An azumamide C analogue without the zinc-binding functionality

Article abstract of DOI:10.1039/c4md00252k

Histone deacetylase (HDAC) inhibitors have attracted considerable attention due to their promise as therapeutic agents. Most HDAC inhibitors adhere to a general cap-linker-Zn²⁺-binding group architecture but recent studies have indicated that potent inhibition may be achieved without a Zn²⁺-coordinating moiety. Herein, we describe the synthesis of an azumamide analogue lacking its native Zn²⁺-binding group and evaluation of its inhibitory activity against recombinant human HDAC1-11. Furthermore, kinetic investigation of the inhibitory mechanism of both parent natural product and synthetic analogue against HDAC3-NCoR2 is reported as well as their activity against Burkitt's lymphoma cell proliferation.

Full text of DOI:10.1039/c4md00252k

MedChemComm
View Article Online
View Journal
CONCISE ARTICLE
An azumamide C analogue without
the zinc-binding functionality†
Cite this: DOI: 10.1039/c4md00252k
Jesper S. Villadsen,a Betul Kitir,‡a Kathrine Wich,‡ Tina Friis,b Andreas S. Madsen‡a
¨
a
*
and Christian A. Olsen‡
Histone deacetylase (HDAC) inhibitors have attracted considerable attention due to their promise as
therapeutic agents. Most HDAC inhibitors adhere to
general “cap-linker-Zn2+-binding group”
architecture but recent studies have indicated that potent inhibition may be achieved without a Zn2+
a
-
coordinating moiety. Herein, we describe the synthesis of an azumamide analogue lacking its native
Zn2+-binding group and evaluation of its inhibitory activity against recombinant human HDAC1–11.
Furthermore, kinetic investigation of the inhibitory mechanism of both parent natural product and
synthetic analogue against HDAC3-NCoR2 is reported as well as their activity against Burkitt's lymphoma
cell proliferation.
Received 13th June 2014
Accepted 11th August 2014
DOI: 10.1039/c4md00252k
www.rsc.org/medchemcomm
propyl group. Not surprisingly, these inhibitors were signi-
cantly less potent than their ethylketone-containing counter-
Introduction
Histone deacetylase (HDAC) enzymes,1,2 or perhaps more
appropriately termed lysine deacylases (KDACs),3–5 have attrac-
ted attention as putative therapeutic targets,6–9 especially due to
their ability to modify the landscape of post-translational
modications (PTMs) on histone proteins in chromatin.6
Indeed, two HDAC inhibitors [i.e., SAHA (vorinostat, marketed
as Zolinza®)10 and romidepsin (FK228, marketed as Istodax®)11]
have received approval by the Food and Drug Administration in
the US for treatment of cutaneous T-cell lymphoma, and several
other compounds are currently in clinical trials for additional
indications.12
parts, but still exhibited Ki-values of just 30–50 nM against
HDACs 1–3.24 Since a wide variety of enzymes rely on metal ions
for catalytic activity, excision of the Zn2+-binding group may
enable development of more selective HDAC inhibitors with
fewer adverse effects.
We found this idea attractive and decided to investigate the
macrocyclic core of the azumamides25–27 (4–8) in a similar
fashion by synthesizing compound 9 (Scheme 1). The azuma-
mides contain relatively weak Zn2+-binding groups (carboxylate
or carboxyamide)28 while still exhibiting high potency as class I
HDAC inhibitors with low nanomolar Ki-values recorded for
In addition to the depsipeptide romidepsin mentioned
above, Nature has provided a variety of potent macrocyclic
HDAC inhibitors such as the trapoxins, apicidins, azumamides,
chlamydocin, and largazole.13–15 The apicidin structure (1)16 has
been the subject of numerous structure–activity relationship
studies,17–23 and recently, analogues without the so-called Zn2+
binding group (the ethylketone moiety) were investigated by
Ghadiri and coworkers (e.g., 2 and 3, Scheme 1).24 The Zn2+
-
-
binding side chain was substituted with various alkyl groups,
and the optimal side chain at this position was a non-branched
aDepartment of Chemistry, Technical University of Denmark, Kemitorvet 207, Kongens
Lyngby, DK-2800, Denmark. E-mail: cao@sund.ku.dk; Tel: +45 45252105
bDepartment of Clinical Biochemistry, Immunology and Genetics (KBIG), Statens
Serum Institut, Artillerivej 5, DK-2300, Copenhagen, Denmark
† Electronic supplementary information (ESI) available: Materials and methods,
supplementary tables and gures, as well as 1H and 13C NMR spectra of the
prepared compounds. See DOI: 10.1039/c4md00252k
‡ Present address: Center for Biopharmaceuticals and Department of Drug Design
and Pharmacology, University of Copenhagen, Universitetsparken 2, DK-2100,
Copenhagen, Denmark.
Scheme 1 Structures of cyclic peptide HDAC inhibitors. Arrows show
the opposite N / C directionality of apicidin and azumamide.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun.

View Article Online
MedChemComm
Concise Article
azumamides C (6) and E (8).29 Since azumamide C was slightly
more potent than azumamide E, we substituted the Zn2+
-
binding side chain of the b-amino acid in compound 6 with a
propyl group to give compound 9. In addition to removal of the
metal-coordinating capability, removal of this negatively
charged carboxylate, may furthermore improve cell perme-
ability of the ligand. We here present the synthesis and evalu-
ation of compound 9 as an HDAC inhibitor.
Scheme 3 Reagents and conditions: (a) HCl (4M in dioxane, 30 equiv.),
MeOH, rt, 22 h; (b) thionyl chloride (3.0 equiv.), MeOH, 0 ꢀC / reflux,
23 h, then thionyl chloride (1.5 equiv.), MeOH, 0 ꢀC / rt, 22 h; (c)
EtOAc, saturated aqueous NaHCO3, benzylchloroformate, rt, 18 h.
Results and discussion
Chemistry
reagent afforded target peptide 9 (Scheme 4). The relatively low
yield is similar to those previously achieved for similar
compounds21–24 and may, at least in part, be attributed to poor
solubility in the water–acetonitrile mobile phase used in
reversed-phase HPLC purication. However, further optimiza-
tion of cyclization efficiency, which is known to be notoriously
troublesome for small peptides,33 may also prove worthwhile.
The three D-amino acids necessary to prepare compound 9 were
purchased and the b-amino acid building block (14, Scheme 2)
was prepared using chiral sulnylimine 12, which was obtained
by condensation of butanal (10) and (S)-2-methylpropane-2-
sulnamide (“Ellman's auxiliary”, 11).30,31 The diastereo-
selective Mannich reaction between imine (12) and the Z-eno-
late of tert-butyl propionate provided the (2S,3R)-b-amino ester
(13) as the major diastereoisomer. Acid mediated cleavage of
the tert-butyl ester and the sulnyl group with both TFA and HCl
to ensure full conversion, afforded the fully deprotected b-
amino acid and subsequent treatment with Fmoc-OSu fur-
nished the desired b-amino acid building block (14).
Unfortunately, we were unable to obtain crystals for deter-
mination of the absolute stereochemistry by X-ray crystallog-
raphy. Thus, the stereochemistry was instead conrmed by
conversion of sulnyl-protected tert-butyl ester 13 to the previ-
ously reported Cbz-protected methyl ester (15, Scheme 3) and
both optical rotation and NMR data were in full agreement with
the previously reported data32 (see also ESI, Table S1†).
Proling of HDAC inhibitory activity
We rst screened the new compound along with a non-cyclized
azumamide C analogue (for structure, see ESI, Fig. S1†), and
SAHA as internal control against the full panel of recombinant
human HDACs 1–11 applying a single inhibitor concentration
(ESI, Fig. S1†). The linear peptide was included to address the
importance of the macrocyclic nature of the azumamides and,
somewhat expectedly, the linear compound was not inhibiting
any of the HDACs potently. Poor inhibitory activities were also
observed for macrocycle 9 against class I HDAC8, class IIa
HDACs (4, 5, 7, and 9), and class IIb HDAC6, and it was there-
fore decided to proceed with full dose–response proling only
against HDACs 1–3, 10, and 11 (ESI, Fig. S2†).
To be able to compare potencies irrespective of applied
substrate identity and concentration in the assays, we deter-
mined Ki-values using the Cheng–Prusoff equation34 and the
Km-values for the selected substrate–enzyme combinations
(Table 1).29,35,36 Not surprisingly, compound 9 was less potent
than its parent compound, azumamide C.
Coupling of the b-amino acid (14) to a resin-bound tripeptide
[H–D-Val–D-Ala–D-Tyr–(2-chlorotrityl-resin)] using HATU, fol-
lowed by removal of the Fmoc group and cleavage from the resin
afforded linear tetrapeptide 16. Macrolactamization, under
dilute conditions (0.4 mM in DMF), using HATU as the coupling
Still, compound 9 exhibited low micromolar inhibition of
class I HDACs even without the ability to coordinate to Zn2+ in the
active site. Interestingly, the edited analogue (9) was either
equipotent or slightly more potent than the corresponding car-
boxamide-containing natural product azumamide B (Table 1 and
ESI, Fig. S2†). The azumamide core, however, could not compete
with the most potent compounds from the more elaborate series
Scheme 2 Reagents and conditions: (a) CuSO4 (5.0 equiv.), anhydrous
CH2Cl2, rt, 20 h; (b) tert-butyl propionate (2.5 equiv.), LDA (2.6 equiv.),
HMPA (5.7 equiv.), anhydrous THF, –78 ꢀC, 30 min; then imine 12; (c)
TFA–CH2Cl2 (1 : 1), rt, 5 h; (d) HCl (4.0 M in dioxane, 3.0 equiv.),
dioxane, rt, 1.5 h; (e) Na2CO3 (4.0 equiv.), Fmoc-OSu (1.2 equiv.), DMF– Scheme 4 Macrolactamization to give compound 9 (7%, based on
H2O, 0 ꢀC / rt, 3 h.
resin loading).
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
Table 1 Ki-values (mM) against HDAC isoforms from different classesa
Class II
Class IV
HDAC11
Class I
Compd
HDAC1
HDAC2
HDAC3b
HDAC10
9
2.4 ꢁ 0.8
5c
1.4 ꢁ 0.7
3c
3.0 ꢁ 0.1
3c
4 ꢁ 1
6 ꢁ 1
AzuB (5)
AzuC (6)
SAHA
—
>5c
0.02 ꢁ 0.01
0.01 ꢁ 0.006
0.018 ꢁ 0.001
0.02 ꢁ 0.005
0.06 ꢁ 0.008
0.01 ꢁ 0.006
0.008 ꢁ 0.004
0.022 ꢁ 0.001
0.04 ꢁ 0.02
0.06 ꢁ 0.02
a
Values were calculated from at least two individual dose–response experiments performed in duplicate using the Cheng–Prusoff equation.
Recombinant human HDAC3 was purchased as a complex with the deacetylase-activating domain of nuclear receptor co-repressor 2 (NCoR2).
From a previous publication.29
b
c
of cyclotetrapeptide HDAC inhibitors without Zn2+-binding investigate whether our compounds could inuence prolifera-
groups previously reported.24
tion of lymphoma cells, we tested their effect on the Epstein Barr
virus (EBV)-infected human Burkitt's lymphoma cell line, EB-3.
Kinetic evaluation of HDACi activity
Since compound 9 was relatively potent against HDAC3 and
since HDAC3 was the only enzyme in complex with its co-
repressor,37 which may affect the inhibitor potency,38 we chose
to investigate this interaction in more detail by kinetic experi-
ments. Performing progression curve inhibition experiments
where the enzymatic catalysis is followed over time in the
presence of various inhibitor concentrations may reveal
insights regarding the mechanism of inhibition. A recent study
demonstrated a slow-binding mechanism of benzamide-based
inhibitors against HDAC3-NCoR2,39 that would not be revealed
by simple dose–response experiments performed in an
endpoint fashion. We thus performed similar experiments to
investigate the mechanism of inhibition exhibited by both
azumamide C (6) and compound 9, with SAHA included as a
control compound (Fig. 1).
In all cases, the linear progression curves revealed a constant
rate of substrate conversion over time, which indicate a stan-
dard fast-on–fast-off mechanism of inhibition. This is in
agreement with previous ndings for SAHA39 and non-Zn2+
-
binding macrocycles,24 but to the best of our knowledge, this is
the rst investigation of the mechanism of inhibition by azu-
mamide natural products.
The Ki-values derived from tting the continuous assay data
were generally 2–4-fold lower than those obtained by applying
the Cheng–Prusoff equation to simple dose–response results
(Table 1). The respective values were in the same range,
however, and the relationship between compound potencies
showed the same trend. Thus, we nd the dose–response/
Cheng–Prusoff method to be sufficient for determination of Ki-
values of compounds with a known fast-on–fast-off inhibition
mechanism.
Effect on cultured cancer cells
Epigenetic silencing of the proapototic Bcl-2-family gene, Bim,
in Burkitt's lymphoma cells by concurrent promoter hyper-
methylation and deacetylation has been demonstrated to be
involved in the mechanism underlying the common chemo-
resistance of Burkitt's lymphoma. This chemoresistance has
Fig. 1 Progression curves and data fitting of the inhibition of HDAC3-
NCoR2. Data were fitted to the equation, vi/v0 ¼ (1 + [I]/(Ki(1 + [S]/
Km)))ꢂ1, using the GraphPad Prism software to derive the Ki-values
been reversed by treatment with the HDAC inhibitor SAHA.40 To shown in the figure above.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun.

View Article Online
MedChemComm
Concise Article
To assess the effects of azumamides B (5) and C (6) as well as CH2Cl2 (7.0 mL). Aer stirring vigorously for 20 hours the
compound 9 along with positive controls SAHA and apicidin, we reaction mixture was ltered through a pad of Celite and the
used the MTT assay, in which cleavage of yellow MTT affords lter cake washed with CH2Cl2. The combined organic phases
purple formazan crystals in functioning mitochondria to give a were concentrated in vacuo and purication by vacuum silica gel
measure of cell viability. Initially, we screened all ve chromatography afforded tert-butanesulnyl imine 12 (0.72 g,
compounds in triplicate at 10 mM and only the positive controls, 92%) as a colorless oil. 1H NMR (300 MHz, CDCl3) d 8.03 (t, J ¼
SAHA and apicidin inhibited cell proliferation. We therefore 4.7 Hz, 1H), 2.46 (td, J ¼ 7.3, 4.7 Hz, 2H), 1.63 (h, J ¼ 7.4 Hz, 2H),
tested the compounds starting at higher concentrations of the 1.16 (s, 9H), 0.95 (t, J ¼ 7.4 Hz, 3H); 13C NMR (75 MHz, CDCl3) d
macrocycles. Since all compound stock solutions were prepared 169.7, 56.5, 38.1, 22.5, 19.0, 13.9. Spectral data were in agree-
in DMSO due to solubility issues, we tested 2-fold dilutions of ment with previously reported data.41
DMSO (from 1%) in parallel with the 2-fold serial dilutions of
(2S,3R)-tert-Butyl
3-((S)-1,1-dimethylethylsulnamido)-2-
the compounds. Based on this, SAHA and apicidin exhibited methylhexanoate (13). A solution of LDA (1.5 M in THF, 5.0 mL,
GI50 values of 2.8 mM and 0.8 mM, respectively. For apicidin, this 7.5 mmol, 2.6 equiv.) was added dropwise over 5 min to a
correlates well with values previously obtained against HeLa, solution of HMPA (1.5 mL, 8.6 mmol, 5.7 equiv.) in dry THF (5.0
Hct-116, MCF-7, KYO-1, and K-562 cells.23 The EB-3 cell prolif- mL) at ꢂ78 C. Aer 15 min, tert-butyl propionate (1.1 mL, 7.3
ꢀ
eration was inhibited at 1% but not at 0.5% DMSO and this mmol, 2.5 equiv.) was added dropwise and aer stirring for
resulted in a maximal reliable concentration of azumamide C of additionally 30 min imine 12 (0.51 g, 2.90 mmol) in anhydrous
25 mM while it was 50 mM for azumamide B and 9. For all three THF (2.0 mL) was added dropwise. Aer 2 hours the reaction
compounds the GI50 values were deemed well above this limit as was quenched with sat. aqueous NH4Cl and warmed to room
no reproducible growth inhibition was observed for any of the temperature. Aqueous HCl (1 M, 20 mL) was added and the
compounds.
mixture extracted with EtOAc (2 ꢃ 60 mL). The combined
organic phases were washed with brine (40 mL), dried (MgSO4),
ltered, and concentrated in vacuo. Diastereoselectivity was
determined by 1H NMR integration of the crude reaction
Conclusions
In the present study, we have synthesized an analogue of the mixture (65 : 25 : 8 : 2). Purication by silica gel chromatog-
naturally occurring HDAC inhibitor, azumamide C, which is raphy afforded b-amino ester 13 (335 mg, 38%) as needle-like
structurally edited to prevent coordination to the zinc ion crystals. 1H NMR (300 MHz, CDCl3) d 3.52 (m, 1H), 3.13 (d, J ¼
present in the HDAC enzyme active sites. This allowed us to 8.1 Hz, 1H), 2.46 (m, 1H), 1.60–1.46 (m, 3H), 1.42 (m, 10H), 1.18
elucidate the binding affinity of the natural product's macro- (s, 9H), 1.07 (d, J ¼ 7.1 Hz, 3H), 0.91 (t, J ¼ 6.9 Hz, 3H); 13C NMR
cyclic core, which proved to be in the low- to sub-micromolar (75 MHz, CDCl3) d 173.8, 80.7, 59.0, 56.3, 45.2, 36.6, 28.2, 22.9,
range. We furthermore, showed that both natural product and 19.4, 13.9, 11.8. HRMS (ESI-TOF) calcd for C15H31NO3H+ [M +
its analogue inhibit HDAC3-NCoR2 via
mechanism by performing continuous assay experiments.
a
fast-on–fast-off H]+ 306.2103, found 306.2110.
(2S,3R)-3-Fmoc-amino-2-methylhexanoic acid (14). TFA (5.0
Not surprisingly, this non-Zn2+-binding analogue was mL) was added to a solution of b-amino ester 13 (224 mg, 0.73
signicantly less potent than its carboxylate-containing parent mmol) in anhydrous CH2Cl2 (5.0 mL). Aer 25 hours the reac-
compound against HDACs. However, it is quite remarkable that tion mixture was concentrated and residual TFA was removed
our synthetic analogue was at least as potent as the corre- by co-evaporation with toluene (3 ꢃ 10 mL). The resulting
sponding carboxamide-containing natural product azumamide residue was dissolved in dioxane (5.0 mL) and HCl in dioxane (4
B (5). It is puzzling that the marine sponge, from where the M, 0.55 mL, 2.2 mmol, 3.0 equiv.) was added. Aer 1.5 hours the
natural products were originally isolated,26 produces these car- mixture was concentrated. The crude fully deprotected b-amino
boxamide-containing compounds that seem to adhere to the acid was suspended in water (5.0 mL) at 0 ꢀC and Na2CO3 (311
general pharmacophore for HDAC inhibition, but apparently mg, 2.93 mmol, 4.0 equiv.) was added followed by Fmoc-O-
contain a redundant Zn2+-binding group. We nally tested the succinimide (297 mg, 0.88 mmol, 1.2 equiv.) in DMF (3.0 mL).
compounds' ability to inhibit proliferation of EB-3 cells in vitro. Cooling was removed and DMF (2.0 mL) was added. Aer 1.5
Contrary to both SAHA and apicidin none of the tested azu- hours, water (5.0 mL) was added and aer additionally 1.5
mamide analogues were potent growth inhibitors, which hours of stirring the reaction mixture was diluted with water (60
underscores that a ne-tuned combination of HDACi potency mL) and extracted with Et2O (15 mL). Aer acidication with
and cell permeability is required for activity in cells.
concentrated HCl, until pH z 1–2, the aqueous layer was
extracted with EtOAc (3 ꢃ 75 mL). The combined organic pha-
ses were dried (MgSO4), ltered, and concentrated in vacuo.
Purication of the crude residue by vacuum silica gel chroma-
tography afforded b-amino acid 14 (66 mg, 25%, 3 steps) as a
Experimental
Synthetic procedures
(S,E)-N-Butylidene-2-methylpropane-2-sulnamide (12).
A
white solid aer co-evaporation of residual acetic acid with
1
solution of butyraldehyde (0.80 mL, 8.91 mmol, 2.0 equiv.) in CH2Cl2–toluene (1 : 1, 3 ꢃ 10 mL); H NMR (300 MHz, DMSO-
anhydrous CH2Cl2 (1.0 mL) was added dropwise to a solution of d6) d 12.19 (s, 1H), 7.86 (d, J ¼ 7.4 Hz, 2H), 7.66 (dd, J ¼ 7.3, 2.9
anhydrous CuSO4 (3.6 g, 22.6 mmol, 5.0 equiv.) and (S)-(ꢂ)-2- Hz, 2H), 7.38 (t, J ¼ 7.4 Hz, 2H), 7.30 (d, J ¼ 7.4, 2H), 7.05 (d, J ¼
methyl-2-propanesulnamide (0.54 g, 4.46 mmol) in anhydrous 9.5 Hz, 1H), 4.31 (m, 2H), 4.18 (t, J ¼ 6.7 Hz, 1H), 3.58 (m, 1H),
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
2.28 (m, 1H), 1.22 (m, 4H), 0.95 (d, J ¼ 6.9 Hz, 3H), 0.79 (t, J ¼ 6.9 ether afforded the TFA salt of the linear peptide 16 (53 mg, 81%
Hz, 3H); 13C NMR (75 MHz, DMSO-d6) d 176.2, 156.2, 143.9, from 2-chlorotrityl chloride resin, based on theoretical loading
140.8, 127.6, 127.1, 127.0, 125.2 (2C), 120.1, 65.0, 52.4, 46.9, of resin) as a white solid. UPLC-MS, tR ¼ 0.84 min, calcd for
44.4, 35.0, 18.8, 14.0, 13.7. HRMS (ESI-TOF) calcd for
22H25NO4H+ [M + H]+ 368.1862, found 368.1863.
C
24H38N4O6H+ [M + H]+ 479.3, found 479.3; 1H NMR (400 MHz,
MeOD) d 7.04 (d, J ¼ 8.2 Hz, 2H), 6.69 (d, J ¼ 8.2 Hz, 2H), 4.58
C
(2S,3R)-Methyl 3-(((benzyloxy)carbonyl)amino)-2-methylhex- (m, 1H), 4.38 (m, 1H), 4.15 (d, J ¼ 7.1 Hz, 1H), 3.38 (m, 1H), 3.07
anoate (15). HCl in dioxane (4 M, 4.7 mL, 18.8 mmol, 30 equiv.) (dd, J ¼ 14.0, 5.3 Hz, 1H), 2.92 (dd, J ¼ 14.0, 7.8 Hz, 1H), 2.79 (m,
was added to a solution of b-amino ester 13 (190 mg, 0.62 mmol) 1H), 2.05 (octet, J ¼ 6.9 Hz, 1H), 1.60 (q, J ¼7.8 Hz, 2H), 1.54–
in MeOH (5.0 mL). Aer stirring for 22 hours the mixture was 1.39 (m, 2H), 1.33 (d, J ¼ 7.1 Hz, 3H), 1.21 (d, J ¼ 7.2 Hz, 3H),
concentrated. 1H NMR of the crude residue showed a mixture of 0.99 (t, J ¼ 7.3 Hz, 3H), 0.93 (d, J ¼ 6.8 Hz, 6H).
the fully deprotected b-amino acid and amine deprotected
Cyclic peptide 9. Linear peptide 16 (50 mg, 0.084 mmol) was
methyl ester. To ensure full conversion to the methyl ester the dissolved in DMF (200 mL z 0.4 mM) and iPr2NEt (73 mL, 0.42
crude product was dissolved in MeOH (5.0 mL) under Ar, cooled mmol, 5.0 equiv.) and HATU (48 mg, 0.13 mmol, 1.5 equiv.) were
to 0 ꢀC, and thionyl chloride (70 mL, 0.94 mmol, 1.5 equiv.) was added. Aer stirring for 19 hours the reaction mixture was
added dropwise. Aer 10 min the reaction mixture was heated concentrated in vacuo. The residue was taken up in CH2Cl2
under reux for 20 hours before additional thionyl chloride (80 mL) and washed with aqueous HCl (1 M, 2 ꢃ 15 mL) and the
(70 mL, 0.94 mmol, 1.5 equiv.) was added. Aer further 2.5 hours aqueous phase was re-extracted with CH2Cl2 (50 mL). The
of stirring the mixture was concentrated and redissolved in combined organics were washed with brine (25 mL), dried
MeOH (5.0 mL). Thionyl chloride (90 mL, 1.23 mmol, 2.0 equiv.) (MgSO4), ltered, and concentrated in vacuo. The resulting
was added at 0 ꢀC and the mixture stirred for 22 hours at room residue was dissolved in DMF (2.5 mL) and puried by
temperature, before concentration afforded the crude amine. preparative HPLC on a [250 mm ꢃ 20 mm, C18 Phenomenex
˚
The crude amine was dissolved in sat. aqueous NaHCO3–EtOAc Luna column (5 mm, 100 A)] using an Agilent 1260 LC system to
(1 : 1, 6.0 mL) and benzylchloroformate (133 mL, 0.93 mmol, 1.5 afford the cyclic peptide 9 (2.7 mg, 7%) as a white solid. A
equiv.) was added. Aer 18 hours of vigorous stirring the phases gradient with eluent III (95 : 5 : 0.1, water–MeCN–TFA) and
were separated and the aqueous layer was diluted with water eluent IV (0.1% TFA in acetonitrile) rising linearly from 0% to
(5 mL) and extracted with ethyl acetate (2 ꢃ 5 mL). The 95% of eluent IV during t ¼ 5–45 min was applied at a ow rate
combined organic phases were washed with aqueous HCl (1 M, of 20 mL minꢂ1. Two conformations are observed in the 1H
6 mL) and brine (6 mL), dried (Na2SO4), ltered, and concen- NMR spectrum. The conformations are present in an 88 : 12
trated in vacuo. Purication by vacuum silica gel chromatog- ratio. Characterization is given for the major conformation. 1H
raphy afforded the b-amino ester 15 (37%, three steps from 13). NMR (500 MHz, DMSO-d6) d 9.19 (br s, 1H), 7.67 (d, J ¼ 8.8 Hz,
[a]2D0 ꢂ39ꢀ (CHCl3, c 1.0), previously reported [a]D20 ꢂ34ꢀ (CHCl3, 1H), 7.55 (d, J ¼ 9.0 Hz, 1H), 7.46 (d, J ¼ 8.4 Hz, 1H), 6.96 (m,
c 1.0);32 1H NMR (400 MHz, CDCl3) d 7.38–7.28 (m, 5H), 5.09 (s, 3H), 6.63 (d, J ¼ 8.4 Hz, 2H), 4.14 (p, J ¼ 7.3 Hz, 1H), 4.07 (p, J ¼
2H), 4.95 (d, J ¼ 9.7 Hz, 1H), 3.93–3.82 (m, 1H), 3.67 (s, 3H), 2.65 8.8 Hz, 1H), 3.99 (m, 1H), 3.67 (t, J ¼ 9.4 Hz 1H), 2.89 (dd, J ¼
(m, 1H), 1.43 (m, 2H), 1.31 (m, 2H), 1.15 (d, J ¼ 7.2 Hz, 3H), 1.15 13.7, 7.2 Hz, 1H), 2.81 (dd, J ¼ 13.7, 9.0 Hz, 1H), 2.54 (m, 1H),
(d, J ¼ 7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3) d 175.0, 156.2, 2.16 (m, 1H), 1.64 (m, 1H), 1.36 (m, 1H), 1.25–1.15 (m, 8H), 0.91–
136.7, 128.6, 128.2 (2C), 66.8, 53.2, 51.9, 44.1, 34.2, 19.6, 13.9, 0.77 (m, 9H); HRMS (ESI-TOF) calcd for C24H36N4O5H+ [M + H]+
13.2. The spectral data were in full accordance with those 461.2719, found 461.2754; HPLC gradient B, tR ¼ 11.14 min
reported in the literature (see ESI, Table S1†).32
(>95%).
Linear peptide 16. Polystyrene 2-chlorotrityl-bound Fmoc-D-
Val–D-Ala–D-Tyr (140 mg, 0.11 mmol) was added to a fritted
syringe and the Fmoc group was removed with piperidine–DMF
(1 : 4, 4 mL, 2 ꢃ 30 min) and DBU–piperidine–DMF (2 : 2 : 96, 4
Biochemical proling
In vitro histone deacetylase inhibition assays. For assaying,
mL, 30 min). The resin was then washed with DMF (ꢃ3), MeOH peptides were reconstituted in DMSO to give 5–10 mM stock
(ꢃ3), and CH2Cl2 (ꢃ3). b-Amino acid 14 (43 mg, 0.12 mmol, 1.1 solutions, the accurate concentrations of which were deter-
equiv.) in DMF (2.0 mL) was preincubated for 5 min with 2,6- mined by UV using the extinction coefficient for tyrosine at 280
lutidine (37 mL, 0.32 mmol, 3.0 equiv.) and HATU (61 mg, 0.16 nm; 3 ¼ 1280 Mꢂ1 ꢃ cmꢂ1. Inhibition of recombinant human
mmol, 1.5 equiv.) before addition to the resin and the reaction HDACs in dose–response experiments with internal controls
was allowed to proceed on a rocking table for 17 hours. Aer was measured in black low binding Corning half-area 96-well
washing with DMF (ꢃ3), MeOH (ꢃ3), and CH2Cl2 (ꢃ3) the Fmoc microtiter plates. The appropriate dilution of inhibitor (5 mL of
group was removed with piperidine–DMF (1 : 4, 4 mL, 2 ꢃ 30 5ꢃ the desired nal concentration, prepared from 5–10 mM
min) and DBU–piperidine–DMF (2 : 2 : 96, 4 mL, 30 min). The DMSO stock solutions) was added to each well followed by
resin was then washed again and treated with TFA–CH2Cl2 substrate in HDAC assay buffer (10 mL). Ac–Leu–Gly–Lys (Ac)–
(1 : 1, 4 mL) for 30 min followed by washing with CH2Cl2 (5 mL). AMC was used at a nal concentration of 20 mM for HDAC1, 2, 3,
A fresh portion of TFA–CH2Cl2 (1 : 1, 4 mL) was added to the 6, and 11. Ac–Leu–Gly–Lys (Tfa)–AMC was used at a nal
resin and aer additional 30 min the resin was drained and all concentration of 20 mM for HDAC4, 120 mM for HDAC5, 40 mM
the fractions were pooled and concentrated in vacuo to provide for HDAC7 and 200 mM for HDAC8, and 80 mM for HDAC9. Ac–
the linear peptide as an oily residue. Trituration with diethyl Arg–His–Lys (Ac)–Lys (Ac)–AMC was used at
a
nal
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun.

View Article Online
MedChemComm
Concise Article
concentration of 5 mM for HDAC10. Finally, a freshly prepared diphenyltetrazolium bromide (MTT, 50 mg) in PBS (10 mL, both
solution of the appropriate HDAC (10 mL) was added and the supplied in the MTT assay kit) and 10 mL of this solution was
plate (containing a nal volume of 25 mL in each well) was added to each well. The cells were incubated for another 4 h at
incubated at 37 ꢀC for 30 min. The nal concentrations of 37 ꢀC in humid 5% CO2 atmosphere for development to take
enzyme was as follows: HDꢂA1C1: 1 ng mLꢂ1, HDAC2: 0.5 ng mLꢂ1
,
place. The purple formazan crystals produced were dissolved by
HDAC3-NCoR2: 0.09 ng mL , HDAC4: 0.04 ng mLꢂ1, HDAC5: 0.2 addition of isopropyl alcohol (100 mL per well) containing HCl
ng mLꢂ1, HDAC6: 2.4 ng mLꢂ1, HDAC7: 0.04 ng mLꢂ1, HDAC8: 0.2 (0.04 M), and the absorbance was measured at 570 nm with
ng mLꢂ1, HDAC9: 0.8 ng mLꢂ1, HDAC10: 4 ng mLꢂ1, HDAC11: 10 background subtraction at 630 nm on a Tecan Sunrise ELISA
ng mLꢂ1. Then trypsin (25 mL, 0.4 mg mLꢂ1) was added and the plate reader (Tecan, Switzerland). All assays were performed three
assay development was allowed to proceed for 15–30 min at times in duplicate.
room temperature, before the plate was read using a Perkin
Elmer Enspire plate reader with excitation at 360 nm and
detecting emission at 460 nm. Each assay was performed in
Acknowledgements
¨
duplicate and repeated at least twice. The data were analyzed by We thank Ms Tina Gustafsson (DTU Chemistry) and Esin Guven
non-linear regression using GraphPad Prism to afford IC50 (KBIG, SSI) for technical assistance. We gratefully acknowledge
values from the dose–response experiments, and Ki values were nancial support from the Danish Independent Research
determined from the Cheng–Prusoff equation, Ki ¼ IC50/(1 + [S]/ Council–Technology and Production Sciences (Sapere Aude
Km), assuming a standard fast-on–fast-off mechanism of inhi- Grant no. 12-132328; A.S.M.), the Danish Independent Research
bition. The applied Km values were previously reported for Council–Natural Sciences (Steno Grant no. 10-080907; C.A.O.),
HDACs 1–3,35 HDAC10,29 and HDAC11.36
the Carlsberg Foundation (C.A.O.), and the Lundbeck Founda-
Continuous assay protocol. The continuous assays of tion (Young Group Leader Fellowship to C.A.O.).
HDAC3-NCoR2 inhibition were performed in 96-well, white,
medium binding, half-area microtiter plates (Greiner Bio One).
The 2- or 3-fold dilution series of inhibitors were prepared as
References
described above. HDAC assay buffer (10 mL) containing
substrate Ac–Leu–Gly–Lys (Ac)–AMC (100 mM; nal concentra-
tion 20 mM) and HDAC assay buffer (10 mL) containing trypsin
(12.5 ng mLꢂ1; nal concentration 2.5 ng mLꢂ1) was added to
each well, followed by the appropriate dilution of inhibitor (20
mL of 2.5ꢃ the desired nal concentration). Finally, a solution of
HDAC3-NCoR2 (10 mL, 0.2 ng mLꢂ1 ¼ 2.5 nM; nal concentra-
tion 0.04 ng mLꢂ1 ¼ 0.5 nM) was added, and uorescence
readings were recorded every 30 seconds for 45 min at 25 ꢀC
using a Perkin Elmer Enspire plate reader with excitation at 360
nm and detecting emission at 460 nm. Each assay was per-
formed twice in duplicate. The rate constants (v) were deter-
mined by linear regression analysis using GraphPad Prism. Ki
values were then determined by non-linear tting to the
1 I. V. Gregoretti, Y. M. Lee and H. V. Goodson, J. Mol. Biol.,
2004, 338, 17–31.
2 G. Blander and L. Guarente, Annu. Rev. Biochem., 2004, 73,
417–435.
3 M. D. Hirschey, Cell Metab., 2011, 14, 718–719.
4 C. A. Olsen, Angew. Chem., Int. Ed., 2012, 51, 3755–3756.
5 C. A. Olsen, ChemMedChem, 2014, 9, 434–437.
6 C. H. Arrowsmith, C. Bountra, P. V. Fish, K. Lee and
M. Schapira, Nat. Rev. Drug Discovery, 2012, 11, 384–400.
7 M. C. Haigis and D. A. Sinclair, Annu. Rev. Pathol.: Mech. Dis.,
2010, 5, 253–295.
8 M. Haberland, R. L. Montgomery and E. N. Olson, Nat. Rev.
Genet., 2009, 10, 32–42.
9 A. G. Kazantsev and L. M. Thompson, Nat. Rev. Drug
Discovery, 2008, 7, 854–868.
following equation, vi/v0 ¼ (1 + [I]/(Ki(1 + [S]/Km)))ꢂ1
.
In vitro cell proliferation assay. Cell proliferation was assessed 10 P. A. Marks and R. Breslow, Nat. Biotechnol., 2007, 25, 84–90.
using the MTT cell growth kit (Millipore, Billerica, MA, USA). EB-3 11 H. M. Prince and M. Dickinson, Clin. Cancer Res., 2012, 18,
cells (ATCC, Manassas, VA, USA) were seeded in 96-well at-
3509–3515.
bottom cell culture plates (Nunc, Thermo Fischer Scientic, 12 A. C. West and R. W. Johnstone, J. Clin. Invest., 2014, 124, 30–
Roskilde, Denmark) with a density of 2 ꢃ 104 cells in 90 mL
39.
culture media composed of RPMI-1640 (ATCC, Manassas, VA, 13 H. Rajak, A. Singh, P. K. Dewangan, V. Patel, D. K. Jain,
USA) supplemented with 10% fetal calf serum (FCS) (Sera Scan-
dia, Hellerup, Denmark) and 1% penicillin/streptomycin (Gibco,
S. K. Tiwari, R. Veerasamy and P. C. Sharma, Curr. Med.
Chem., 2013, 20, 1887–1903.
¨
Invitrogen, Taastrup, Denmark). Aer cultivation overnight at 37 14 F. F. Wagner, M. Weıwer, M. C. Lewis and E. B. Holson,
ꢀC in humid 5% CO2 atmosphere, 10 mL of culture medium
Neurotherapeutics, 2013, 10, 589–604.
containing appropriate concentrations of the ve compounds or 15 P. Bertrand, Eur. J. Med. Chem., 2010, 45, 2095–2116.
the vehicle (DMSO) as control were added. The 2-fold dilution 16 S. J. Darkin-Rattray, A. M. Gurnett, R. W. Myers, P. M. Dulski,
series were prepared from DMSO stock solutions (5 mM for
azumamide C and 10 mM for the remaining compounds) to give
nal assay concentrations starting from 50 mM for azumamide C,
SAHA, and apicidin while 100 mM was applied for azumamide B
T. M. Crumley, J. J. Allocco, C. Cannova, P. T. Meinke,
S. L. Colletti, M. A. Bednarek, S. B. Singh, M. A. Goetz,
A. W. Dombrowski, J. D. Polishook and D. M. Schmatz,
Proc. Natl. Acad. Sci. U. S. A., 1996, 93, 13143–13147.
and compound 9. Aer incubation for 3 days, MTT solution was 17 P. T. Meinke, S. L. Colletti, G. Doss, R. W. Myers,
freshly prepared by dissolving 3-(4,5-dimethylthiazol-2-yl)-2,5-
A. M. Gurnett, P. M. Dulski, S. J. Darkin-Rattray,
Med. Chem. Commun.
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
J. J. Allocco, S. Galuska, D. M. Schmatz, M. J. Wyvratt and 29 J. S. Villadsen, H. M. Stephansen, A. R. Maolanon, P. Harris
M. H. Fisher, J. Med. Chem., 2000, 43, 4919–4922. and C. A. Olsen, J. Med. Chem., 2013, 56, 6512–6520.
18 S. L. Colletti, R. W. Myers, S. J. Darkin-Rattray, A. M. Gurnett, 30 T. P. Tang and J. A. Ellman, J. Org. Chem., 2002, 67, 7819–
P. M. Dulski, S. Galuska, J. J. Allocco, M. B. Ayer, C. Li, J. Lim, 7832.
T. M. Crumley, C. Cannova, D. M. Schmatz, M. J. Wyvratt, 31 M. T. Robak, M. A. Herbage and J. A. Ellman, Chem. Rev.,
M. H. Fisher and P. T. Meinke, Bioorg. Med. Chem. Lett.,
2001, 11, 107–111.
19 S. L. Colletti, R. W. Myers, S. J. Darkin-Rattray, A. M. Gurnett,
2010, 110, 3600–3740.
32 J. E. Wilson, A. D. Casarez and D. W. MacMillan, J. Am. Chem.
Soc., 2009, 131, 11332–11334.
P. M. Dulski, S. Galuska, J. J. Allocco, M. B. Ayer, C. Li, J. Lim, 33 C. J. White and A. K. Yudin, Nat. Chem., 2011, 3, 509–524.
T. M. Crumley, C. Cannova, D. M. Schmatz, M. J. Wyvratt, 34 Y. Cheng and W. H. Prusoff, Biochem. Pharmacol., 1973, 22,
M. H. Fisher and P. T. Meinke, Bioorg. Med. Chem. Lett.,
2001, 11, 113–117.
20 W. S. Horne, C. A. Olsen, J. M. Beierle, A. Montero and
M. R. Ghadiri, Angew. Chem., Int. Ed., 2009, 48, 4718–4724.
3099–3108.
35 J. E. Bradner, N. West, M. L. Grachan, E. F. Greenberg,
S. J. Haggarty, T. Warnow and R. Mazitschek, Nat. Chem.
Biol., 2010, 6, 238–243.
21 A. Montero, J. M. Beierle, C. A. Olsen and M. R. Ghadiri, J. 36 A. S. Madsen and C. A. Olsen, Angew. Chem., Int. Ed., 2012,
Am. Chem. Soc., 2009, 131, 3033–3041. 51, 9083–9087.
22 C. A. Olsen and M. R. Ghadiri, J. Med. Chem., 2009, 52, 7836– 37 P. J. Watson, L. Fairall, G. M. Santos and J. W. Schwabe,
7846. Nature, 2012, 481, 335–340.
23 C. A. Olsen, A. Montero, L. J. Leman and M. R. Ghadiri, ACS 38 M. Bantscheff, C. Hopf, M. M. Savitski, A. Dittmann,
Med. Chem. Lett., 2012, 3, 749–753.
P. Grandi, A. M. Michon, J. Schlegl, Y. Abraham, I. Becher,
¨
24 C. J. Vickers, C. A. Olsen, L. J. Leman and M. R. Ghadiri, ACS
Med. Chem. Lett., 2012, 3, 505–508.
25 I. Izzo, N. Maulucci, G. Bifulco and F. De Riccardis, Angew.
Chem., Int. Ed., 2006, 45, 7557–7560.
G. Bergamini, M. Boesche, M. Delling, B. Dumpelfeld,
D. Eberhard, C. Huthmacher, T. Mathieson, D. Poeckel,
V. Reader, K. Strunk, G. Sweetman, U. Kruse, G. Neubauer,
N. G. Ramsden and G. Drewes, Nat. Biotechnol., 2011, 29,
255–265.
26 Y. Nakao, S. Yoshida, S. Matsunaga, N. Shindoh, Y. Terada,
K. Nagai, J. K. Yamashita, A. Ganesan, R. W. M. van Soest 39 C. J. Chou, D. Herman and J. M. Gottesfeld, J. Biol. Chem.,
and N. Fusetani, Angew. Chem., Int. Ed., 2006, 45, 7553–7557. 2008, 283, 35402–35409.
27 N. Maulucci, M. G. Chini, S. D. Micco, I. Izzo, E. Cafaro, 40 J. A. Richter-Larrea, E. F. Robles, V. Fresquet, E. Beltran,
A. Russo, P. Gallinari, C. Paolini, M. C. Nardi,
A. Casapullo, R. Riccio, G. Bifulco and F. De Riccardis, J.
Am. Chem. Soc., 2007, 129, 3007–3012.
A. J. Rullan, X. Agirre, M. J. Calasanz, C. Panizo,
J. A. Richter, J. M. Hernandez, J. Roman-Gomez, F. Prosper
and J. A. Martinez-Climent, Blood, 2010, 116, 2531–2542.
28 D. Wang, P. Helquist and O. Wiest, J. Org. Chem., 2007, 72, 41 F. A. Davis, M. B. Nolt, Y. Wu, K. R. Prasad, D. Li, B. Yang,
5446–5449.
K. Bowen, S. H. Lee and J. H. Eardley, J. Org. Chem., 2005,
70, 2184–2190.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun.