Cyclic tetrapeptides with thioacetate tails or intramolecular disulfide bridge as potent inhibitors of histone deacetylases

Article abstract of DOI:10.1016/j.bmcl.2012.03.004

Two thioacetate tails were introduced to the chlamydocin- and CHAP31-related cyclic tetrapeptides. An intramolecular disulfide bridge could be formed in the CHAP31-related cyclic peptides. Both the thioacetate-tailed and disulfide-bridged peptides were po

Full text of DOI:10.1016/j.bmcl.2012.03.004

Bioorganic & Medicinal Chemistry Letters 22 (2012) 6770–6772
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Cyclic tetrapeptides with thioacetate tails or intramolecular disulfide bridge
as potent inhibitors of histone deacetylases
Md. Ashraful Hoque a, Toru Arai b, Norikazu Nishino a, , Hyun-Jung Kim c, Akihiro Ito d, Minoru Yoshida d
⇑
a Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology, Kitakyushu 808-0196, Japan
b Department of Applied Chemistry, Kyushu Institute of Technology, Kitakyushu 804-8550, Japan
c BioRunx Co Ltd, Seoul 110-749, Republic of Korea
d RIKEN, Saitama 351-0198, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Two thioacetate tails were introduced to the chlamydocin- and CHAP31-related cyclic tetrapeptides. An
intramolecular disulfide bridge could be formed in the CHAP31-related cyclic peptides. Both the thioace-
tate-tailed and disulfide-bridged peptides were potent histone deacetylase inhibitors in the presence of
sulfhydryl compound. Potent p21 promoter inducing activity was also observed in vivo.
Ó 2012 Published by Elsevier Ltd.
Received 23 December 2011
Revised 2 March 2012
Accepted 5 March 2012
Available online 8 March 2012
Keywords:
HDAC inhibitor
Cyclic peptide
Disulfide bridge
Chlamydocin
FK228
Histone deacetylases (HDACs) play a prominent role in the reg-
ulation of gene transcription by histone deacetylation.1 Unusual
deacetylation of histone is sometimes linked to carcinogenesis.2
Therefore, novel HDAC inhibitor may lead to a novel cancer drug
such as vorinostat.3 Current challenges include the discovery of
the selective inhibitor to target cancer.4,5
the ‘LDLD’ amino acid configuration of CHAP31.9 The linear tetra-
peptide Boc- -Ab7- -Aob7- -Phe-
-Pro-OtBu (1, Ab7, 2-amino-7-
bromoheptanoic acid; Aob7, 2-amino-4-(2-bromoethoxy)butyric
L
D
L
D
The naturally occurring FK-228 (Fig. 1) displays impressive anti-
cancer and anti-angiogenesis activities through the HDAC inhibi-
tion,6 and is now approved as a cancer drug (IstodaxÒ). FK228 is
a bicyclic depsipeptide with intramolecular disulfide bridge, which
is reductively activated after uptake into the cells generating the
sulfhydryl groups.7 Inspired by the stability of FK228 due to its
prodrug nature, several HDAC inhibitors with the disulfide bond
have been proposed with more simplified structure for the ease
of synthetic study.8 We have reported sulfur containing cyclic tet-
rapeptides with disulfide bonds (SCOPs, Fig. 1) based on CHAP31
(cyclo(-L-Asu(NHOH)-D-Tyr(Me)-L-Ile-D-Pro-); Asu, 2-aminosuberic
acid) skeleton (Fig. 1).9 Here we report the synthesis of a bicyclic
tetrapeptidyl inhibitor with an intramolecular disulfide bridge.
These cyclic peptides with disulfide bonds are activated in a similar
manner to FK228 and inhibit HDACs by the coordination of the
sulfhydryl tails to Zn2+ ions.
First, a cyclic tetrapeptide 3 (Fig. 2) with the thioacetate side-
chains in the neighboring amino acids was designed, based on
⇑
Corresponding author. Tel./fax: +81 93 695 6061.
E-mail address: nishino@life.kyutech.ac.jp (N. Nishino).
Figure 1. Natural and synthetic HDAC inhibitors with cyclic peptide structure.
0960-894X/$ - see front matter Ó 2012 Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.bmcl.2012.03.004

Md. A. Hoque et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6770–6772
6771
Figure 3. HPLC profiles of the reaction mixture of the I2-oxidation of (a) 3 (–X = –
SH) and (b) 9 (–X = –SH).
synthesized in a similar manner to 1. TFA deprotection and HATU
cyclyzation under high dilution conditions (4.0 mmol Lꢀ1) afforded
the cyclic peptides 7 and 8 in fair yields (74% and 60%, respectively)
after silica-gel column chromatography. Compounds 7 and 8 were
transformed to 9 and 10 bearing the thioacetate sidechains, respec-
tively. Unfortunately, the methylamine treatment and successive
I2-oxidation did not afforded the bicyclic peptide with the intramo-
lecular disulfide bridges. All the new compounds were character-
ized by 1H, 13C NMR and FAB-MS including the high resolution
MS. The purity of the compounds was checked by HPLC and was
>98%.
Figure 3 shows the HPLC analyses of the reaction mixture of I2-
oxidation. The CHAP31-related 30 (–X = –SH) generated 4 with
some oligomeric byproducts. In 3, the sidechains of
L-amino acid
and that of
D
-amino acid in the neighboring positions may come
close with each other. Contrary, the oxidation of 100 (–X = –SH)
formed only oligomers. In 9 and 10, the sulfur tails at the opposite
positions may spread from the cyclic peptides, therefore, the sulfur
groups are separated with each other.
Figure 2. Synthesis of bicyclic tetrapeptide 4 with the ‘LDLD’ configuration and
cyclic tetrapeptides 9 and 10 with the ‘LAibLD’ configuration.
1H NMR spectra of 3 and 4 (Fig. 4) indicates that the cyclic pep-
tide framework little changed upon the intramolecular disulfide
bridging. The amide-NH protons of 3 (d 7.11, Aob7; 6.58, Phe;
acid) was synthesized from D
-Pro-OtBu in the solution phase, that
is, the stepwise deprotections of the N-termini and DCC (N,N-dicy-
clohexylcarbodiimide)-HOBt (1-hydroxybenzotriazol) couplings.
6.30, Ab7) slightly shifted in 4 (d 7.15, 6.71, 6.40). The C
and 4 were also similar with each other. (Note: slight shift of the
H of Aob7 and Ab7 are caused by the change of –SAc to –SS–.)
The J values of NH/C H, which reflect the dihedral angles of the
aH of 3
Ca
After TFA deprotection of 1, the obtained free peptide (H-
-Aob7- -Phe- -Pro-OH) TFA salt was treated with HATU (1H-7-
azobenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluoro-
L-Ab7-
a
D
L
D
phosphate) in DMF-diisopropylethylamine under high dilution
conditions (4.0 mmol Lꢀ1 peptide concentration). The desired cyc-
lic peptide 2 was obtained in fair yield (72%) after silica-gel column
chromatography. Such a high cyclization yield suggested that the
peptide conformation of 2 with the ‘LDLD’ amino acid configura-
tion was not strained.
The bromoalkyl sidechains of 2 were transformed to thioace-
tates (3) by KSAc in DMF (71%). The reaction of 3 with methyl-
amine in methanol generated the sulfhydryl groups, which were
successively I2-oxidized to yield the desired bicyclic peptide 4. This
intramolecular disulfide formation took place with moderate
yield (36%) without extreme dilution (50 mmol Lꢀ1 peptide
concentration).
Next, cyclic tetrapeptides with the thioacetate sidechains in the
opposite amino acids (Fig. 3) were synthesized, based on the
‘LAibLD’ amino acid configuration of chlamydocin (cyclo(-
Aib- -Phe- -Pro-); Aoe, 2-amino-8-oxo-9,10-epoxydecanoic acid;
Aib, 2-amino-2-methypropionic acid).10 Linear tetrapeptides Boc-
-Ab7-Aib- -XXX-
-Pro-OtBu; XXX = Ab7 for 5, Aob7 for 6) were
L-Aoe-
L
D
L
L
D
Figure 4. 1H NMR spectra of (a) 3 and (b) 4 in CDCl3.

6772
Md. A. Hoque et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6770–6772
Table 1
In vitro and in vivo HDAC inhibitory activities of cyclic peptide
Compound
Configuration
HDAC inhibitory activity (IC50
/
l
mol Lꢀ1
)
P21 promoter assay (EC1000/l )
mol Lꢀ1
HDAC1
HDAC4
HDAC6
Trichostatin A
FK228
FK228 + DTT
3
3 + DTT
4
4 + DTT
9
9 + DTT
10
10 + DTT
0.023
0.036
0.001
2.40
0.19
>100
0.010
1.8
0.034
0.51
Not tested
0.99
0.099
>100
0.0059
1.2
0.065
>500
0.62
>100
6.50
>100
0.33
12
0.020
0.0031
–
0.16
–
0.15
–
0.017
–
LDLD
LDLD (bicyclic)
LAibLD
0.15
4.2
0.21
0.10
0.9
0.12
1.7
11
2.0
LAibLD
0.036
–
amino acid residues,11 were all 10 Hz in 3 and 4. Thus, no confor-
mational change was detected in the reaction of 3 to 4 in 1H NMR.
In the CD measurements, 3 and 4 also showed a similar spectra
(0.1 mmol Lꢀ1 in methanol, data not shown). Positive Cotton effect
at 253 nm, negative one at 243 nm, positive one at 228 nm and
negative one at 212 nm. This fact also supported that the intramo-
sion was measured.9 The cyclic peptide 3 and the bicyclic peptide 4
showed similar EC1000 (effective concentration of 10 times in-
creased induction) values. These prodrugs were transformed to
the sulfhydryl derivative by the glutathione-reduction or prote-
ase-hydrolysis in the cell, therefore, they showed similar activity
under the cellular assay conditions. In this assay, the cyclic pep-
tides with the chlamydocin framework, 9 and 10, were more effi-
cient inhibitors than the CHAP31-based peptides, 3 and 4. The
relationship between HDAC inhibitory activity and p21 promotion
activity should be studied in future.
In summary, the bicyclic tetrapeptide 4 based on the CHAP31
framework with an intramolecular disulfide bridge was for the first
time successfully synthesized. In the chlamydocin-related cyclic
peptides, the intramolecular disulfide bridges were not formed.
Biochemical assays showed that 4 is a potent inhibitor of HDAC
in the presence of DTT in vitro. Interestingly, the precursor cyclic
peptides with the thioacetate ligands (3, 9, 10) moderately inhib-
ited the action of HDAC. Compounds 3, 4, 9 and 10 were also active
in vivo p21 promoter inducing activity.
lecular disulfide bridging in
distortion.
4 caused little conformational
Table 1 shows the HDAC inhibitory activities of the bicyclic pep-
tide with the intramolecular disulfide bridge (4) and the cyclic pep-
tides with the thioacetate groups (3, 9, 10). The in vitro activities of
peptides were tested using the enzymes (HDAC1, HDAC4 and
HDAC6) and a fluorophore-modified substrate. For HDAC1 and
HDAC4, the cyclic peptides 3, 9 and 10 were moderately effective
inhibitors with l
mol Lꢀ1-range IC50 (50% inhibitory concentration).
Probably, the thioacetate branches of these peptides weakly coor-
dinated to Zn2+ ions in the enzyme active sites. No tail existed in
the bicyclic peptide 4 that can coordinate to Zn2+ ion, therefore,
4 did not inhibit HDAC1, HDAC4 and HDAC6 in the absence of
dithiothreitol (DTT).
The bicyclic peptide 4 showed excellent inhibitory activities
l
mol Lꢀ1 for HDAC1 and
References and notes
after treatment with DTT, IC50 of 0.010
0.0059
l
mol Lꢀ1 for HDAC4, while DTT showed no effect in HDAC
1. Grozinger, C. M.; Schreiber, S. L. Chem. Biol. 2002, 9, 3.
2. Cress, W. D.; Seto, E. J. Cell. Physiol. 2000, 184, 1.
3. Mann, B. S.; Johnson, J. R.; Cohen, M. H.; Justice, R.; Pazdur, R. Oncologist 2007,
12, 1247.
4. Yoshida, M.; Kijima, M.; Akita, M.; Beppu, T. J. Biol. Chem. 1990, 265, 17174.
5. Yoshida, M.; Matsuyama, A.; Komatsu, Y.; Nishino, N. Curr. Med. Chem. 2003, 10,
2351.
6. Lee, Y. M.; Kim, S.-H.; Kim, H.-S.; Son, M. J.; Nakajima, H.; Kwon, H. J.; Kim, K.-
W. Biochem. Biophys. Res. Commun. 2003, 300, 241.
7. Furumai, R.; Matsuyama, A.; Kobashi, N.; Lee, K.-H.; Nishiyama, M.; Nakajima,
H.; Tanaka, A.; Komatsu, Y.; Nishino, N.; Yoshida, M.; Horinouchi, S. Cancer Res.
2002, 62, 4916.
8. Suzuki, T.; Hisakawa, S.; Itoh, Y.; Maruyama, S.; Kurotaki, M.; Nakagawa, H.;
Miyata, N. Bioorg. Med. Chem. Lett. 2007, 17, 1558.
inhibition. Reduction of 4 by DTT generated the sulfhydryl groups,
which inserted into the enzyme active site pocket. The inhibition of
4 in the presence of DTT is still less than CHAP31 (IC50 3.32 nM)12
or chlamydocin (IC50 0.15 nM).13 However, Compound 4 is proba-
bly stable in the blood (relatively oxidative conditions) due to its
disulfide bridge. Such a potent inhibitor with nmol Lꢀ1 level activ-
ity with improved stability is greatly beneficial to the practical
application as a cancer prodrug. Interestingly, the addition of DTT
also ca. 10 times increased the inhibitory activities of 3, 9 and
10, suggesting that the thioacetate groups may be in some extent
hydrolyzed under the assay conditions to generate sulfhydryl
groups. The less effective nature of 3, 9, and 10 compared to 4 in
the presence of DTT suggests the incomplete generation of sulfhy-
dryl groups. It may be noted that these cyclic peptide inhibitors (3,
4, 9 and 10) were less effective for HDAC6, even in the presence of
DTT. Because HDAC4 is more frequently expressed in tumor tis-
sues, this target sensitivity is valuable.14
9. Nishino, N.; Jose, B.; Okamura, S.; Ebisusaki, S.; Kato, T.; Sumida, Y.; Yoshida, M.
Org. Lett. 2003, 5, 5079.
10. De Schepper, S.; Bruwiere, H.; Verhulst, T.; Steller, U.; Andries, L.; Wouters, W.;
Janicot, M.; Arts, J.; Van Heusden, J. J. Pharmacol. Exp. Ther. 2003, 304, 881.
11. Pardi, A.; Billeter, M.; Wüthrich, K. J. Mol. Biol. 1984, 180, 741.
12. Komatsu, Y.; Tomizaki, K.-y.; Tsukamoto, M.; Kato, T.; Nishino, N.; Sato, S.;
Yamori, T.; Tsuruo, T.; Furumai, R.; Yoshida, M.; Horinouchi, S.; Hayashi, H.
Cancer Res. 2001, 61, 4459.
13. Nishino, N.; Jose, B.; Shinta, R.; Kato, T.; Komatsu, Y.; Yoshida, M. Bioorg. Med.
Chem. 2004, 12, 5777.
14. Kozikowski, A. P.; Butler, K. V. Curr. Pharm. Des. 2008, 14, 505.
The cyclic peptides were examined for in vivo p21 promoter
assay, in which the effect of the peptide inhibitors on a cell expres-