Izenamides A and B, Statine-Containing Depsipeptides, and an Analogue from a Marine Cyanobacterium

Article abstract of DOI:10.1021/acs.jnatprod.8b00417

Izenamides A, B, and C (1-3), new linear depsipeptides, were isolated from a taxonomically distinct marine cyanobacterium. Izenamides A and B contain a statine moiety [(3S,4S)-4-amino-3-hydroxy-6-methylheptanoic acid] and inhibited the activity of cathepsin D, an aspartic peptidase. Meanwhile, izenamides did not show growth-inhibitory activity against HeLa, HL60, or MCF-7 cells at up to 10 μM.

Full text of DOI:10.1021/acs.jnatprod.8b00417

Article
pubs.acs.org/jnp
Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX
Izenamides A and B, Statine-Containing Depsipeptides, and an
Analogue from a Marine Cyanobacterium
Yuki Kanamori,† Arihiro Iwasaki,† Shimpei Sumimoto,† Teruhiko Matsubara,‡ Toshinori Sato,‡
and Kiyotake Suenaga*,†
†Department of Chemistry, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Ko-hoku-ku, Yokohama, Kanagawa
223-8522, Japan
‡Department of Biosciences and Informatics, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku,
Yokohama, Kanagawa 223-8522, Japan
S
* Supporting Information
ABSTRACT: Izenamides A, B, and C (1−3), new linear depsipeptides, were
isolated from a taxonomically distinct marine cyanobacterium. Izenamides A and
B contain a statine moiety [(3S,4S)-4-amino-3-hydroxy-6-methylheptanoic acid]
and inhibited the activity of cathepsin D, an aspartic peptidase. Meanwhile,
izenamides did not show growth-inhibitory activity against HeLa, HL60, or
MCF-7 cells at up to 10 μM.
arine cyanobacteria are precious sources of biologically
active secondary metabolites. Among these metabolites,
M
many peptides have been isolated from marine cyanobacteria
and biosynthesized by nonribosomal peptide synthases (NRPS),
polyketide synthases (PKS), and ribosomes. They are note-
worthy not only because of their remarkable structures, such as
unusual amino acid or fatty acid moieties, but also because of
their biological activities.1 Therefore, compounds from marine
cyanobacteria can be expected to lead to the discovery of new
drugs. In fact, several peptides possessing potent cytotoxicity,
such as dolastatin 10,2 apratoxin,3 bisebromoamide,4 and
kurahyne,5 have been discovered from marine cyanobacteria,
and the new drug “brentuximab vedotin”, which is used in the
treatment of anaplastic large cell lymphoma and Hodgkin’s
lymphoma, was developed based on the structure of dolastatin
10.6 Furthermore, some peptides derived from cyanobacteria are
potent protease inhibitors, such as urumamide7 (chymotrypsin),
lyngbyastatins 4−108 (elastase), gallinamide A9 (cathepsin L),
and grassystatins A−F10 (cathepsins D/E). Proteases have been
recognized as pathogenic factors in many human diseases such
as cancer,11 Alzheimer’s disease,12 and HIV infection;13 thus the
therapeutic modulation of proteolytic activity by drugs may
represent an approach to treatment.
RESULTS AND DISCUSSION
■
The marine cyanobacterium 1605-514 (1700 g, wet weight) was
collected at Izena Island, Okinawa, Japan. This sample was
morphologically classified into Lyngbya (Figure S39). However,
based on a 16S rRNA sequence analysis, the cyanobacterium
was taxonomically distinct from the Lyngbya genus (Figure
S40). The sample was extracted with MeOH. The extract was
filtered, concentrated, and partitioned between EtOAc and
H2O. The EtOAc-soluble material was further partitioned
Herein, we describe the isolation, structure determination,
and biological evaluation of three linear depsipeptides,
izenamides A−C (1−3), from a marine cyanobacterium.
Izenamides A and B (1 and 2) are new inhibitors of cathepsin
D and contain a statine moiety [(3S,4S)-4-amino-3-hydroxy-6-
methylheptanoic acid, Sta].
Received: May 24, 2018
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 1. NMR Spectra for Izenamide A (1) at 400 MHz (1H) and 100 MHz (13C) in CD3OD
residue
position
δC, type
δH, mult (J in Hz)
COSY
HMBC (H→C)
selected NOESY
a
Pro-O-Me
1
173.8,
C
2
3a
3b
4a
4b
5a
60.9, CH
4.42, dd (7.4, 7.0)
2.27, m
1.86, m
1.96, m
1.86, m
3a, 3b
2, 3b, 4a, 4b
2, 3a, 4a
3a, 3b, 4b, 5a, 5b
3a, 4a, 5a, 5b
4a, 4b, 5b
1, 3, 4
30.0, CH2
26.2, CH2
48.5, CH2
1, 2, 4, 5
1, 2, 4, 5
3
2, 3
3, 4
3.46, m
8
5b
6
7
3.40, m
3.72, s
4a, 4b, 5a
3, 4
1
52.6, CH3
170.3, C
N-Me-Phe
8
9a
57.2, CH
35.6, CH2
5.70, dd (10.0, 5.8)
3.11, dd (14.4, 5.8)
2.93, dd (14.4, 10.0)
9a, 9b
8, 9b
8, 9a
7, 9, 10, 16, 17
8, 10, 11/15
7, 8, 10, 11/15
5a
9b
10
11/15
12/14
13
16
17
18
19
20
21
22
23a
23b
24
25
26
27
28
29
30a
30b
31
32
33
34
35
36
37a
37b
38
39
40
41
42
43
44
138.3, C
130.8, CH
129.2, CH
127.6, CH
31.0, CH3
174.6, C
46.6, CH
16.6, CH3
173.0, C
7.24, m
7.24, m
7.16, m
3.07, s
9, 10, 12/14, 13, 15/11
10, 11/15, 13, 14/12
11/15, 12/14
13
12/14
8, 17
18
16
Ala
Ile
4.69, q (7.0)
0.82, d (7.0)
19
18
17, 19
17, 18
58.6, CH
38.3, CH
25.7, CH2
4.21, d (7.4)
1.76, m
1.49, m
1.15, m
0.87, t (7.4)
0.87, d (7.0)
22
20, 22, 23, 25
21, 23, 24, 25
21, 25
21, 22, 24, 25
22, 23
27
21, 23a, 23b, 25
22, 23b, 24
22, 23a, 24
23a, 23b
22
11.5, CH3
15.8, CH3
21, 22, 23
a
Sta
173.9,
C
b
41.4, CH2
71.3, CH
52.2, CH
2.38, m
3.99, m
3.99, m
1.57, m
1.33, m
1.56, m
0.93, d (6.1)
0.89, d (6.3)
28
27
26, 28, 29
26, 27
34
29, 31
31
29, 30, 32
30, 31, 33
30, 31, 32
21
30a, 30b
29, 30b
29, 30a
32, 33
31
b
41.3, CH2
26.0, CH
23.8, CH3
22.2, CH3
171.7, C
80.6, CH
31.7, CH
18.1, CH3
31
valic acid-2
4.72, d (5.8)
2.14, m
1.01, d (6.7)
36
34, 36, 37, 40
34, 35, 37, 39
35, 36, 39
35, 37a, 38
36
c
19.25, CH3
1.01, d (6.7)
36
35, 36, 37a
valic acid-1
175.8, C
76.4, CH
33.3, CH
4.13, d (4.5)
2.10, m
1.01, d (6.7)
0.93, d (6.7)
42
41, 43, 44
42
42
40, 42, 43, 44
40, 41, 43, 44
41, 42, 44
c
19.30, CH3
17.2, CH3
41, 42, 43
a−c
These signals are interchangeable.
between 90% aqueous MeOH and hexane to give an aqueous
MeOH portion. This portion was subjected to fractionation with
reversed-phase column chromatography (ODS silica gel,
MeOH−H2O) and repeated reversed-phase HPLC to give
izenamides A (1, 34.0 mg), B (2, 2.2 mg), and C (3, 6.1 mg).
The molecular formula of izenamide A (1) was determined to
be C43H69N5O11 based on HRESIMS data. Perusal of the 1H and
13C NMR spectra revealed that 1 was a depsipeptide (Table 1)
with proton signals of one N-methyl amide substituent (δH
3.07), one O-methyl ester substituent (δH 3.72), α-protons (δH
4−5) of amides, aromatic protons (δH 7.1−7.3), and a methine
adjacent to an acylated oxygen atom (δH/C 4.72/80.6). Further
1
analyses of the H NMR, 13C NMR, COSY, HMQC, HMBC,
and NOESY spectra in CD3OD (Table 1) revealed the presence
of a statine (Sta), four α-amino acid (Pro-O-Me, N-Me-Phe, Ala,
and Ile), and two valic acid residues. Based on analyses of the
HMBC and NOESY spectra, two structural fragments, [(valic
acid-1)-(valic acid-2)-Sta-Ile] and [Ala-(N-Me-Phe)-(Pro-O-
B
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 1. 2D NMR data for izenamides A−C (1−3).
Figure 2. Selected MS fragmentation cleavages for izenamides A−C (1−3).
Me)], were established. The continuous sequence of these two
fragments was clarified to be [(valic acid-1)-(valic acid-2)-Sta-
Ile-Ala-(N-Me-Phe)-(Pro-O-Me)] by ESI-ion trap MS (ESI-
ITMS) fragmentation analysis (Figures 1 and 2).
27a, δH 2.41, dd (J = 14.8, 8.8 Hz)) showed a large coupling to
H-28, and the more shielded α-proton signal (H-27b, δH 2.35,
dd (J = 14.8, 5.2 Hz)) showed a small coupling to H-28 (Table
S1, Supporting Information). These observations indicated that
the relative configuration was anti (Figure S32). Therefore, the
absolute configuration of Sta in 1 was determined to be (3S, 4S),
and the absolute configuration of izenamide A (1) was
established as shown in formula 1.
To establish the absolute configuration of 1, the acid
hydrolysate of 1 was analyzed by chiral-phase HPLC. The
result revealed the presence of L-Ile, L-Ala, N-Me-D-Phe, L-Pro,
and D/L-valic acid residues in izenamide A (1). To distinguish
the D/L-valic acid residues, 1 was hydrolyzed under basic
conditions, and we obtained the valic acid derived from the N-
terminus of 1.14 Chiral-phase HPLC analysis for this valic acid
determined that the N-terminal valic acid was the L-form. On the
other hand, the other fragment without the L-valic acid residue
was hydrolyzed under acidic conditions, and the hydrolysate was
analyzed by chiral-phase HPLC. As expected, only D-valic acid
was detected in this analysis. Regarding the Sta moiety, the
absolute configuration of the hydroxy group at C-28 was
determined to be S by a modified Mosher’s method15 (S43,
Supporting Information). According to a previous paper, the
relative configuration of a γ-amino-β-hydroxy acid, such as a Sta
unit, can be easily determined by examination of the coupling
constants and the 1H NMR chemical shifts of the α-methylene
signals.16 In CDCl3, the more deshielded α-proton signal (H-
The molecular formula of izenamide B (2) was determined to
be C44H71N5O11 based on the HRESIMS data, which is larger
1
than that of 1 by 14 MS units (CH2). The H and 13C NMR
spectra revealed that the structure of 2 was very similar to that of
1. Further NMR analyses revealed all of the partial structures in
2: Pro-O-Me, N-Me-Phe, Ala, Ile, valic acid, isoleucic acid, and
Sta residues. Next, we determined their sequence based on
NMR data (Table 2). The presence of the fragment [Ala-(N-
Me-Phe)-(Pro-O-Me)] was determined on the basis of HMBC
and NOESY data. The remaining sequence was clarified to be
[valic acid-isoleucic acid-Sta-Ile-Ala-(N-Me-Phe)-(Pro-O-Me)]
by ESI-ITMS fragmentation analysis (Figures 1 and 2).
To establish the absolute configuration, the acid hydrolysate
of 2 was analyzed by chiral-phase HPLC. The result revealed the
presence of L-Ile, L-Ala, N-Me-D-Phe, and L-Pro in 2. However,
C
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 2. NMR Spectra for Izenamides B (2) and C (3) at 400 MHz (1H) and 100 MHz (13C) in CD3OD
izenamide B (2)
izenamide C (3)
δC, type
residue
position
δC, type
δH, mult (J in Hz)
residue
position
δH, mult (J in Hz)
a
Pro-O-Me
1
2
3a
173.8,
C
Pro-O-Me-
1
2
3a
174.0, C
60.6, CH
29.9, CH2
61.0, CH
30.0, CH2
4.42, dd (7.0, 7.0)
2.27, m
4.37, dd (8.8, 5.4)
2.20, m
3b
4a
4b
5a
5b
6
7
1.86, m
1.96, m
1.86, m
3.46, m
3.40, m
3.72, s
3b
4a
4b
5a
5b
6
7
1.85, m
1.93, m
1.85, m
3.45, m
3.40, m
3.71, s
26.2, CH2
48.5, CH2
26.0, CH2
48.1, CH2
52.7, CH3
52.6, CH3
170.4, C
c
N-Me-Phe
N-Me-Phe
170.2, C
8
9a
57.2, CH
35.6, CH2
5.70, dd (10.0, 5.8)
3.11, dd (14.4, 5.8)
2.94, dd (14.4, 10.0)
8
9a
57.8, CH
35.6, CH2
5.58, dd (7.6, 7.6)
3.19, dd (13.7, 7.6)
2.84, dd (13.7, 7.6)
9b
10
11/15
12/14
13
16
17
18
19
20
21
22
23a
23b
24
25
26
27
28
29
30a
30b
31
32
33
34
35
36
37a
37b
38
39
40
41
42
43
44
9b
10
11/15
12/14
13
16
17
18a
18b
19
20
21
22a
22b
23
24
25
26
27
28a
28b
29
30
31
32
33
34
35
138.3, C
138.6, C
130.8, CH
129.2, CH
127.6, CH
31.0, CH3
174.6, C
46.6, CH
16.6, CH3
173.0, C
7.24, m
7.24, m
7.16, m
3.07, s
130.5, CH
129.4, CH
127.6, CH
30.5, CH3
7.24, m
7.24, m
7.17, m
3.03, s
c
Ala
Ile
Gly
Ile
170.1, C
4.68, q (7.0)
0.83, d (7.0)
41.9, CH2
4.16, d (17.1)
3.91, d (17.1)
173.4, C
58.9, CH
37.9, CH
25.7, CH2
58.7, CH
38.3, CH
25.7, CH2
4.21, d (7.4)
1.76, m
1.49, m
1.15, m
0.87, t (7.4)
0.87, d (6.7)
4.34, d (7.0)
1.92, m
1.47, m
1.14, m
0.89, dd (7.4, 7.4)
0.93, d (7.0)
11.5, CH3
15.8, CH3
11.5, CH3
16.1, CH3
172.2, C
77.9, CH
38.5, CH
27.0, CH2
a
Sta
173.9,
C
isoleucic acid
b
41.4, CH2
71.4, CH
52.3, CH
2.38, m
3.99, m
3.99, m
1.57, m
1.33, m
1.56, m
0.93, m
0.89, m
5.07, d (4.5)
1.94, m
1.44, ddq (13.9, 7.6, 7.6)
1.29, ddq (13.9, 7.6, 7.6)
0.96, t (7.6)
b
41.3, CH2
11.9, CH3
14.7, CH3
175.1, C
76.4, CH
33.3, CH
19.4, CH3
16.9, CH3
26.0, CH
23.8, CH3
22.2, CH3
171.8, C
78.3, CH
40.2, CH
27.0, CH2
0.96, d (7.0)
valic acid
4.12, d (4.3)
2.11, qqd (7.0, 6.7, 4.3)
1.01, d (7.0)
isoleucic acid
4.93, m
1.91, m
1.45, m
1.31, m
0.96, m
0.98, m
0.91, d (6.7)
11.9, CH3
14.8, CH3
175.8, C
76.0, CH
33.3, CH
19.3, CH3
17.1, CH3
valic acid
4.14, d (4.5)
2.10, m
1.01, m
0.92, m
a−c
These signals are interchangeable.
we could not clarify the absolute configurations of the two α-
hydroxy acid moieties due to epimerization during acid
hydrolysis. To prevent epimerization, 2 was hydrolyzed under
basic conditions to give the valic acid, which was analyzed by
chiral-phase HPLC and determined to be in the L-form.
Meanwhile, chiral-phase HPLC analysis of the acid hydrolysate
of the other fragment without the L-valic acid revealed the
presence of D-allo-isoleucic acid in 2. Regarding the Sta moiety,
the relative configuration was determined to be anti based on the
same analysis as described above (Table S1, Figure S32). To
determine the absolute configuration at C-28, a modified
Mosher’s method was carried out. Although the (R)-MTPA
ester of 2 was obtained as a single isomer, the (S)-MTPA ester
was obtained as two diastereomers probably due to epimeriza-
tion during the derivatization reaction.17 Despite several
attempts, epimerization could not be avoided. Therefore, we
D
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
calculated Δδ (Δδ(S)‑MTPA − Δδ(R)‑MTPA) values by using each
diastereomer of the (S)-MTPA esters, respectively. Fortunately,
in both cases, the distributions of the positive and negative Δδ
values were consistent with each other, which indicated that the
absolute configuration at C-28 was S (S46). Therefore, the
absolute configuration of Sta in 2 was determined to be (3S, 4S),
the same configuration as that in 1, and we clarified the structure
of izenamide B (2).
assembly was corroborated by an ESI-ITMS fragmentation
(Figure 2).
To establish the absolute configuration, a portion of 3 was
hydrolyzed under acidic conditions, and the hydrolysate was
analyzed by chiral-phase HPLC. The result revealed the
presence of L-valic acid, D-allo-isoleucic acid, L-Ile, N-Me-D-
Phe, and L-Pro in 3, and thus the absolute configuration of 3 was
established.
The molecular formula of izenamide C (3) was determined to
be C35H54N4O9 based on HRESIMS data. Detailed analyses of
the 1H NMR, 13C NMR, COSY, HMQC, HMBC, and NOESY
spectra in CD3OD of 3 (Table 3) revealed the presence of one
The structures of izenamides A (1) and B (2) were similar to
those of the known depsipeptides grassystatins10 and tasiamides
B and F.18 These known compounds possess a γ-amino acid
moiety as a common structure and have been known to inhibit
some proteases such as cathepsin D. Although this unusual
amino acid moiety is clarified to be responsible for the protease-
inhibitory activity, detailed structure−activity relationship
(SAR) studies beyond the γ-amino acid part are needed.
We assessed the biological activities of izenamides (1−3).
First, we evaluated the cell-growth-inhibitory activity by MTT
assay and found that 1−3 did not exhibit any cytotoxicity against
HeLa, HL60, or MCF-7 cells up to 10 μM. Next, we focused on
the characteristic γ-amino acid moiety, the Sta residue, in 1 and
2. The Sta unit was first described as a partial structure of a very
potent cathepsin D inhibitor, pepstatin A, from an Actinomyces
sp.19 The inhibitory mechanism of pepstatin A against cathepsin
D was elucidated based on a cocrystal structure analysis, which
found that the hydroxy group in a Sta unit interacts with the two
aspartate residues in the active site of the enzyme, Asp33 and
Table 3. IC50 Values of Izenamides A−C (1−3) against
Cathepsin D
b
IC50 values
izenamide A (1)
izenamide B (2)
izenamide C (3)
380 90 nM
270 150 nM
>10 μM
a
pepstatin A
50 48 pM
a
b
Positive control. Values are shown SD.
valic acid, one isoleucic acid, and four α-amino acid (Pro-O-Me,
N-Me-Phe, Gly and Ile) residues. The sequence of 3 was clarified
to be [valic acid-isoleucic acid-Ile-Gly-(N-Me-Phe)-(Pro-O-
Me)] based on the HMBC and NOESY spectra (Figure 1). This
Figure 3. Conformational analyses of izenamide A (1)/cathepsin D and pepstatin A/cathepsin D. (A) Docking conformation of izenamide A (1,
yellow) with cathepsin D in the global minimum energy. (B) Plausible hydrogen bonds between izenamide A (1) with cathepsin D in the global
minimum energy. (C) Cocrystal structure of pepstatin A with cathepsin D (1LYB). (D) Plausible hydrogen bonds between pepstatin A with cathepsin
D in the cocrystal, 1LYB.
E
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Asp231.20 Cathepsin D, an aspartic endopeptidase, has attracted
attention due to its association with several diseases. For
example, cathepsin D is associated with cancer in that it
promotes proliferation, invasion, and metastasis and is clinically
used as a tumor marker of breast cancer.21 In addition, it has
been suggested that cathepsin D plays a significant role in the
infection of humans by influenza viruses.22 Therefore, the
discovery of a new cathepsin D inhibitor is significant for the
development of new drugs. Against this background, izenamides
A−C (1−3) were subjected to a cathepsin D-inhibition assay. As
expected, 1 and 2 exhibited similar potent inhibitory activities
against cathepsin D. Compound 3, without a Sta residue, did not
show this activity (Table 3).
Compared with the IC50 values against cathepsin D, the
activities of both 1 and 2 were much weaker than that of
pepstatin A (Table 3), although they possessed a Sta unit. To
elucidate the cause of this difference at the molecular level, we
calculated a plausible docking structure of izenamide A (1) and
cathepsin D using a molecular modeling method (Macro-
Model)23 as follows. The crystal structure of cathepsin D with
pepstatin A (1LYB) was used as a starting point. Based on the
cocrystal structure, izenamide A (1) was manually docked into
the pepstatin A-binding pocket on the enzyme. Next, we
performed a conformational search for 1 and the amino acid
residues around 1 by a torsional sampling method. As a
constraint, the Sta moiety, which blocks the active site of
cathepsin D, was placed within 4 Å from the active site of
cathepsin D. Optimization of the obtained conformations
afforded the docking pose in the global minimum energy
(Figure 3A) along with those in several local energy minima.
Considering the Boltzmann distribution at room temperature,
we found that the conformation at the global minimum energy
accounted for 93% of the total (Table S2). Therefore, we
regarded this conformation to be the plausible docking pose of
izenamide A (1) at the cathepsin D active site (Figure 3A).
Before starting detailed analyses of the docking structure, we
validated our calculation approach as follows. We carried out
molecular modeling of pepstatin A and cathepsin D by using the
same method as described above and successfully regenerated
the crystal structure of 1LYB (Figure S38). On the basis of this
result, we confirmed the adequacy of this approach.
On the basis of the docking structure (Figure 3A) and the
cocrystal structure (Figure 3C), we assessed several factors that
affect the stability of the complex structures of izenamide A (1)/
cathepsin D and pepstatin A/cathepsin D by MacroModel and
found that two factors significantly destabilize the docking
structure of 1 (Table S3).
The first destabilizing factor is the protrusion of a hydro-
phobic group into the hydrophilic environment. The C-terminal
residue of pepstatin A (Figure 3C) is wrapped by the binding
pocket of cathepsin D. On the other hand, regarding izenamide
A (1) (Figure 3A), the hydrophobic methylene chain of Pro-O-
Me protrudes into the solvent region (a water region). This
unfavorable interaction could destabilize the ligand−receptor
complex.
The second destabilizing factor is the difference in the
hydrogen-bonding ability with the Ser80 residue between
pepstatin A and izenamide A (1). We assessed the contribution
of every hydrogen bond to stabilization of the ligand/receptor
complex by using MacroModel (Figure 3B and D and Table S4).
As a result, the stabilizing effect of the hydrogen bond from
Ser80 was especially different between pepstatin A and
izenamide A (1). While pepstatin A formed two hydrogen
bonds with the Ser residue, 1 formed only one hydrogen bond
with it. We analyzed the reason for this difference. The carbonyl
group and amide-NH group of valine at the P2 residue in
pepstatin A each formed a hydrogen bond with the Ser80 residue
(Figure 3D). However, 1 possessed valic acid at the
corresponding position instead of valine and could not form
two hydrogen bonds with Ser80, probably due to the lack of an
amide-NH (Figure 3B).
On the basis of these considerations, we conjectured the
origin of the difference in the cathepsin D-inhibitory activity
between 1 and pepstatin A. To date, a SAR study of cathepsin D
and Sta-containing acyclic peptides has only been carried out
using grassystatins.10 Grassystatins were isolated from marine
cyanobacteria and were shown to exhibit cathepsin D-inhibitory
activity. They possess a hydrophilic amino acid residue at the P2
position, and the previous study clarified that this hydrophilicity
reduces the affinity toward cathepsin D. Furthermore, it has
been suggested that the presence of a N,N-dimethyl amino acid
residue at the N-terminus potentiates the inhibitory activity. Our
study on izenamides (1−3) adds new insights to the SAR study
of cathepsin D and Sta-containing peptides.
In summary, we have reported the isolation and the structure
determination of new linear depsipeptides, izenamides A−C
(1−3), from a marine cyanobacterium collected at Okinawa,
Japan. According to our identification study, this cyanobacte-
rium is taxonomically independent of cyanobacteria in other
genera including izenamide-analogue-producing cyanobacteria
(Figure S40). The structures of 1−3 were elucidated by a
combination of spectroscopic analyses and degradation/
derivatization reactions. Izenamides A (1) and B (2), which
possess a Sta moiety, inhibited cathepsin D without exhibiting
cytotoxicity against HeLa, HL60, or MCF-7 cells. The cathepsin
D-inhibitory activities of 1 and 2 were much weaker than that of
pepstatin A, a structurally related known inhibitor. On the basis
of a molecular modeling method, we suggested a cause for this
difference at the molecular level. Recently, cathepsin D has
attracted attention due to its association with several diseases,
and this study provides new knowledge for the development of
cathepsin D inhibitors.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured with a JASCO DIP-1000 polarimeter. IR spectra were
recorded on a JASCO FT/IR-4200 instrument. All NMR data were
1
recorded on a JEOL JNM-ECX400 spectrometer for H (400 MHz)
and 13C (100 MHz). 1H NMR chemical shifts (referenced to residual
CHD2OD observed at δH 3.31 and CHCl3 observed at δH 7.26) were
assigned using a combination of data from COSY and HMQC
experiments. Similarly, 13C NMR chemical shifts (referenced to
CD3OD observed at δC 49.00 and CDCl3 observed at δC 77.16) were
assigned based on HMBC and HMQC experiments. HRESIMS spectra
were obtained on an LCT Premier XE time-of-flight (TOF) mass
spectrometer (LCT Premier XE, Waters). ESI-ITMS spectra were
recorded with an amaZone SL mass spectrometer (Bruker Daltonics).
Chromatographic analyses were performed using an HPLC system
consisting of a pump (model PU-2080, JASCO) and a UV detector
(model UV-2075, JASCO). All chemicals and solvents used in this
study were the best grade available (special grade) from a commercial
source.
Identification of the Marine Cyanobacterium. Small pieces of
the collected marine cyanobacterium, 1605-5, were preserved for
genetic analysis in RNAlater (Qiagen). To remove the cyanobacterial
sheaths, the cyanobacterial sample was treated with 0.1% (w/v)
sarkosyl solution before the isolation of single filaments. Single
filaments of marine cyanobacteria were isolated by the pipet-washing
F
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
method24 under a light microscope. The isolated single filament was
crushed with a pipet on a slide glass under a light microscope and
dissolved in 10 μL of sterilized H2O. The sterilized H2O containing the
broken filament was used as a DNA template for PCR amplification.
The 16S rRNA genes were PCR-amplified in two stages by using two
cyanobacterial primer sets: CYA 106F25 and 23S 30R26 for the first
amplification, and CYA 359F25 and 16S 1371R27 for the second
amplification. The first PCR reaction contained 10 μL of DNA
template, 0.5 μL of KOD-Multi & Epi- (Toyobo), 12.5 μL of 2× PCR
buffer for KOD-Multi & Epi- (Toyobo), and 1.0 μL of CYA 106F and
23S 30R (10 pM). The PCR reaction was carried out with a T100
thermal cycler (Bio-Rad) as follows: initial denaturation at 94 °C for 2
min, amplification for 40 cycles of 10 s at 98 °C, and final elongation at
66 °C for 1.5 min. The obtained PCR products were diluted 1000-fold
with sterilized H2O. The second PCR reaction contained 10 μL of the
diluted PCR product, 0.5 μL of KOD-Multi & Epi- (Toyobo), 12.5 μL
of 2× PCR buffer for KOD-Multi & Epi- (Toyobo), and 1.0 μL of CYA
359F and 16S 1371R (10 pM). The second PCR reaction was
performed under the same conditions as the first PCR.
PCR products were analyzed on agarose gel (1%) in TBE buffer,
visualized by ethidium bromide staining, and purified by a PCR
Advanced PCR cleanup system (Viogene). Sequences were determined
with CYA 359F and 16S 1371R primers by a commercial firm
(Macrogen Japan Corp.). These sequences are available in the DDBJ/
EMBL/GenBank databases under accession number LC315181.
Morphological Characterization. Morphological observation
was performed using a phase contrast microscope (Eclipse Ti-S,
Nikon). The mean cell size and standard deviation of 50 cells were
measured. The cell length was 1.7 0.5 μm, and the width was 17.3
0.6 μm. The mean sheath size and standard deviation of 25 sheaths were
measured. The thickness of the sheath was 1.2 0.2 μm. The filaments
were unbranched and the sheaths contained only one trichome, which
were consistent with the morphological features of Lyngbya. The cell
color was olive-green, and the sheath was colorless and unlamellated.
Morphologically, the cyanobacterium was very close to Lyngbya
confervoide. However, although the sheath of L. confervoide is
lamellated,28 that of 1605-5 was not. 1605-5 was compared with
Okeania,29 Moorea30 and Caldora penicillata,31 but was not consistent
with any of these genera (Table S5, Figure S39).
Phylogenetic Analysis. The nucleotide sequence of the 16S rRNA
gene obtained in this study was used for phylogenetic analysis with the
sequences of related cyanobacterial 16S rRNA genes that have been
previously reported. All sequences were aligned by MEGA732 with a
default setting of ClustalW. Gaps in the alignment data were manually
deleted from the sequences, and 188 nucleotides were used for
cladistics analysis. JModeltest (version 2.1.7)33,34 with the default
settings was used to select the best model of DNA substitution for the
maximum likelihood (ML) analysis and Bayesian analysis according to
the Akaike information criterion. The ML analysis was performed with
RAxML35 (version 8), using the GTR+I+G model. Bootstrap
resampling was performed on 1000 replicates. The ML tree was
visualized with Njplot (version 2.3).36 The Bayesian analysis was
conducted with MrBayes (version 3.2.6)37 using the GTR+I+G model.
The Markov chain Monte Carlo process was set at two chains, and
5 000 000 generations were conducted. The sampling frequency was
assigned to be every 500 generations. After analysis, the first 500 000
trees were deleted as burn-in, and the consensus tree was constructed.
The Bayesian tree was visualized with FigTree (version 1.4.3, http://
tree.bio.ed.ac.uk/software/figtree/).
Collection, Extraction, and Isolation. The cyanobacterium
1605-5 was collected at Izena Island, Okinawa Prefecture, Japan, at a
depth of 0−1 m in May 2016. The collected cyanobacterium (1700 g,
wet weight) was extracted twice with MeOH (3 L). The extract was
filtered and concentrated. The residue was partitioned between EtOAc
(3 × 300 mL) and H2O (500 mL). The material obtained from the
organic layer was partitioned between hexane (3 × 300 mL) and 90%
aqueous MeOH (300 mL) to give an aqueous MeOH fraction (1.39 g).
This fraction was separated by column chromatography on ODS gel
(Cosmosil 75C18-OPN 14.0 g) eluting with 40% aqueous MeOH, 60%
aqueous MeOH, 80% aqueous MeOH, MeOH, and CHCl3−MeOH
(1:1). The 80% MeOH fraction (450 mg) was subjected to HPLC
[conditions for HPLC separation, Cosmosil 5C18-AR-II (ϕ 20 mm ×
250 mm); solvent 80% MeOH; flow rate 5 mL/min; detection UV 215
nm] in 12 batches to give several fractions containing izenamide C (3,
29.9 mg, tR = 26.3 min), izenamide A (1, 52.3 mg, tR = 29.1 min), and
izenamide B (2, 67.4 mg, tR = 36.0 min). These fractions containing
izenamides A−C were further separated by HPLC [Cosmosil Cholester
(ϕ 20 mm × 250 mm); solvent 51% MeCN; flow rate 5 mL/min;
detection UV 215 nm] in two batches to give izenamide A (1, 34.0 mg,
tR = 49.8 min), HPLC [Cosmosil Cholester (ϕ 20 mm × 250 mm);
solvent 59% MeCN; flow rate 5 mL/min; detection UV 215 nm] in one
batch to give izenamide B (2, 1.9 mg, tR = 36.8 min), and HPLC
[Cosmosil Cholester (ϕ 20 mm × 250 mm); solvent 55% MeCN; flow
rate 5 mL/min; detection UV 215 nm] in one batch to give izenamide C
(3, 6.1 mg, tR = 38.2 min).
Izenamide A (1): colorless oil; [α]23 −17 (c 0.31, MeOH); IR
D
(neat) νmax 3316, 2962, 2875, 1746, 1632, 1537, 1453, 1367, 1265,
1198, 1034, 753 cm−1; 1H NMR, 13C NMR, COSY, HMQC, HMBC,
and selected NOESY data in CD3OD, Table 1; 1H NMR, COSY, and
TOCSY in CDCl3, Table S1; HRESIMS m/z 832.5048 [M + H]+
(calcd for C43H70N5O11, 832.5072).
Izenamide B (2): colorless oil; [α]23 −11 (c 0.11, MeOH); IR
D
(neat) νmax 3313, 2962, 1743, 1633, 1533, 1446, 1198 cm−1; 1H NMR,
13C NMR data, Table 2; 1H NMR, COSY, and TOCSY in CDCl3, Table
S1; HRESIMS m/z 846.5237 [M + H]+ (calcd for C44H72N5O11,
846.5228).
Izenamide C (3): colorless oil; [α]23 +24 (c 0.21, MeOH); IR
D
(neat) νmax 3313, 2964, 2877, 1746, 1635, 1521, 1446, 1199, 1177,
1135, 1031, 753 cm−1; 1H NMR, 13C NMR data, Table 2; HRESIMS
m/z 675.3956 [M + H]+ (calcd for C35H55N4O9, 675.3970).
Determination of the Configuration of Izenamide A (1).
Izenamide A (1, 0.2 mg) was dissolved in 100 μL of 9 M HCl in a sealed
tube and heated at 110 °C for 24 h. The hydrolyzed products were
evaporated to dryness and could be separated into each component by
HPLC [conditions for HPLC separation: column, Cosmosil 5C18- PAQ
(ϕ 4.6 × 250 mm); flow rate, 1.0 mL/min; detection at 215 nm; solvent
H2O]. Retention times (tR min) of components: Ala (2.9), Pro (3.3),
Ile (4.6), N-Me-Phe (12.7).
Izenamide A (1, 0.2 mg) was dissolved in 150 μL of MeOH−0.5 M
NaOH (1:1) and reacted at room temperature for 20 min. The solution
was neutralized by the addition of 37.5 μL of 1 M HCl. The product was
evaporated to dryness and separated to obtain the terminal valic acid
(valic acid-1) by HPLC [conditions for HPLC separation: column, PBr
(ϕ 4.6 × 250 mm); flow rate, 1.0 mL/min; detection at 215 nm; solvent
MeCN−H2O−trifluoroacetic acid (TFA) (5:95:0.1)]. Retention time
(tR min) of component: valic acid (10.8)]. The other product without
the terminal residue was dissolved in 100 μL of 9 M HCl in a sealed tube
and heated at 110 °C for 24 h. The hydrolyzed products were
evaporated to dryness and separated to obtain another valic acid (valic
acid-2) [conditions for HPLC separation: column, PBr (ϕ 4.6 × 250
mm); flow rate, 1.0 mL/min; detection at 215 nm; solvent MeCN−
H2O−TFA (5:95:0.1)]. Retention time (tR min) of component: valic
acid (11.9).
The configurations of the amino and hydroxy acids in 1 were
determined by chiral-phase HPLC [DAICEL CHIRALPAK (MA+) (ϕ
4.6 × 50 mm); flow rate, 1.0 mL/min; detection at 254 nm; solvent 2.0
mM CuSO4, 2.0 mM CuSO4−MeCN (90:10), 2.0 mM CuSO4−
MeCN (85:15)]. With 2.0 mM CuSO4 as a solvent, the retention times
of Ala, Pro, and Ile matched those of the authentic standards of L-Ala
(2.0 min; D-Ala, 1.6 min), L-Pro (6.3 min; D-Pro, 3.3 min), L-Ile (23.9
min; D-allo-Ile, 8.6 min, D-Ile, 11.0 min, L-allo-Ile, 16.8 min). With 2.0
mM CuSO4−MeCN (90:10) as a solvent, the retention times of N-Me-
Phe matched those of the authentic standards of N-Me-D-Phe (7.3 min;
N-Me-L-Phe, 8.2 min). With 2.0 mM CuSO4−MeCN (85:15) as a
solvent, the retention times of the terminal valic acid-1 and valic acid-2
matched those of the authentic standards of L-valic acid (9.3 min; D-
valic acid, 6.1 min) and D-valic acid, respectively.
To a stirred solution of izenamide A (1) (0.8 mg, 0.96 μmol) in
pyridine (100 μL) were added 2 drops of (R)-MTPACl and 4-
(dimethylamino)pyridine (0.3 mg, 2.5 μmol). The mixture was stirred
G
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
at room temperature (rt) for 22 h and concentrated to give an oil, which
was purified by HPLC [conditions for HPLC separation: column,
Cosmosil 5C18-MS-II (ϕ 4.6 × 250 mm); flow rate, 5.0 mL/min;
detection at 215 nm; solvent MeOH−H2O (90:10); tR = 36.3 min] to
afford the (S)-MTPA ester of 1 (0.5 mg) as a colorless oil: HRESIMS
m/z 1264.5901 [M + H]+ (calcd for C63H84F6N5O15, 1264.5868).
To a stirred solution of 1 (1.3 mg, 1.6 μmol) and 4-
(dimethylamino)pyridine (0.2 mg, 1.6 μmol) in pyridine (100 μL)
were added 2 drops of (S)-MTPACl. The mixture was stirred at rt for 6
h and concentrated to give an oil, which was purified by HPLC
[conditions for HPLC separation: column, Cosmosil 5C18-AR-II (ϕ 4.6
× 250 mm); flow rate, 5.0 mL/min; detection at 215 nm; solvent
MeOH−H2O (90:10); tR = 35.2 min] to afford the (R)-MTPA ester of
1 (0.9 mg) as a colorless oil: HRESIMS m/z 1264.5862 [M + H]+
(calcd for C63H84F6N5O15, 1264.5868).
Determination of the Configuration of Izenamide B (2).
Izenamide B (2, 0.5 mg) was dissolved in 100 μL of 9 M HCl in a sealed
tube and heated at 110 °C for 24 h. The hydrolyzed products were
evaporated to dryness and could be separated into each component by
HPLC [conditions for HPLC separation: column, Cosmosil 5C18- PAQ
(ϕ 4.6 × 250 mm); flow rate, 1.0 mL/min; detection at 215 nm; solvent
H2O]. Retention times (tR min) of components: Ala (2.9), Pro (3.4),
Ile (4.5), N-Me-Phe (12.6).
Izenamide B (2, 0.2 mg) was dissolved in 150 μL of MeOH−0.5 M
NaOH (1:1) and reacted at rt for 20 min. The solution was neutralized
by the addition of 37.5 μL of 1 M HCl. The product was evaporated to
dryness and separated to obtain the terminal valic acid by HPLC
[conditions for HPLC separation: column, PBr (ϕ 4.6 × 250 mm); flow
rate, 1.0 mL/min; detection at 215 nm; solvent MeCN−H2O−TFA
(5:95:0.1)]. Retention time (tR min) of component: valic acid (10.9).
The product with the terminal residue removed was dissolved in 100 μL
of 6 M HCl in a sealed tube and heated at 110 °C for 24 h. The
hydrolyzed products were evaporated to dryness and separated to
obtain isoleucic acid [conditions for HPLC separation: column, ODS
HG-5 (ϕ 4.6 × 250 mm); flow rate, 1.0 mL/min; detection at 215 nm;
solvent MeCN−H2O−TFA (15:85:0.1)]. Retention time (tR min) of
component: isoleucic acid (12.0).
The configurations of the amino and hydroxy acids in 2 were
determined by chiral-phase HPLC [DAICEL CHIRALPAK (MA+) (ϕ
4.6 × 50 mm); flow rate, 1.0 mL/min; detection at 254 nm; solvent 2.0
mM CuSO4, 2.0 mM CuSO4−MeCN (90:10), 2.0 mM CuSO4−
MeCN (85:15)]. With 2.0 mM CuSO4 as a solvent, the retention times
of Ala, Pro, and Ile matched those of the authentic standards of L-Ala
(2.3 min; D-Ala, 1.7 min), L-Pro (7.2 min; D-Pro, 3.7 min), and L-Ile
(23.9 min; D-allo-Ile, 8.6 min, D-Ile, 11.0 min, L-allo-Ile, 16.8 min). With
2.0 mM CuSO4−MeCN (90:10) as a solvent, the retention times of N-
Me-Phe matched those of authentic standards of N-Me-D-Phe (9.3 min;
N-Me-L-Phe, 10.4 min). With 2.0 mM CuSO4−MeCN (85:15) as a
solvent, the retention times of valic acid and isoleucic acid matched
those of authentic standards of L-valic acid (9.4 min; D-valic acid, 6.2
min) and D-allo-isoleucic acid (12.3 min; D-isoleucic acid, 14.4 min, L-
allo-isoleucic acid, 18.8 min, L-isoleucic acid, 23.2 min).
colorless oil: HRESIMS m/z 1300.5844 [M + Na]+ (calcd for
C64H85F6N5O15Na, 1300.5844).
Determination of the Configuration of Izenamide C (3).
Izenamide C (3, 0.7 mg) was dissolved in 100 μL of 9 M HCl in a sealed
tube and heated at 110 °C for 24 h. The hydrolyzed products were
evaporated to dryness and could be separated into each component by
HPLC. Conditions for HPLC separation: column, Cosmosil 5C18-
PAQ (ϕ 4.6 × 250 mm); flow rate, 1.0 mL/min; detection at 215 nm;
solvent H2O. Retention times (tR min) of components: Pro (3.3), Ile
(4.7), N-Me-Phe (12.4). Conditions for HPLC separation: column,
PBr (ϕ 4.6 × 250 mm); flow rate, 1.0 mL/min; detection at 215 nm;
solvent MeCN−H2O−TFA (5:95:0.1). Retention time (tR min) of
component: valic acid (10.6). Conditions for HPLC separation:
column, ODS HG-5 (ϕ 4.6 × 250 mm); flow rate, 1.0 mL/min;
detection at 215 nm; solvent MeCN−H2O−TFA (15:85:0.1).
Retention time (tR min) of component: isoleucic acid (13.1).
The configurations of the amino and hydroxy acids in 3 were
determined by chiral-phase HPLC [DAICEL CHIRALPAK (MA+) (ϕ
4.6 × 50 mm); flow rate, 1.0 mL/min; detection at 254 nm; solvent 2.0
mM CuSO4, 2.0 mM CuSO4−MeCN (90:10), 2.0 mM CuSO4−
MeCN (85:15)]. With 2.0 mM CuSO4 as a solvent, the retention times
of Pro and Ile matched those of authentic standards of L-Pro (5.1 min;
D-Pro, 2.8 min) and L-Ile (16.6 min; D-allo-Ile, 6.7 min, D-Ile, 8.4 min, L-
allo-Ile, 12.2 min). With 2.0 mM CuSO4−MeCN (90:10) as a solvent,
the retention times of N-Me-Phe and valic acid matched those of
authentic standards of N-Me-D-Phe (7.3 min; N-Me-L-Phe, 8.2 min)
and L-valic acid (17.6 min; D-valic acid, 11.3 min). With 2.0 mM
CuSO4−MeCN (85:15) as a solvent, the retention times of isoleucic
acid matched those of authentic standards of D-allo-isoleucic acid (14.1
min; D-isoleucic acid, 16.4 min, L-allo-isoleucic acid, 21.3 min, L-
isoleucic acid, 26.1 min).
Cell Growth Inhibition Assay. HeLa and MCF-7 cells were
cultured at 37 °C with 5% CO2 in Dulbecco’s modified Eagle’s medium
(Nissui) supplemented with 10% heat-inactivated fetal bovine serum
(FBS), 100 units/mL penicillin, 100 μg/mL streptomycin, 0.25 μg/mL
amphotericin, 300 μg/mL L-glutamine, and 2.25 mg/mL NaHCO3.
HL60 cells were cultured at 37 °C with 5% CO2 in RPMI (Nissui)
supplemented with 10% heat-inactivated FBS, 100 units/mL penicillin,
100 μg/mL streptomycin, 0.25 μg/mL amphotericin, 300 μg/mL L-
glutamine, and 2.25 mg/mL NaHCO3. HeLa and MCF-7 cells were
seeded at 4 × 103 cells/well in 96-well plates (Iwaki) and cultured
overnight. HL60 cells were seeded at 2 × 104 cells/well in 96-well
plates. Ten and 1 μM of compounds 1−3 in methanol were then added,
and cells were incubated for 72 h. Cell proliferation was measured by
the MTT assay.
Cathepsin D Inhibition Assay. Cathepsin D-inhibitory activities
of izenamides A−C (1−3) were assessed using a commercial assay kit
provided by Biovision (#K148-100) according to the manufacturer’s
protocol. The assay was attempted at the following concentrations:
1000, 100, 10 nM (1 and 2, n = 3), 10, 1, 0.1 μM (3, n = 3), and 250, 25,
2.5 pM (pepstatin A, n = 2).
Molecular Modeling of Izenamide A (1) with Cathepsin D.
The molecular modeling of izenamide A (1) and cathepsin D was
carried out using MacroModel (version 11.3).24 The crystal structure of
cathepsin D with pepstatin A (PDB: 1LYB) was used as a starting point.
Based on the cocrystal structure, izenamide A (1) was manually docked
into the binding pocket of cathepsin D, and the hydroxy group in the Sta
unit in 1, which was expected to interact with the active site of the
enzyme, was placed within 4 Å of the two aspartate residues in the active
site, Asp33 and Asp231.21 Torsional sampling by the Monte Carlo
Multiple Minimum method (1000 steps) was used to perform a
conformational search, and the obtained complex was minimized using
an OPLS3 force field and the Polak−Ribier conjugate gradient
algorithm (maximum derivative less than 0.1 kJ/mol) with water as a
solvent. To perform a conformational search, all bonds in izenamide A
(1) and amino acid residues within 3 Å of 1 were considered as torsional
sampling targets, except for amide bonds. In addition, residues within 6
Å of the ligand were set to be free, those between 6 and 10 Å were
constrained by a force of 200 kJ/(mol·Å2), and those between 10 and 20
Å were set to be frozen. To select nonredundant conformers, we
To a stirred solution of izenamide B (2) (0.3 mg, 0.36 μmol) in
pyridine (100 μL) were added 1 drop of (R)-MTPACl and 4-
(dimethylamino)pyridine (0.1 mg, 0.82 μmol). The mixture was stirred
at rt for 24 h and concentrated to give an oil, which was purified by
HPLC [conditions for HPLC separation: column, Cosmosil 5C18-MS-
II (ϕ 4.6 × 250 mm); flow rate, 5.0 mL/min; detection at 215 nm;
solvent MeOH−H2O (90:10); tR = 42.6, 43.5 min] to afford two kinds
of (S)-MTPA esters of 2 (A, 0.3 mg; B, 0.2 mg, respectively) as colorless
oils: HRESIMS m/z 1300.5841 for A, 1300.5851 for B [M + Na]+
(calcd for C64H85F6N5O15Na, 1300.5844).
To a stirred solution of 2 (0.3 mg, 0.36 μmol) in pyridine (100 μL)
were added 1 drop of (S)-MTPACl and 4-(dimethylamino)pyridine
(0.1 mg, 0.82 μmol). The mixture was stirred at rt for 24 h and
concentrated to give an oil, which was purified by HPLC [conditions for
HPLC separation: column, Cosmosil 5C18-MS-II (ϕ 4.6 × 250 mm);
flow rate, 5.0 mL/min; detection at 215 nm; solvent MeOH−H2O
(90:10); tR = 40.2 min] to afford the (R)-MTPA ester of 2 (0.3 mg) as a
H
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
eliminated redundant conformers by setting a 1.0 Å root-mean-square
deviation minimum cutoff for saving structure. To validate our
calculation approach, we carried out the same calculation for the
complex of pepstatin A and cathepsin D and successfully regenerated
the structure of 1LYB (Figure S38). The Boltzmann distribution of the
obtained conformations was calculated by MacroModel (Table S2).
The factors that affected the stability of the ligand/receptor complex
were analyzed based on the results of the calculation using XP (extra
precision) mode (Table S3). The contribution of each amino acid
residue to stabilization of the complex was estimated based on the
results of the same calculation (Table S4). The hydrogen bonds were
defined as having donor−acceptor distances within 2.5 Å from the
protein to the inhibitor.
(9) Miller, B.; Friedman, A. J.; Choi, H.; Hogan, J.; McCammon, J. A.;
Hook, V.; Gerwick, W. H. J. Nat. Prod. 2014, 77, 92−99.
(10) (a) Kwan, J. C.; Eksioglu, E. A.; Liu, C.; Paul, V. J.; Luesch, H. J.
Med. Chem. 2009, 52, 5732−5747. (b) Al-Awadhi, F. H.; Law, B. K.;
Paul, V. J.; Luesch, H. J. Nat. Prod. 2017, 80, 2969−2986.
(11) Koblinskia, J. E.; Ahrama, M.; Sloane, B. F. Clin. Chim. Acta 2000,
291, 113−135.
(12) Strooper, B. D. Physiol. Rev. 2010, 90, 465−494.
(13) Brik, A.; Wong, C. H. Org. Biomol. Chem. 2003, 1, 5−14.
(14) The reaction was achieved under the same conditions (0.5 M
NaOH− MeOH, 1:1, rt) as for the basic hydrolysis of grassystatin A.9
(15) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem.
Soc. 1991, 113, 4092−4096.
(16) Preciado, A.; Williams, P. G. J. Org. Chem. 2008, 73, 9228−9234.
(17) The exact position of epimerization in the (S)-MTPA ester could
not be identified because of the scarcity of the samples. We assumed
that the epimerization could occur at other acylated oxymethines, C-35
or C-41.
(18) (a) Williams, P. G.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J. J.
Nat. Prod. 2003, 66, 1006−1009. (b) Sun, T.; Zhang, W.; Zong, C.;
Wang, P.; Li, Y. J. Pept. Sci. 2010, 16, 364−374. (c) Al-Awadhi, F. H.;
Ratnayake, R.; Paul, V. J.; Luesch, H. Bioorg. Med. Chem. 2016, 24,
3276−3282.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.8b00417.
■
S
NMR spectra and MS/MS spectra for all new
compounds; determination of the absolute configuration
of izenamides A, B, and C; molecular modeling of
pepstatin A and cathepsin D; identification of the
cyanobacterium (PDF)
(19) (a) Morishima, H.; Takita, T.; Aoyagi, T.; Takeuchi, T.;
Umezawa, H. J. Antibiot. 1970, 23, 263−265. (b) Umezawa, H.; Aoyagi,
T.; Morishima, H.; Matsuzaki, M.; Hamada, M.; Takeuchi, T. J.
Antibiot. 1970, 23, 259−262.
(20) Baldwin, E. T.; Bhat, T. N.; Gulnik, S.; Hosur, M. V.; Sowder, R.
C.; Cachau, R. E.; Collins, J.; Silva, A. M.; Erickson, J. W. Proc. Natl.
Acad. Sci. U. S. A. 1993, 90, 6796−6800.
AUTHOR INFORMATION
Corresponding Author
*E-mail: suenaga@chem.keio.ac.jp. Tel: +81-45-566-1819.
ORCID
Arihiro Iwasaki: 0000-0002-3775-5066
Teruhiko Matsubara: 0000-0002-8006-4324
Toshinori Sato: 0000-0002-4429-6101
Kiyotake Suenaga: 0000-0001-5343-5890
■
(21) Benesa, P.; Vetvickab, V.; Fusekc, M. Crit. Rev. Oncol Hematol
2008, 68, 12−28.
(22) Matarrese, P.; Nencioni, L.; Checconi, P.; Ciarlo, L.;
Gambardella, L.; Ascione, B.; Sgarbanti, R.; Garaci, E.; Malorni, W.;
Palamara, A. T. J. Cell. Physiol. 2011, 226, 3368−3377.
̈
(23) MacroModel; Schrodinger LLC, New York, NY, 2017.
(24) Pringsheim, E. G. Pure Cultures of Algae; Cambridge University
Press: London, 1949.
̈
(25) Nubel, U.; Garcia-Pichel, F.; Muyzer, G. Appl. Environ. Microbiol.
1997, 63, 3327−3332.
Notes
The authors declare no competing financial interest.
(26) Taton, A.; Grubisic, S.; Brambilla, E.; Wit, R. D.; Wilmotte, A.
Appl. Environ. Microbiol. 2003, 69, 5157−5169.
(27) Murakami, A.; Miyashita, H.; Iseki, M.; Adachi, K.; Mimuro, M.
Science 2004, 303, 1633.
ACKNOWLEDGMENTS
■
This work was supported in part by Grants-in-Aid from the
Ministry of Education, Culture, Sports, Science and Technology
of Japan (16H03285 and 17H05450) and Keio Leading-edge
Laboratory of Science, and Technology (KLL)-specified
research projects.
́
(28) Komarek, J.; Anagnostidis, K. In Suesswasserflora von
Mitteleuropa, Bd. 19/2: Cyanoprokaryota: Bd. 2/Part 2: Oscillatoriales;
Spektrum Akademischer Verlag: Stuttgart, 2008.
(29) Engene, N.; Paul, V. J.; Byrum, T.; Gerwick, W. H.; Thor, A.;
Ellisman, M. H. J. Phycol. 2013, 49, 1095−1106.
̌
́
(30) Engene, N.; Rottacker, E. C.; Kastovsky, J.; Byrum, T.; Choi, H.;
REFERENCES
■
́
Ellisman, M. H.; Komarek, J.; Gerwick, W. H. Int. J. Syst. Evol. Microbiol.
(1) Shah, S. A. A.; Akhter, N.; Auckloo, B. N.; Khan, I.; Lu, Y.; Wang,
K.; Wu, B.; Guo, Y. W. Mar. Drugs 2017, 15, 354−383.
(2) (a) Pettit, G. R.; Kamano, Y.; Herald, C. L.; Tuinman, A. A.;
Boettner, F. E.; Kizu, H.; Schmidt, J. M.; Baczynskyj, L.; Tomer, K. B.;
Bontems, R. J. J. Am. Chem. Soc. 1987, 109, 6883−6885. (b) Luesch, H.;
Moore, R. E.; Paul, V. J.; Mooberry, S. L.; Corbett, T. H. J. Nat. Prod.
2001, 64, 907−910.
2012, 62, 1171−1178.
(31) Engene, N.; Tronholm, A.; Salvador-Reyes, L. A.; Luesch, H.;
Paul, V. J. J. Phycol. 2015, 51, 670−681.
(32) Kumar, S.; Stecher, G.; Tamura, K. Mol. Biol. Evol. 2016, 33,
1870−1874.
(33) Darriba, D.; Taboada, G. L.; Doallo, R.; Posada, D. Nat. Methods
2012, 9, 772.
(3) Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J.; Corbett, T. H.
J. Am. Chem. Soc. 2001, 123, 5418−5423.
(34) Guindon, S.; Gascuel, O. Syst. Biol. 2003, 52, 696−704.
(35) Stamatakis, A. Bioinformatics 2014, 30, 1312−1313.
(4) Teruya, T.; Sasaki, H.; Fukazawa, H.; Suenaga, K. Org. Lett. 2009,
11, 5062−5065.
̀
(36) Perriere, G.; Gouy, M. Biochimie 1996, 78, 364−369.
(37) Ronquist, F.; Teslenko, M.; van der Mark, P.; Ayres, D. L.;
(5) Iwasaki, A.; Ohno, O.; Sumimoto, S.; Suda, S.; Suenaga, K. RSC
Adv. 2014, 4, 12840−12843.
̈
Darling, A.; Hohna, S.; Larget, B.; Liu, L.; Suchard, M. A.; Huelsenbeck,
J. P. Syst. Biol. 2012, 61, 539−542.
(6) Senter, P. D.; Sievers, E. L. Nat. Biotechnol. 2012, 30, 631−637.
(7) Kanamori, Y.; Iwasaki, A.; Sumimoto, S.; Suenaga, K. Tetrahedron
Lett. 2016, 57, 4213−4216.
(8) (a) Taori, K.; Matthew, S.; Rocca, J. R.; Paul, V. J.; Luesch, H. J.
Nat. Prod. 2007, 70, 1593−1600. (b) Kwan, J. C.; Taori, K.; Paul, V. J.;
Luesch, H. Mar. Drugs 2009, 7, 528−538.
I
DOI: 10.1021/acs.jnatprod.8b00417
J. Nat. Prod. XXXX, XXX, XXX−XXX