Structure-activity relationship of marinostatin, a serine protease inhibitor isolated from a marine organism

Article abstract of DOI:10.1002/psc.1244

A 12-residue MST isolated from a marine organism is a potent serine protease inhibitor that has a double cyclic structure composed of two ester linkages formed between the β-hydroxyl and β-carboxyl groups, Thr ³-Asp⁹ and Ser⁸

Full text of DOI:10.1002/psc.1244

Research Article
Received: 1 March 2010
Revised: 26 March 2010
Accepted: 26 March 2010
Published online in Wiley Interscience: 14 May 2010
(www.interscience.com) DOI 10.1002/psc.1244
Structure–activity relationship of
marinostatin, a serine protease inhibitor
isolated from a marine organism
Misako Taichi,a,b Tohimasa Yamazaki,c Kazuki Kawahara,d
Daisuke Motooka,d Shota Nakamura,e Shusaku Harada,d
Tadashi Teshima,a,b Tadayasu Ohkubo,d Yuji Kobayashif and
Yuji Nishiuchia,b∗
A 12-residue MST isolated from a marine organism is a potent serine protease inhibitor that has a double cyclic structure
composed of two ester linkages formed between the β-hydroxyl and β-carboxyl groups, Thr3-Asp9 and Ser8-Asp11. MST was
synthesized by a regioselective esterification procedure employing two sets of orthogonally removable side-chain protecting
groups for the Asp and Ser/Thr residues. In the MST molecule, there were no significant changes observed in yield by changing
the order of esterification. SAR study of MST revealed that the minimum required structure for expressing the inhibitory activity
is the sequence (1–9) in a monocyclic structure where Pro7 located in the ring plays a crucial role in keeping the structural
rigidity. By applying the structural motif of MST, we rationally designed protease inhibitory specificities that differ from those
c
of the natural product. Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd.
Keywords: ester linkage; MST; NMR analysis; peptide synthesis; SAR study; serine protease inhibitor; solution structure
Introduction
MST isolated from extracellular products of the marine organism,
Pseudoaltermonus sagamiensis, is a serine protease inhibitor
∗
Correspondence to: Yuji Nishiuchi, SAITO Research Center, Peptide Institute,
Inc., 7-2-9 Saito-Asagi, Ibaraki-shi, Osaka 567-0085, Japan.
consisting of 12 amino acids with two internal ester linkages
that are formed between the β-hydroxyl and β-carboxyl groups,
Thr3-Asp9 and Ser8-Asp11 (Figure 1) [1–3]. MST strongly inhibits
serine proteases such as subtilisin, chymotrypsin, and elastase at
an enzyme–inhibitor ratio of 1 : 1, but not trypsin [2]. To confirm
the reported primary structure of MST including ester linkages,
it was synthesized by a regioselective esterification procedure
employing two sets of orthogonally removable protecting groups
at the side-chains of Asp and Ser/Thr. Spectral and analytical
data of the synthetic MST were in good agreement with those
in the literature [4]. The inhibitory potency of the synthetic
MST against subtilisin (Ki, 0.59 0.05 nM) was found to be
comparable with a reported value for native MST (Ki, 1.5 nM)
[3]. Serine protease inhibitor loops generally show similar
canonical conformations with a hydrophobic core arising from
the disulfide linkages across the families, which is responsible
for the direct binding to the catalytic sites of proteases [5].
In the case of MST, however, it is considered that its ester
linkages may play a crucial role in expressing the canonical
inhibitory activity and that its inhibitory potential is attributable to
the hydrogen bond linking between the backbone NH proton
of Arg5 and the carbonyl oxygen atom of the ester linkage
with Thr3-Asp9, which protects the scissile bond of Met4-Arg5
[6].
E-mail: yuji@peptide.co.jp
a
SAITO Research Center, Peptide Institute, Inc., Ibaraki, Osaka 567-0085, Japan
b
Graduate School of Science, Osaka University, Toyonaka, Osaka 560-0043,
Japan
c
d
e
f
Protein Research Unit, National Institute of Agrobiological Science, Tsukuba,
Ibaraki 305-8602, Japan
Graduate School of Pharmaceutical Science, Osaka University, Suita, Osaka
565-0871, Japan
Research Institute for Microbial Diseases, Osaka University, Suita, Osaka 567-
0871, Japan
Osaka University of Pharmaceutical Sciences, Takatsuki, Osaka 569-1094,
Japan
Abbreviations used: 2-PySH, 2-mercaptopyridine; 4-MeBzl, 4-methylbenzyl;
6-Cl-HOBt, 6-chloro-1-hydroxybenzotriazole; All, allyl; AMC, 7-amino-4-
methylcoumarin;BrZ, 2-bromobenzyloxycarbonyl;CARL, subutilisinCarlsberg;
cHx, cyclohexyl; Cl2Bzl, 2,4-dichlorobenzyl; Dap, 2,3-diaminopropionic
acid; DQF-COSY, double quantum filtered-correlated spectroscopy;
EDC·HCl, 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride;
HCTU,
1-[bis(dimethylamino)methylene]-5-chloro-1H-benzotriazolium
3-oxide hexafluorophosphate; HFIP, hexafluoroisopropanol; HOHAHA,
homo-nuclear hartmann-hahn experiment; HOOBt, 3,4-dihydro-3-hydroxy-
4-oxo-benzotriazine; MNBA, 2-methyl-6-nitrobenzoic anhydride; MST,
marinostatin; NMP, 1-methyl-2-pyrrolidinone; OMTKY3, turkey ovomucoid
third domain; Pen, 3-pentyl; TASF, tris(dimethylamino)sulfonium difluoro-
The relatively small and unique structure of MST prompted us
to explore the characteristics including such a strong inhibitory
potency. In the present study, we examined the effect of the order
trimethylsilicate; TBS,
2-chlorotrityl.
t-butyldimethylsilyl; Tos,
tosyl;
Trt(2-Cl),
c
J. Pept. Sci. 2010; 16: 329–336
Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd.

TAICHI ET AL.
FastMoc protocols of coupling with Fmoc amino acid/HCTU/6-
Cl-HOBt/DIEA (4/4/4/8 equiv.) in NMP (single coupling, 20 min).
The protected peptide resin (0.86 g, 0.25 mmol) was treated with
HFIP/CHCl3 (1/4, v/v, 20 ml) for 1 h. The resin was filtered and the
filtrate was concentrated in vacuo to give an oily residue, which
was precipitated with ether; yield 0.51 g (quant.); ESI-MS: 1002.6
([M+2H]2+), theoretical value: 1002.5.
Met4
S
Phe1
H2N
Ala2
Thr3
O
O
N
H
N
H
Arg5
NH2
O
HN
O
N
H
N
H
O
N
H
O
O
Asp9
O
Ser10
H
N
Z-Phe-Ala-Thr(TBS)-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser(tBu)-Asp(OAll)-
Ser(Bzl)-Asp(OtBu)-Glu(OcHx)-OBzl
HO
NH
OH
O
NH
Tyr6
O
O
O
To
a solution of Z-Phe-Ala-Thr(TBS)-Met-Arg(Tos)-Tyr(Pen)-
N
Asp11
O
HN
Pro-Ser(tBu)-Asp(OAll)-Ser(Bzl)-Asp(OtBu)-OH (0.50 g, 0.25 mmol),
Glu(OcHx)-OBzl·HCl (93 mg, 0.26 mmol), and HOOBt (43 mg,
0.26 mmol)inDMF(5 ml)at0 ◦CwasaddedEDC(46 µl, 0.26 mmol).
The reaction mixture was stirred for 1.5 h at room temperature
and poured into an iced water to give a residue that was extracted
with CHCl3. The organic layer was washed successively with 5%
NaHCO3, water, 5% citric acid, water, and brine, and dried over
anhydrous MgSO4. Precipitation from CHCl3 and ether yielded
0.51 g (90%). ESI-MS: 1153.6 ([M+2H]2+), theoretical value: 1153.1.
NH
O
Pro7
HN
O
O
HO
Glu12
Ser8
O
HO
Figure 1. Structure of MST (1).
Z-Phe-Ala-Thr(TBS)-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser(tBu)-Asp-
Ser(Bzl)-Asp(OtBu)-Glu(OcHx)-OBzl
of esterification on the efficiency to form the intramolecular ester
linkage by changing its combination of the side-chain protecting
groups for Asp and Ser/Thr. Next, we synthesized a series of
MST analogs and measured their inhibitory activities to elucidate
the structure–activity relationship (SAR) by employing the same
synthetic strategy as that for MST. Furthermore, we rationally
designed protease inhibitory specificities that are different from
those of the natural product by applying the structural motif
of MST.
To
a solution of Z-Phe-Ala-Thr(TBS)-Met-Arg(Tos)-Tyr(Pen)-
Pro-Ser(tBu)-Asp(OAll)-Ser(Bzl)-Asp(OtBu)-Glu(OcHx)-OBzl (0.50 g,
0.22 mmol) in DMF (5 ml) under an argon atmosphere were added
Pd(PPh3)4 (50 mg, 43 µmol) and morpholine (0.19 ml, 2.2 mmol)
at room temperature. The reaction mixture was stirred for 1 h,
quenched with water at 0 ◦C, and then extracted with CHCl3.
The organic layer was washed successively with 5% citric acid,
water and brine, and dried over anhydrous MgSO4. The filtrate
was concentrated in vacuo. Purification of the resulting residue
by flash column chromatography [silica gel 30 g, CHCl3/methanol
100/0–10/1 (v/v)] gave 0.34 g (68%). ESI-MS: 1133.6 ([M+2H]2+),
theoretical value: 1133.1.
Materials and Methods
Boc, Fmoc, and Z-amino acids were obtained from the Peptide
Institute, Inc. (Osaka, Japan) or Watanabe Chemical Industries
(Hiroshima, Japan). All other chemicals and solvent of special
grade were obtained from Nacalai Tesque (Kyoto, Japan) and were
used without further purification. Syntheses were carried out on
a 433A peptide synthesizer (Applied Biosystems, Foster City, CA,
USA). Amino acid analysis was carried out on a Hitachi L-8800
(Ibaraki, Japan) after hydrolysis of the peptide with 6 N HCl at
110 ◦C for 22 h. HPLC was performed on an HP G1600A (Palo Alto,
VA, USA) and a Shimadzu liquid chromatograph Model LC-8A or
LC-20AT (Kyoto, Japan). Molecular weights were measured with
an ESI-MS (HP1100 LC/MSD) and a MALDI-TOF MS (Voyager-DE
STR, Applied Biosystems, Framingham, MA, USA).
Z-Phe-Ala-Thr-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser(tBu)-Asp-Ser(Bzl)-
Asp(OtBu)-Glu(OcHx)-OBzl
To
a
solution of Z-Phe-Ala-Thr(TBS)-Met-Arg(Tos)-Tyr(Pen)-
(0.33 g,
Pro-Ser(tBu)-Asp-Ser(Bzl)-Asp(OtBu)-Glu(OcHx)-OBzl
0.15 mmol) in DMF (3 ml) was added TASF (0.20 g, 0.73 mmol) at
room temperature. After being stirred for 50 min, the product
was precipitated with water at 0 ◦C, and collected by filtration.
The product was purified by reprecipitation from CHCl3/methanol
(3/1, v/v) and ether, and dried in vacuo: yield 0.29 g (93%); ESI-MS:
1076.2 ([M+2H]2+), theoretical value: 1076.0.
Z-Phe-Ala-Thr-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser(tBu)-Asp-Ser(Bzl)-
Asp(OtBu)-Glu(OcHx)-OBzl (with the ester linkage of Thr3-Asp9)
Synthesis of MST (1)
Z-Phe-Ala-Thr(TBS)-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser(tBu)-Asp(OAll)-
Ser(Bzl)-Asp(OtBu)-OH
To a stirred solution of MNBA (29 mg, 84 µmol) and DMAP
(4.3 mg, 35 µmol) in anhydrous DCM (13 ml) were added a solution
ofZ-Phe-Ala-Thr-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser(tBu)-Asp-Ser(Bzl)-
Asp(OtBu)-Glu(OcHx)-OBzl (0.15 g, 70 µmol) in anhydrous DCM
(5.6 ml) and a solution of TEA (21 µl, 0.15 mmol) in anhydrous DCM
(5.6 ml) simultaneously using mechanical driven syringes for 3.5 h.
After being stirred for 20 h at room temperature, the reaction
mixture was quenched with H2O at 0 ◦C and then extracted with
chloroform. The organic layer was washed successively with 5%
NaHCO3, water, 5% citric acid, water, and brine, and then dried
The protected MST(1–11) was synthesized with an ABI 433A
peptide synthesizer using Fmoc strategy on an Asp(OtBu)-Trt(2-Cl)
resin (0.32 g, 0.25 mmol) [7]. The following side-chain-protected
Fmoc amino acids were employed: Arg(Tos), Asp(OAll), Asp(OtBu),
Ser(tBu), Ser(Bzl), Thr(TBS), and Tyr(Pen) [8]. Z-Phe was used
for N-terminal incorporation. Deprotection of Nα-Fmoc groups
was performed using 20% morpholine (3 min ×5) to reduce the
aspartimide formation [9]. The peptide chain was elongated using
c
www.interscience.com/journal/psc Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2010; 16: 329–336

SAR STUDY OF MARINOSTATIN
over anhydrous MgSO4. The filtrate was concentrated in vacuo.
The product was precipitated with ether, and the precipitates
were collected by filtration and dried in vacuo: yield 0.12 g (81%);
ESI-MS: 1067.1 ([M+2H]2+), theoretical value: 1067.0.
resin which was assembled onto a Fmoc-Asp[OTrt(2-Cl) resin]-
Glu(OcHx)-OBzl (0.39 g, 0.25 mmol) as described above for MST (1)
to yield 0.41 g (70%). The ester formation of Ser8-Asp11, followed
by that of Ser3-Asp9, was performed as reported previously [4] to
obtain 39 mg (60%). ESI-MS: 1368.5 ([M+H]+), theoretical value:
1368.6; amino acid analysis: Asp 1.98 (2), Ser 2.62 (3), Glu 1.00 (1),
Ala 1.02 (1), Met 0.89 (1), Tyr 0.96 (1), Phe 0.97 (1), Arg 0.97 (1), Pro
1.00 (1).
Z-Phe-Ala-Thr-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser-Asp-Ser(Bzl)-Asp-
Glu(OcHx)-OBzl (with the ester linkages of Thr3-Asp9 and
Ser8-Asp11
)
[Ser8-Asp11]-MST(1–12) (3)
Z-Phe-Ala-Thr(tBu)-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser-Asp(OtBu)-
Ser(Bzl)-Asp-Glu(OcHx)-OBzl with the ester linkage of Thr3-Asp9
(0.29 g, 0.14 mmol) was dissolved in TFA (100 ml) at 0 ◦C, and
then allowed to react for 1 h at room temperature. After removal
of excess TFA in vacuo, the product was precipitated with
diethylether, and collected by filtration, and dried to yield 0.27 g
(97%); ESI-MS: 1011.4 ([M+2H]2+), theoretical value: 1010.9. A
solution of this peptide (50 mg, 25 µmol) and TEA (3.4 µl, 25 µmol)
in anhydrous DCM/DMF (5/1, v/v, 2.4 ml) was added to a solution
of MNBA (10 mg, 30 µmol), DMAP (1.5 mg, 12 µmol), and TEA
(4.1 µl, 25 µmol) in anhydrous DCM/DMF (5/1, v/v, 3.6 ml) with a
mechanical driven syringe for 3.5 h. After being stirred for 19 h
at room temperature, the reaction mixture was quenched with
water at 0 ◦C, and then extracted with chloroform. The organic
layer was washed successively with 5% NaHCO3, water, 5% citric
acid, water, and brine, and then dried over anhydrous MgSO4. The
filtrate was concentrated in vacuo. The product was precipitated
with ether and the precipitates were collected by filtration, and
dried to yield 40 mg (81%). ESI-MS: 1002.5 ([M+2H]2+), theoretical
value: 1001.9.
This peptide was prepared as described above for [Ser3]-
MST(1–12) (2) to yield 39 mg (60%). ESI-MS: 1400.5 ([M+H]+),
theoretical value: 1400.6; amino acid analysis: Asp 2.02 (2), Thr 0.98
(1), Ser 1.82 (2), Glu 1.02 (1), Ala 1.01 (1), Met 0.94 (1), Tyr 1.00 (1),
Phe 1.01 (1), Arg 1.00 (1), Pro 1.03 (1).
Linear MST(1–12) (4)
The protected peptide was synthesized with an ABI 433A peptide
synthesizer on a Boc-Glu(OcHx)-PAM resin (0.59 g, 0.5 mmol)
using in situ neutralization protocols of coupling with Boc-amino
acid/HCTU/6-Cl-HOBt/DIEA (4/4/4/6 eq) in NMP (single coupling,
acetylation after each coupling step). The following side-chain-
protected Boc-amino acids were employed: Asp(OcHx), Arg(Tos),
Ser(Bzl),Thr(Bzl),andTyr(BrZ).Thepeptideresin(0.96 g,0.25 mmol)
was treated with HF (12 ml) in the presence of p-cresol (3.0 ml,
28 mmol) and 2-PySH (0.14 g, 1.3 mmol) at −2 to −5 ◦C for 1 h to
give a crude product, which was purified by RP-HPLC as described
above to obtain 0.21 g (60%). ESI-MS: 1418.2 ([M+H]+), theoretical
value: 1418.5; amino acid analysis: Asp 1.99 (2), Thr 0.97 (1), Ser
1.79 (2), Glu 1.00 (1), Ala 0.99 (1), Met 0.98 (1), Tyr 0.99 (1), Phe 0.99
(1), Arg 0.99 (1), Pro 1.04 (1).
MST (1)
Z-Phe-Ala-Thr-Met-Arg(Tos)-Tyr(Pen)-Pro-Ser-Asp-Ser(Bzl)-Asp-
Glu(OcHx)-OBzl with the ester linkages of Thr3-Asp9 and
Ser8-Asp11 (70 mg, 35 µmol) was treated with HF (2 ml) in the
presence of p-cresol (0.50 ml, 4.8 mmol) and 2-PySH (20 mg,
0.18 mmol) at −2 to −5 ◦C for 1 h to give a crude product,
which was purified by RP-HPLC using a YMC-Pak ODS column
(30 × 250 mm) at a flow rate of 20 ml/min. The RP-HPLC run was
eluted with a linear gradient of MeCN in 0.1% TFA (10–30%,
80 min) to obtain 21 mg (44%). MALDI-TOF MS: 1383.38 ([M+H]+),
theoretical value: 1383.46 ([M+H]+); amino acid analysis: Asp 1.98
(2), Thr 0.98 (1), Ser 1.99 (2), Glu 1.00 (1), Ala 1.04 (1), Met 0.94
[Thr3-Asp9]-MST(1–12) (5)
This peptide was prepared as described above for MST (1) to yield
36 mg (45%). ESI-MS: 1400.5 ([M+H]+), theoretical value: 1400.6;
amino acid analysis: Asp 2.00 (2), Thr 0.96 (1), Ser 1.80 (2), Glu 1.01
(1), Ala 1.00 (1), Met 0.94 (1), Tyr 0.98 (1), Phe 1.00 (1), Arg 1.00 (1),
Pro 1.02 (1).
[Thr3-Asp9]-MST(1–9) (6)
Protected peptide was synthesized with an ABI 433A peptide
synthesizer using Fmoc chemistry on an Fmoc-Asp(Wang resin)-
OBzl(0.40 g, 0.25 mmol). Thefollowing side-chain-protectedFmoc
amino acids were employed: Arg(Tos), Ser(Bzl), Thr(tBu), and
Tyr(Cl2Bzl). Z-Phe was used for N-terminal incorporation. The
chain assembly was performed by employing the same conditions
as those for MST (1) except for the Fmoc deprotection which was
carried out using 20% piperidine/NMP (2.5 min ×4). This peptide
was prepared as described above for [Ser3]-MST(1–12) (2) to yield
26 mg (52%). ESI-MS: 1069.4 ([M+H]+), theoretical value: 1069.5;
amino acid analysis: Asp 1.00 (2), Thr 0.97 (1), Ser 0.93 (1), Ala 1.01
(1), Met 0.90 (1), Tyr 0.99 (1), Phe 0.99 (1), Arg 1.00 (1), Pro 1.00 (1).
1
(1), Tyr 1.00 (1), Phe 0.99 (1), Arg 0.97 (1), Pro 1.00 (1); H-NMR
[H2O/D2O (v/v, 9/1)]: Phe1 δ 4.24 (Hα), 3.24, 3.09 (Hβ), 7.26 (2, 6H),
7.35 (3, 5H), 7.32 (4H). Ala2 8.64 (NH), 4.47 (Hα), 1.35 (Hβ). Thr3
8.72 (NH), 4.58 (Hα), 5.52 (Hβ), 1.35 (Hγ ). Met4 8.95 (NH), 4.66 (Hα),
2.32, 1.81 (Hβ), 2.58, 2.44 (Hγ ). Arg5 7.72 (NH), 4.10 (Hα), 1.79, 1.70
(Hβ) 1.70, 1.53 (Hγ ), 3.26, 3.18 (Hδ), 7.36 (NH), 6.94, 6.48 (NH2).
Tyr6 8.52 (NH), 4.52 (Hα), 2.96, 2.86 (Hβ), 7.11 (2, 6H), 6.83 (3, 5H).
Pro7 3.49 (Hα), 1.76, 1.63 (Hβ), 1.65, 1.56 (Hγ ), 3.62, 3.34 (Hδ). Ser8
7.19 (NH), 4.66 (Hα), 4.92, 4.25 (Hβ). Asp9 9.29 (NH), 4.47 (Hα), 2.97,
2.97 (Hβ). Ser10 8.59 (NH), 4.55 (Hα), 3.84, 3.79 (Hβ). Asp11 7.83
(NH), 4.78 (Hα), 2.72, 2.63 (Hβ). Glu12 8.36 (NH), 4.23 (Hα), 2.14,
1.93 (Hβ), 2.41, 2.41 (Hγ ).
[Thr3-Asp9]-MST(3–9) (7)
This peptide was prepared as described above for [Thr3-Asp9]-
MST(1–9) (6) to yield 26 mg (52%). ESI-MS: 851.4 ([M+H]+),
theoretical value: 851.4; amino acid analysis: Asp 1.00 (1), Thr
0.95 (1), Ser 0.88 (1), Met 0.92 (1), Tyr 0.97 (1), Arg 0.99 (1), Pro
1.00 (1).
Synthesis of MST Analogs
[Ser3]-MST(1–12) (2)
Z-Phe-Ala-Ser(tBu)-Met-Arg(Tos)-Tyr(Pen)-Ser(TBS)-Asp(OtBu)-
Ser(Bzl)-Asp-Glu(OcHx)-OBzl was detached from the peptide
c
J. Pept. Sci. 2010; 16: 329–336 Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. www.interscience.com/journal/psc

TAICHI ET AL.
[Cys3-Cys9, Cys8-Cys11]-MST(1–12) (8)
1.0 mM CaCl2 for 1 min at 30 ◦C. The reaction was started by
the addition of 0.1 ml of the corresponding substrate solution. The
releaseofAMCwasmonitoredbymeasuringfluorescenceintensity
at 440 nm with excitation at 350 nm. The residual enzyme activity
was plotted against the concentration of the peptides to calculate
the Ki value [7].
This peptide was prepared as reported previously [3]. ESI-MS:
1408.4 ([M+H]+), theoretical value: 1408.5; amino acid analysis:
Ser 0.92 (1), Glu 1.00 (1), Ala 1.01 (1), 1/2(Cys)2 2.70 (4), Met 0.94
(1), Tyr 0.98 (1), Phe 0.98 (1), Arg 0.98 (1), Pro 0.99 (1).
[Dap3-Asp9, Dap8-Asp11]-MST(1–12) (9)
NMR Spectroscopy
Protected peptide was synthesized with an ABI 433A peptide
synthesizer using Fmoc chemistry on an Fmoc-Asp[OTrt(2-Cl)
resin]-Glu(cHx)-OBzl (0.56 g, 0.25 mmol). The following side-chain-
protectedFmocaminoacidswereemployed:Dap(Boc),Dap(Alloc),
Arg(Tos), Asp(OtBu), Ser(Bzl), and Tyr(Cl2Bzl). Z-Phe was used for
N-terminal incorporation. The chain assembly was performed by
employing the same conditions as those for MST (1) to obtain
0.80 g. After removal of the Alloc group on Dap8 by treatment with
Pd(PPh3)4/Me2NH·BH3 (0.24 g, 4.1 mmol) in DCM, the protected
peptide was detached from the Trt(2-Cl) resin by treatment with
CHCl3/HFIP (20 ml, v/v = 4/1) to yield 0.32 g (68%). To a solution
of the protected peptide (0.55 g, 0.24 mmol) in DMF (83 ml) was
added EDC·HCl (93 mg, 0.49 mmol) and HOBt (66 mg, 0.49 mmol).
After being stirred at room temperature for 19 h, the reaction
mixture was concentrated in vacuo to give the residue, which was
precipitated from CHCl3/methanol and ether to yield 0.51 g (93%).
After treating this peptide (0.26 g, 0.12 mmol) with TFA (5 ml) for
1 h, the second amide linkage with Dap3-Asp9 was formed by
adding EDC (25 µl, 0.14 mmol) and HOBt (19 mg, 0.14 mmol) in
DMF(37 ml). Thereactionmixturewasstirredatroomtemperature
for22 h, andthenconcentratedinvacuotoyieldtheresidue, which
was precipitated from CHCl3/methanol and ether to yield 0.24 g
(97%). The HF treatment of the peptide (0.15 g, 72 µmol) was
followed by purification on RP-HPLC as described above to give
48 mg (48%). ESI-MS: 1366.5 ([M+H]+), theoretical value: 1366.6;
amino acid analysis: Asp 1.98 (2), Ser 0.87 (1), Glu 1.00 (1), Ala 1.00
(1), Met 0.89 (1), Tyr 0.95 (1), Phe 0.97 (1), Arg 0.97 (1), Pro 0.96 (1).
NMR samples were prepared by dissolving the synthetic peptides
in 200 µl of 100% D2O or 90% H2O/10% D2O solution to give
final concentrations of approximately 4 mM (6 mg/ml). All NMR
spectra were acquired at 10 ◦C on a Bruker DMX750 spectrometer
equipped with pulse-field gradients. Complete sequence-specific
1H resonance assignments were obtained according to the
standard strategy of the spin system identification followed by
sequential assignments [8] using 2D DQF-COSY spectrum, 2D
HOHAHA spectrum with a mixing time of 70 ms, and 2D NOESY
spectrum with a mixing time of 150 ms. 13C resonances were
assigned using 2D 1H-13C HSQC spectrum together with the
results of 1H resonance assignments. 3J coupling constants for
nonoverlapping signals were measured using 1D 1H spectra
with high S/N and digital resolution, where possible, and from
separations between antiphase absorptive and dispersive peak
extrema observed in the DQF-COSY spectra. All NMR spectra were
processed using XWINNMR (Bruker Instruments, Rheinstetten,
Germany) and the NMRPipe software [9], and were analyzed using
the PIPP software [10]. 1H and 13C chemical shifts were referenced
indirectly to TSP [sodium 3-(trimethylsilyl)propionate].
Structure Calculations
NOE-derived interproton distance restraints were classified into
three ranges: 1.8–2.7 Å, 1.8–3.5 Å, 1.8–5.0 Å, corresponding to
strong, medium, and weak NOEs. An additional 1.5 Å was added
to the upper bound for NOEs involving methyl protons and
methylene protons that were not assigned stereospecifically
[11]. Torsion angle restraints on φ and ψ were derived from
3JHNHα coupling constants and short-range NOEs. The side-
chain χ1 and χ2 angle restraints were derived from short-range
[Ala7]-MST(1–12) (10)
This peptide was prepared as described above for [Ser3]-
MST(1–12) (2) to yield 21 mg (27%). ESI-MS: 1356.5 ([M+H]+),
theoretical value: 1356.6; amino acid analysis: Asp 2.05 (2), Thr 0.99
(1), Ser 1.90 (2), Glu 1.04 (1), Ala 2.05 (2), Met 0.97 (1), Tyr 0.96 (1),
Phe 1.01 (1), Arg 1.00 (1).
3
3
NOEs combined with JHαHβ and JHβHγ coupling constants,
respectively. Structures of MST analogs were calculated using the
hybrid distance geometry-dynamic simulated annealing method
as contained in X-PLOR 3.1 [12]. For structure calculations of MST
(1), 167 interproton distance restraints (comprising 73 intra- and
94 inter-residue restraints) obtained from 2D NOESY spectra were
employed along with two distance restraints for one hydrogen
bond between the NH proton of Arg5 and the carbonyl oxygen
of the ester linkage with Thr3-Asp9. In addition, 19 dihedral angle
restraints (10 φ, 2 ψ, 6 χ1 and 1 χ2) were included in the structure
calculation. A peptide bond between Tyr6 and Pro7 was set to a cis
configuration, i.e. ω ≈ 0◦, based on observations of an extremely
strong sequential Hα –Hβ NOE between these two residues and
a large difference between Pro7Cβ and Cγ chemical shifts. For
structure calculations of MST[P7A] (10), 137 interproton distance
restraints (65 intra- and 72 inter-residue restraints) and 24 dihedral
angle restraints (10 φ 6 ψ, 7 χ1 and 1χ2) were used. For both cases,
a final set of 20 lowest energy structures were selected from 100
calculations. None of them had NOE or dihedral angle violations
of >0.5 Å or 5◦, respectively.
[Arg4, Met5]-MST(1–12) (11)
This peptide was prepared as described above for [Ser3]-
MST(1–12) (2) to yield 17 mg (33%). ESI-MS: 1382.5 ([M+H]+),
theoretical value: 1382.6; amino acid analysis: Asp 1.97 (2), Thr 0.95
(1), Ser 1.78 (2), Glu 1.00 (1), Ala 0.99 (1), Met 0.89 (1), Tyr 0.95 (1),
Phe 0.98 (1), Arg 0.98 (1), Pro 1.00 (1).
Protease Inhibitory Activities
The inhibitory activity of synthetic peptides against subtilisin or
trypsin was measured according to a literature procedure [3]. The
substrate (Suc-Ala-Ala-Pro-Phe-MCA for subtilisin, Bz-Arg-MCA for
trypsin)obtainedfromthePeptideInstitute,Inc.(Osaka,Japan)was
dissolved in DMSO at a concentration of 25 µM. CARL (Calbiochem,
Germany, 110 nM) or bovine trypsin (Sigma-Aldrich, St. Louis,
MO, USA, 110 nM) was incubated with an appropriate amount of
peptides in 2.0 ml of 25 mM phosphate buffer (pH 7.0) containing
c
www.interscience.com/journal/psc Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2010; 16: 329–336

SAR STUDY OF MARINOSTATIN
Figure 2. Regioselective formation of intramolecular ester linkages for MST. Route A, the previous work; route B, this work. Analogs 2, 3, 9, 10, and 11,
and analogs 5 and 8 were synthesized according to route A and B, respectively. The substitution of Ser for Thr3 was done for analog 2. In the case of 8
and 9, Cys3,8-Cys9,11 and Dap3,8-Asp9,11, respectively, were substituted for Thr3/Ser8-Asp9,11
.
using Fmoc chemistry and detached from the resin by treatment
with CHCl3/HFIP (v/v, 4/1) to obtain in the form of the fully pro-
tected segment with a free α-carboxyl group [14]. During the chain
assembly using repeated Fmoc deprotection performed by 20%
piperidine/NMP (2.5 min ×4), the Asp(OAll)9-Ser(tBu)10 sequence
quantitatively converted to the aspartimide one, Asi-Ser(tBu) [15].
The Asi formation could be significantly reduced (∼18%) by sub-
stituting 20% morpholine/NMP for 20% piperidine/NMP, although
prolonged deprotection steps (3 min ×5) were required for com-
plete removal of the Fmoc group.
Table 1. Effect of the order of esterification reaction mediated by the
MNBA/DMAP method on the ratio of products
Ratio of products (monomer : dimera : Asi-peptide)
First esterification
Second esterification
Route A
Route B
94 : 3 : 3b
93 : 5 : 2c
92 : 6 : 2c
90 : 2 : 8b
a Dimer refers to cyclic and linear dimerized peptides.
b Asi11-peptide.
c Asi9-peptide.
To evaluate the efficacy to produce the intramolecular ester
linkages by changing the order of ester formation precisely in the
MST molecule, the same reaction conditions, as those used for
the previous study, were employed. That is, a pseudo-high dilution
procedure, involving a slow addition of the linear-peptide into
the reaction mixture, with the aid of MNBA/DMAP [16], which
proved to be essential not only for accelerating intramolecular
esterification but also for preventing intermolecular esterification
and the Asi formation at Asp9 or Asp11. The formation of the first
ester linkage with Thr3-Asp9 after successive removal of the All
group and the TBS group, followed by that of the second linkage
with Ser8-Asp11 after removal of the tBu groups, was sequentially
carried out to produce the protected MST. In the MST molecule,
however, no significant changes in terms of the efficiency yielding
individual ester linkages were observed by changing the order
of esterification as shown in Table 1. The protected MST thus
obtained was treated by HF in the presence of p-cresol and 2-PySH
[17] at −2 to −5 ◦C for 1 h to give a crude product, which was
purified by RP-HPLC and characterized by amino acid analysis,
MALDI-TOF MS and NMR.
Results and Discussion
Effect of Changing the Order of Esterification on the Synthesis
of MST
MST was synthesized by a regioselective esterification procedure
employing two sets of orthogonally removable side-chain pro-
tecting groups for the Asp and Ser/Thr residues as previously
reported [4]. This strategy can allow esterification to take place
sequentially. In this synthesis, MST was constructed by forming
the first ester linkage between Ser8 and Asp11 after removal of
TBS and Trt(2-Cl) groups from the respective side-chain functions,
followed by the second one between Thr3 and Asp9, after re-
t
moval of their Bu groups (Route A in Figure 2). In the present
study, to examine whether the order of esterification can affect
the efficiency to form the intramolecular ester linkage, we also
synthesized the alternative starting linear-peptide, in which the
side-chain functional groups of Ser8 and Asp11 were protected
t
with Bu groups, and those of Thr3 and Asp9 with TBS and allyl
(All) groups, respectively (Route B in Figure 2). This combination
of the protecting groups for Asp and Ser/Thr allowed for changing
the order of esterification. The other side-chain protecting groups
and the N/C-terminal ones must be compatible with both Boc and
Fmoc chemistry and must be readily removed by the final HF treat-
ment. Considering this, the 3-pentyl (Pen) group was chosen to
protect the phenolic hydroxyl group of Tyr6, as it is a base-resistant
protecting group and cleavable by HF without the formation of
any significant amounts of the alkyltyrosine rearrangement prod-
uct [13]. The linear-peptide was assembled in solution from the
N-terminal segment (1–11) and the C-terminal Glu(OHex)-OBzl.
The former was elongated on an Fmoc-Asp(OtBu)-Trt(2-Cl) resin
Structure–Activity Relationship of MST
It is noted that the residues of OMTKY3 in the OMTKY3–CARL
complex and the corresponding residues of the solution structure
of MST were well superimposed as previously reported [6]. In
addition, the Ramachandran angles of the MST residues coincide
with those of OMTKY3 binding loop [18]. Thus, it is expected that
MST binds to an active pocket of CARL in the same manner as
OMTKY3. To achieve a presumption of the interaction mechanism
between MST and proteases, we developed a MST–CARL docking
model using the crystal structure of OMTKY3–CARL complex as a
template. The distances for each contact are summarized in
c
J. Pept. Sci. 2010; 16: 329–336 Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. www.interscience.com/journal/psc

TAICHI ET AL.
(A)
Table 3. MST analogs and their Ki values against CARL
Peptide
Ki (nM)a
Conformation of AA7
MST (1)
0.59 0.05
0.87 0.20
>50 000
cis
2
cis
3
–
4
>50 000
–
5
14.0 1.6
17.9 1.8
>50 000
cis : trans = 85: 15
cis : trans = 85: 15
cis : trans = 85: 15
cis : trans = 70: 30
cis : trans = 70: 30
trans
6
7
8
92.7 3.8
23.4 0.8
84.3 1.2
9
10
(B)
a Mean of 4–6 determinations SD.
Figure 3. (A) Active site representation of the MST–CARL docking model.
The docking model was constructed manually, by superimposing the Cα
atoms from P3ꢁ to P4residues of MSTto those of OMTKY3bounded to CARL
(PDB code, IYU6). The six residues of MST (carbon, white; nitrogen, blue;
oxygen, red)areshownasstickmodel. TheresiduesofCARL(carbon, green;
nitrogen, blue; oxygen, red) interacting with MST and those of catalytic
triad are also shown. (B) Active site representation of the OMTKY3–CARL
complex crystal structure. The six residues of OMTKY3 (carbon, cyan;
nitrogen, blue; oxygen, red) are shown as stick model. These images were
made with PyMol.
Table 2. Distances within 4 Å between CARL and OMTKY3/MST
CARL–OMTKY3a
(Å)
CARL–MST
(model, Å)
Contact
P1 (N)–Ser 125 (CO)
P1 (N)–Ser 221 (O)
P1 (CO)–Asn 155 (N)
P1 (CO)–Ser 221 (N)
P2 (N)–Gly 100 (CO)
P2ꢁ (N)–Asn218 (CO)
P3 (N)–Gly127 (CO)
P3 (CO)–Gly127 (N)
P4 (N)–Gly 102 (CO)
P4 (CO)–Gly 102 (N)
3.30
2.96
2.71
3.21
3.64
2.83
3.10
2.95
3.07
2.89
3.73
3.33
–
2.84
2.88
1.94
3.32
2.81
2.66
3.36
Figure 4. Structures of MST (1) and its analogs 2–10.
Ser8 and Asp11 (3, 4, and 5), truncating the peptide chain (6 and
7), substituting the disulfide (8) or amide bonds (9) for the ester
linkages, and substituting Ala for Pro7 (10). These analogs were
synthesized as in the case of MST by applying the regioselective
ester, amide or disulfide formation with two sets of orthogonally
removable side-chain protecting groups for Ser/Thr and Asp, Dap
and Asp, or Cys and Cys, respectively. The peptides were evaluated
for the potency inhibiting protease activity of CARL (Table 3). The
inhibitory potency was not affected when the MST structure was
modified by replacement of Thr3 with Ser (2) to remove the
β-methyl group. Opening of the ester linkage between Thr3 and
a Data from literature [18].
Table 2. This indicated that the six residues MST(1–6) are involved
in direct contacts with CARL (Figure 3). To elucidate the structural
requirements of MST including the ester linkages for expressing
the protease inhibitory potential, we designed its analogs (2–10)
for SAR study (Table 3), and their structures are illustrated in
Figure 4. These analogs were based on substituting Ser for Thr3
(2), opening the ester linkage(s) between Thr3 and Asp9 and/or
c
www.interscience.com/journal/psc Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2010; 16: 329–336

SAR STUDY OF MARINOSTATIN
(A)
(B)
Figure 5. NMR solution structures of (A) MST (1) and (B) [Ala7]-MST (10). A dotted line in (A) represents a hydrogen bond formed between the NH proton
of Arg5 and the carbonyl oxygen of the ester linkage with Thr3-Asp9.
Asp9 resultedincompletelossofthepotency(3), whiletheanalogs
(5) and (6) having this ester linkage retained significant inhibitory
activity even with removal of the other one with Ser8-Asp11 and
the C-terminal Ser10-Asp11-Glu12. However, removal of the N-
terminal Phe1-Ala2 resulted in complete loss of the potency (7).
It is suggested that the subtilisin–substrate association consists
mainly of hydrogen bonds producing an antiparallel β-sheet
between the main chains, and hydrophobic interactions between
the side-chains at the P1 and P4 positions and the corresponding
binding pockets at the S1 and S4 subsites, respectively [19].
Therefore,theN-terminalPhe1-Ala2 (i.e.P4–P3)shouldbeessential
for addressing the MST molecule to proteases. Thus, the minimum
size required for expressing the inhibitory activity is the sequence
(1–9) with a monocyclic structure, i.e. the analog (6). Pro7 located
in this ring structure was found to take a cis-conformation from
the NMR analysis [4,6]. The cis-isomer would be the conformation
boundtoproteasesalthoughthecis-Pro7 doesnotinteractdirectly
with the protease surface. This may contribute to a rigid type VI
β-turn structure composed of Tyr6-Pro7, which promotes the
internal hydrogen bond between the NH proton of Arg5 and the
carbonyl oxygen atom of the ester linkage with Thr3-Asp9 to fix
the scissile bond of Met4-Arg5. When substituting Ala for Pro7,
the analog (10) in which Ala should take a trans-conformation
displayed 2 orders of magnitude less potency than the original
level. The analog (10) assumes almost the same structure for
the smaller ring as the wild-type, but a significantly different
structure for the larger ring in which Arg5-Tyr6 segment forms a
type II β-turn (Figure 5). Distortion of the ring structure in (10)
determined by NMR should be caused by the trans-conformation
at position 7, leading to loss of internal hydrogen bond even
though two ester linkages are retained. The distances between
the NH proton of Arg5 and the carbonyl oxygen atom of the
ester linkage with Thr3-Asp9 are 2.2 and 6.8 Å for 1 and 10,
respectively. On the other hand, substitution of the amide bonds
or disulfide bridges for the ester linkages altered the population
of the backbone conformation at Pro7 (8 and 9, cis/trans =
70/30). Besides this, the latter analog, in particular, resulted in
reduction of the potency to 2 orders of magnitude less than the
original level due to no chance to form the internal hydrogen
bond as in the case of (10). The cis-Pro7 conformation involved
in promoting the internal hydrogen bond could be responsible
for enhancing the inhibitory potency. Furthermore, monocyclic
analogs (5 and 6) having the ester linkage with Thr3-Asp9 which
is one of the requirements for expressing the inhibitory potential
possessed the ratio between cis- and trans-conformation at Pro7
of 85 to 15. Thus, the other ester linkage with Ser8-Asp11 is
likely to assist to stabilize the cis-conformation at Pro7. From
these observations, we concluded that the ester linkage with
Thr3-Asp9, the cis-conformation at Pro7, and the N-terminal Phe1-
Ala2 may play a critical role in making it such a small-sized
protease inhibitor for the canonical inhibitory activity. Recently,
we determined the crystal structure of MST–CARL complex. This
would be providing an interpretation to confirm the findings
deduced from the SAR study. The details of the structure will be
reported elsewhere.
Rational Design of Protease Inhibitory Specificities
By applying the structural motif of MST, we tried to rationally
design protease inhibitory specificities that differ from those of
the natural products. The canonical inhibitors of serine proteases
are involved in the substrate-like interaction. The recognition of
bothsubstrateandinhibitorbymostserineproteasesisperformed
by accommodating the solvent-exposed primary binding residue
P1. Therefore, the characteristic of the side-chain function on P1 is
critical for the protease–inhibitor association and specificity. MST
has a strong inhibitory potential against subtilisin, chymotrypsin,
and elastase, but not trypsin. Amino acid preference at the P1
position for subtilisin is Met/Leu, while that for trypsin is Arg/Lys.
As expected, [Arg4, Met5]-MST (11), in which P1 and P1ꢁ residues
of native MST (i.e. Met4-Arg5) were mutually replaced, resulted
in complete loss of activity against subtilisin. And this gave a
stronger inhibitory potency against trypsin than that of leupeptin,
a well-known trypsin inhibitor (Table 4). Thus, the structural motif
of MST is a useful tool for rational design of various protease
inhibitors.
c
J. Pept. Sci. 2010; 16: 329–336 Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. www.interscience.com/journal/psc

TAICHI ET AL.
temporary inhibition and structure of disulfide linkage analogs
of marinostatin, a natural ester-linked protein protease inhibitor.
J. Pept. Res. 2005; 66: 49–58.
Table 4. KivaluesofMST(1)and[Arg4,Met5]-MST(11)againsttrypsin
and CARL
4 Taichi M, Yamazaki T, Kimura T, Nishiuchi Y. Total synthesis of
marinostatin, a serine protease inhibitor isolated from the marine
bacterium Pseudoallteromonas sagamiensis. Tetrahedron Lett. 2009;
50: 2377–2380.
Ki (nM)a
Peptide
Trypsin
CARL
5 Bode W, Huber R. Natural protein proteinase inhibitors and their
interaction with proteinases. Eur. J. Biochem. 1992; 204: 433–451.
6 Kanaori K, Kamei K, Taniguchi M, Koyama T, Yasui T, Tanaka R,
Imada C. Solution structure of marinostatin, a natural ester-linked
protein protease inhibitor. Biochemistry 2005; 44: 2462–2468.
7 Green NM, Work E. Pancreatic trypsin inhibitor. II. Reaction with
trypsin. Biochem. J. 1953; 54: 347–352.
MST (1)
>50 000
5.3 0.8
42.3 7.5
0.59 0.05
378 15
–
[Arg4, Met5]-MST (11)
Leupeptin
a Mean of 3–6 determinations SD.
8 Wu¨thrich K. NMRofProteinandNucleicacids. John Wiley & Sons: New
York, 1986.
9 Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A. NMR,
NMRPipe: a multidimensional spectral processing system based
on Unix pipes. J. Biol. 1995; 16: 277–293.
10 Garrett DS, Powers R, Gronenborn AM, Clore GM. A common sense
approach to peak picking in two-, three-, and four-dimensional
spectra usinh automatic computer analysis of contour diagrams.
J. Magn. Reson. 1991; 95: 214–220.
11 Wu¨thrich K, Billetter M, Braun W. Pseudo-structures for the 20
common amino acids for use in studies of protein conformations by
measurements of intramolecular proton-proton distance constrains
with nuclear magnetic resonance. J. Mol. Biol. 1983; 169: 949–961.
12 Bru¨nger AT. X-PLOR Manual Version 3.1. Yale University: New Haven,
1993.
13 Bo´di J, Nishiuchi Y, Nishio H, Inui T, Kimura T. 3-Pentyl (Pen) group
as a new base-resistant side chain protecting group for tyrosine.
Tetrahedron Lett. 1998; 39: 7117–7120.
14 Barlos K, Chatzi O, Gatos D, Stavropoulos G. 2-Chlorotrityl chloride
resin. Studies on anchoring of Fmoc-amino acids and peptide
cleavage. Int. J. Pept. Protein Res. 1991; 37: 513–520.
15 Do¨lling R, Beyermann M, Haenel J, Kernchen F, Krause E, Franke P,
Brudel M, Bienert M. Piperidine-mediated side product formation
for Asp(OBut)-containing peptides. J. Chem. Soc. Chem. Commun.
1994; 853–854.
Conclusions
MST was synthesized by a regioselective esterification procedure
employing two sets of orthogonally removable side-chain protect-
ing groups for the Asp and Ser/Thr residues to form two internal
ester linkages sequentially. In the MST molecule, there were no
significant changes in terms of the efficiency yielding individual
ester linkages by changing the order of esterification as long as a
pseudo-high dilution procedure with the aid of MNBA/DMAP was
appliedfortheesterformation. WealsosynthesizedaseriesofMST
analogs to elucidate SAR by employing the same strategy as that
for MST. This revealed that the ester linkage with Thr3-Asp9, the
cis-conformation at Pro7 and the N-terminal Phe1-Ala2 are essen-
tial for inhibitory activity. Furthermore, we demonstrated that the
structural motif of MST is a useful tool for rational design of various
inhibitors by changing the MST serine protease specificities from
subtilisin into trypsin.
Acknowledgements
16 Shiina I, Ibuka R, Kubota M. A new condensation reaction for the
synthesis of carboxylic esters from nearly equimolar amounts
of carboxylic acids and alcohols using 2-methyl-6-nitrobenzoic
anhydride. Chem. Lett. 2002; 286–287.
This work was supported partly by the Program for Promotion
of Fundamental Studies in Health Science of National Institute of
Biomedical Innovation (NIBL).
17 Taichi M, Kimura T, Nishiuchi Y. Suppression of side reactions during
final deprotection employing a strong acid in Boc chemistry:
regeneration of methionyl residues from their sulfonium salts. Int. J.
Pept. Res. Ther. 2009; 15: 247–253.
References
18 Maynes JT, Cherney MM, Qasim MA, Laskowski M, James MNG.
Structure of the subtilisin Carlsberg-OMTKY3 complex reveals two
different ovomucoid conformations. ActaCryst. 2005; D61: 580–588.
19 Bech LM, Sørensen SB, Breddam K. Significance of hydrophobic S4-
P4 interactions in subtilisin 309 from Bacillus lentus. Biochemistry
1993; 32: 2845–2852.
1 Imada C, Maeda M, Hara S, Taga N, Simidu U. Purification and
characterization of subtilisin inhibitors ‘marinostatin’ produced by
marine Alteromonas sp. J. Appl. Bacteriol. 1986; 60: 469–476.
2 Imada C, Hara S, Maeda M, Simidu U. Amino acid sequences of
marinostatins C-1 and C-2 from marine Alteromonas sp. Bull. Jpn.
Soc. Sci. Fish 1986; 52: 1455–1459.
3 Taniguchi M, Kamei K, Kanaori K, Koyama T, Yasui T, Takano R,
Harada S, Tajima K, Imada C, Hara S. Relationship between
c
www.interscience.com/journal/psc Copyright ꢀ 2010 European Peptide Society and John Wiley & Sons, Ltd. J. Pept. Sci. 2010; 16: 329–336