Extracellular signal-regulated kinases (ERK) inhibitors from Aristolochia yunnanensis

Article abstract of DOI:10.1021/np300887d

Six new sesquiterpenoids, aristoyunnolins A-F (1-6), an artifact of isolation [7-O-ethyl madolin W (7)], and 12 known analogues were isolated from stems of Aristolochia yunnanensis. The structures were determined by combined chemical and spectral methods, and the absolute configurations of compounds 2, 3, 5-7, 9, 14, and 17 were determined by the modified Mosher's method and CD analysis. Compounds 1-19 were screened using a bioassay system designed to evaluate the effect on mitogen-activated protein kinases (MAPKs) signaling pathways. Among three MAPKs (ERK1/2, JNK, and p38), compounds 1, 4, 10-13, 16, 18, and 19 exhibited selective inhibition of the phosphorylation of ERK1/2. Compounds 16 and 19 were more active than the positive control PD98059, a known inhibitor of the ERK1/2 signaling pathway.

Full text of DOI:10.1021/np300887d

Article
pubs.acs.org/jnp
Extracellular Signal-Regulated Kinases (ERK) Inhibitors from
Aristolochia yunnanensis
Zhong-Bin Cheng,† Wei-Wei Shao,† Ye-Na Liu, Qiong Liao, Ting-Ting Lin, Xiao-Yan Shen,
and Sheng Yin*
School of Pharmaceutical Sciences, Sun Yat-sen University, Guangzhou, Guangdong 510006, People’s Republic of China
S
* Supporting Information
ABSTRACT: Six new sesquiterpenoids, aristoyunnolins A−F (1−6),
an artifact of isolation [7-O-ethyl madolin W (7)], and 12 known
analogues were isolated from stems of Aristolochia yunnanensis. The
structures were determined by combined chemical and spectral
methods, and the absolute configurations of compounds 2, 3, 5−7,
9, 14, and 17 were determined by the modified Mosher’s method and
CD analysis. Compounds 1−19 were screened using a bioassay system
designed to evaluate the effect on mitogen-activated protein kinases
(MAPKs) signaling pathways. Among three MAPKs (ERK1/2, JNK,
and p38), compounds 1, 4, 10−13, 16, 18, and 19 exhibited selective
inhibition of the phosphorylation of ERK1/2. Compounds 16 and 19
were more active than the positive control PD98059, a known
inhibitor of the ERK1/2 signaling pathway.
itogen-activated protein kinase (MAPK) signaling path-
19). Biological tests verified that compounds 1, 4, 10−13, 16,
Mways are evolutionarily conserved in multicellular 18, and 19 were responsible for the inhibition of the
phosphorylation of ERK1/2. In particular, 16 and 19 had
much stronger activity than the positive control PD98059, a
known inhibitor of the ERK1/2 signaling pathway. Herein,
details of the isolation, structural elucidation, and ERK1/2
phosphorylation inhibitory activity of these compounds are
described.
organisms and play a key role in physiological events such as
gene expression, mitosis, movement, metabolism, and pro-
grammed death.1−3 These pathways are a phosphorelay system
composed of kinases MAPKKK, MAPKK, and MAPK. The
output of these pathways is transduced via MAPK family
members, which phosphorylate and regulate a wide series of
substrates, including other protein kinases, phospholipases,
transcription factors, and cytoskeletal proteins.1 Classic MAPKs
consist of the extracellular signal-regulated kinases (ERK), the
c-Jun NH2-terminal kinases (JNK), and p38.2 ERK1/2 function
in the control of cell division and inhibitors of these enzymes
are being explored as anticancer agents. JNKs are critical
regulators of transcription, and JNK inhibitors may be effective
in the control of rheumatoid arthritis. The p38 MAPKs are
activated by inflammatory cytokines and environmental stresses
and may contribute to diseases such as asthma and auto-
immunity.3 Thus, the MAPK inhibitors are regarded as
potential leads in drug development for multiform diseases.
Aristolochia yunnanensis Franch. (Aristolochiaceae), endemic
to Yunnan Province of China, is known as “Nan Mu Xiang” in
traditional Chinese medicine for the treatment of gastro-
intestinal diseases, trichomoniasis, and rheumatic pain.4
Previous chemical investigation of this plant revealed four
sesquiterpene lactones that showed activity in anticancer
research.5 In our screening program aimed at the discovery of
MAPK inhibitors from medicinal plants, a fraction of the
ethanolic extract of A. yunnanensis showed significant inhibitory
activity. Subsequent chemical investigation led to the isolation
of seven new sesquiterpenoids (1−7) and 12 known ones (8−
RESULTS AND DISCUSSION
■
The air-dried powder of stems of A. yunnanensis was extracted
with 95% EtOH at room temperature to give a crude extract,
which was suspended in H2O and successively partitioned with
petroleum ether, EtOAc, and n-BuOH. Various column
chromatographic separations of the EtOAc extract afforded
compounds 1−19.
Compound 1, a white powder, had the molecular formula
C15H21NO5, as determined by HREIMS and ESIMS. The IR
spectrum exhibited absorption bands for OH (3309 cm−1),
lactone (1753 cm−1), and nitro (1547 and 1368 cm−1)
1
functionalities. The H NMR spectrum showed two methyl
singlets [δH 1.06 (H3-14) and 1.80 (H3-12)], a terminal double
bond [δH 4.86 (1H, br s, H-13a) and 4.87 (1H, br s, H-13b)],
three protons bonded to carbons bearing heteroatoms [δH 4.21
(s, H-5), 4.61 (s, H-6), and 4.63 (dd, J = 8.0, 8.0, H-1)], and a
series of aliphatic methylene multiplets. The 13C NMR
spectrum, in combination with DEPT experiments, resolved
15 carbon resonances attributable to one carbonyl, one sp2
Received: December 20, 2012
Published: April 9, 2013
© 2013 American Chemical Society and
American Society of Pharmacognosy
664
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671

Journal of Natural Products
Article
unsaturation required that 1 was tricyclic. The above-
mentioned information was similar to that of versicolactone
C (8), the first example of a hydroazulene-type sesquiterpene
previously isolated from A. versicolar,6 except for the presence
of a nitro group in 1. HMBC correlations from CH3-14, H2-3,
and H2-9 to a severely downfield-shifted carbon (δC 95.5)
revealed that the nitro group was located at C-1, as the C-1 of 8
bearing an OH resonated at δC 82.2. Detailed 2D analyses
1
(HSQC, H−1H COSY, and HMBC) supported the planar
structure of 1 as depicted. The relative configuration of 1 was
established by analysis of the 1D NOE difference spectra. The
interactions of H-5 (after irradiation) with H-1, H-6, H-7, and
H-9β indicated that the skeletal rings were cis-fused, and these
protons were β-oriented. The interactions of H3-14 (after
irradiation) with H-2α, H-3α, H-8α, and H-9α indicated H3-14
was α-oriented (S70, Supporting Information). Compound 1,
possessing a rare nitro group, was only the second example of a
hydroazulene-type sesquiterpene found in nature, and it was
given the trivial name aristoyunnolin A. Moreover, the initial
assignments of the NMR data (at C-1, C-4, C-5, and C-6) of
versicolactone C (8) were revised by detailed analyses of its 2D
NMR data in this study (see Tables 1 and 2).
quaternary carbon, one sp2 methylene, two methyls, four sp3
methines (three bearing heteroatoms), four sp3 methylenes,
and two sp3 quaternary carbons. As three of the six degrees of
unsaturation were accounted for by a nitro group, a double
bond, and a lactone, the remaining three degrees of
Compound 2, a colorless oil, had the molecular formula
C15H20O4, as established by HRESIMS. The 1H and 13C NMR
a
1
Table 1. H NMR Data of Compounds 1−8
b
b
b
b
b
b
b
c
no.
1
1
2
3
4
5
6
7
8
4.63, dd (8.0, 3.09, d (10.6) 5.14, dd (11.0,
9.38, s
10.15, s
9.30, s
3.79, dd (10.3,
7.6)
8.0)
5.1)
2α
2β
3α
2.70, m
2.14, m
2.31, m
2.31, m
1.87, m
a 2.57, m
2.24, m
2.15, m
1.93, m
1.67, m
a 1.93, m
2.79, ddd (12.0,
3.5, 3.5)
6.62, s
6.29, d (9.2)
5.93, d (10.8)
6.10, d (11.4)
3β
2.31, m
b 2.40, m
2.04, m
b 2.12, m
4
4.53, dddd (9.6, 9.6, 6.20, ddd (10.8, 10.8, 1.93, m
3.8, 3.8)
4.3)
5α
5β
4.21, s
7.06, s
6.20, d (9.1)
a 2.65, d
(8.8)
2.57, m
2.28, m
1.28, m
4.15, s
b 2.65, d
(8.8)
2.22, m
2.51, m
0.90, t (5.4)
6
7
4.61, s
5.10, s
2.36, dd (9.8, 9.1)
4.52, s
2.35, m
2.69, dd
(12.0) 4.1)
1.73, ddd (10.3,
9.8, 2.2)
4.85, dd (6.4, 4.96, dd (7.3, 7.3)
6.4)
4.99 br d (10.8)
2.43, d (8.4)
2.41, m
8α
8β
1.55, m
1.73, m
1.86, m
0.89, m
2.32, m
1.95, m
2.10, m
2.10, m
1.97, m
2.30, m
a 2.03, m
b 1.66, m
1.58, m
1.55, m
1.66, dd (7.5, 1.85, m
3.7)
9α
2.02, dd (15.1, 2.41, m
7.5)
2.10, m
2.18, m
1.97, m
2.00, m
1.99, m
a 2.31, m
1.50, m
1.75, m
9β
1.76, m
2.17, m
2.00, m
2.16, m
b 2.02, m
11
4.70, dd (6.8, 4.73, dd (7.7, 7.7)
6.8)
4.77, dd (11.3, 3.0)
5.24 br d (10.0)
12a 1.80, s
12b
4.90, s
4.99, s
1.85, s
α 2.57, m
β 2.09, m
2.55, m
2.55, m
2.11, m
2.30, m
α 2.03, m
β 2.36, m
2.04, m
2.15, m
2.78, m
1.80, s
13a 4.86 br s
1.41, s
α 1.72, ddd (12.8,
12.8, 3.5)
4.83, s
4.85, s
0.92, s
13b 4.87 br s
2.29, m
2.11, m
β 2.78, ddd (12.8, 3.9, 2.05, m
3.9)
14
15
1.06, s
1.23, s
9.30, s
1.25, s
1.63, s
1.43, s
1.57, s
1.34, s
4-OAc:
1.55, s
1.31, s
1.20, s
1.44, s
11-COOCH3: 3.70, s
2.04, s −OCH2CH3: a 3.56, qdd (7.0,
7.0, 2.0)
b 3.43, qdd (7.0,
7.0, 2.0)
−OCH2CH3: 1.19, t (7.0)
a
b
c
Data were recorded at 400 MHz, chemical shifts are in ppm, coupling constant J is in Hz. In CDCl3. In CD3OD.
665
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671

Journal of Natural Products
Article
Table 2. 13C NMR (100 MHz) Data (δ) of Compounds 1−8
a
a
a
a
a
a
a
b
no.
1
2
3
4
5
6
7
8
1
95.5, CH
25.3, CH2
23.8, CH2
58.8, C
78.3, CH
30.6, CH2
23.4, CH2
134.2, C
124.9, CH
27.7, CH2
23.9, CH2
144.8, C
171.2, C
195.9, CH
143.0, C
191.2, CH
141.3, C
194.1, CH
143.0, C
82.2, CH
29.7, CH2
22.6, CH2
58.6, C
2
135.2, C
3
148.8, CH
106.3, C
154.7, CH
68.3, CH
46.8, CH2
129.8, C
145.0, CH
65.7, CH
45.2, CH2
128.4, C
158.1, CH
25.9, CH
26.8, CH2
28.9, C
4
5
74.1, CH
88.0, CH
50.5, CH
21.8, CH2
37.2, CH2
50.5, C
150.1, CH
82.4, CH
48.4, CH
25.2, CH2
66.2, CH
65.1, C
150.4, CH
30.8, CH
36.3, CH
23.3, CH2
39.2, CH2
134.4, C
46.9, CH2
128.7, C
75.4, CH
89.8, CH
51.4, CH
23.0, CH2
36.7, CH2
49.8, C
6
7
131.7, CH
25.2, CH2
39.2, CH2
135.6, C
128.5, CH
24.8, CH2
38.4, CH2
134.0, C
130.7, CH
25.0, CH2
39.6, CH2
135.1, C
89.4, CH
29.4, CH2
39.3, CH2
135.8, C
8
9
10
11
12
13
14
15
145.2, C
145.5, C
27.1, C
124.7, CH
24.1, CH2
25.4, CH2
18.3, CH3
14.9, CH3
125.1, CH
25.0, CH2
25.1, CH2
18.6, CH3
15.4, CH3
−OAc:
125.1, CH
25.8, CH2
31.7, CH2
15.9, CH3
14.9, CH3
21.2, CH3
170.1, C
126.5, CH
27.4, CH2
23.0, CH2
14.3, CH3
15.2, CH3
148.0, C
21.9, CH3
112.7, CH2
17.8, CH3
177.3, C
113.1, CH2
21.5, CH3
11.2, CH3
173.2, C
175.6, C
22.1, CH3
112. 2, CH2
15.6, CH3
182.4, C
10.0, CH3
193.9, CH
17.2, CH3
52.3, CH3
−OCH3:
−OCH2CH3:
−OCH2CH3:
64.6, CH2
15.7, CH3
a
b
In CDCl3. In CD3OD.
spectra of 2 showed signals for an α,β-unsaturated-γ-lactone [δC
173.2, 150.1, 134.2, and 82.4], an isopropenyl group [δC 145.5,
113.1, and 21.5; δH 4.90 (1H, br s), 4.99 (1H, br s), and 1.85
(3H, s)], a tertiary methyl [δH 1.23 (3H, s)], and three
oxygenated carbons [δC 78.3, 66.2, and 65.1]. These data
showed high similarity to those of versicolactone B (9)6 except
for the presence of two additional oxygenated carbon signals
[δC 65.1 (s) and 66.2 (d)] in 2 instead of signals for Δ9 in 9,
indicating that 2 was a 9,10-epoxy derivative of 9. HMBC
correlations from CH3-14 to these two oxygenated carbons (C-
9 and C-10) and from H-8 to C-10 confirmed the location of
the epoxy ring. The relative configuration of 2 was determined
by a NOESY experiment. NOESY correlations of H-5/H-6, H-
7, and H-9 and H-1/H-9 indicated that these protons were
cofacial and designated as β-oriented, while the correlations of
CH3-14/H-3α and H-8α assigned CH3-14 to be α-oriented
(S70, Supporting Information). The chemical transformation of
9 to 2 by 3-chloroperbenzoic acid oxidation confirmed the
structure of 2. As the absolute configuration at C-1 in 9 was
assigned as S by the modified Mosher’s method in our current
study (Figure 1), the absolute configuration of 2 was thus
determined as depicted. Compound 2 was given the trivial
name aristoyunnolin B.
for the absence of a tertiary methyl in 12 and the presence of a
methoxycarbonyl group [δH 3.70 (3H, s); δC 175.6 and 52.3] in
3. HMBC correlations from a tertiary methyl [δH 1.41 (3H, s),
CH3-13] to the carbonyl (δC 175.6, C-12) and a quaternary
carbon (δC 27.1, C-11) and from an OCH3 [δH 3.70 (3H, s)] to
C-12 indicated that one geminal methyl of C-11 in 12 was
replaced by a methoxycarbonyl group in 3. The strong NOESY
correlations of H-1/H-9, H3-15/H-2, H-14/H-5, and H-3/H-6
indicated that both Δ1,10 and Δ4 had an E configuration. The
9.8 Hz coupling constant and NOESY correlations between H-
6 and H-7 indicated the cis configuration of the cyclopropane
moiety.8 The CH3-13 was β-oriented on the basis of NOESY
correlations of CH3-13/H-5 and H-8β (S70, Supporting
Information). The absolute configurations at C-6 and C-7 in
3 were assigned to be the same as those of 12 (6S and 7R), as
the CD curve of 3 [λmax (Δε) 202 (+0.86), 247 (+1.72), 330
(+0.11) nm] matched well with that of 12 [λmax (Δε) 203
(+1.06), 256 (+2.88), 327 (+0.44) nm] (S69, Supporting
Information). In a similar way, the absolute configuration of the
known analogue madolin K (14) [λmax (Δε) 203 (+0.25), 254
(+1.99), 327 (+0.20) nm] was determined in the current study
as depicted. Compound 3 was named aristoyunnolin C.
Compound 4 was isolated as a white powder with the
molecular formula C15H20O3, as determined by HREIMS. IR
absorption bands at 3396 and 1744 cm−1 indicated the
Compound 3 had a molecular formula of C16H22O3, as
1
established by HRESIMS. The H and 13C NMR spectra of 3
were similar to those of (+)-isobicyclogermacrenal (12)7 except
presence of OH and carbonyl groups. The H NMR of 4
1
displayed signals for two olefinic methyl singlets [δH 1.63 (s,
H3-14) and 1.43 (s, H3-15)] and three olefinic protons [δH 6.62
(s, H-3), 4.85 (dd, J = 6.4, 6.4 Hz, H-7), and 4.70 (dd, J = 6.8,
6.8 Hz, H-11)]. The 15 carbon resonances were classified by
DEPT experiments as two methyls, one lactone, three
trisubstituted double bonds, five sp3 methylenes, and one sp3
quaternary carbon. This information was very similar to those
of manshurolide (16)9 except for the absence of signals for C-4
(δH 5.10, δC 80.1) in 16 and the presence of a highly
oxygenated quaternary carbon (δC 106.3, C-4) in 4. As the
molecular formula of 4 showed one more O atom than that of
16, compound 4 was proposed to be a 4-hydroxylated
derivative of 16. This was supported by HMBC correlations
Figure 1. ΔδH values of (S)- and (R)-MTPA esters of 5, 9, and 17,
respectively.
666
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671

Journal of Natural Products
Article
from H-5 and H-3 to C-4. Configurations of the Δ6 and Δ10
double bonds in 4 were both assigned as E by NOESY
experiments and comparison of the NMR data with those of
16. As compound 4 showed no optical activity and gave a flat
line in its CD spectrum, the only chiral center (C-4) was
racemic. Enantiomeric separation of 4 on a chiral column failed,
as the hemiketal nature of 4 resulted in a racemic equilibrium
under HPLC conditions. Thus, compound 4 was 4-hydrox-
ymanshurolide and was given the trivial name aristoyunnolin D.
Compound 5 had the molecular formula C15H22O2 (by
HREIMS). The UV, IR, and NMR data of 5 were similar to
those of madolin L8 except for the absence of the methoxy
group and the upfield-shifted carbon signal of C-4 (δC 68.3 in 5
and δC 77.0 in madolin L), implying that 5 was a 4-O-demethyl
derivative of madolin L. This was supported by the molecular
formula and confirmed by detailed 2D NMR analyses of 5, in
which H-4 (δH 4.53) showed correlations with H-3 and H-5 in
compounds 3, 6, 9, and 14 impaired the viability of HeLa cells
to varying degrees at 50 μM (Figure 2). The remaining
Figure 2. Viability of HeLa cells treated with compounds 1−19. The
viability of HeLa cells was inhibited by 3, 6, 9, and 14, whereas it was
promoted by 4. HeLa cells were treated with 50 μM of each
compound for 48 h. Cell viability was detected by a WST-8-based
colorimetric assay (n = 3, *p < 0.05, **p < 0.01, ANOVA).
1
the H−1H COSY spectrum. The configurations of the double
bonds at Δ6, Δ10, and Δ2 were assigned to be the same as those
of madolin L by NOESY experiment and by comparison of
their NMR data. The absolute configuration of C-4 in 5 was
assigned as S by the modified Mosher’s method (Figure 1).
Compound 5 was given the trivial name aristoyunnolin E.
compounds were examined for inhibitory activity on the
phosphorylation of three key MAPKs (ERK1/2, JNK, and p38
MAPK) by immunoblotting. As shown in Figure 3, cells treated
with 1, 4, 10−13, 15, 16, 18, and 19 at a dose of 20 μM
induced a significant decrease in phosphorylation of ERK1/2,
but not of p38 and JNK, suggesting a selective inhibitory
activity of these compounds on ERK1/2 activation. To evaluate
the potency of the compounds, PD98059, a well-known
selective inhibitor of the ERK1/2 signaling pathway, was used
as the positive control for confirmation testing. Compounds 16
and 19 showed stronger activity than PD98059, while 1, 4, 10−
13, and 18 exhibited activity comparable to the positive control
(Figure 4). Further investigation revealed that compounds 16
and 19 inhibited the phosphorylation of ERK1/2 in a dose-
dependent manner ranging from 2.5 to 40 μM, being more
active than the positive control at concentrations below 10 μM
(Figure 5).
The MAPK signaling pathways have drawn considerable
attention, especially the ERK pathway, which is becoming a
promising molecular target for anticancer drug develop-
ment.1−3 The present findings represent examples of natural
products as selective ERK1/2 inhibitors. However, the
mechanism of inhibition on the ERK1/2 phosphorylation and
the potential usage of these compounds in anticancer drug
development require further investigation.
1
Compound 6 had the molecular formula C17H24O3. The H
and 13C NMR spectra of 6 resembled those of madolin M,8
with the only difference being due to the presence of a C-4
acetoxyl in 6 [δH 2.04 (3H, s); δC 21.2 and 170.1] instead of a
C-4 OCH3 in madolin M. The C-4 acetoxyl in 6 was confirmed
by the HMBC correlation from H-4 (δH 6.20, ddd, J = 10.8,
10.8, 4.3 Hz) to the carbonyl carbon (δC 170.1). The
configurations of the double bonds at Δ6, Δ10, and Δ2 were
determined to be the same as those of madolin M by NOESY
experiment, and comparison of the 13C NMR data with those
of madolin M, especially the NOESY correlations of H-4/H-1
and H-3/H-13, assigned the Z configuration of Δ2. The
absolute configuration of the only chiral center (C-4) in 6 was
proposed to be R based on its positive optical rotation ([α]20
D
+ 113), which was opposite that of 5. Compound 6 was given
the trivial name aristoyunnolin F.
Compound 7, obtained as a colorless oil, had the molecular
formula C17H26O2, as determined by HRESIMS. The UV, IR,
and NMR spectra of 7 were very similar to those of madolin W
(17)10 except for the presence of an ethoxy group and the
downfield-shifted C-7 carbon signal at δC 89.4 in 7 (δC 82.4 in
17), indicating that 7 was a 7-O-ethyl derivative of 17.
Chemical transformation of 17 to 7 using the Williamson
reaction confirmed the structure of 7. Compound 7 was
considered to be an artificial product, which could not be
detected in the crude extract. Moreover, the absolute
configuration of madolin W (17) was first determined by the
modified Mosher’s method in the current research (Figure 1).
The known compounds versicolactone C (8),6 versicolac-
tone B (9),6 madolin U (10),10 aristolactone (11),8
(+)-isobicyclogermacrenal (12),7 volvalerenal D (13),11
madolin K (14),8 madolin T (15),12 manshurolide (16),9
madolin W (17),10 madolin A (18),13 and 1(10)-aristolen-2-
one (19)14 were identified by comparison of their NMR data
with those in the literature.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured on a Perkin-Elmer 341 polarimeter, and CD spectra were
obtained on an Applied Photophysics Chirascan spectrometer. UV
spectra were recorded on a Shimadzu UV-2450 spectrophotometer. IR
spectra were determined on a Bruker Tensor 37 infrared
spectrophotometer with KBr disks. NMR spectra were measured on
a Bruker AM-400 spectrometer at 25 °C. EIMS and HREIMS (70 eV)
were recorded on a Finnigan MAT 95 mass spectrometer, and ESIMS
was carried out on a Finnigan LC QDECA instrument. A Shimadzu LC-
20 AT equipped with an SPD-M20A PDA detector was used for
HPLC, and a YMC-pack ODS-A column (250 × 10 mm, S-5 μm, 12
nm) was used for semipreparative HPLC separation. A chiral column
(Phenomenex Lux, cellulose-2, 250 × 4.6 mm, 5 μm) was used for
chiral analysis. Silica gel (300−400 mesh, Qingdao Haiyang Chemical
Co., Ltd.), C18 reversed-phase silica gel (12 nm, S-50 μm, YMC Co.,
Ltd.), Sephadex LH-20 gel (Amersham Biosciences), and MCI gel
(CHP20P, 75−150 μm, Mitsubishi Chemical Industries Ltd.) were
Compounds 1−19 were screened in a bioassay system
designed to evaluate the effect on MAPK signaling pathways.
To exclude cytotoxic compounds, which may also cause
variations of this pathway, compounds 1−19 were first tested
for cytotoxicity against HeLa cells. The data showed that
667
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671

Journal of Natural Products
Article
Figure 3. Effects of nontoxic compounds on the phosphorylation of MAPKs. HeLa cells were treated with 20 μM single compounds for 3 h. Proteins
(20 μg) extracted from total lysates of HeLa cells were subjected to SDS-PAGE followed by Western blot with indicated antibodies. The blots were
then quantified by densitometry (n = 3, **p < 0.001, ANOVA). Compound 7 was not tested in this assay due to the limited amount and the artificial
product nature.
Figure 4. Inhibitory potency of active compounds on the phosphorylation of ERK1/2. PD98059 was used as a positive control. HeLa cells were
treated with 20 μM single compounds for 3 h. Proteins (20 μg) extracted from total lysates of HeLa cells were subjected to SDS-PAGE followed by
Western blot with indicated antibodies. The blots were then quantified by densitometry (n = 3, *p < 0.01 vs DMSO, **p < 0.001 vs DMSO, #p <
0.01 vs PD98059, ##p < 0.001 vs PD98059, ANOVA).
used for column chromatography (CC). All solvents used were of
analytical grade (Guangzhou Chemical Reagents Company, Ltd.).
WST-8-based colorimetric assay used a commercial kit (Cell Counting
Kit-8, CCK-8) from Beyotime Institute of Biotechnology. Antibodies
against p-ERK, p-JNK, p-p38, total-ERK, and α-tublin were from Cell
Signaling Technology. Anti-mouse and anti-rabbit second antibodies
were from Promega, and PD98059 was from Sigma. The HeLa cell
line was obtained from the China Center for Type Culture Collection
of Chinese Academy of Sciences.
Plant Material. Stems of Aristolochia yunnanensis Franch.
(synonym: Aristolochia griffithii) were collected in October 2010
from Yunnan Province, P. R. China, and were identified by Prof. You-
Kai Xu of Xishuangbanna Tropical Botanical Garden, Chinese
Academy of Sciences. A voucher specimen (accession number:
668
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671

Journal of Natural Products
Article
Figure 5. Compounds 16 and 19 inhibited the phosphorylation of ERK1/2 in a dose-dependent manner. HeLa cells were treated with indicated
concentrations of single compounds for 3 h. Proteins (20 μg) extracted from total lysates of HeLa cells were subjected to SDS-PAGE followed by
Western blot with indicated antibodies. The blots were then quantified by densitometry (n = 3, ANOVA). PD98059 was used as the positive control.
NMX201010) has been deposited at the School of Pharmaceutical
Sciences, Sun Yat-sen University.
Na]+; negative ESIMS m/z 263.1 [M − H]−; HRESIMS m/z 287.1249
[M + Na]+ (calcd for C15H20O4Na, 287.1254).
Aristoyunnolin C (3): white powder; [α]20 +166 (c 0.09, CHCl3);
Extraction and Isolation. The air-dried powder of the stems of A.
yunnanensis (4.5 kg) was extracted with 95% EtOH (3 × 10 L) at room
temperature (rt) to give 214 g of crude extract. The extract was
suspended in H2O (1 L) and successively partitioned with petroleum
ether (PE, 3 × 1 L), EtOAc (3 × 1 L), and n-BuOH (3 × 1 L). The
EtOAc extract (120 g) was subjected to MCI gel CC eluted with a
MeOH/H2O gradient (3:7 → 10:0) to afford five fractions (I−V).
Fraction I (10.5 g) was chromatographed over C18 reversed-phase
(RP-18) silica gel eluted with MeOH/H2O (5:5 → 10:0) to afford five
fractions (Ia−Ie). Fraction Ie (1.6 g) was separated by silica gel CC
(PE/acetone, 3:1), followed by Sephadex LH-20 CC using EtOH to
give 8 (22 mg). Fraction II (12 g) was subjected to silica gel CC (PE/
EtOAc, 5:1 → 1:2) to give four fractions (IIa−IId). Fraction IIb (3.4
g) was purified on silica gel CC (PE/EtOAc, 5:1) to obtain 9 (155
mg) and 10 (12 mg). Fraction IIc (0.9 g) was applied to silica gel CC
(CH2Cl2/acetone, 25:1 → 5:1) to yield 2 (46 mg). Fraction III (28 g)
was subjected to silica gel CC (PE/EtOAc, 5:1 → 1:1) to give 10
fractions (IIIa−IIIj). Fraction IIIc (1.9 g) was subjected to RP-18 silica
gel CC (MeOH/H2O, 6:4 → 10:0), followed by silica gel CC (PE/
acetone, 20:1 → 1:1), to afford 4 (14 mg), 5 (106 mg), and 13 (5 mg).
Fraction IIIf (230 mg) was chromatographed over silica gel (CHCl3/
MeOH, 200:1) to yield 1 (13 mg). Fraction IV (35 g) was
chromatographed over an MCI gel column eluted with a gradient of
MeOH/H2O (v/v from 6:4 to 10:0) to give eight fractions (IVa−IVh).
Fraction IVa (2.3 g) was separated by RP-18 CC using a gradient of
MeOH/H2O (v/v from 6:4 to 10:0) to yield 17 (104 mg) and 14 (48
mg). Fraction IVb (8.6 g) was subjected successively to silica gel CC
(PE/EtOAc, 50:1 → 5:1), RP-18 silica gel CC (MeOH/H2O, 7:3 →
10:0), and Sephadex LH-20 eluted with EtOH to yield 18 (4 mg), 11
(1.2 g), 19 (65 mg), and 16 (32 mg). Fraction IVc (2.7 g) was applied
to silica gel CC (PE/EtOAc, 40:1 → 1:1) to give three fractions
(IVc1−IVc3). Fraction IVc1 (840 mg) was separated by silica gel CC
(PE/EtOAc, 50:1 → 30:1) to give 3 (8 mg) and 15 (46 mg). Further
purification of IVc2 (1.3 g) by silica gel CC (CHCl3/PE, 3:1) afforded
7 (2 mg), 12 (150 mg), and 6 (4 mg). The purity of compounds 1−19
D
UV (MeOH) λmax (log ε) 250 (4.03) nm; CD (c 5 × 10−4 M,
CH3CN), λmax (Δε) 202 (+0.86), 247 (+1.72), 330 (+0.11); IR (KBr)
1
νmax 2927, 1717, 1677, 1626, 1213, 1137, and 1112 cm−1; H and 13
C
NMR data, see Tables 1 and 2; positive ESIMS m/z 263.2 [M + H]+;
HRESIMS m/z 285.1453 [M + Na]+ (calcd for C16H22O3Na,
285.1461).
Aristoyunnolin D (4): white powder; [α]20D 0 (c 0.12, CHCl3); UV
(MeOH) λmax (log ε) 205 (2.88) nm; IR (KBr) νmax 3396, 2917, 2851,
1744, 1435, 1130, 977 cm−1; 1H and 13C NMR data, see Tables 1 and
2; EIMS m/z 248 [M]+ (28), 230 (69), 205 (100), 175 (56); HREIMS
m/z 248.1410 [M]+ (calcd for C15H20O3, 248.1407).
Aristoyunnolin E (5): colorless oil; [α]20 −117 (c 0.34, CHCl3);
D
UV (MeOH) λmax (log ε) 225 (3.95) nm; IR (KBr) νmax 3428, 2920,
2855, 1688, 1441, 1094 cm−1; 1H and 13C NMR data, see Tables 1 and
2; EIMS m/z 234 [M]+ (29), 216 (57), 191 (100), 187 (68), 173
(68); HREIMS m/z 234.1612 [M]+ (calcd for C15H22O2, 234.1614).
Aristoyunnolin F (6): white powder; [α]20 +113 (c 0.04, CHCl3);
D
UV (MeOH) λmax (log ε) 227 (3.96) nm; IR (KBr) νmax 2921, 2854,
1741, 1681, 1439, 1370, 1232, 1019, 961 cm−1; 1H and 13C NMR data,
see Tables 1 and 2; positive ESIMS m/z 299.2 [M + Na]+; HRESIMS
m/z 299.1606 [M + Na]+ (calcd for C17H24O3Na, 299.1618).
7-O-ethylmadolin W (7): colorless oil; [α]20 −187 (c 0.07,
D
CHCl3); UV (MeOH) λmax (log ε) 254 (3.90) nm; IR (KBr) νmax
3445, 2926, 2855, 1682, 1628, 1125 cm−1; 1H and 13C NMR data, see
Tables 1 and 2; positive ESIMS m/z 263.2 [M + H]+; HRESIMS m/z
285.1832 [M + Na]+ (calcd for C17H26O2Na, 285.1825).
Preparation of (R)- and (S)-MTPA Esters of 5, 9, and 17.
Sesquiterpenoid (5−6 mg, 0.02 mmol) was dissolved in 1 mL of
pyridine and stirred at rt for 10 min. An excess of (R)- or (S)-MTPA-
Cl (6−7 mg, 0.025 mmol) was added, and the reaction was left to stir
overnight at rt. The reaction mixture was evaporated in vacuo to
provide the reaction crude, which was purified by C18 RP-HPLC using
a gradient of MeOH/H2O (80:20 → 100:0, 20 min, 3.0 mL/min, tR of
most derivatives was ca. 19 min).
1
was estimated to be greater than 95%, as determined by H NMR
1
(S)-MTPA ester of 5 (5a): H NMR (CDCl3, 400 MHz) δH 9.46
spectra (Supporting Information).
Aristoyunnolin A (1): white powder; [α]20D −50.0 (c 0.08, CHCl3);
IR (KBr) νmax 3309, 2926, 2873, 1753, 1547, 1461, 1368, 1207, 1187,
(1H, s, H-1), 6.28 (1H, d, J = 10.5 Hz, H-3), 5.77 (1H, td, J = 10.5, 3.9
Hz, H-4), 5.07 (1H, t, J = 7.5 Hz, H-7), 4.82 (1H, d, J = 7.1 Hz, H-11),
2.58 (1H, m, H-5α), 2.33 (2H, m, H-12), 2.30 (1H, m, H-5β), 2.19
(2H, m, H-13), 2.13 (2H, m, H-8), 2.05 (2H, m, H-9), 1.63 (3H, s, H-
14), 1.37 (3H, s, H-15); 13C NMR (CDCl3, 100 MHz) δC 15.2 (CH3,
C-14), 18.3 (CH3, C-15), 24.7 (CH2, C-8), 24.9 (CH2, C-13), 25.1
(CH2, C-12), 38.2 (CH2, C-9), 43.0 (CH2, C-5), 71.8 (CH, C-4),
124.9 (CH, C-11), 128.2 (C, C-6), 129.8 (CH, C-7), 133.8 (C, C-10),
146.0 (C, C-2), 147.1 (CH, C-3), 194.9 (CH, C-1); ESIMS m/z 473.1
[M + Na]+; ESIMS m/z 449.0 [M − H]+.
1
958, and 899 cm−1; H and 13C NMR data, see Tables 1 and 2;
positive ESIMS m/z 296.1 [M + H]+; negative ESIMS m/z 341.2 [M
+ NO2]−; HREIMS m/z 295.1416 [M]+ (calcd for C15H21NO5,
295.1414).
Aristoyunnolin B (2): colorless oil; [α]20D +76 (c 0.02, CHCl3); UV
(MeOH) λmax (log ε) 219 (3.92) nm; IR (KBr) νmax 3474, 2963, 2872,
1
1749, 1446, 1076, 1033, 907, and 758 cm−1; H and 13C NMR data,
see Tables 1 and 2; positive ESIMS m/z 265.2 [M + H]+, 287.1 [M +
669
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671

Journal of Natural Products
Article
1
(R)-MTPA ester of 5 (5b): H NMR (CDCl3, 400 MHz) δH 9.38
Cell Viability Assay. The cells were cultured in 96-well culture
plates. Cell viability was analyzed with the CCK-8. A 100 μL amount
of phenol red-free DMEM containing 10% WST-8 was added to each
well. The optical density of each well at 450 nm was measured after 1
h of incubation at 37 °C. The effect of compounds on cell viability was
calculated from the mean values of three wells.
Total Protein Extraction. Cells were exposed to tested
compounds at the concentration of 20 μM for 3 h. Total protein
was extracted in lysis buffer (50 mM Tris (pH 7.4), 150 mM NaCl, 1%
NP-40, 0.5% sodium deoxycholate, 0.1% SDS) with protease and
phosphatase inhibitors. The protein contents of the cell lysis were
measured by the bicinchoninic acid (BCA) method.
(1H, s, H-1), 6.13 (1H, d, J = 10.0 Hz, H-3), 5.76 (1H, ddd, J = 11.2,
10.0, 4.2 Hz, H-4), 5.08 (1H, t, J = 7.5 Hz, H-7), 4.81 (1H, d, J = 7.2
Hz, H-11), 2.69 (1H, m, H-5α), 2.41 (1H, m, H-5β), 2.35 (2H, m, H-
12), 2.20 (2H, m, H-13), 2.14 (2H, m, H-8), 2.05 (2H, m, H-9), 1.63
(3H, s, H-14), 1.37 (3H, s, H-15); 13C NMR (CDCl3, 100 MHz) δC
15.3 (CH3, C-14), 18.4 (CH3, C-15), 24.6 (CH2, C-8), 24.8 (CH2, C-
13), 25.1 (CH2, C-12), 38.3 (CH2, C-9), 43.1 (CH2, C-5), 72.0 (CH,
C-4), 124.9 (CH, C-11), 128.2 (C, C-6), 129.8 (CH, C-7), 133.7 (C,
C-10), 146.3 (C, C-2),147.0 (CH, C-3), 194.8 (CH, C-1); ESIMS m/z
449.0 [M − H]+.
1
(S)-MTPA ester of 9 (9a): H NMR (CDCl3, 400 MHz) δH 6.80
Western Blotting Assay. Proteins (20 μg) from cell lysates were
denatured and loaded into 8−10% acrylamide gels and separated by
sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-
PAGE). The separated proteins were then electrotransferred to
nitrocellulose membranes. Nonspecific binding sites on the mem-
branes were blocked by 5% skimmed milk powder in TBST for 1 h at
room temperature. Blocked membranes were incubated by specific
primary antibodies overnight at 4 °C. After rinsing with TBST three
times, secondary antibodies were chosen to bind on the membranes
for 1 h at room temperature. Protein bands were detected by
chemiluminescence and exposed to GE ImageQuant LAS 4000 Mini.
(1H, t, J = 1.3 Hz, H-5), 5.50 (1H, dd, J = 11.8, 2.4 Hz, H-1), 5.10
(1H, d, J = 9.5 Hz, H-9), 5.09 (1H, d, J = 1.3 Hz, H-6), 5.01 (1H, s, H-
12b), 4.91 (1H, t, J = 1.4 Hz, H-12a), 2.73 (1H, m, H-8α), 2.69 (2H,
m, H-3), 2.54 (1H, dd, J = 11.8, 4.4 Hz, H-7), 2.34 (1H, m, H-2α),
2.21 (1H, m, H-8β), 1.89 (3H, s, H-13), 1.62 (1H, m, H-2β), 1.58
(3H, t, J = 1.6 Hz, H-14); 13C NMR (CDCl3, 100 MHz) δC 10.8
(CH3, C-14), 21.2 (CH3, C-13), 22.5 (CH2, C-2), 23.6 (CH2, C-3),
28.0 (CH2, C-8), 45.6 (CH, C-7), 82.6 (CH, C-1), 83.4 (CH, C-6),
112.6 (CH2, C-12), 131.5 (C, C-4), 131.5 (C, C-10),134.5 (CH, C-9),
146.4 (C, C-11), 150.0 (CH, C-5), 172.8 (C, C-15); ESIMS m/z 487.1
[M + Na]+; ESIMS m/z 463.1 [M − H]+.
1
(R)-MTPA ester of 9 (9b): H NMR (CDCl3, 400 MHz) δH 6.80
(1H, s, H-5), 5.49 (1H, dd, J = 11.9, 1.5 Hz, H-1), 5.09 (1H, d, J = 1.3
Hz, H-6), 5.08 (1H, d, J = 9.7 Hz, H-9), 5.01 (1H, s, H-12b), 4.91
(1H, s, H-12a), 2.71 (1H, m, H-3), 2.70 (1H, m, H-8α), 2.54 (1H, dd,
J = 11.8, 4.2 Hz, H-7), 2.42 (1H, m, H-2α), 2.20 (1H, m, H-8β), 1.88
(3H, s, H-13), 1.71 (1H, m, H-2β), 1.47 (3H, s, H-14); 13C NMR
(CDCl3, 100 MHz) δC 10.5 (CH3, C-14), 21.1 (CH3, C-13), 22.7
(CH2, C-2), 23.7 (CH2, C-3), 28.0 (CH2, C-8), 45.6 (CH, C-7), 82.7
(CH, C-1), 83.4 (CH, C-6), 112.5 (CH2, C-12), 131.4 (C, C-4), 131.4
(C, C-10), 134.5 (CH, C-9), 146.5 (C, C-11), 150.0 (CH, C-5), 172.8
(C, C-15); ESIMS m/z 487.1 [M + Na]+.
ASSOCIATED CONTENT
■
S
* Supporting Information
1D and 2D NMR spectra of 1−7, 1H and 13C NMR spectra of
1
known compounds (8−19), and H NMR spectra of Mosher’s
esters of 5, 9, and 17 are available free of charge via the Internet
at http://pubs.acs.org.
AUTHOR INFORMATION
■
(S)-MTPA ester of 17 (17a): 1H NMR (CDCl3, 400 MHz) δH 9.32
(1H, s, H-1), 6.05 (1H, d, J = 11.4 Hz, H-3), 5.31 (1H, dd, J = 9.2, 5.8
Hz, H-11), 4.39 (1H, d, J = 8.2 Hz, H-7), 2.82 (1H, m, H-13a), 2.24
(2H, m, H-9), 2.22 (1H, m, H-8a), 2.18 (1H, m, H-4), 2.09 (1H, m,
H-13b), 2.08 (2H, m, H-12), 1.57 (1H, m, H-8b), 1.47 (3H, s, H-15),
1.33 (1H, dd, J = 8.7, 5.1 Hz, H-5α), 1.24 (3H, s, H-14), 0.85 (1H, t, J
= 5.1 Hz, H-5β); 13C NMR (CDCl3, 100 MHz) δC 14.5 (CH3, C-14),
14.9 (CH3, C-15), 22.9 (CH2, C-13), 25.4 (CH2, C-5), 27.3 (CH2, C-
12), 27.6 (CH, C-4), 28.3 (C, C-6), 28.9 (CH2, C-8), 38.3 (CH2, C-
9), 86.1 (CH, C-7), 127.4 (CH, C-11), 134.5 (C, C-10), 143.4 (C, C-
2), 155.2 (CH, C-3), 193.6 (CH, C-1); ESIMS m/z 451 [M + H]+.
(R)-MTPA ester of 17 (17b): 1H NMR (CDCl3, 400 MHz) δH 9.31
(1H, s, H-1), 6.02 (1H, d, J = 11.3 Hz, H-3), 5.32 (1H, dd, J = 9.5, 5.3
Hz, H-11), 4.39 (1H, d, J = 10.1 Hz, H-7), 2.82 (1H, m, H-13a), 2.28
(1H, m, H-8a), 2.27 (2H, m, H-9), 2.16 (1H, m, H-4), 2.09 (1H, m,
H-13b), 2.07 (2H, m, H-12), 1.69 (1H, m, H-8b), 1.47 (3H, s, H-15),
1.29 (1H, dd, J = 8.7, 5.2 Hz, H-5α), 1.16 (3H, s, H-14), 0.76 (1H, t, J
= 5.2 Hz, H-5β); 13C NMR (CDCl3, 100 MHz) δC 14.2 (CH3, C-14),
14.8 (CH3, C-15), 22.8 (CH2, C-13), 25.1 (CH2, C-5), 27.3 (CH, C-
4), 27.3 (CH2, C-12), 28.0 (C, C-6), 28.9 (CH2, C-8), 38.2 (CH2, C-
9), 86.1 (CH, C-7), 127.3 (CH, C-11), 134.3 (C, C-10), 143.3 (C, C-
2), 155.4 (CH, C-3), 193.5 (CH, C-1); ESIMS m/z 451 [M + H]+.
Chemical Transformation of 9 to 2. To a stirred solution of 9 (5
mg, 0.02 mmol) in CH2Cl2 (1 mL) at 0 °C was added 3-
chloroperbenzoic acid (10 mg). The mixture was stirred at rt for 2
h and then evaporated. The resulting crude product was subjected to
Corresponding Author
*Tel: +86-20-39943090. Fax: +86-20-39943090. E-mail:
yinsh2@mail.sysu.edu.cn.
Author Contributions
†Z.-B. Cheng and W.-W. Shao contributed equally to this paper.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank the National Natural Science Foundation of
China (No. 81102339) and Guangdong Natural Science
Foundation (No. S2011040002429) for providing financial
support to this work.
REFERENCES
■
(1) Johnson, G. L.; Lapadat, R. Science 2002, 298, 1911−1912.
(2) Raman, M.; Chen, W.; Cobb, M. H. Oncogene 2007, 26, 3100−
3112.
(3) English, J. M.; Cobb, M. H. Trends Pharmacol. Sci. 2002, 23, 40−
45.
(4) Song, L.R. Chinese Materia Medica (Zhonghua BenCao); Shanghai
Science and Technology Press: Shanghai, 1999; Vol. 8, pp 486−487.
(5) (a) Ming, C. W.; Mayer, R.; Zimmermann, H.; Rucker, G.
Phytochemistry 1989, 28, 3233−3234. (b) Wang, C. Y.; Tsai, A. C.;
Peng, C. Y.; Chang, Y. L.; Lee, K. H.; Teng, C. M.; Pan, S. L. PLoS One
2012, 7, 1−12. (c) Choi, Y. K.; Seo, H. S.; Choi, H. S.; Choi, H. S.;
Kim, S. R.; Shin, Y. C.; Ko, S. G. Mol. Cell. Biochem. 2012, 363, 119−
128.
̈
silica gel CC (CHCl3/MeOH, 40:1) to afford 2 (4 mg), which was
1
identified by the H NMR and [α]20
.
Chemical Transformation of 17Dto 7. To a stirred solution of 17
(2.55 mg, 0.01 mmol) in DMF (1 mL) was added 2 mg of NaH. The
mixture was stirred at rt for 0.5 h, and 3 μL of bromoethane (0.04
mmol) was added. The reaction was left to stir at rt for 2 h and
quenched by adding 1 mL of H2O. After workup, the crude product
was purified by C18 RP-HPLC using a gradient of MeOH/H2O (90:10
→ 100:0, 20 min, 3.0 mL/min, tR 15 min) to afford 7 (1.5 mg), which
was identified by NMR and MS data.
(6) Zhang, J.; He, L. X. Acta Pharm. Sin. 1986, 21, 273−278.
(7) Rucker, G.; Mayer, R.; Wiedenfeld, H.; Chung, B. S.; Gullmann,
̈
̈
A. Phytochemistry 1987, 26, 1529−1530.
(8) Wu, T. S.; Chan, Y. Y.; Leu, Y. L. Chem. Pharm. Bull. 2000, 48,
357−361.
670
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671

Journal of Natural Products
Article
(9) Rucker, G.; Ming, C. W.; Mayer, R.; Will, G.; Guellmann, A.
̈
̈
Phytochemistry 1990, 29, 983−985.
(10) Wu, T. S.; Chan, Y. Y.; Leu, Y. L. J. Nat. Prod. 2001, 64, 71−74.
(11) Wang, P. C.; Ran, X. H.; Chen, R.; Luo, H. R.; Liu, Y. Q.; Zhou,
J.; Zhao, Y. X. J. Nat. Prod. 2010, 73, 1563−1567.
(12) Wu, T. S.; Chan, Y. Y.; Leu, Y. L. J. Chin. Chem. Soc. 2000, 47,
957−960.
(13) Wu, T. S.; Chan, Y. Y.; Leu, Y. L. J. Nat. Prod. 1998, 61, 511−
514.
(14) De Pascual Teresa, J.; Urones, J. G.; Fernandez, A.
Phytochemistry 1983, 22, 2753−2754.
671
dx.doi.org/10.1021/np300887d | J. Nat. Prod. 2013, 76, 664−671