Cyclic tetrapeptides with-SS-bridging between amino acid side chains for potent histone deacetylases' inhibition

Article abstract of DOI:10.1007/s00726-013-1527-8

Cyclic depsipeptide FK228 with an intramolecular disulfide bond is a potent inhibitor of histone deacetylases (HDAC). FK228 is stable in blood because of its prodrug function, whose -SS- bond is reduced within the cell. Here, cyclic peptides with -SS- bridges between a variety of amino acids were synthesized and assayed for HDAC inhibition. Cyclic peptide 3, cyclo(-l-amino acid-l-amino acid-l-Val-d-Pro-), with an -SS- bridge between the first and second amino acids, was found to be a potent HDAC inhibitor. Cyclic peptide 7, cyclo(-l-amino acid-d-amino acid-l-Val-d-Pro-), with an -SS- bridge between the first and second amino acids, was also a potent HDAC inhibitor.

Full text of DOI:10.1007/s00726-013-1527-8

Amino Acids
DOI 10.1007/s00726-013-1527-8
ORIGINAL ARTICLE
Cyclic tetrapeptides with –SS– bridging between amino acid side
chains for potent histone deacetylases’ inhibition
•
Toru Arai Md. Ashraful Hoque Norikazu Nishino
•
•
•
Hyun-Jung Kim Akihiro Ito Minoru Yoshida
•
Received: 14 March 2013 / Accepted: 24 May 2013
Ó Springer-Verlag Wien 2013
Abstract Cyclic depsipeptide FK228 with an intramo-
lecular disulfide bond is a potent inhibitor of histone
deacetylases (HDAC). FK228 is stable in blood because of
its prodrug function, whose –SS– bond is reduced within
the cell. Here, cyclic peptides with –SS– bridges between a
variety of amino acids were synthesized and assayed for
HDAC inhibition. Cyclic peptide 3, cyclo(-L-amino acid-L-
amino acid-L-Val-D-Pro-), with an –SS– bridge between the
first and second amino acids, was found to be a potent
HDAC inhibitor. Cyclic peptide 7, cyclo(-L-amino acid-D-
amino acid-L-Val-D-Pro-), with an –SS– bridge between the
first and second amino acids, was also a potent HDAC
inhibitor.
1994; Masuoka et al. 2001). These depsipeptides potently
inhibit histone deacetylase (HDAC) after the reductive
cleavage of the –SS– bonds. The –SH group coordinates to
zinc ion at the active site of class I and II HDACs (Furumai
et al. 2002). For practical use, the anticancer drugs should
be stable in the body before their delivery to the target
sites. But the digestive system and blood deactivate some
HDAC inhibitors, although they are potent in vitro. FK228
is useful prodrug because of the stable disulfide group.
Other natural cyclic peptidyl HDAC inhibitors include
chlamydocin, trapoxins (Kijima et al. 1993), and HC-toxins
(Miller et al. 2003), tethering epoxyketone moieties that
interact with the active site. We have discovered that cyclic
tetrapeptides with artificial zinc ligating groups are potent
and specific HDAC inhibitors (Komatsu et al. 2001; Islam
et al. 2011). Here, we present cyclic tetrapeptides that
contain disulfide moieties as prodrug candidates, which are
easier to synthesize than FK228 and analogs (Hoque et al.
2012).
Keywords Cyclic peptide Á Histone deacetylase
inhibitor Á Disulfide bridge Á FK228 analog
Introduction
FK228 (FR901228) and some natural cyclic depsipeptides
contain intramolecular disulfide bridges (Fig. 1; Ueda et al.
Results and discussion
Cyclo(-L-Am7(Ac)-Aib-L-Am7(Ac)-D-Pro-) (1)
with a chlamydocin framework (-L*LD-) and attempted
intramolecular disulfide bond formation
T. Arai (&)
Department of Applied Chemistry, Kyushu Institute
of Technology, Kitakyushu 804-8550, Japan
e-mail: arai@che.kyutech.ac.jp
The aminoisobutylic acid (Aib) in chlamydocin, cyclo(-L-
Aoe-Aib-L-Phe-D-Pro-) (Aoe, (2S,9S)-2-amino-8-oxo-9,10-
epoxydecanoic acid), preferably has a b-turn structure and
fixes cyclic peptide framework. Using this -L*LD- peptide
sequence and using hydroxamic acid as a zinc-ligating
group, a chlamydocin–hydroxamic acid analog was suc-
cessfully synthesized as a potent HDAC inhibitor (Fig. 2,
Nishino et al. 2004). An intramolecular disulfide bridge
Md. Ashraful Hoque Á N. Nishino
Graduate School of Life Science and Systems Engineering,
Kyushu Institute of Technology, Kitakyushu 808-0196, Japan
H.-J. Kim
BioRunx Co Ltd, Seoul 110-749, Republic of Korea
A. Ito Á M. Yoshida
Riken, Saitama 351-0198, Japan
123

T. Arai et al.
Fig. 3 Unnatural amino acids prepared for –SS–
Fig. 1 Cyclic depsipeptidyl HDAC inhibitor FK228 with –SS– bond
and cyclic tetrapeptidyl HDAC inhibitors with epoxyketone
was added between the facing amino acids, i.e., between
the –SH groups written in italics in -L*LD-. Cyclo(-L-
Am7(Ac)-Aib-L-Am7(Ac)-D-Pro-) (1) was synthesized
expecting –SS– formation between the two Am7(H)s
(Figs. 2, 3).
The linear tetrapeptide Boc-L-Ab7-Aib-L-Ab7-D-Pro-OtBu
was prepared by the stepwise elongation of D-Pro-OtBu
(Scheme 1), taking advantage of the fact that the cyclic imino
acid (Pro) hardly racemizes during couplings. Coupling Boc-
L-Ab7-OH ((S)-N-tbutoxycarbonyl-2-amino-7-bromohepta-
noic acid, Watanabe et al. 2004) and D-Pro-OtBu with N,N’-
dicyclohexylcabodiimide
(DCC)/1-hydroxybenzotriazole
(HOBt) gave Boc-L-Ab7-D-Pro-OtBu. Boc was selectively
removed with HCl/dioxane without cleaving the –COOtBu
(Han et al. 2008), then the residue was coupled with Z-Aib-
OH to afford Z-Aib-L-Ab7-D-Pro-OtBu. Z-removal with
H2-Pd/C and coupling the product with Boc-L-Ab7-OH gave
Boc-L-Ab7-Aib-L-Ab7-D-Pro-OtBu. After TFA treatment, the
product TFAÁH–D-Ab7-Aib-L-Ab7-D-Pro-OH was cyclized
in DMF ([peptide] = 4.1 mM) with O-(7-azabenzotriazol-
1-yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate
(HATU)/diisopropylethylamine (DIEA). The precursor
cyclic tetrapeptide cyclo(-L-Ab7-Aib-L-Ab7-D-Pro-) was
obtained in 74 % yield based on the linear tetrapeptide after
silica gel column chromatography. The bromoalkyl side
chains were transformed to AcS– with potassium thioacetate
(KSAc) in DMF to yielded cyclo(-L-Am7(Ac)-Aib-L-
Am7(Ac)-D-Pro-) (1). All the new compounds were charac-
terized by FAB-MS, HR-FAB-MS, and 1H-NMR. C18 HPLC
was used to determine the peptides purity (Table 1).
The –SAc side chains of 1 were deprotected using
excess MeNH2 in MeOH (r.t. 2 h). The cyclic tetrapep-
tide with –SH groups (cyclo(-L-Am7(H)-Aib-L-Am7(H)-D-
Pro-), not isolated) was mixed with I2 in DMF-EtOH
([peptide] = 5.0 mM, r.t. 30 min). However, the expected
Fig. 2 HDAC inhibitor with chlamydocin framework (-L*LD-)
123

Cyclic tetrapeptides with –SS– bridging between amino acid
Scheme 1 Reagents and
conditions: (a) Boc-L-Ab7-OH,
DCC/HOBt; (b) HCl/dioxane;
(c) Z-Aib-OH, DCC/HOBt;
(d) Pd/C,H2; (e) TFA;
(f) HATU/DIEA; (g) KSAc
Table 1 Yields and HPLC retention times for the intramolecular –SS– bridging of the cyclic peptide inhibitors
Model
Chlamydocin
Trapoxin
B
HC-toxin
I
Configuration
–SAc precursor
HPLCa
-L*LD-
-LLLD-
-LDLD-
-LDLD-
1
2
4
6
8
10
12
14
16
7.28
7.45
3
8.33
5
7.82
7
6.50
9
7.98
11
5.84
13
7.18
15
7.51
17
–SS–
Not obtained
formation
Yield/%b
HPLCb
a
–
–
31
23
45
28
28
34
36
36
6.58
7.90
7.03
5.33
7.22
4.92
6.43
7.04
Retention time/min, C18 column (see experimental)
b
Based on –SAc precursor
the chlamydocin framework (-L*LD-) appeared to be
inappropriate for intramolecular –SS– formation.
Cyclo(-L-Am7(Ac)-L-Am7(Ac)-L-Val-D-Pro-) (2)
with a trapoxin B (-LLLD-) framework
and intramolecularly disulfide bridged cyclo(-L-Am7(-)-
L-Am7(-)-L-Val-D-Pro-) (3)
Next, the trapoxin B -LLLD- sequence, cyclo(-L-Aoe-L-
Phe-L-Phe-D-Pro-), was used to arrange two L-Am7(Ac)
residues at neighboring positions in cyclo(-L-Am7(Ac)-L-
Am7(Ac)-L-Val-D-Pro-) (2, Fig. 5). The third amino acid
L-Phe in trapoxin B was replaced by L-Val. The linear
tetrapeptide Boc-L-Ab7-L-Ab7-L-Val-D-Pro-OtBu was syn-
thesized from H-D-Pro-OtBu, Z-L-Val-OH, and Boc-L-Ab7-
OH (two times), deprotected, cyclized with HATU/DIEA
(77 % yield from the linear tetrapeptide), and transformed
to 2 with KSAc. Treatment with MeNH2/MeOH followed
by I2 oxidation ([peptide] = 5.0 mM) gave the intramo-
lecularly –SS– bridged cyclic tetrapeptide cyclo(-L-
Am7(-)-L-Am7(-)-L-Val-D-Pro-) (3) with some oligomeric
byproducts. LH-20 gel filtration chromatography with
DMF afforded 3 in 31 % yield based on 2. The –SS– bond
formation between neighboring amino acids in 2 seems to
be easier than that between facing amino acids in 1. The
–SS–bridged 3 eluted earlier than 2 in the HPLC, probably
because of its compact molecular structure.
Fig. 4 HPLC profiles of the reaction mixture of the I2 oxidation of 1
(a) and 10 (b)
intramolecular –SS– bridged product was not formed. The
C18 HPLC chromatogram of the reaction mixture (Fig. 4a),
using H2O-CH3CN as the mobile phase, showed no signals
corresponding to the bicyclic tetrapeptide, but did show
oligomeric byproducts with the intermolecular –SS–
bridges (determined by MS). For comparison, Fig. 4b
shows the intramolecular –SS– bridging from 10 to 11,
which was successful (see below) and led to a large 11
signal appearing. The two –SH groups facing each other in
123

T. Arai et al.
Fig. 5 HDAC inhibitors with trapoxin and HC-toxin frameworks
Cyclo(-L-Am7(Ac)-D-Tic-L-Val-D-Am7(Ac)-) (4)
with an HC-toxin (-LDLD-) framework
and the intramolecularly disulfide bridged cyclo
(-L-Am7(-)-D-Tic-L-Val-D-Am7(-)-) (5)
Fig. 6 HDAC inhibitors with -LDLD- configuration
tested instead of 2 (Fig. 6). The linear tetrapeptide Boc-L-
Ab7-D-Ab7-L-Val-D-Pro-OtBu was deprotected, cyclized
with HATU/DIEA (95 % yield), and transformed to 6 with
KSAc. Treatment with MeNH2/MeOH and I2 gave the
intramolecularly –SS– bridged cyclic peptide 7 in 45 %
yield after LH-20 purification. This was the highest yield
for the intramolecular –SS– formation found in this study
(Table 1). The structure of 7 with an -LDLD- configuration
might be less restricted than the structure of 3 (-LLLD-
configuration); 6 and 7 with -LDLD- configurations eluted
later than 2 and 3 with -LLLD- configuration in C18 HPLC
analyses, respectively, although they contained the same
amino acid residues. The later C18 HPLC elution suggests
that 6 and 7 were less polar than 2 and 3, respectively,
possibly because of the effective intramolecular hydrogen
bonds in 6 and 7.
The third cyclic tetrapeptide motif was the -LDLD-
sequence in the HC-toxin, cyclo(-L-Aoe-D-Pro-L-Ala-
D-Ala-). Cyclo(-L-Am7(Ac)-D-Tic-L-Val-D-Am7(Ac)-) (4;
Tic = 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid)
was designed with increased hydrophobicity by modifying
-D-Pro-L-Ala- to -D-Tic-D-Val- (Fig. 5). The linear tetra-
peptide Boc-L-Val-D-Ab7-L-Ab7-D-Tic-OtBu was prepared,
deprotected, and cyclized with HATU/DIEA to give cyclo(-L-
Ab7-D-Tic-L-Val-D-Ab7-) in a modest yield (58 %), which
was then transformed to cyclo(-L-Am7(Ac)-D-Tic-L-Val-D-
Am7(Ac)-) (4). Deprotecting 4 with MeNH2/MeOH and
oxidation with I2/EtOH afforded cyclo(-L-Am7(-)-D-Tic-L-
Val-D-Am7(-)-) (5) in a rather low yield (23 %).
Cyclo(-L-Am7(Ac)-D-Am7(Ac)-L-Val-D-Pro-) (6)
with an -LDLD- configuration and the intramolecularly
disulfide bridged cyclo(-L-Am7(-)-D-Am7(-)-L-Val-D-
Pro-) (7)
Effect of amino acid hydrophobicity and ether-linked
side chains on the analogs of 6 and 7 with -LDLD-
configuration
The intramolecular –SS– formation of the trapoxin-mod-
eled peptide with an -LLLD- configuration 2, cyclo(-L-
Am7(Ac)-L-Am7(Ac)-L-Val-D-Pro-), gave a moderate yield
of 3 (31 %). Cyclo(-L-Am7(Ac)-D-Am7(Ac)-L-Val-D-Pro-)
(6), a cyclic peptide with an -LDLD- configuration, was
Using the -LDLD- configuration, the effect of hydropho-
bicity was examined by introducing L-Ala and L-Phe
instead of L-Val in 6 and 7. Therefore, cyclo(-L-Am7(Ac)
123

Cyclic tetrapeptides with –SS– bridging between amino acid
-D-Am7(Ac)-L-Ala-D-Pro-) (8), cyclo(-L-Am7(-)-D-Am7(-)-
L-Ala-D-Pro-) (9), cyclo(-L-Am7(Ac)-D-Am7(Ac)-L-Phe-D-
Pro-) (10), and cyclo(-L-Am7(-)-D-Am7(-)-L-Phe-D-Pro-)
(11) were synthesized (Fig. 6). The HPLC retention times
showed that the order of hydrophobicities of these products
was 10 and 11 (Phe) [ 6 and 7 (Val) [ 8 and 9 (Ala).
We also introduced a new amino acids Aob7 (2-amino-
5-oxy-7-bromoheptanoic acid) and Aom7(Ac) (2-amino-5-
oxy-7-acetylmercaptoheptanoic acid), which has an ether
moiety in the side chains (Fig. 3). This ether linkage may
affect the solubility, synthetic yield, and biological activity.
Cyclo(-L-Am7(Ac)-D-Aom7(Ac)-L-Ala-D-Pro-) (12), cyclo
(-L-Am7(-)-D-Aom7(-)-L-Ala-D-Pro-) (13), cyclo(-L-Am7
(Ac)-D-Aom7(Ac)-L-Phe-D-Pro-) (14), cyclo(-L-Am7(-)-D-
Aom7(-)-L-Phe-D-Pro-) (15), cyclo(-L-Aom7(Ac)-D-Am7
(Ac)-L-Phe-D-Pro-) (16), and cyclo(-L-Aom7(-)-D-Am7(-)-
L-Phe-D-Pro-) (17) were synthesized. These ether-linked
peptides were more hydrophilic and eluted from the HPLC
faster than the corresponding non-ether-containing pep-
tides. For instance, 12 and 13 eluted faster than 8 and 9,
respectively.
Fig. 7 1H-NMR spectra of 8 (a) and 9 (b) in CDCL3
-LDLD- configuration gave a higher activity than 3 with an -
LLLD- configuration. The cyclic peptides with –SAc tails (1,
2, 4, 6, 12, and 14) inhibited the enzymes almost one order
of magnitude less effectively than 3 did. It should be noted
here that these cyclic peptides were hardly effective in the
absence of the reductant DTT, indicating that the active
moiety for inhibition is –SH (Table 2).
1H-NMR investigation into the peptide structure
before and after the –SS– bridging reaction
1
Figure 7 shows the H-NMR spectra (in CDCl3, low-field
The cyclic peptides with –SAc and –SS– groups were
both ineffective inhibitors of HDAC6. For instance, the
IC50 for 7 was *1,000 times higher for HDAC6 than that
for HDAC1. However, the effect of 3 was less different on
HDAC1 and HDAC6. Selectivity for HDAC1 over
HDAC4 was not observed probably because of their similar
surface binding regions. The p21 promoter assay was
performed in vivo without DTT. The cyclic peptides with –
SAc tails, 2 and 6, had higher activity (EC1000 below
0.1 lM). The disulfide bond might be reduced to –SH in
the cell medium in this assay. The reason why some pep-
tides with –SAc tails (2 and 6) showed higher activity than
those with –SS– (3 and 7) is not yet clear. The confor-
mation and the hydrogen-bondings in the cyclic peptides
may have some influence on the membrane permeability of
the peptides, for instance. The effect of amino acid struc-
ture, for instance replacing Val (6 and 7) with Ala (8 and
9), did not show clear effect in the p21 promoter activity.
The effect of ether linkage (12 and 13) in the alkyl side
chain gave little difference compared to 8 and 9.
region) of (a) 8, which has -SAc tails, and (b) 9, with an
–SS– bridge. The amide NH protons were slightly low-field
shifted comparing 9 (d 7.12, d, 10 Hz, L-Am7(-); d 6.42, d,
10 Hz, Ala, and d 6.14, d, 10 Hz, D-Am7(-)) with 8 (d 7.09,
d, 10 Hz; d 6.30, d, 10 Hz; d 6.05, d, 10 Hz). The 10 Hz
coupling constant of every amide protons in 8 and 9 indi-
cated that the cyclic peptide conformation was little altered
by the –SS– bridge (Pardi et al. 1984). The J value (cou-
pling constant) between NH and CaH reflects the dihedral
angle of the amino acid. The Ala and D-Pro Ca protons in 8
(d 4.93, dq, Ala; d 4.71, dd, D-Pro) were not shifted in 9,
but the Ca protons of L-Am7(Ac) (d 4.35, dt) and D-
Am7(Ac) (d 4.29, dt) in 8 were shifted to d 4.46 (m,
intensity 2 H), due to the change from SAc to –SS–.
Enzyme inhibition and biological activity
The cyclic peptides were assayed for HDAC activity in vitro
using three enzymes, HDAC1, HDAC4, and HDAC6, which
were prepared using 293T cells and a substrate attaching a
fluorophore (Furumai et al. 2002). In the presence of 0.1 mM
dithiothreitol (DTT), the cyclic peptides with the intramo-
lecular –SS– bridges (3, 5, 7, 13, and 15) showed modest
activities with IC50 concentrations lower than 0.1 lM for
HDAC1 and HDAC4. Compound 7 gave an IC50 of
0.0081 lM (8.1 nM), making it more potent than trichostatin
A or FK228. It is interesting that the cyclic peptide 7 with an
Experimental
General
All solvents and reagents were reagent grade. Silica gel 60
(230–400 mesh) eluting with CHCl3-1 % MeOH (v/v) or
123

T. Arai et al.
Table 2 Activity of cyclic peptides for the HDAC inhibition in vitro and in vivo
Compound/active tail
HDAC inhibition (IC50/lM)a
P21 promoter activity
HDAC1
HDAC4
0.034
HDAC6
EC1000/lMb
EC10000/lMb
Trichostatin Ac
FK228
0.023
0.001
0.065
0.62
0.020
0.031
1/–SAc
2/–SAc
3/–SS–
4/–SAc
5/–SS–
6/–SAc
7/–SS–
8/–SAc
9/–SS–
12/–SAc
13/–SS–
14/–SAc
15/–SS–
a
0.017
0.036
0.23
0.23
0.14
0.042
0.41
0.14
0.29
0.26
0.19
0.16
0.15
0.049
0.071
0.53
0.43
0.23
0.079
0.60
0.39
0.86
0.47
0.33
0.29
0.32
0.26
0.18
0.92
0.083
3.7
0.0139
0.15
0.0093
0.20
0.027
0.30
0.037
0.16
0.59
47
0.0081
0.0083
7.3
0.31
0.13
5.4
0.016
0.19
0.011
0.099
0.0059
0.32
6.5
0.010
0.33
50 % inhibitory concentration, in the presence of 0.1 mM dithiothreitol
b
c
EC1000; effective concentration of 109 increased induction, EC10000; 1009 increased induction
In the absence of dithiothreitol
LH-20 (gel filtration) eluting with DMF was used for
chromatography. Merck Chromolith Performance RP-18e
column (4.6 9 100 mm) was used for HPLC, with a sol-
vent gradient of (H2O-0.1 % TFA)/(CH3CN-0.1 % TFA)
over 15 min, 2.0 mL/min flow rate, and 220 nm detection
to note the retention time (Rt). FAB-mass spectra including
the high resolution spectra (HR-MS) were measured on a
JEOL JMS-SX 102A. NMR spectra were recorded on a
JEOL JNM a-500 in CDCl3.
evaporation, Boc-D-Aob7-OEt was extracted with diethyl
ether under basic conditions. After acidification, Boc-L-
Aob7-OH was extracted with AcOEt as colorless oil (2.5 g,
7.7 mmol, 46 % based on Boc-D-L-Aob7-OEt). Hydrolysis
of Boc-D-Aob7-OEt (1 M NaOH) afforded Boc-D-Aob7-
OH as colorless oil (2.5 g, 7.8 mmol, 48 %).
Cyclo(-L-Am7(Ac)-Aib-L-Am7(Ac)-D-Pro-) (1)
The mixture of Boc-L-Ab7-OH (1.7 g, 5.2 mmol), H-D-
Pro-OtBu (0.89 g, 5.2 mmol), DCC (1.3 g, 6.2 mmol),
HOBtÁH2O (0.80 g, 5.2 mmol) was mixed in 11 mL DMF
(0 °C to r.t. 5 h). After evaporation, the residue taken up in
AcOEt was filtered, washed, concentrated, and chromato-
graphed (silica gel) to yield Boc-L-Ab7-D-Pro-OtBu as an
oil (2.1 g, 4.4 mmol, 85 %). This dipeptide (4.4 mmol)
was dissolved in ice-cold 4 M HCl/dioxane (11 mL) and
was kept at room temperature (30 min) to obtain H-L-Ab7-
D-Pro-OtBu as a syrup (1.5 g, 4.0 mmol, 91 %), which was
reacted with Z-Aib-OH (1.1 g, 4.8 mmol) to give Z-Aib-L-
Ab7-D-Pro-OtBu as an oil (1.6 g, 2.7 mmol, 67 %). This
tripeptide (2.7 mmol) was stirred with 0.13 g Pd/C in
13 mL CH3COOH under H2 overnight. After filtration and
evaporation, the residue was dissolved in AcOEt, washed,
and concentrated to obtain H-Aib-L-Ab7-D-Pro-OtBu
(1.2 g, 2.6 mmol, 95 %), which was reacted with Boc-L-
Ab7-OH (0.83 g, 2.6 mmol) to afford Boc-L-Ab7-Aib-L-
Ab7-D-Pro-OtBu as white foam (1.3 g, 1.7 mmol, 66 %).
This tetrapeptide (0.63 g, 0.82 mmol) was dissolved in
Boc-L-2-amino-5-oxy-7-bromoheptanoic acid (Boc-L-Aob7-
OH) and Boc-D-2-amino-5-oxy-7-bromoheptanoic acid
(Boc-D-Aob7-OH)
Na (0.46 g, 20 mmol) were dissolved in EtOH (20 mL), to
which diethyl (Boc-amino) malonate (5.5 g, 20 mmol) was
added and refluxed for 30 min. Bis(2-bromoethyl) ether
(14 g, 60 mmol) was added and refluxed for further 6 h.
Oily diester was obtained by diethyl ether extraction, which
was monohydrolysed in 20 mL EtOH/20 mL of 1 M
aqueous NaOH (0 °C, 5 h). After evaporation and acidifi-
cation (citric acid), the monoacid monoester product was
extracted with AcOEt to give an oil. Refluxing the oil in
20 mL toluene for 7 h yielded Boc-DL-Aob7-OEt
(16 mmol, 82 %) as a solid. This solid was suspended in
DMF (17 mL)/H2O (51 mL) at 38 °C with pH 7–8 (1 M
NH3). Subtilisin Carlsberg from B. licheniformis (Sigma,
17 mg, 1 mg enzyme per mmol substrate) was added with
adding 1 M NH3 to maintain pH 7–8 for 3 h. After
123

Cyclic tetrapeptides with –SS– bridging between amino acid
Cyclo(-L-Am7(-)-L-Am7(-)-L-Val-D-Pro-) (3)
4 mL TFA (0 °C) and kept at room temperature (3 h) to
yield TFAÁH-L-Ab7-Aib-L-Ab7-D-Pro-OH as
a solid
The cyclic peptide 2 (0.15 g, 0.25 mmol) in 2.0 mL DMF
was mixed with MeNH2 (2.6 mmol) in 0.27 mL MeOH
under Ar (r.t. 2 h). After evaporation, the residue was
dissolved in 52 mL DMF and I2 (67 mg, 0.26 mmol) in
EtOH was added dropwise (r.t. 30 min). The oxidized
peptide was purified by LH-20 chromatography to obtain
cyclo(-L-Am7(-)-L-Am7(-)-L-Val-D-Pro-) (3) as a solid
(40 mg, 78 lmol, 31 %, HPLC: Rt 6.58 min). HR-MS,
calcd for C24H41N4O4S2 ([M ? H]?) 513.2569, found
(0.58 g, 0.82 mmol, 100 %). This TFA salt (0.82 mmol),
HATU (0.47 g, 1.2 mmol), and DIEA (0.56 mL,
2.1 mmol) were added in two portions to DMF (0.20 L) in
every 30 min. After evaporation, the residue was washed
and chromatographed (silica gel) to yield cyclic tetrapep-
tide, cyclo(-L-Ab7-Aib-L-Ab7-D-Pro-) (0.36 g, 0.61 mmol,
74 %). This cyclic tetrapeptide (0.59 mmol) was reacted
with KSAc (0.20 g, 1.8 mmol) in 5.0 mL DMF (r.t. 7 h) to
obtain cyclo(-L-Am7(Ac)-Aib-L-Am7(Ac)-D-Pro-) (1) as
heavy oil after silica gel chromatography (0.21 g, 61 %,
HPLC: Rt 7.28 min). HR-MS, calcd for C27H45N4O6S2
1
513.2558. H NMR: d 7.19 (d, J = 10 Hz, 2H), 6.45 (d,
J = 5 Hz, 1H), 4.80 (dd, J = 2 Hz, 1H), 4.55 (t,
J = 10 Hz, 1H), 4.28 (m, 1H), 3.92 (m, 1H), 3.5 (m, 2H),
2.86 (m, 4H), 2.3 (m, 4H), 1.9 (m, 1H), 1.8 (m, 2H), 1.7 (m,
5H), 1.3 (m, 9H), 1.00 (d, 3H), 0.91 (d, 3H).
1
([M ? H]?) 585.2781, found 585.2792. H NMR: d 7.35
(d, J = 6 Hz, 1H), 7.14 (dd, J = 3 Hz, 1H), 5.95 (d,
J = 3 Hz, 1H), 4.84 (m, 1H), 4.75 (m, 1H), 4.19 (m, 1H),
3.92 (m, 1H), 3.54 (m, 1H), 2.67 (m, 4H), 2.39 (m, 1H),
2.26 (m, 1H), 1.85 (m, 8H), 1.66 (m, 10H), 1.37 (m, 12H).
Cyclo(-L-Am7(Ac)-D-Tic-L-Val-D-Am7(Ac)-) (4)
Boc-L-Ab7-D-Tic-OtBu was synthesized from Boc-L-Ab7-
OH, H–D-Tic-OtBu, DCC/HOBt in DMF (78 % after silica
gel chromatography). Treatment with 4 M HCl/dioxane
gave H-L-Ab7-D-Tic-OtBu (70 %), which was reacted with
Boc-D-Ab7-OH to obtain Boc-D-Ab7-L-Ab7-D-Tic-OtBu as
a white foam (72 % after silica gel chromatography).
Treatment again with 4 M HCl/dioxane gave H-D-Ab7-L-
Ab7-D-Tic-OtBu (75 %), which was condensed with Boc-
L-Val-OH to obtain Boc-L-Val-D-Ab7-L-Ab7-D-Tic-OtBu as
a white foam (60 % after silica gel chromatography). This
linear tetrapeptide was deprotected with TFA to yield
TFAÁH-L-Val-D-Ab7-L-Ab7-D-Tic-OH (91 %). This tetra-
peptide (0.58 g, 0.73 mmol), HATU (0.42 g, 1.1 mmol),
and DIEA (0.32 mL, 1.8 mmol) were added in two por-
tions to 0.18 L DMF in every 30 min. After work-up and
silica gel chromatography, cyclo(-L-Ab7-D-Tic-L-Val-D-
Ab7-) was obtained as a white foam (0.28 g, 0.42 mmol,
58 %). This cyclic tetrapeptide (0.42 mmol) was reacted
with KSAc (1.3 mmol) to yield cyclo(-L-Am7(Ac)-D-Tic-L-
Val-D-Am7(Ac)-) (4) as a heavy oil (0.19 g, 69 % after
silica gel chromatography, HPLC: Rt 8.33 min). HR-MS,
calcd for C33H49N4O6S2 ([M ? H]?) 661.3094, found
661.2859. 1H NMR: d 7.2 (m, 1H), 7.14 (d, J = 8 Hz, 1H),
6.84 (d, J = 10 Hz, 1H), 6.50 (d, J = 10 Hz, 1H), 6.14 (d,
J = 10 Hz, 1H), 5.1 (m, 2H), 4.93 (d, J = 15 Hz, 1H), 4.4
(m, 2H), 3.89 (t, J = 10 Hz, 1H), 3.4 (m, 1H), 3.0 (m, 1H),
2.8 (m, 4H), 2.31(d, J = 5 Hz, 6H), 2.2 (m, 1H), 1.9 (m,
2H), 1.7 (m, 4H), 1.5 (m, 6H), 1.4 (m, 6H), 0.96 (d,
J = 10 Hz, 3H), 0.90 (d, J = 5 Hz, 3H).
Cyclo(-L-Am7(Ac)-L-Am7(Ac)-L-Val-D-Pro-) (2)
Z-L-Val-D-Pro-OtBu was synthesized from Z-L-Val-OH
(3.3 g, 13 mmol), HClÁH-D-Pro-OtBu (2.5 g, 12 mmol),
DCC (3.0 g, 14 mmol), HOBtÁH2O (1.8 g, 12 mmol), and
triethylamine (12 mmol, 1.7 mL) to yield 4.3 g (11 mmol,
88 %) solid after chromatography. This dipeptide
(11 mmol) was deprotected with H2-Pd/C to yield H-L-Val-
D-Pro-OtBu (2.5 g, 9.3 mmol, 88 %), which was reacted
with Boc-L-Ab7-OH (3.0 g, 9.3 mmol) to obtain Boc-L-
Ab7-L-Val-D-Pro-OtBu as white foam (4.2 g, 7.3 mmol,
78 %). This tripeptide (7.3 mmol) was treated with 4 M
HCl/dioxane to generate H-L-Ab7-L-Val-D-Pro-OtBu
(2.8 g, 5.8 mmol, 80 %), which was again coupled with
Boc-L-Ab7-OH (1.9 g, 5.8 mmol) to obtain the tetrapeptide
Boc-L-Ab7-L-Ab7-L-Val-D-Pro-OtBu as white foam after
silica gel chromatography (3.6 g, 4.6 mmol, 80 %). After
TFA treatment, TFAÁH-L-Ab7-L-Ab7-L-Val-D-Pro-OH
(4.6 mmol), HATU (2.6 g, 7.0 mmol), and DIEA (2.0 mL,
12 mmol) were added in five portions to 0.77 L DMF in
every 10 min. After work-up and silica gel chromatogra-
phy, cyclo(-L-Ab7-L-Ab7-L-Val-D-Pro-) was obtained
(0.48 g, 0.78 mmol, 77 %). This cyclic peptide
(0.78 mmol) was reacted with KSAc (0.27 g, 2.4 mmol) in
6.2 mL DMF to afford cyclo(-L-Am7(Ac)-L-Am7(Ac)-L-
Val-D-Pro-) (2) as a solid after silica gel chromatography
(0.40 g, 0.67 mmol, 85 %, HPLC: Rt 7.45 min). HR-MS,
calcd for C28H47N4O6S2 ([M ? H]?) 599.2915, found
1
599.2977. H NMR: d 7.10 (d, J = 10 Hz, 2H), 6.41 (d,
J = 5 Hz, 1H), 4.75 (dd, J = 2 Hz, 1H), 4.43 (t,
J = 10 Hz, 1H), 4.2 (m, 1H), 3.9 (m, 1H), 3.5 (m, 2H),
2.84 (t, 4H), 2.4 (m, 1H), 2.3 (m, 8H), 2.2 (m, 1H), 1.9 (m,
1H), 1.69 (m, 2H), 1.6 (m, 5H), 1.3 (m, 9H), 0.99 (d, 3H),
0.90 (d, 3H).
Cyclo(-L-Am7(-)-D-Tic-L-Val-D-Am7(-)-) (5)
The solution of 4 (85 mg, 0.13 mmol) in 2.0 mL DMF was
treated with MeNH2 (0.13 mL, 1.3 mmol) in MeOH (r.t.
123

T. Arai et al.
5 h). After evaporation, the residue dissolved in 25 mL
DMF was reacted with I2 (33 mg, 0.13 mmol) in EtOH (r.t.
30 min). After LH-20 chromatography, cyclo(-L-Am7(-)-D-
Tic-L-Val-D-Am7(-)-) (5) was obtained as a solid (17 mg,
30 lmol, 23 %, HPLC: Rt 7.90 min). HR-MS, calcd for
C29H43N4O4S2 ([M ? H]?) 575.2726, found 575.2608. 1H
NMR: d 7.2 (m, 1H), 6.75 (d, J = 10 Hz, 1H), 6.53 (d,
J = 10 Hz, 1H), 6.20 (d, J = 10 Hz, 1H), 5.2 (m, 2H), 4.9
(m, 1H), 4.6 (m, 2H), 3.85 (t, J = 10, Hz, 1H), 3.39 (dd,
J = 6 Hz, 1H), 2.97 (dd, J = 8 Hz, 1H), 2.7 (m, 4H), 2.2
(m, 2H), 2.1 (m, 1H), 1.7 (m, 6H), 1.5 (m, 11H), 0.95 (d,
3H), 0.89 (d, 3H).
NMR: d 7.12 (d, J = 6 Hz, 1H), 6.42 (d, J = 10 Hz, 1H),
6.14 (d, J = 10 Hz, 1H), 5.0 (m, 1H), 4.71 (d, J = 8 Hz,
1H), 4.5 (m, 2H), 4.0 (m, 1H), 3.5 (m, 1H), 2.7 (m, 4H), 2.3
(m, 2H), 2.18 (t, 1H), 2.00 (m, 1H), 1.8 (m, 3H), 1.7 (m,
4H), 1.4 (m, 12H).
Cyclo(-L-Am7(Ac)-D-Am7(Ac)-L-Phe-D-Pro-) (10)
95 % yield from cyclo(-L-Ab7-D-Ab7-L-Phe-D-Pro-),
HPLC: Rt 7.98 min, HR-MS, calcd for C33H47N4O6S2
1
([M ? H]?), 647.2937, found 647.2960. H NMR: d 7.21
(m, 5H), 7.13 (d, J = 10 Hz, 1H), 6.47 (d, J = 10 Hz, 1H),
6.06 (d, J = 10 Hz, 1H), 5.08 (m, 1H), 4.63 (m, 1H), 4.38
(m, 1H), 4.27 (m, 1H), 3.50 (m, 2H), 3.07 (m, 2H), 2.85 (m,
4H), 2.32 (m, 6H), 1.97 (m, 2H), 1.70 (m, 4H), 1.56 (m,
4H), 1.34 (m, 10H).
Cyclo(-L-Am7(Ac)-D-Am7(Ac)-L-Val-D-Pro-) (6)
This compound was synthesized similarly to 2, using Boc-
D-Ab7-OH instead of Boc-L-Ab7-OH in the synthesis of
tripeptide. 95 % yield from cyclo(-L-Ab7-D-Ab7-L-Val-D-
Pro-), HPLC: Rt 7.82 min, HR-MS, calcd for
C28H47N4O6S2 ([M ? H]?) 599.2937, found 599.2977. 1H
NMR: d 7.16 (d, J = 10 Hz, 1H), 6.28 (d, J = 10 Hz, 1H),
6.06 (d, J = 10 Hz, 1H), 4.71 (dd, J = 3 Hz, 1H), 4.3 (m,
2H), 4.2 (m, 1H), 4.0 (m, 1H), 3.5 (m, 1H), 2.84 (t,
J = 8 Hz, 4H), 2.4 (m, 1H), 2.3 (m, 8H), 2.2 (m, 1H), 1.9
(m, 1H), 1.8 (m, 2H), 1.76 (s, 1H), 1.5 (m, 5H), 1.3 (m,
8H), 0.99 (s, 3H), 0.89 (s, 3H).
Cyclo(-L-Am7(-)-D-Am7(-)-L-Phe-D-Pro-) (11)
28 % yield from 10, HPLC: Rt 7.22 min, HR-MS, calcd for
C28H41N4O4S2 ([M ? H]?) 561.2569, found 561.2546. 1H
NMR: d 7.21 (m, 5H), 7.15 (d, J = 10 Hz, 1H), 6.56 (d,
J = 10 Hz, 1H), 6.10 (d, J = 10 Hz, 1H), 5.08 (m, 1H),
4.63 (m, 1H), 4.46 (m, 2H), 3.48 (m, 2H), 3.06 (m, 2H),
2.75 (m, 2H), 2.65 (m, 2H), 2.02 (m, 2H), 1.96 (m, 2H),
1.84 (m, 2H), 1.47 (m, 2H), 1.71 (m, 6H), 1.54 (m, 2H),
1.42 (m, 2H), 1.34 (m, 2H).
Cyclo(-L-Am7(-)-D-Am7(-)-L-Val-D-Pro-) (7)
Cyclo(-L-Am7(Ac)-D-Aom7(Ac)-L-Ala-D-Pro-) (12)
This compound was synthesized similarly to 3. 45 % yield
from 4, HPLC: Rt 7.03 min, HR-MS, calcd for
C24H41N4O4S2 ([M ? H]?) 513.2569, found 513.2585. 1H
NMR: d 7.18 (d, J = 10 Hz, 1H), 6.38 (d, J = 10 Hz, 1H),
6.09 (d, J = 10 Hz, 1H), 4.72 (dd, J = 2, 2 Hz, 1H), 4.5
(m, 1H), 4.4 (m, 1H), 4.37 (t, J = 10 Hz, 1H), 3.9 (m, 1H),
3.5 (m, 1H), 2.8 (m, 2H), 2.6 (m, 2H), 2.3 (m, 2H), 2.1 (m,
2H), 2.0 (m, 1H), 1.8 (m, 2H), 1.7 (m, 2H), 1.6 (m, 3H), 1.4
(m, 9H), 0.98 (d, 3H), 0.88 (d, 3H).
77 % yield from cyclo(-L-Ab7-D-Aob7-L-Ala-D-Pro-),
HPLC: Rt 5.84 min, HR-MS, calcd for C25H41N4O7S2
1
([M ? H]?) 573.2417, found 573.2424. H NMR: d 7.08
(d, J = 10 Hz, 1H), 6.42 (d, J = 10 Hz, 1H), 6.29 (d,
J = 10 Hz, 1H), 4.9 (m, 1H), 4.71 (dd, 1H), 4.6 (m, 1H),
4.3 (m, 1H), 4.0 (m, 1H), 3.5 (m, 5H), 3.07 (t, 2H), 2.85 (t,
2H), 2.33 (d, 6H), 2.1 (m, 1H), 1.9 (m, 1H), 1.8 (m, 3H),
1.75 (s, 1H), 1.6 (m, 3H), 1.3 (m, 8H).
Cyclo(-L-Am7(Ac)-D-Am7(Ac)-L-Ala-D-Pro-) (8)
Cyclo(-L-Am7(-)-D-Aom7(-)-L-Ala-D-Pro-) (13)
85 % yield from cyclo(-L-Ab7-D-Ab7-L-Ala-D-Pro-),
HPLC: Rt 6.50 min, HR-MS, calcd for C26H43N4O6S2
34 % yield from 12, HPLC: Rt 4.92 min, HR-MS, calcd for
C21H35N4O5S2 ([M ? H]?) 487.2049, found 487.2045. 1H
NMR: d 7.12 (d, J = 10 Hz, 1H), 6.57 (d, J = 10 Hz, 1H),
6.45 (d, J = 10 Hz, 1H), 4.9 (m, 1H), 4.7 (m, 2H), 4.4 (m,
1H), 3.9 (m, 1H), 3.6 (m, 5H), 2.9 (m, 2H), 2.8 (m, 1H), 2.7
(m, 1H), 2.4 (m, 1H), 2.3 (m, 1H), 2.1 (m, 1H), 2.0 (m,
4H), 1.8 (m, 2H), 1.6 (m, 3H), 1.4 (m, 5H).
1
([M ? H]?), 571.2624, found 571.2663. H NMR: d 7.09
(d, J = 10 Hz, 1H), 6.30 (d, J = 10 Hz, 1H), 6.05 (d,
J = 10 Hz, 1H), 4.9 (m, 1H), 4.70 (dd, J = 3 Hz, 1H), 4.3
(m, 2H), 4.0 (m, 1H), 3.5 (m, 1H), 2.84 (m, 4H), 2.3 (m,
8H), 1.9 (m, 5H), 1.5 (m, 6H), 1.3 (m, 10H).
Cyclo(-L-Am7(-)-D-Am7(-)-L-Ala-D-Pro-) (9)
Cyclo(-L-Am7(Ac)-D-Aom7(Ac)-L-Phe-D-Pro-) (14)
28 % yield from 8, HPLC: Rt 5.33 min, HR-MS, calcd for
C22H37N4O4S2 ([M ? H]?) 485.2256, found 485.2278. 1H
71 % from cyclo(-L-Ab7-D-Aob7-L-Phe-D-Pro-), HPLC: Rt
7.18 min, HR-MS, calcd for C31H45N4O7S2 ([M ? H]?)
123

Cyclic tetrapeptides with –SS– bridging between amino acid
1
649.2730, found 649.2756. H NMR: d 7.27 (m, 5H), 7.11
obtained by substituting Val in 6 with Ala and Phe (8 and
10, respectively), and by incorporating an ether linkage
into the side chains of 6 (to give 12, 14, and 16). Com-
pounds 3, 7, and 15 potently inhibited HDAC1 and
HDAC4, which were generated by the oxidation of 2, 6,
and 14, respectively.
(d, J = 10 Hz, 1H), 6.58 (d, J = 10 Hz, 1H), 6.30 (d,
J = 10 Hz, 1H), 5.09 (m, 1H), 4.63 (m, 2H), 4.27 (m, 1H),
3.5 (m, 6H), 3.1 (m, 2H), 3.0 (m, 2H), 2.85 (m, 2H),
2.3 * 2.0 (m, 10H), 1.8 (m, 2H), 1.7 (m, 4H), 1.33 (m,
4H).
Conflict of interest The authors declare that they have no conflict
of interest.
Cyclo(-L-Am7(-)-D-Aom7(-)-L-Phe-D-Pro-) (15)
36 % yield from 14, HPLC: Rt 6.43 min, HR-MS, calcd for
C27H39N4O5S2 ([M ? H]?) 563.2362, found 563.2315. 1H
NMR: d 7.27 (m, 5H), 7.15 (d, J = 10 Hz, 1H,), 6.71 (d,
J = 10 Hz, 1H), 6.40 (d, J = 10 Hz, 1H), 5.09 (m, 1H),
4.73 (m, 1H), 4.63 (dd, 1H), 4.42 (m, 1H), 3.83 and 3.17
(two m, 2H), 3.57 (m, 2H), 3.47 (m, 2H), 3.0 (m, 2H), 2.88
(m, 2H), 2.7 (m, 2H), 2.1 * 1.9 (m, 6H), 1.8 (m, 4H), 1.5
(m, 2H), 1.3 (m, 2H).
References
Furumai R, Matsuyama A, Kobashi N, Lee K-H, Nishiyama M,
Nakajima H, Tanaka A, Komatsu Y, Nishino N, Yoshida M,
Horinouchi S (2002) FK228 (depsipeptide) as a natural prodrug
that inhibits class
62:4916–4921
Han G, Tamaki M, Hruby VJ (2008) Fast, efficient and selective
deprotection of the tert-butoxycarbonyl (Boc) group using HCl/
dioxane (4 M). J Peptide Res 58:338–341
Hoque MA, Arai T, Nishino N, Kim H-J, Ito A, Yoshida M (2012)
Cyclic tetrapeptides with thioacetate tails or intramolecular
disulfide bridge as potent inhibitors of histone deacetylases.
Bioorg Med Chem Lett 22:6770–6772
Islam MS, Bhuiyan MPI, Islam MN, Nsiama TK, Oishi N, Kato T,
Nishino N, Ito A, Yoshida M (2011) Evaluation of functional
groups on amino acids in cyclic tetrapeptides in histone
deacetylase inhibition. Amino Acids 42:2103–2110
Kijima M, Yoshida M, Sugita K, Horinouchi S, Beppu T (1993)
Trapoxin, an antitumor cyclic tetrapeptide, is an irreversible
inhibitor of mammalian histone deacetylase. J Biol Chem
268:22429–22435
I histone deacetylases. Cancer Res
Cyclo(-L-Aom7(Ac)-D-Am7(Ac)-L-Phe-D-Pro-) (16)
64 % from cyclo(-L-Aob7-D-Ab7-L-Phe-D-Pro-), HPLC: Rt
7.51 min, HR-MS, calcd for C31H45N4O7S2 ([M ? H]?)
1
649.2730, found 649.2695. H NMR: d 7.24 (d, 5H), 7.16
(d, J = 10 Hz, 1H), 6.48 (d, J = 10 Hz, 1H), 6.17 (d,
J = 10 Hz, 1H), 5.09 (m, 1H), 4.63 (m, 1H), 4.54 (m, 1H),
4.38 (m, 1H), 3.5 (m, 6H), 3.2 * 2.9 (m, 6H), 2.3 * 2.0
(m, 10H), 1.8 (m, 2H), 1.7 (m, 4H), 1.34 (m, 4H).
Cyclo(-L-Aom7(-)-D-Am7(-)-L-Phe-D-Pro-) (17)
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 (2001) Cyclic hydroxamic-acid-containing peptide
31, a potent synthetic histone deacetylase inhibitor with antitu-
mor activity. Cancer Res 61:4459–4466
Masuoka Y, Nagai A, Shin-ya K, Furihata K, Nagai K, Suzuki K,
Hayakawa Y, Seto Y (2001) Spiruchostatins A and B, novel gene
expression-enhancing substances produced by Pseudomonas sp.
Tetrahedron Lett 42:41–44
36 % from 16, HPLC: Rt 7.04 min, HR-MS, calcd for
C27H39N4O5S2 ([M ? H]?) 563.2362, found 563.2354. 1H
NMR: d 7.24 (m, 5H), 7.16 (d, J = 10 Hz, 1H), 6.61 (d,
J = 10 Hz, 1H), 6.15 (d, J = 10 Hz, 1H), 5.09 (m, 1H),
4.70 (m, 1H), 4.63 (m, 1H), 4.42 (m, 1H), 3.5 (m, 6H), 3.1
(m, 2H), 2.8 (m, 4H), 2.0 (m, 6H), 1.7 (m, 4H), 1.52 (m,
2H), 1.38 (m, 2H).
Miller TA, Witter DJ, Belvedere S (2003) Histone deacetylase
inhibitors. J Med Chem 46:5097–5116
Nishino N, Jose N, Shinta R, Kato T, Komatsu Y, Yoshida M (2004)
Chlamydocin–hydroxamic acid analogues as histone deacetylase
inhibitors. Bioorg Med Chem 12:5777–5784
Conclusion
Pardi A, Billeter M, Wu¨thrich K (1984) Calibration of the angular
dependence of the amide proton-Ca proton coupling constants,
3JHNa, in a globular protein. J Mol Biol 180:741–751
Ueda H, Nakajima H, Hori Y, Fujita T, Nishimura M, Goto T,
Okuhara M (1994) 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
47:301–310
Cyclic tetrapeptides with intramolecular –SS– bridges were
synthesized by the oxidation of –SH side chains, which
were generated from –SAc precursors. Cyclo(-L-Am7(Ac)-
Aib-L-Am7(Ac)-D-Pro-) (1) did not afford intramolecular
–SS– bridged products, whereas cyclo(-L-Am7(Ac)-L-Am7
(Ac)-L-Val-D-Pro-) (2), cyclo(-L-Am7(Ac)-D-Tic-L-Val-D-
Am7(Ac)-) (4), and cyclo(-L-Am7(Ac)-D-Am7(Ac)-L-Val-
D-Pro-) (6) did afford –SS– bridged products. Various
cyclic peptides with intramolecular –SS– bridges were
Watanabe LA, Jose B, Kato T, Nishino N, Yoshida M (2004)
Synthesis of L-a-amino-x-bromoalkanoic acid for side chain
modification. Tetrahedron Lett 45:491–494
123