Suberoylanilide hydroxamic acid, a histone deacetylase inhibitor, induces the production of anti-inflammatory cyclodepsipeptides from Beauveria felina

Article abstract of DOI:10.1021/np400143j

The addition of the histone deacetylase inhibitor suberoylanilide hydroxamic acid to a culture of the filamentous fungus Beauveria felina significantly changed its secondary metabolite profile and led to the isolation of eight compounds, including three new cyclodepsipeptides, desmethylisaridin E (1), desmethylisaridin C2 (2), and isaridin F (3), along with five known cyclodepsipeptide compounds. Isaridin F (3) possesses a cyclodepsipeptide ring with N-methylbutyric acid, which is rare in natural peptides. Absolute configurations of the new cyclodepsipeptides were achieved by Marfey's method. The anti-inflammatory activity of the isolated compounds was investigated through evaluating their effect on superoxide anion production and elastase release by FMLP-induced human neutrophils. Among the tested compounds, desmethylisaridin E (1) inhibited superoxide anion production and desmethylisaridin C2 (2) inhibited elastase release, with IC₅₀ values of 10.00 ± 0.80 and 10.01 ± 0.46 μM, respectively.

Full text of DOI:10.1021/np400143j

Article
pubs.acs.org/jnp
Suberoylanilide Hydroxamic Acid, a Histone Deacetylase Inhibitor,
Induces the Production of Anti-inflammatory Cyclodepsipeptides
from Beauveria felina
Yu-Ming Chung,† Mohamed El-Shazly,†,‡ Da-Wei Chuang,† Tsong-Long Hwang,§ Teigo Asai,⊥
Yoshiteru Oshima,⊥ Mohamed L. Ashour,‡ Yang-Chang Wu,*,†,∥,∇,○ and Fang-Rong Chang*,†
†Graduate Institute of Natural Products, College of Pharmacy, Kaohsiung Medical University, Kaohsiung 807, Taiwan, Republic of
China
‡Department of Pharmacognosy and Natural Products Chemistry, Faculty of Pharmacy, Ain-Shams University, Organization of
African Unity Street 11566, Abassia, Cairo, Egypt
§Graduate Institute of Natural Products, College of Medicine, Chang Gung University, Taoyuan 33302, Taiwan, Republic of China
⊥Graduate School of Pharmaceutical Sciences, Tohoku University, Aoba-yama, Aoba-ku, Sendai 980-8578, Japan
∥School of Pharmacy, College of Pharmacy, China Medical University, Taichung 404, Taiwan, Republic of China
∇Chinese Medicine Research and Development Center, China Medical University Hospital, Taichung 404, Taiwan, Republic of
China
○Center for Molecular Medicine, China Medical University Hospital, Taichung 404, Taiwan, Republic of China
S
* Supporting Information
ABSTRACT: The addition of the histone deacetylase
inhibitor suberoylanilide hydroxamic acid to a culture of the
filamentous fungus Beauveria felina significantly changed its
secondary metabolite profile and led to the isolation of eight
compounds, including three new cyclodepsipeptides, desme-
thylisaridin E (1), desmethylisaridin C2 (2), and isaridin F (3),
along with five known cyclodepsipeptide compounds. Isaridin
F (3) possesses a cyclodepsipeptide ring with N-methylbutyric
acid, which is rare in natural peptides. Absolute configurations of the new cyclodepsipeptides were achieved by Marfey’s method.
The anti-inflammatory activity of the isolated compounds was investigated through evaluating their effect on superoxide anion
production and elastase release by FMLP-induced human neutrophils. Among the tested compounds, desmethylisaridin E (1)
inhibited superoxide anion production and desmethylisaridin C2 (2) inhibited elastase release, with IC50 values of 10.00 0.80
and 10.01 0.46 μM, respectively.
diseases.5 Biologically active compounds have been obtained
eutrophils play a pivotal role in the defense of the human
Nbody against infections. However, abnormal activation of directly from fungal culture medium cultivated under normal
growth conditions. However, focusing on pathways active only
under normal growth conditions may be relatively unproduc-
tive, because most of the major active metabolites may have
been already isolated.6 Recently, the addition of histone
deacetylase (HDAC) and/or DNA methyltransferase inhibitors
to a fungal culture medium as a chemical epigenetic modifier
proved to be an effective technique for stimulating the
transcription of silent biosynthetic gene clusters, resulting in
the production of a variety of novel secondary metabolites.7−9
We have applied this strategy to filamentous fungi through the
concomitant addition of SBHA (suberic bis-hydroxamic acid,
an HDAC inhibitor) and RG-108 (a DNMT inhibitor) to the
fungal cultural medium. Two new highly oxidized ergosterols,
five new isariotin analogues from Gibellula formosana, and one
neutrophils releases high concentrations of reactive oxygen
species (ROS) and elastase, which mediate immunological
insults and oxidative damage.1 In response to diverse stimuli,
activated neutrophils secrete a series of cytotoxins, such as the
superoxide anion, a precursor for ROS. The formation of
superoxide anion by NADPH oxidase in human neutrophils is
directly or indirectly linked to the damage and destruction of
the surrounding tissues.2,3 Neutrophil elastase is a major
secreted product of the stimulated neutrophils and a major
contributor to the destruction of tissue in chronic inflammatory
disease.4 It is crucial to restrain the neutrophil oxidative burst in
inflammatory reactions to minimize tissue damage. Finding new
natural anti-inflammatory agents that inhibit the release of
neutrophil inflammatory mediators with high efficiency is our
goal in the current study.
Microbial natural products have provided a myriad of
Received: February 18, 2013
therapeutically active agents effective in targeting life-threating
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
dx.doi.org/10.1021/np400143j | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
novel skeletal polyketide from Isaria tenuipes were obtained.10,11
These results encouraged us to examine the effect of epigenetic
modifying agents on the secondary metabolites profile of other
filamentous fungi. Screening secondary metabolites produced
by filamentous fungi grown in the presence of these inhibitors
showed that the addition of 0.5 mM SAHA (suberoylanilide
hydroxamic acid, an HDAC inhibitor) to the culture medium of
Beauveria felina induced significant changes in the metabolite
profile (Figure 1). The experiment was scaled-up, and further
cyclodepsipeptides, desmethylisaridin E (1), desmethylisaridin
C2 (2), and isaridin F (3), together with five known
cyclodepsipeptides. Isaridin F (3) possesses a cyclodepsipeptide
ring with an α-N-methylbutyric acid, which is rare in natural
peptides (Figure 2). The isolated compounds were evaluated
for their anti-inflammatory activity on superoxide anion
production and elastase release in FMLP-induced human
neutrophils. Desmethylisaridin E (1) showed selective activity
on superoxide anion production with an IC50 value of 10.00
0.80 μM, whereas desmethylisaridin C2 (2) displayed selective
activity on elastase release with an IC50 value of 10.01 0.46
μM.
RESULTS AND DISCUSSION
■
Desmethylisaridin E (1) was isolated as a colorless oil. Its
molecular formula, C34H51N5O7, was determined by HRESIMS
at m/z 664.3692 [M + Na]+ (calcd for C34H51N5O7Na,
664.3686), which corresponds to 12 degrees of unsaturation.
The UV spectrum showed absorption at 208 nm. The IR
spectrum indicated the presence of amine (3357 cm−1), ester
carbonyl (1724 cm−1), amide carbonyl (1642 cm−1), and
aromatic (1536 cm−1) functionalities. Analysis of the 13C NMR
and DEPT spectra of 1 indicated the presence of six carbonyls,
one quaternary aromatic carbon, five tertiary aromatic carbons,
one tertiary oxymethine carbon, four tertiary nitrogenated
methine carbons, seven methylene carbons, three methine
carbons, a single N-methyl, and six methyls, suggesting a cyclic
hexapeptide-type compound (Table 1). The 1H NMR
spectrum revealed the presence of three amide NH protons
(δH 8.06, 7.12, and 6.50), five aromatic protons (δH 7.27−7.18
for H-25 to H-29), five alpha protons [δH 5.20 (H-2), 4.62 (H-
22), 4.50 (H-17), 4.30 (H-11), 4.12 (H-31)], one N-methyl
group (δH 2.91, H3-15), and six methyl groups (δH 0.89−0.99
for H3-5, H3-6, H3-13, H3-14, H3-19, and H3-20), indicating a
cyclic hexadepsipeptide with a phenylalanine unit.12
Figure 1. HPLC profile of B. felina EtOAc extract cultivated in YM
medium containing 0.5 mM SAHA (a) and control (b) as detected by
UV absorption at 215 nm.
The assignment of the proton and carbon signals to the
amino acid residues was achieved using 2D NMR (1H−1H
COSY, HSQC, and HMBC experiments). From the HSQC
investigation of the culture broth treated with SAHA led to the
isolation of eight cyclodepsipeptides, including three new
Figure 2. Structures of the isolated compounds 1−8.
B
dx.doi.org/10.1021/np400143j | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
a
1
Table 1. H and 13C NMR Data in CDCl3 for Compounds 1−3
1
2
3
position
δH mult. (J in Hz)
δC
δH mult. (J in Hz)
δC
position
δH mult. (J in Hz)
HMPA
HMPA
HMPA
1
169.8
73.2
38.9
170.1
73.2
39.1
1
2
5.20 dd (11.2, 2.2)
1.96 m
5.15 dd (11.0, 2.0)
1.93 m
2
5.20 dd (11.0, 2.0)
1.97 m
3a
3b
4
3a
3b
4
1.29 m
1.24 m
1.28 m
1.92 m
24.8
23.3
20.9
1.97 m
24.8
23.3
21.0
1.95 m
5
0.99 d (6.5)
0.96 d (6.5)
β-Ala
0.99 d (6.7)
0.97 d (6.5)
β-Ala
5
0.98 d (6.6)
0.96 d (6.6)
β-Ala
6
6
7
173.5
34.9
173.4
34.9
7
8a
8b
9a
9b
NH
2.65 dt (16.0, 3.0)
2.53 dt (12.2, 3.6)
4.09 m
2.66 dt (15.8, 3.6)
2.54 m
8a
8b
9a
9b
NH
2.61 m (overlapped)
2.61 m (overlapped)
4.14 m
35.2
4.11 m
35.2
3.21 t (12.7)
7.12 d (9.2)
N-Me-Val
3.22 m
3.20 m
7.09 d (9.2)
N-Me-Val
7.51 d (8.4)
α-N-Me-BA
10
11
12
13
14
15
168.4
66.5
27.0
19.5
19.4
29.1
168.4
66.5
27.0
19.5
19.4
29.1
10
11
12
13
14
4.30 d (10.7)
2.46 m
4.26 d (10.7)
2.46 m
4.73 m
2.82 m, 1.31 m
0.93 t (7.2)
2.87 s
0.95 d (6.5)
0.90 d (6.5)
2.92 s
0.95 d (6.9)
0.89 d (6.5)
2.91 s
N-Me-Val
Val
Val
15
16
17
18
19
20
16
17
18
19
20
NH
171.7
54.7
31.5
20.2
18.7
171.6
54.6
31.5
20.1
18.8
5.10 d (10.8)
2.33 m
4.50 t (9.2)
2.12 m
4.50 m
2.11 m
0.91 d (6.6)
0.88 d (6.6)
3.16 s
0.89 d (6.2)
0.96 d (6.2)
6.50 d (9.2)
Phe
0.88 d (6.7)
0.99 d (6.5)
6.20 d (9.2)
Phe
Phe
21
21
172.5
55.2
37.2
172.3
55.2
37.2
22
4.75 m
22
4.62 m
4.53 m
23a
23b
24
3.04 dd (14.4, 4.8)
2.93 dd (13.8, 10.8)
23a
23b
24
3.07 dd (13.7, 5.9)
2.95 m
3.05 dd (13.8, 6.2)
2.98 dd (10.6, 3.1)
136.3
128.8
128.9
127.2
136.2
128.8
128.9
127.2
25, 29
26, 28
27
7.20 m
7.29 m
7.24 m
8.08 d (7.8)
Pro
25, 29
26, 28
27
7.18 d (6.8)
7.27 m
7.17 d (6.8)
7.28 m
7.25 m
7.24 m
NH
NH
8.06 d (8.2)
Pro
7.98 d (8.2)
β-Me-Pro
30
30
171.6
61.0
32.0
171.3
68.0
39.3
31
4.10 dd (7.8, 1.2)
2.24 m
31
4.12 d (8.7)
2.27 dd (12.9, 6.4)
2.08 m
3.70 d (2.0)
2.62 m
32a
32b
33a
33b
34a
34b
32a
32b
33a
33b
34a
34b
35
2.12 m
1.78 m
1.74 m
21.8
47.0
1.42 m (overlapped)
1.42 m (overlapped)
3.69 m
28.6
45.1
19.1
1.39 m
2.00 m
3.50 m (overlapped)
3.50 m (overlapped)
3.47 m (overlapped)
3.47 m (overlapped)
3.37 m
1.08 d (7.1)
a
Assignments were based on HSQC and HMBC experiments.
1H−1H COSY, indicating the presence of a proline unit. The
spin systems O-CH-CH2-CH(CH3)2, NH-CH2-CH2, CH3-N-
CH-CH(CH3)2, and NH-CH-CH(CH3)2 were also identified
and confirmed by 1H−1H COSY and HMBC spectra,
suggesting the presence of 2-hydroxy-4-methylpentanoic acid
(HMPA), β-alanine, N-Me-valine, and valine residues (Figure
3). The sequence of the five amino acid units was defined by
alpha protons and amide NH signals to the carbonyl functional
spectrum, seven methylenes [δH 1.29, 1.96 (H2-3), 2.53, 2.65
(H2-8), 3.21, 4.09 (H2-9), 2.95, 3.07 (H2-23), 2.08, 2.27 (H2-
32), 1.74, 2.00 (H2-33), 3.47 (H2-34)] were correlated to δC
38.9 (C-3), 34.9 (C-8), 35.2 (C-9), 37.2 (C-23), 32.0 (C-32),
21.8 (C-33), and 47.0 (C-34), respectively. The 1H−1H COSY
and HMBC correlations from amide NH protons led to the
identification of the amino acid side chains. The independent
spin system of N-CH-CH2-CH2-CH2-N was defined using
C
dx.doi.org/10.1021/np400143j | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
1
Figure 3. H−1H COSY and HMBC correlations and key NOE effects of 1 and 2.
groups based on the following HMBC cross-peaks: H-2/C-1,
C-4; H2-8/C-7; H2-9/C-8, C-9; NH (δH 7.12)/C-10; H-11/C-
10, C-12, C-15, C-16; H-17/C-16, C-18, C-19, C-21; H-22/C-
21; NH (δH 8.06)/C-23, C-30; H-31/C-30, C-33, C-34 (Figure
3). Upon extensive analysis of these data, desmethylisaridin E
(1) was assigned as a cyclic hexapeptide containing one
equivalent of Phe1, Pro2, HMPA3, β-Ala4, N-Me-Val5, and Val6,
which was consistent with the 12 degrees of unsaturation. The
closest known compound to this new cyclodepsipeptide is
isaridin E (4), which is also a major metabolite of B. felina.
However, the major difference in structure is the replacement
of the N-Me-Val residue in 4 with the Val6 residue in 1.
Additionally, it was found that 1 fulfills the description of
isaridins reported by Balaram et al. in 2007, in which it was
shown that the cyclodepsipeptides belonging to the isaridin
type possess an α-hydroxy acid and a β-amino acid, with a
preponderance of N-alkylated residues.13
The sequence of the amino acids in 1 was further confirmed
on the basis of EIMS experiments. The molecular ion m/z 641
[M]+ of 1 was deduced by EIMS, and the fragmentation was
followed by MSn. The EIMS spectrum of 1 showed preferential
opening of the macrocyclic ring at the N-Me-Val5-Val6 amide
bond, followed by a series of major fragmentation pathways
(Figure 4). One ion series started with the loss of 99 amu due
The relative configurations of Phe1, Pro2, HMPA3, N-Me-
Val5, and Val6 alpha protons were assigned by the analysis of
NOESY correlations (Figure 3). In the NOESY spectrum of 1,
the alpha proton of HMPA3 (δH 5.20, H-2) and Pro2 (δH 4.12,
H-31) showed NOE effects, indicating that these protons
possess the same orientation with respect to the macrocyclic
ring. We found that Phe1, Val6, and N-Me-Val5 also possess the
same relative configuration based on the NOE data [Phe1 (δH
4.62, H-22)/Val6 (δH 7.29, NH); Val6 (δH 7.29, NH)/Val6 (δH
2.12, H-18); Val6 (δH 4.50, H-17)/N-Me-Val5 (δH 4.30, H-11)].
The configuration of Pro protons was assigned as cis, which was
confirmed by the large chemical shift difference (10.2 Hz)
between the β- and γ-carbons (Δδβγ: trans < 6 Hz, cis > 8
Hz).14
The absolute configurations of the four alpha amino acids in
1 were determined through a combination of Marfey’s method
and the use of reversed-phase HPLC.15 A small sample of 1 was
hydrolyzed with 6 N HCl, and the hydrolysate was derivatized
with 1-fluoro-2,4-dinitrophenyl-5-L-alanine amide (FDAA).
The HPLC analysis of the acid hydrolysate derivative of 1
resulted in the same retention times as those prepared from L-
Phe, L-Pro, N-Me-L-Val, and L-Val authentic samples. The four
amino acid units (Phe, Pro, N-Me-Val, and Val) in 1 were
assigned as L-configured. On the basis of the relative
configuration of alpha protons established between Pro and
HMPA by NOESY data, the absolute configuration of C-2 in
HMPA was deduced as S. On the basis of the aforementioned
analysis, compound 1 was confirmed as a new cyclodepsipep-
tide and was named desmethylisaridin E.
Desmethylisaridin C2 (2) was obtained as a colorless oil with
a molecular formula, C35H53N5O7, determined by HRESIMS at
m/z 678.3848 [M + Na]+ (calcd for C35H53N5O7Na,
678.3842), which was larger than desmethylisaridin E (1) by
14 mass units. This indicated that an additional methyl group is
present in 2. The 1H and 13C NMR spectra of 2 were similar to
those of 1, except for the presence of an additional methyl
group at δc 19.1/δH 1.08 (3H, d, J = 7.1 Hz, H3-35) (Table 1).
1
The cross-peaks in the H−1H COSY and HMBC spectra
Figure 4. EIMS fragments of 1.
clarified the sequence of all amino acid derivatives in the
macrocyclic ring. The HMBC correlations of H-31/C-32, C-33,
C-34, C-35; H3-35/C-32 suggested that the methyl group is
attached on the β position of the Pro unit (Figure 3).
Moreover, the EIMS fragmentation pattern of 2 was also
identical to that of 1, which established the sequence of amino
acids as cyclo(-Phe1-β-Me-Pro2-HMPA3-β-Ala4-N-Me-Val5-
Val6-) (Figure 5). The structure of 2 was similar to isaridin
C2 isolated from Isaria felina, except for the replacement of
valine6 residue in isaridin C2 by N-Me-Val residue in 2.13
Determination of the relative configurations of the amino
acids alpha protons in 2 was based on NOESY analysis. The
to Val, leaving m/z 542 (Phe1-Pro2-HMPA3-β-Ala4-N-Me-Val5),
which then lost 28 amu (carbonyl group of Phe), affording m/z
514. Another pathway left a series of fragments at m/z 528,
457, 343, corresponding to the successive loss of N-Me-Val, β-
Ala, and HMPA, respectively, from the N-terminal to the C-
terminal. The base ion peak observed at m/z 556 was obtained
by cleaving the bond between the carbonyl and the α-carbon on
N-Me-Val. Moreover, another pathway was also observed
leaving major fragments at m/z 626 and 598 due to the loss of
15 amu (CH3 of N-Me-Val) and 43 amu (i-Pr of Val),
respectively.
D
dx.doi.org/10.1021/np400143j | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
systems established six moieties containing Phe1, Pro2, HMPA3,
β-Ala4, α-N-methylbutyric acid (α-N-Me-BA5), and N-Me-Val6.
The structure of 3 was assigned as cyclo-(-Phe1-Pro2-HMPA3-
β-Ala4-α-N-Me-BA5-N-Me-Val6-). This is a new cyclic hexadep-
sipeptide characterized by the presence of an α-N-Me-BA
residue, unprecedented in the isaridin series.
Due to limited isolated compound, 13C NMR, HSQC, and
HMBC experiments could not be conducted. EIMS was utilized
for the confirmation of the amino acid sequence of 3. The
EIMS cleavage fragment patterns of 3 were similar to those of
1. In the EIMS of 3, a molecular ion peak at m/z 641 [M]+ was
observed, with a preferential opening of the macrocycle at the
α-N-Me-BA5-N-Me-Val6 amide bond, followed by a series of
major fragmentation pathways (Figure 7). Cleavage of the bond
Figure 5. EIMS fragments of 2.
key NOE effects of 2 [Phe1 (δH 4.53, H-22)/Val6 (δH 6.20,
NH); Val6 (δH 4.50, H-17)/ N-Me-Val5 (δH 4.26, H-11); N-
Me-Val5 (δH 4.26, H-11)/β-Ala4 (δH 7.09, NH); HMPA3 (δH
5.15, H-2)/β-Me-Pro2 (δH 3.70, H-31)] were similar to those of
1, indicating that the five subunits, HMPA, β-Me-Pro, Phe, Val,
and N-Me-Val, have the same orientation with respect to the
macrocyclic ring. Moreover, the NOE correlation between the
methyl (δH 1.08, H3-35) and alpha proton (δH 3.70, H-31) in
the β-Me-Pro residue was observed. The absolute configuration
of each alpha proton in 2 was determined by the Marfey’s
method. The configuration of the four amino acid units (Phe,
Pro, N-Me-Val, and Val) in 2 was assigned as L according to the
results obtained from the Marfey’s experiment. Additionally,
the absolute configuration of both the methyl group in the β
position of β-Me-Pro and C-2 in HMPA was assigned as S. On
the basis of the forgoing data, the structure of compound 2 was
deduced and named desmethylisaridin C2 (Figure 2).
Figure 7. EIMS fragments of 3.
between the carbonyl and α-carbon of the fifth amino acid
easily produced the base ion peak, which was observed at m/z
570, supporting the presence of an α-N-Me-BA moiety. The
EIMS data were consistent with those of 1 and 2. On the other
hand, the fragmentation of the lesser abundant linear peptide
generated after cleavage of the i-Pr bond on N-Me-Val6 left a
fragment at m/z 598 corresponding to [M − i-Pr]+, which was
followed by a successive loss of m/z 113, 142, and 99 amu,
leaving m/z 528 [M − N-Me-Val6]+, 499 [M − N-Me-Val −
Phe − Pro − HMPA − CO]+, and 542 [M − α-N-Me-BA]+,
respectively. Hence, the structure of compound 3 was deduced
and found to possess a rare α-N-methylbutyric acid unit (Figure
2). The literature indicated that only one cyclodepsipeptide,
enniatin K1, isolated from Fusarium scirpi, possesses this partial
moiety.16
Compound 3 was isolated as a colorless oil. It showed a
pseudomolecular ion peak at m/z 664.3691 [M + Na]+ (calcd
for C34H51N5O7Na, 664.3686) based on HRESIMS, corre-
sponding to the molecular formula C34H51N5O7 with 12
degrees of unsaturation. IR absorption revealed the presence of
NH (3348 cm−1), ester carbonyl (1724 cm−1), amide carbonyl
(1669 cm−1), and aromatic (1550 cm−1) groups. The 1H NMR
of 3 suggested a structural similarity to 1, except for the absence
of a methyl group and the presence of an additional N-methyl
moiety at δH 3.16 (H3-20). The spectral data revealed that 3 is
1
also an isaridin-type compound. Detailed analysis of the H
NMR spectrum of 3 revealed the presence of two amide
protons (δH 7.51 and 8.08), five aromatic protons (δH 7.20−
7.29, H-25 to H-29), five alpha protons [δH 5.20 (H-2), 5.10
(H-16), 4.75 (H-22), 4.73 (H-11), 4.10 (H-31)], two N-methyl
group protons [δH 3.16 (H3-20), 2.87 (H3-14)], and five methyl
group protons [δH 0.98 (H3-5), 0.96 (H3-6), 0.93 (H3-13), 0.91
The five known metabolites isolated in the current study
were identified by comparing their physical and spectroscopic
data with the corresponding authentic samples or literature
values; they were isaridin E (4),17 isaridin C2 (5),13 destruxin A
(6),18 resetoxin B (7),19 and roseocardin (8).20
1
(H3-18), 0.88 (H3-19)]. H−1H COSY data established the
Neutrophils accumulate at sites of inflammation and
immunological reaction in response to locally existing chemo-
tactic mediators. Bacterial N-formyl peptides, such as formyl-L-
methionyl-L-leucyl-L-phenylalanine (FMLP), are some of the
first identified and most potent chemoattractants for
neutrophils.21 The anti-inflammatory effect of the isolated
cyclodepsipeptides was examined, by evaluating their suppres-
sive action on FMLP-induced superoxide anion generation and
elastase release in human neutrophils. Genistein was used as the
positive control. Unfortunately, due to the limited quantity of
isaridin F (3), its biological activity was not tested. As shown in
Figure 8, compounds 1, 2, 4, and 5 inhibited superoxide anion
production, and compounds 2, 4, 5, and 8 inhibited elastase
release. The IC50 values of the tested compounds are shown in
spin systems in 3, leading to the identification of hydroxy and
amino acid residues (Figure 6). The most prominent spin
1
Figure 6. H−1H COSY correlations of 3.
E
dx.doi.org/10.1021/np400143j | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
of some compounds was achieved by a preparative HPLC (Shimadzu
LC-10AT, Shimadzu Inc., Kyoto, Japan) system equipped with a
Thermo ODS Hypersil column (10 × 250 mm, 5 μm, Thermo Fisher
1
Scientific Inc., Rockford, IL, USA). H NMR and 13C NMR spectra
were recorded on Varian Inova 600 and Varian Unity Plus 400 MHz
(Varian Inc., Palo Alto, CA, USA) spectrometers using TMS as an
internal standard. Chemical shifts were reported in ppm (parts per
million), and coupling constants (J) were expressed in Hz. HRESIMS
measurements were performed on a Shimadzu LCMS-IT-TOF liquid
chromatography mass spectrometer (Shimadzu Inc.) with ESI
interface. Optical rotations were measured utilizing a JASCO DIP-
370 digital polarimeter (JASCO Inc., Tokyo, Japan). UV spectra were
obtained on a Hitachi 220-20 spectrophotometer (Hitachi Co., Tokyo,
Japan). IR spectra were measured using a Hitachi 260-30
spectrophotometer (Hitachi Co., Tokyo, Japan). SAHA was
synthesized as shown in Scheme S1 in the Supporting Information.
Fungal Material. B. felina was purchased from Bioresource
Collection and Research Center (BCRC) with the following voucher
number: 32873. The fungal strain was maintained in potato dextrose
agar media at 25 °C for 7 days and directly inoculated into 250 mL
Erlenmeyer flasks containing 75 mL of fermentation YM media (1%
glucose, 0.5% peptone, 0.3% malt extract, and 0.3% yeast extract)
treated either with SAHA (0.5 mM) or without inhibitor (control).
The flasks were incubated on a rotary shaker at 150 rpm at 25 °C.
After 12 days of cultivation, the culture media were extracted by
EtOAc. The EtOAc extracts of each condition were analyzed by
reversed-phase HPLC on a Cosmosil reversed-phase column (250 ×
4.6 mm, 5 μm, 1.0 mL/min, Nacalai Tesque, Kyoto, Japan) with water
and acetonitrile as the mobile phase (0−5 min: 80:20, 5−20 min: from
80:20 to 0:100, 20−30 min: 0:100).
Figure 8. Inhibitory effect of isolated compounds from B. felina on
ROS production and elastase release by human neutrophils in
response to FMLP. Results are presented as percentage of inhibition
at 10 μg/mL concentration (n = 3, 4). **p < 0.01, ***p < 0.001
compared to control. Genistein was used as a positive control.
Table 2. It was found that the presence of a Val6 residue
(compounds 1 and 2) led to a selective inhibitory activity on
Table 2. Effect of Compounds 1−8 on Superoxide Anion
Production and Elastase Release in FMLP-Induced Human
Neutrophils
Fermentation, Extraction, and Isolation. For the scale-up
procedures, B. felina was cultivated using 80 Erlenmeyer flasks (500
mL), each flask containing 150 mL of fermentation YM media in the
presence of 0.5 mM SAHA, which was incubated on a rotary shaker at
150 rpm at 25 °C. After 12 days, 12.0 L of the whole broth was filtered
to separate the broth filtrate from the mycelia. The filtrate was
extracted three times with ethyl acetate and concentrated under
reduced pressure to obtain a dry extract (1.10 g). This extract was
chromatographed on a Sephadex LH-20 column eluted with MeOH to
yield five fractions (F1−F5). Fraction F2 (561.1 mg) was purified
using silica gel flash column with CH2Cl2−MeOH (19:1) as the eluent
to obtain eight subfractions (F2-1 to F2-5). Fraction F2-2 (314.6 mg)
was subjected to a silica gel column eluted with a stepwise gradient
solvent system of increasing polarity starting from 20% EtOAc and
80% dichloromethane to 100% EtOAc, which yielded five subfractions
(F2-2-1 to F2-2-5). Subfraction F2-2-2 was purified using reversed-
phase HPLC on a Thermo ODS Hypersil column (250 × 10 mm, 5
μm, 2.5 mL/min) with acetonitrile and water (70:30 v/v) as the
eluent, yielding compound 5 (50.9 mg). Subfraction F2-2-4 was also
purified by reversed-phase HPLC using acetonitrile and water (50:50
v/v) as the eluent to yield compounds 1 (3.5 mg), 2 (2.5 mg), 3 (0.97
mg), and 4 (121.2 mg). Compounds 6 (21.5 mg), 7 (12.0 mg), and 8
(3.9 mg) were obtained from subfraction F2-2-5 using a silica gel flash
column eluted with CH2Cl2−EtOAc (6:1) and purified by reversed-
phase HPLC with acetonitrile and water (50:50 v/v) as the eluent.
Desmethylisaridin E (1): colorless oil; [α]2D5 −106.9 (c 0.54,
CH3OH); UV (CH3OH) λmax (log ε) 208 (4.48); IR (neat) νmax 3486
(br), 3357 (br), 3284 (br), 2963, 1724, 1674, 1642, 1536, 1174 cm−1;
1H NMR (CDCl3, 600 MHz) and 13C NMR (CDCl3, 150 MHz) data
are given in Table 1; HRESIMS m/z 664.3692 [M + Na]+ (calcd for
C34H51N5O7Na, 664.3686).
a
a
superoxide anion
elastase release
b
b
compound
IC50 (μM)
IC50 (μM)
d
1
2
3
4
5
6
7
8
10.00 0.80
10.90 0.59
N.D.
10.01 0.46
d
d
N.D.
N.D.
12.21 0.98
10.09 0.83
12.76 1.00
12.12 0.72
d
d
N.D.
N.D.
d
d
N.D.
N.D.
d
N.D.
15.09 0.28
c
genistein
0.69 0.02
6.72 1.00
a
b
Results are presented as mean SEM (n = 3, 4). Concentration
c
necessary for 50% inhibition (IC50). Genistein was used as a positive
control. Not determined.
d
FMLP-induced superoxide anion production and elastase
release in neutrophils. The presence of allylic functionality in
compounds 6 and 7 resulted in a significant reduction of anti-
inflammatory activity. Additionally, the cell viability was
measured by LDH release (see Supporting Information). The
LDH release data showed that the active anti-inflammatory
compounds 1, 2, 4, 5, and 8 did not increase LDH release
compared to the control, indicating that these five cyclo-
depsipeptides exhibited anti-inflammatory activity without
toxicity toward human neutrophils.
Desmethylisaridin C2 (2): colorless oil; [α]2D5 −111.3 (c 0.55,
CH3OH); UV (CH3OH) λmax (log ε) 208 (4.47); IR (neat) νmax 3348
EXPERIMENTAL SECTION
■
1
(br), 3284 (br), 2963, 1724, 1678, 1642, 1541, 1174 cm−1; H NMR
General Experimental Procedures. TLC was performed on
silica gel 60, F254 (0.20 nm, Merck KGaA, Darmstadt, Germany), and
compounds were detected using ultraviolet light at 254 and 365 nm
and stained by spraying with 50% H2SO4 followed by heating on a hot
plate. For column chromatography, silica gel (Kieselgel 60, 70−230
and 230−400 mesh, Merck KGaA) and Sephadex LH-20 (Pharmacia
Fine Chemicals AB, Uppsala, Sweden) were used. Further purification
(CDCl3, 600 MHz) and 13C NMR (CDCl3, 150 MHz) data are given
in Table 1; HRESIMS m/z 678.3848 [M + Na]+ (calcd for
C35H53N5O7Na, 678.3842).
Isaridin F (3): colorless oil; [α]2D5 −159.0 (c 0.55, CH3OH); UV
(CH3OH) λmax (log ε) 208 (4.41); IR (neat) νmax 3339 (br), 3282
(br), 2956, 1723, 1676, 1619, 1523, 1171 cm−1; 1H NMR (CDCl3, 600
F
dx.doi.org/10.1021/np400143j | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
MHz) data are given in Table 1; HRESIMS m/z 664.3691 [M + Na]+
(calcd for C34H51N5O7Na, 664.3686).
AUTHOR INFORMATION
Corresponding Author
*Tel: +886-4-220-57153. Fax: +886-4-220-60248. E-mail:
yachwu@mail.cmu.edu.tw (Y.-C. Wu); aaronfrc@kmu.edu.tw
(F.-R. Chang).
■
Marfey’s Analysis of Desmethylisaridin E (1) and Desme-
thylisaridin C2 (2). A sample (ca. 0.5 mg) of the target compound (1
or 2) was hydrolyzed by heating in 6 N HCl (1 mL) at 110 °C for 24
h. The solvent was removed under reduced pressure, and the resulting
material was subjected to further derivatization. After dissolving the
hydrolysate mixture in 0.1 mL of H2O, 0.1 mL of 1 M NaHCO3 and
0.2 mL of 1% Marfey’s reagent (FDAA) in acetone were added, and
the mixture was incubated at 40 °C for 1 h. The reaction mixture was
quenched with 10 μL of 1 N HCl and analyzed by reversed-phase
HPLC. Standard L- and D-amino acids were also derivatized with
FDAA following the same procedures.22 The analysis was performed
under the following conditions: Cosmosil RP-18 column (250 × 4.6
mm, 5 μm, 1.0 mL/min) with a linear gradient of acetonitrile and
water containing 0.05% trifluoroacetic acid (0−10 min: 20:80, 10−40
min: from 20:80 to 100:0, 40−60 min: 100:0), UV detection at 340
nm. The retention times of the authentic acid FDAA derivatives were
as follows: D-Phe (18.1 min), L-Phe (15.2 min), D-Pro (14.6 min), L-
Pro (16.3 min), N-Me-L-Val (17.6 min), N-Me-D-Val (19.3 min), L-Val
(22.9 min), D-Val (24.5 min); the hydrolysate product gave peaks with
retention times of 15.2, 16.4, 17.6, and 23.0 min, corresponding to L-
Phe, L-Pro, N-Me-L-Val, and L-Val, respectively.
Human Neutrophil Superoxide Anion Generation. Human
neutrophils were obtained from venous blood of healthy, adult
volunteers (20−28 years old) by means of dextran sedimentation and
Ficoll centrifugation. Superoxide anion production was assayed by
monitoring the superoxide dismutase-inhibitable reduction of
ferricytochrome c.23 In brief, after supplementation with 0.5 mg/mL
ferricytochrome c and 1 mM Ca2+, neutrophils were equilibrated at 37
°C for 2 min and incubated with the test compound for 5 min. Cells
were incubated with cytochalasin B (CB) (1 μg/mL) for 3 min, before
activation by the tripeptide N-formyl-L-methionyl-L-leucyl-L-phenyl-
alanine (100 nM) for 10 min. Changes in the absorbance with the
reduction of ferricytochrome c at 550 nm were continuously
monitored in a double-beam, six-cell positioner spectrophotometer
under constant stirring. Calculations were based on the differences in
reaction with and without superoxide dismutase (100 U/mL) divided
by the extinction coefficient for the reduction of ferricytochrome c (ε
21.1 mM/10 mm).
Measurement of Elastase Release. Degranulation of azurophilic
granules in human neutrophils was determined by elastase release as
described previously. Experiments were performed using MeO-Suc-
Ala-Ala-Pro-Val-p-nitroanilide as the elastase substrate. After supple-
mentation with MeO-Suc-Ala-Ala-Pro-Val-p-nitroanilide (100 μM),
neutrophils (6 × 105/mL) were equilibrated at 37 °C for 2 min and
incubated with each test compound for 5 min. Cells were activated by
FMLP (100 nM)/CB (0.5 μg/mL), and changes in absorbance at 405
nm were monitored continuously for elastase release. The results were
expressed as the percentage of the initial rate of elastase.24
Lactate Dehydrogenase (LDH) Release. Release of LDH to the
cell medium indicates cell membrane damage. LDH release was
determined by a commercially available method (Promega, Madison,
WI, USA). Cytotoxicity was represented by LDH release in the cell-
free medium as a percentage of the total LDH release. The total LDH
release was determined by lysing cells with 0.1% Triton X-100 for 30
min at 37 °C.25
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by grants from National Science
Council, Taiwan, awarded to Y.-C.W. and F.-R.C. (NSC 98-
2323-B-037-00). The authors would like to thank the Center
for Research Resources and Development (CRRD), Kaohsiung
Medical University, for technical support and services in LC-
MS and NMR analysis.
REFERENCES
■
283.
(1) Fung, Y. L.; Silliman, C. C. Transfus. Med. Rev. 2009, 23, 266−
(2) El-Benna, J.; Dang, P. M.; Gougerot-Pocidalo, M. A.; Marie, J. C.;
Braut-Boucher, F. Exp. Mol. Med. 2009, 41, 217−225.
(3) Nathan, C. Nat. Rev. Immunol. 2006, 6, 173−182.
(4) Pham, C. T. N. Nat. Rev. Immunol. 2006, 6, 541−550.
(5) Berdy, J. J. Antibiot. 2005, 58, 1−26.
(6) Brakhage, A. A.; Schroeckh, V. Fungal Genet. Biol. 2011, 48, 15−
22.
(7) Henrikson, J. C.; Hoover, A. R.; Joyner, P. M.; Cichewicz, R. H.
Org. Biomol. Chem. 2009, 7, 435−438.
(8) Cichewicz, R. H. Nat. Prod. Rep. 2010, 27, 11−22.
(9) Wang, X.; Sena Filho, J. G.; Hoover, A. R.; King, J. B.; Ellis, T. K.;
Powell, D. R.; Cichewicz, R. H. J. Nat. Prod. 2010, 73, 942−948.
(10) Asai, T.; Chung, Y.-M.; Sakurai, H.; Ozeki, T.; Chang, F.-R.;
Wu, Y.-C.; Yamashita, K.; Oshima, Y. Tetrahedron 2012, 68, 5817−
5823.
(11) Asai, T.; Chung, Y.-M.; Sakurai, H.; Ozeki, T.; Chang, F.-R.;
Yamashita, K.; Oshima, Y. Org. Lett. 2011, 14, 513−515.
(12) Vera, B.; Vicente, J.; Rodriguez, A. D. J. Nat. Prod. 2009, 72,
1555−1562.
(13) Sabareesh, V.; Ranganayaki, R. S.; Raghothama, S.; Bopanna, M.
P.; Balaram, H.; Srinivasan, M. C.; Balaram, P. J. Nat. Prod. 2007, 70,
715−729.
(14) Siemion, I. Z.; Wieland, T.; Pook, K.-H. Angew. Chem., Int. Ed.
1975, 14, 702−703.
(15) Bhushan, R.; Bruckner, H. Amino Acids 2004, 27, 231−247.
(16) Pohanka, A.; Capieau, K.; Broberg, A.; Stenlid, J.; Stenstrom, E.;
Kenne, L. J. Nat. Prod. 2004, 67, 851−857.
(17) Langenfeld, A.; Blond, A.; Gueye, S.; Herson, P.; Nay, B.;
Dupont, J.; Prado, S. J. Nat. Prod. 2011, 74, 825−830.
(18) Jegorov, A.; Sedmera, P.; Mat’ha, V. Phytochemistry 1993, 33,
1403−1405.
(19) Engstrom, G. W.; DeLance, J. V.; Richard, J. L.; Baetz, A. L. J.
Agric. Food. Chem. 1975, 23, 244−253.
(20) Tsunoo, A.; Kamijo, M.; Taketomo, N.; Sato, Y.; Ajisaka, K. J.
Antibiot. 1997, 50, 1007−1013.
(21) Le, Y.; Yang, Y.; Cui, Y.; Yazawa, H.; Gong, W.; Qiu, C.; Wang,
J. M. Int. Immunopharmacol. 2002, 2, 1−13.
(22) Marfey, P. Carlsberg Res. Commun. 1984, 49, 591−596.
(23) Chung, Y.-M.; Lan, Y.-H.; Hwang, T.-L.; Leu, Y.-L. Molecules
2011, 16, 1917−1927.
(24) Sklar, L. A.; McNeil, V. M.; Jesaitis, A. J.; Painter, R. G.;
Cochrane, C. G. J. Biol. Chem. 1982, 257, 5471−5475.
(25) Hwang, T.-L.; Fang, C.-L.; Al-Suwayeh, S. A.; Yang, L.-J.; Fang,
J.-Y. Toxicol. Lett. 2011, 203, 172−180.
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Synthesis of SAHA; UV, IR, HRESIMS, EIMS, H NMR,
COSY data for compounds 1, 2, and 3; and 13C NMR, NOESY,
HSQC, HMBC data for compounds 1 and 2 are available free
of charge via the Internet at http://pubs.acs.org.
G
dx.doi.org/10.1021/np400143j | J. Nat. Prod. XXXX, XXX, XXX−XXX