthesis for most of our compounds followed the route
described previously (Scheme 1).14 Briefly, esterification of
N-Boc-protected amino acid 12 with chloromethyl ketones
(prepared in two steps from the corresponding carboxylic acids)
was followed by cyclization of the intermediate ketoesters by
heating for 5 min at 150 °C (microwave oven) in o-xylene in
the presence of excess ammonium acetate to afford N-Boc
protected imidazole intermediates (14a-f). Deprotection of the
amino function and standard derivatization furnished the desired
final products (15-22, 24). Alternatively, in cases where
esterification via alkylation with halomethyl ketones was
unsuccessful, the final compounds were prepared from advanced
intermediate 25 by Suzuki-Miyaura cross-coupling and imi-
dazole deprotection (Scheme 2).23 On a larger scale, compound
32 was prepared via the route shown in Scheme 3. Ortho-
lithiation of 2-methoxyquinoline and reaction with TBS-
protected hydroxyacetaldehyde gave alcohol 28, which was
oxidized and deprotected under mild conditions to afford
hydroxymethyl ketone 29 in good yield. This intermediate was
coupled to amino acid 12 using EDC, giving rise to a keto ester
intermediate 30 which was treated with excess of ammonium
acetate in refluxing toluene to obtain imidazole intermediate 31.
The azetidine-3-carboxamide moiety was installed by coupling
with N-Boc-azetidine-3-carboxylic acid, Boc-removal, and
reductive alkylation. Direct amide coupling with N-methylaze-
tidine-3-carboxylic acid gave low yields.
a IC50 and CC50 values are averaged from multiple determinations
(n g 2). Standard deviations are <30% of the mean. b r ) rat; d ) dog;
h ) human. c Units: (mL/min)/kg. d Methyl ketone analogue.
with respect to 11 but is more stable in rat liver microsomes.
Unfortunately, its in vivo clearance in rats remains very high
(Cl ) 155 (mL/min)/kg). Isoquinolin-2-yl analogue 16 is 10-
fold more potent than 21 in the proliferation assay but has a
similar high microsomal turnover. Even less stable is the
quinolin-3-yl analogue 17. Quinolin-6-yl analogue 18 shows a
similar stability in rat liver microsomes as 15 and was not tested
in vivo.
After the initial exploration of quinoline isomers our attention
turned to quinolines with additional substituents or naphthyl
replacements containing two heteroatoms.
The 4-methoxy-quinolin-2-yl analogue 19 is largely stabilized
in liver microsomes from all three tested species compared to
11 and 15. This improvement is reflected in reduced clearance
in vivo in rats (Cl ) 36 (mL/min)/kg). Despite this encouraging
With the aim to improve the metabolic stability in rats and
to maintain the good activity of analogues containing a naphthyl
group like 11 we first explored heteroatom containing aryl
groups like quinoline isomers. The results are shown in Table
1 (15-18). Quinolin-2-yl analogue 15 loses 2-fold in activity

Letters
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 11 3455
Scheme 1a
Table 2. Activity of 32, 1, and 3 against HDAC Isoforms (IC50, nM)
and Cancer Cell Lines (CC50, nM)a
isoform
32
1
3
HDAC 1
HDAC 2
HDAC 3
HDAC 4b
HDAC 5
HDAC 6
HDAC 7
HDAC 8
13
18
12
>10 µM
>1 µM
680
30
82
57
>10 µM
>10 µM
43
120
250
400
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
>10 µM
1700
a Reagents and conditions: (a) (i) Cs2CO3, EtOH (ii) 13a-f, DMF, (iii)
o-xylene, NH4OAc, 150 °C, 5 min, microwave; (b) DCM, TFA; (c) EDC-
HCl, HOBt, Et3N, DMF, R2CO2H.
cell line
32
1
3
cervical HeLa
colon HCT-116
lung A549
kidney G401
ovarian A2780
breast MCF-7
leukemia U937
135
220
490
370
740
750
240
460
1800
700
1200
1300
3000
nd
1000
1800
1000
2600
3000
1050
Scheme 2a
nd
a IC50 and CC50 values are averaged from multiple determinations
(n g 2), standard deviations are <30% of the mean; b refers to HDAC 4 wt
as described in ref.14
Table 3. PK Profile of 32
a Reagents and conditions: (a) R1B(OH)2, S-Phos, Pd(OAc)2, K3PO4,
1-butanol, 80 °C; (b) TFA.
rat
19
7.9
7
dog
Cl ((mL/min)/kg)
Vdss (L/kg)
t1/2 (h)
4
3.5
14
31
Scheme 3a
F (%)a
73
a Dosed as L-tartrate salt.
and a 3-fold improvement in the proliferation assay compared
to 22, maintaining good stability in liver microsomes. A further
increase in antiproliferative activity is obtained with the
2-fluoroquinolin-3-yl analogue 26 (CC50 ) 190 nM), and
subsequent replacement of the fluoro substituent by a methoxy
group yields 32 with a CC50 of 135 nM, good liver microsome
stability, and a low clearance when dosed in rats.
The 2-methoxyquinoline 32 was profiled on the various
HDAC isoforms, with 1 and 3 (Table 2). The ketone 32 has
low nanomolar IC50 values against the HDAC class I isoforms
1, 2, and 3 and is inactive on all other tested isoforms with the
exception of HDAC 6 where there is a 30-fold selectivity. In
comparison, 1 displays weaker activity on HDACs 1-3 and
high activity on HDAC 6 (IC50 ) 43 nM). It also displays some
activity against HDAC 8. Compound 3 shows a similar degree
of selectivity for HDACs 1-3 as 32 but is 10- to 30-fold less
active (Table 2).
a Reagents and conditions: (a) mesityllithium, TBSOCH2CHO, THF, 0
°C; (b) DMP, DCM; (c) AcOH/ H2O/ THF, 3:1:1; (d) EDC-HCl, HOBt,
DMAP, DMF, 12; (e) toluene, NH4OAc, reflux; (f) DCM, TFA; (g) EDC-
HCl, HOBt, DMF, N-Boc-azetidine-3-carboxylic acid; (h) H2CO, NaOAc,
MeOH, H2O, NaBH3CN.
In proliferation assays against a wide panel of cancer cell
lines 32 proved to be 3- to 4-fold more active compared to 1
and 2- to 10-fold more active than 3 (Table 2).
result, the activity of 19 was not deemed sufficient (CC50
)
770 nM) and further SAR at R2 was explored with the aim to
improve enzymatic and cellular activity. The N,N-dimethylami-
noethyl analogue 20 is slightly more active in the proliferation
assay but displays a higher clearance in rats (Cl ) 59 (mL/
min)/kg). Compound 21 has a 2-fold increased activity but
unfortunately is unstable in liver microsomes of all species.
Analogue 22 bearing a N-methylazetidin-3-yl as R2 captured
our interest because of its increased stability in liver microsomes
and in rats (Cl ) 20 (mL/min)/kg). At this point we decided to
turn our attention to further SAR exploration of R1, keeping
the N-methylazetidin-3-yl constant.
The pharmacokinetic properties of 32 as the L-tartrate were
determined in rats and dogs (Table 3). In both species low
clearance and long terminal half-life are observed. The oral
bioavailability is excellent in rats and acceptable in dogs. This
is noteworthy, considering the presence of the long and very
flexible alkyl ketone chain (>7 freely rotatable bonds). After
oral doses to rats (3 mg/kg) and dogs (2 mg/kg) good levels of
32 (>100 nM) are detected in plasma out to 24 h. The compound
also displays excellent solubility of 13 mg/mL in buffered
aqueous solution at pH 4.1, suggesting that 32 is suitable both
for iv and oral dosing in man.
The quinoxalin-6-yl analogue 23 demonstrates a 4-fold higher
enzyme activity compared to 22 and very good stability in liver
microsomes. Unfortunately, activity in the proliferation assay
remains in the micromolar range. Quinoxalin-2-yl analogue 24
exhibits a further 5-fold increase in activity against HDAC 1
In a preliminary efficacy study in a subcutaneous HCT116
xenograft mouse model 32 displays comparable antitumor
activity to 1 when dosed 5 days per week at the maximum
tolerated dose (MTD). This activity of 32 is observed at 1/10 of

3456 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 11
Letters
(7) Grant, S.; Easley, C.; Kirkpatrick, P. Vorinostat. Nat. ReV. Drug
DiscoVery 2007, 6, 21–22.
(8) Olsen, E. A.; Kim, Y. H.; Kuzel, T. M.; Pacheco, T. R.; Foss, F. M.;
Parker, S.; Frankel, S. R.; Chen, C.; Ricker, J. L.; Arduino, J. M.;
Duvic, M. Phase IIB multicenter trial of vorinostat in patients with
persistent, progressive, or treatment refractory cutaneous T-cell
lymphoma. J. Clin. Oncol 2007, 25 (21), 3109–3115.
(9) Suzuki, T.; Ando, T.; Tsuchiya, K.; Fukazawa, N.; Saito, A.; Mariko,
Y.; Yamashita, T.; Nakanishi, O. Synthesis and histone deacetylase
inhibitory activity of new benzamide derivatives. J. Med. Chem. 1999,
42, 3001–3003.
(10) Vaisburg, A. Discovery and Development of MGCD0103, an Orally
Active HDAC Inhibitor in Human Clinical Trials. Presented at the
19th International Symposium on Medicinal Chemistry, Istanbul,
Turkey, Aug 2006; Paper L57.
(11) Ueda, H.; Nakajima, H.; Hori, Y.; Fujita, T.; Nishimura, M.; Goto,
T.; Okuhara, M. FR901228, a novel antitumor bicyclic depsipeptide
produced by Chromobacterium Violaceum no. 968. I. Taxonomy,
fermentation, isolation, physico-chemical and biological properties,
and antitumor activity. J. Antibiot. 1994, 47, 301–310.
Figure 3. Human HCT116 colon cancer xenograft model in nude mice
(10 per group). Compounds: 32 (2) 10 mg/kg, 1 (×) 100 mg/kg, and
vehicle (0), administered ip once a day, 5 days per week.
(12) Plumb, J. A.; Finn, P. W.; Williams, R. J.; Bandara, M. J.; Romero,
M. R.; Watkins, C. J.; La Thangue, N. B.; Brown, R. Pharmacody-
namic response and inhibition of growth of human tumor xenografts
by the novel histone deacetylase inhibitor PXD101. Mol. Cancer Ther.
2003, 2, 721–728.
(13) Maiso, P.; Carvajal-Vergara, X.; Ocio, E. M.; Lo´pez-Pe´rez, R.; Mateo,
G.; Gutie´rrez, N.; Atadja, P.; Pandiella, A.; San Miguel, J. F. The
histone deacetylase inhibitor LBH589 is a potent antimyeloma agent
that overcomes drug resistance. Cancer Res. 2006, 66, 5781–5789.
(14) Jones, P.; Altamura, S.; De Francesco, R.; Gonzelez Paz, O.; Kinzel,
O.; Mesiti, G.; Monteagudo, E.; Pescatore, G.; Rowley, M.; Verdirame,
M.; Steinku¨hler, C. A novel series of potent and selective ketone
histone deacetylase inhibitors with antitumer activity in vivo. J. Med.
Chem. 2008, 51, 2350–2353.
(15) Witter, D. J.; Harrington, P.; Wilson, K. J.; Chenard, M.; Fleming,
J. C.; Haines, B.; Kral, A. M.; Secrist, J. P.; Miller, T. A. Optimization
of biaryl selective HDAC1&2 inhibitors (SHI-1:2). Bioorg. Med.
Chem. Lett. 2008, 18 (2), 726–731.
(16) Moradei, O. M.; Mallais, T. C.; Frechette, S.; Paquin, I.; Tessier, P. E.;
Leit, S. M.; Fournel, M.; Bonfils, C.; Trachy-Bourget, M.-C.; Liu, J.;
Yan, T. P.; Lu, A.-H.; Rahil, J.; Wang, J.; Lefevre, S.; Li, Z.; Vaisburg,
A. F.; Besterman, J. M. Novel aminophenyl benzamide-type histone
deacetylase inhibitors with enhanced potency and selectivity. J. Med.
Chem. 2007, 50 (23), 5543–5546.
the dose of 1, and no adverse effects or significant body weight
loss is observed (Figure 3).
Counterscreen on a panel of 150 enzymes or receptors (MDS
Pharma Services) reveals no significant off-target activities, with
32 displaying at least 100-fold selectivity. No significant
CYP3A4 inhibition or induction is found (although weak
inhibition of CYP2C9, IC50 ) 5.5 µM).
When 32 was dosed via continuous infusion to dogs, no major
effect on hemodynamics and cardiac electrical activity was
observed with plasma levels reaching 30 µM. In a microbial
mutagenicity test (Ames) 32 is negative.
In summary, starting from lead 11, optimization of the two
capping groups R1 and R2 of the inhibitors has led to the
identification of 32, a potent, selective HDACs 1, 2, and 3
inhibitor that has excellent pharmacokinetic properties in
preclinical species. The alkyl ketone 32 displays antitumor
activity comparable to that of clinical candidates and is devoid
of off-target activities.
(17) Recent reviews on isoform selective HDAC inhibitors: (a) Kozikowski,
A. P.; Butler, K. V. Chemical origin of selectivity in histone
deacetylase inhibitors. Curr. Pharm. Des. 2008, 14, 505–528. (b)
Suzuki, T.; Itoh, Y.; Miyata, N. Isoform-selective histone deactylase
inhibitors. Curr. Pharm. Des. 2008, 14, 529–544.
Acknowledgment. The authors thank Sergio Serafini,
Ottavia Cecchetti, and Stephane Spieser for their support of this
work.
(18) Sandhu, P.; Andrews, P. A.; Baker, M. P.; Koeplinger, K. A.; Soli,
E. D.; Miller, T.; Baillie, T. A. Disposition of vorinostat, a novel
histone deacetylase inhibitor and anticancer agent, in preclinical
species. Drug Metab. Lett. 2007, 1 (2), 153–161.
(19) 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. Apicidin: a novel antipro-
tozoal agent that inhibits parasite histone deacetylase. Proc. Natl. Acad.
Sci. U.S.A. 1996, 93, 13143–13147.
(20) Jones, P.; Altamura, S.; Chakravarty, P. K.; Cecchetti, O.; De
Francesco, R.; Gallinari, P.; Ingenito, R.; Meinke, P. T.; Petrocchi,
A.; Rowley, M.; Scarpelli, R.; Serafini, S.; Steinkuhler, C. A series of
novel, potent, and selective histone deacetylase inhibitors. Bioorg. Med.
Chem. Lett. 2006, 16, 5948–5952.
(21) Pescatore, G.; Kinzel, O.; Attenni, B.; Cecchetti, O.; Fiore, F.; Fonsi,
M.; Rowley, M.; Schultz-Fademrecht, C.; Serafini, S.; Steinkuehler,
C.; Jones, P. Optimization of a series of potent and selective ketone
histone deacetylase inhibitors. Bioorg. Med. Chem. Lett. 2008, 18 (20),
5528–5532.
Supporting Information Available: Complete experimental
procedures and characterization data. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) Santos-Rosa, H.; Caldas, C. Chromatin modifier enzymes, the histone
code and cancer. Eur. J. Cancer 2005, 41, 2381–2402.
(2) Yoo, C. B.; Jones, P. A. Epigentic therapy of cancer: past, present
and future. Nat. ReV. Drug DiscoVery 2006, 5, 37–50.
(3) Histone Deacetylases, Transcriptional Regulation and Other Cellular
Functions; Verdin, E., Ed.; Humana Press: Totowa, NJ, 2006.
(4) de Ruijter, A. J. M.; van Gennip, A. H.; Caron, H. N.; Kemp, S.; van
Kuilenburg, A. B. P. Histone deacetylases (HDACs): characterization
of the classical HDAC family. Biochem. J. 2003, 370, 737–749.
(5) Bolden, J. E.; Peart, M. J.; Johnstone, R. W. Anticancer activities of
histone deacetylase inhibitors. Nat. ReV. Drug DiscoVery 2006, 5, 769–
784.
(6) Recent review on HDAC inhibitors: Paris, M.; Porcelloni, M.; Binaschi,
M.; Fattori, D. Histone deacetylase inhibitors: from bench to clinic.
J. Med. Chem. 2008, 51 (6), 1505–1529. (b) Rasheed, W.; Bishton,
M.; Johnstone, R. W.; Prince, H. M. Histone deacetylase inhibitors in
lymphoma and solid malignancies. Expert ReV. Anticancer Ther. 2008,
8 (3), 413–432.
(22) Fonsi, M.; Fiore, F.; Jones, P.; Kinzel, O.; Rowley, M.; Laufer, R.;
Monteagudo, E. Unpublished results.
(23) Schultz-Fademrecht, C.; Kinzel, O.; Marko´, I. E.; Pospisil, T.; Pesci,
S.; Rowley, M.; Jones, P. Unpublished results.
JM9004303