Identification of key structural characteristics of schisandra chinensis lignans involved in P-glycoprotein inhibition

Article abstract of DOI:10.1021/np500521v

The aim of the present study was to determine the structural requirements for dibenzocyclooctadiene lignans essential for P-glycoprotein inhibition. Altogether 15 structurally related lignans isolated from Schisandra chinensis or prepared by modification of their backbone were investigated, including three pairs of enantiomers. P-Glycoprotein inhibition was quantified using a doxorubicin accumulation assay in human promyelotic leukemia HL60/MDR cells overexpressing P-glycoprotein. A preliminary quantitative structure-activity relationship analysis revealed three main structural features involved in P-glycoprotein inhibition: a 1,2,3-trimethoxy moiety, a 6-acyloxy group, and the absence of a 7-hydroxy group. The most effective inhibitors, (-)-gomisin N (1) and (+)-deoxyschizandrin [(+)-2], were selected for further evaluation of their effects. Both these lignans restored the cytotoxic effect of doxorubicin in HL60/MDR cells and when combined with a subtoxic concentration of this compound increased the proportion of G2/M cells significantly, which is a usual response to treatment with this anticancer drug.

Full text of DOI:10.1021/np500521v

Article
pubs.acs.org/jnp
Identification of Key Structural Characteristics of Schisandra chinensis
Lignans Involved in P‑Glycoprotein Inhibition
†
‡,∥
Jirí Slanina,†,▽ Gabriela Pachnikova,‡,▽ Martina Carnecka, Ludmila Porubova Koubíkova,
̌
̌
́
́
́
́
́
†
§
⊥
,‡
̌
Lenka Adamkova, Otakar Humpa, Karel Smejkal, and Iva Slaninova*
́
́
́
†Department of Biochemistry, Faculty of Medicine, Masaryk University, Kamenice 5, Building A16, 625 00 Brno, Czech Republic
‡Department of Biology, Faculty of Medicine, Masaryk University, Kamenice 5, Building A6, 625 00 Brno, Czech Republic
§Central European Institute of Technology, Masaryk University, Kamenice 5, Building A4, 625 00 Brno, Czech Republic
⊥Department of Natural Drugs, Faculty of Pharmacy, University of Veterinary and Pharmaceutical Sciences Brno, Palackeh
́
o 1-3, 612
42 Brno, Czech Republic
S
* Supporting Information
ABSTRACT: The aim of the present study was to determine
the structural requirements for dibenzocyclooctadiene lignans
essential for P-glycoprotein inhibition. Altogether 15 structur-
ally related lignans isolated from Schisandra chinensis or
prepared by modification of their backbone were investigated,
including three pairs of enantiomers. P-Glycoprotein inhibition
was quantified using a doxorubicin accumulation assay in
human promyelotic leukemia HL60/MDR cells overexpressing
P-glycoprotein. A preliminary quantitative structure−activity
relationship analysis revealed three main structural features
involved in P-glycoprotein inhibition: a 1,2,3-trimethoxy moiety, a 6-acyloxy group, and the absence of a 7-hydroxy group. The
most effective inhibitors, (−)-gomisin N (1) and (+)-deoxyschizandrin [(+)-2], were selected for further evaluation of their
effects. Both these lignans restored the cytotoxic effect of doxorubicin in HL60/MDR cells and when combined with a subtoxic
concentration of this compound increased the proportion of G2/M cells significantly, which is a usual response to treatment with
this anticancer drug.
lignans, has been identified as inhibitors of P-gp and MRPs
he ability of cancer cells to be cross-resistant to
Tstructurally and functionally unrelated anticancer drugs (multidrug resistance-related proteins).5−9 Such lignans are
found almost exclusively in species of the two plant genera
Schisandra and Kadsura.10 The main source of these lignans are
the fruits of Schisandra chinensis (Turcz.) K. Koch (Schisan-
draceae), which are used widely in traditional Chinese medicine
as a tonic and antitussive agent and in the treatment of lung,
heart, and kidney disorders and used in health food products.
Schisandra extracts are also clinically prescribed in mainland
China for the treatment of hepatitis.11,12
The interaction of several dibenzocyclooctadiene lignans
with P-gp was reported using competitive inhibition of
photoaffinity labeling.13 Wan et al. observed reversion of P-
gp-mediated multidrug resistance by gomisin A through
noncompetitive inhibition of substrate−P-gp association.8
Schisandrin A (also known as deoxyschizandrin) showed the
ability to reverse multidrug resistance in cancer cells by
inhibition of P-gp expression.9 Our previous observations on
lung cancer cells COR-L23/R overexpressing MRP1 protein
confirmed the ability of dibenzocyclooctadiene lignans to
overcome the resistance to doxorubicin.14
is known as multidrug resistance. Overexpression of the ATP
binding cassette (ABC) transporters is responsible for most
cases of cancer multidrug resistance (MDR) and chemotherapy
failure.1,2 The most typical ABC transporter is P-glycoprotein
(P-gp; 170 kDa), coded by the ABCB1 (MDR1) gene, which is
responsible for the transportation out of cells of various
xenobiotics including cytostatic compounds. A promising
method for overcoming the resistance to anticancer drugs is
the coadministration of compounds showing low toxicity and
an ability to inhibit ABC transporters, i.e., the use of MDR
inhibitors. In the last several decades, the search for effective
and clinically applicable MDR inhibitors has occurred, and
three generations of these agents were introduced into clinical
trials. However, this has not led to any clear successes,
especially because of the toxicity of these agents at doses
necessary to block ABC transporters.3
As a result of unfavorable clinical outcome to previous MDR
inhibitors, research has focused also on searching for naturally
occurring MDR inhibitor candidates.3 Promising inhibitory
activity of natural products on P-gp has been shown using plant
phenols, especially curcumoids and flavonoids.4 Recently,
another group of plant phenols, the dibenzocyclooctadiene
Received: June 26, 2014
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
In contrast to verapamil, these lignans reveal antioxidative
properties and show cardioprotective and hepatoprotective
activities, which were demonstrated both on a cellular level and
in mice.15−17 The cardioprotective effect of schisandrin B was
shown using H9c2 cardiomyocytes.15 The experiments on mice
demonstrated a protective effect of this compound against
carbon tetrachloride-induced hepatotoxicity.16,17
The current knowledge of the structural aspects of P-gp
inhibition by dibenzocyclooctadiene lignans is limited due to
the small number of lignans that have been evaluated thus far.
In the present study several such compounds were isolated
from S. chinensis, and three pairs of enantiomers were prepared
to better understand their structure−activity relationships of P-
gp inhibition.
activity) for both resistant HL60/MDR and parental HL60 cells
were defined using a WST assay. The IC50 value of HL60/
MDR cells was approximately 200 times higher than that of
parental HL60 cells (IC50 4.0 vs 0.02 μM).
In order to confirm overexpression of MDR1 and to
investigate expression of two other ABC transporters, MRP1
and MRP2 in both HL60/MDR and parental HL60 cells, RT-
PCR analysis was carried out. β-Actin was used as a control.
The results showed that only the MDR1 gene is expressed in
HL60/MDR cells. Neither MDR1 nor MRP1 and/or MRP2
genes were expressed in parental HL60 cells (Figure 2).
RESULTS AND DISCUSSION
■
Dibenzocyclooctadiene Lignans Investigated. The
following dibenzocyclooctadiene lignans were isolated from
the seeds of S. chinensis: gomisin N (1), (+)-deoxyschizandrin
[(+)-2], tigloylgomisin P (3), ( )-γ-schizandrin [( )-4],
angeloylgomisin H (5), schisantherin C (6), gomisin G (7),
gomisin J (8), wuweizisu C (9), gomisin A (10), and
schizandrin (11), as described in the Experimental Section.
Compounds ( )-4 and 8 were resolved into their enantiomers
by chiral chromatography to give (+)-4, (−)-4, (+)-8, and
(−)-8. The lignan (−)-2 was prepared by methylation of (−)-8.
The structures of the lignans investigated are shown in Figure
1.
Figure 2. RT-PCR analysis of ABC transporter expression: MDR1 (P-
gp), MRP1, MRP2.
The presence of P-gp on the surface of HL60/MDR cells was
demonstrated by flow cytometry analysis using a specific anti P-
gp antibody (Figure 3).
Initial Biological Evaluation. The effect of dibenzocy-
clooctadiene lignans on doxorubicin accumulation and their
ability to increase cytotoxic action of doxorubicin on resistant
HL60/MDR cells were studied. IC50 values of doxorubicin (i.e.,
the concentration that induces a 50% inhibition of metabolic
Figure 3. P-gp on the surface of HL60/MDR cells. (A) Isotype control
[PE/IgG1 (IOTest), Beckman Coulter]; (B) mouse monoclonal anti
P-gp antibody conjugated with PE [anti-P-gp PE (IOTest) CD243,
Beckman Coulter].
The inhibition of P-gp was studied by accumulation of the
fluorescent substrate doxorubicin within resistant HL60/MDR
cells. After 15 min of preincubation of the cells with the test
lignans (25 μM), doxorubicin (10 μM) was added and the cells
were incubated for another 60 min. The intensity of
doxorubicin fluorescence within the cells was measured by
flow cytometry.
Previous studies have revealed that the lignan (+)-2 is a
highly effective P-gp inhibitor.13,18 This finding was also
supported by the present investigation, which indicated a
high accumulation of doxorubicin within HL60/MDR cells
treated with (+)-2. Therefore, (+)-2 was used to determine the
most suitable concentration of the test lignans for the
assessment of their P-gp inhibition potential. The doxorubicin
accumulation was analyzed at various concentrations of (+)-2
as plotted in Figure 4. The doxorubicin accumulation was
increased with rising concentrations of (+)-2 up to 25 μM
almost linearly. Above this concentration, the doxorubicin
accumulation increased at a slower pace. Accordingly, a 25 μM
concentration of the test lignans was selected for the
doxorubicin accumulation experiments.
Figure 1. Structures of compounds [tig = tigloyl (E-2-methyl-2-
butenoyl), bnz = benzoyl, ang = angeloyl (Z-2-methyl-2-butenoyl)].
B
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
7-hydroxy group, the second best descriptor, significantly
improved the correlation (R = 0.674, p = 0.036, Table 2). The
negative impact of a hydroxy group at C-7 was evident from the
partial correlation coefficient for this structural parameter in the
one-parameter model (R = −0.324, Table 2). The three-
parameter models were derived from the best two-parameter
model.
The use of two three-parameter models was found to
improve the relationship determined. They included either the
M-biaryl conformation (R = 0.807, p = 0.012) or 6-acyloxy
group (R = 0.838, p = 0.006) as a third parameter. Since the
latter model described the relationship more tightly, the
optimal three-parameter model involved the following struc-
tural features: a 1,2,3-trimethoxy moiety, a 7-hydroxy group,
and a 6-acyloxy group. The combination of four descriptors
(1,2,3-trimethoxy, 7-hydroxy, 6-acyloxy, and M-biaryl) did not
improve significantly the optimal three-parameter model. The
correlation coefficient was only weakly improved (R = 0.853 vs
0.838), but the four-parameter model was less significant (p =
0.012 vs 0.006), and particularly, this model made two
descriptors insignificant, the 7-hydroxy and 6-acyloxy groups,
with p = 0.14 and 0.37, respectively. Therefore, the final model
is represented by eq 1.
Figure 4. Relationship between concentration of (+)-2 and the ability
to potentiate the accumulation of doxorubicin in HL60/MDR cells.
The accumulation of doxorubicin in combination with
dibenzocyclooctadiene lignans related to the accumulation of
doxorubicin alone is presented in Table 1. The accumulation
was increased significantly on the effects of the lignans with a
high degree of certainty according to the Dunnett test (p <
0.0001), but the effects of ( )-4 and 11 were less conclusive (p
< 0.05). Under the conditions used, some lignans were more
effective than verapamil, namely, 1 (p < 0.001), (−)-2 (p <
0.01), and (+)-2 (p < 0.05).
Although several articles have reported the inhibition of P-gp
transport activity by dibenzocyclooctadiene lignans,5−9 none of
these investigations studied their quantitative structure−activity
relationships (QSAR). To identify the structural characteristics
of lignans essential for potent P-gp inhibition, the results were
analyzed by a stepwise multiple regression. The data given in
Table 1 were used for construction of a QSAR model. The
structural characteristics of interest were the biaryl config-
uration and the 1,2,3-trimethoxy, 2,3-methylenedioxy, 3,12-
dihydroxy, 12,13,14-trimethoxy, 12,13-methylenedioxy, 6-acy-
loxy, and 7-hydroxy groups. First, a one-parameter model for
each descriptor was calculated. The partial correlation
coefficients and p-values for one-parameter models are given
in Table 2. The best correlation was obtained with a 1,2,3-
trimethoxy moiety (R = 0.444, p = 0.112). Subsequently, two-
parameter models were constructed on the base of the best
one-parameter model. For two-parameter models, addition of a
% doxorubicin accumulation
= 181 ( 14) + 74 ( 18) 1,2,3‐tri‐OCH3 − 92 ( 22)
7‐OH + 73 ( 25) 6‐acyloxy
(1)
The relation between the P-gp inhibition measured (Table 1)
and the inhibition calculated according to eq 1 is shown in
Figure 5. The least-squares regression showed a very good
correlation between the calculated and measured data (R =
0.838). The square of the correlation coefficient (R2 = 0.702)
indicated that 70% of the variation may be explained by three
significant variables.
The five most effective dibenzocyclooctadiene lignans found
have two of three important structural characteristics, i.e., the
presence of a 1,2,3-trimethoxy moiety, a 6-acyloxy group, and/
or the absence of a 7-hydroxy group, but none of them met all
Table 1. Percentage of Doxorubicin (10 μM) Accumulation in HL60/MDR Cells after Pretreatment with Lignans (25 μM) and
a
the Correlation Matrix for Multiple Regression
%
M-
biaryl
1,2,3-
trimethoxy
2,3-
3,12-
dihydroxy
12,13,14-
trimethoxy
12,13-
methylenedioxy
6-
acyloxy
7-
compound
accumulation
methylenedioxy
hydroxy
verapamil
239 30
283 33
277 24
264 28
234 14
232 38
152 24
202 12
201 19
199 27
191 27
156 14
187 21
162 13
125 17
1
0
0
1
0
1
0
1
0
0
0
1
0
1
1
1
1
1
1
1
0
1
1
0
0
0
0
1
1
0
0
0
0
0
1
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
1
1
0
0
1
0
0
1
0
0
0
0
1
1
0
0
1
1
0
0
1
0
0
0
1
1
0
0
0
0
1
0
0
0
1
1
0
0
0
0
0
0
0
0
1
0
0
1
1
1
0
0
0
1
1
(−)-2
(+)-2
3
(+)-4
(−)-4
5
6
7
(−)-8
(+)-8
9
10
11
a
Accumulation of racemic ( )-4 was 124 10%. The values of accumulation (in %) are means SD of at least four independent experiments
performed in triplicate.
C
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
a
Table 2. Partial Correlation Coefficient (R) and p-Values for Three Consecutive Multiparameter Models
one-parameter model
two-parameter model
three-parameter model
substituent
R
p
R
p
R
p
1,2,3-trimethoxy
7-hydroxy
0.444
−0.324
0.075
0.112
0.259
0.798
0.355
0.321
0.347
0.453
0.946
1st parameter
1st parameter
0.674
0.448
0.630
0.450
0.450
0.449
0.444
0.036
0.291
0.062
0.297
0.297
0.289
0.299
2nd parameter
0.838
6-acyloxy
0.006
M-biaryl
−0.267
−0.286
−0.272
0.219
0.807
0.012
0.080
0.080
0.092
0.096
2,3-methylenedioxy
3,12-dihydroxy
12,13-methylenedioxy
12,13,14-trimethoxy
0.690
0.690
0.678
−0.020
0.674
a
One-parameter model, two-parameter model, and three-parameter model describing the effect of lignans on doxorubicin accumulation in HL60/
MDR overexpressing P-glycoprotein. The three-parameter model: % accumulation = 181 ( 14) + 74 ( 18) 1,2,3-tri-OCH3 − 92 ( 22) 7-OH + 73
( 25) 6-acyloxy.
(Table 2, R = −0.020). A methylenedioxy group attached to the
B ring correlated better with increased accumulation of
doxorubicin than the trimethoxy group, but the correlation
was found to be statistically insignificant (Table 2, R = 0.219, p
= 0.45). Four pairs of lignans [10 vs 11, (−)-4 vs 9, (+)-2 vs
(+)-4, and 1 vs (−)-2], differing from each other only in the
substitution on the aromatic B ring, were selected for direct
comparison. Compounds with a methylenedioxy group
attached to the B ring showed increased doxorubicin
accumulation for three of these four pairs, with only (+)-2,
having a trimethoxy group, being more effective than (+)-4. A
paired Student’s t test was applied to the accumulation
Figure 5. Correlation between observed versus calculated doxorubicin
accumulation in HL60/MDR cells induced by the test compounds.
The calculated data were obtained by the final three-parameter QSAR
percentages of these four pairs of compounds, and no
significant differences were found (p = 0.539).
Interestingly, compound 5 differs from the substantially less
model. The squared correlation coefficient was 0.702.
active 11 (202% vs 125% doxorubicin accumulation) only in
having a nonpolar, quite bulky angeloyl group attached to C-14
three structural features necessary for increasing doxorubicin
accumulation. Thus, the use of a QSAR method can predict
that there are enhanced inhibitors that meet all structural
requirements.
of the B ring. Lignan 5 was found to be the most active among
those lignans containing only one of the three parameters
enhancing doxorubicin accumulation. This demonstrates that
substitution at C-14 may markedly enhance the effectiveness of
P-gp inhibition.
Lignan 10 was previously isolated from the fruits of S.
chinensis by activity-guided fractionation as a compound that
effectively blocked the function of P-gp.8 However, only
moderate or weak P-gp inhibition by 10 was found in the
present work, which is also consistent with some studies
published previously.9,13
This QSAR work showed that the absence of a 7-hydroxy
group is important for the potent activity of the lignans
investigated. This effect is not surprising since the absence of
the hydroxy group results in an increase of the hydrophobicity
of inhibitors, which is essential for their uptake into the lipid
bilayer. Compounds 3 and 5, containing a 7-hydroxy group, are
relatively polar lignans, but they remain effective due to the
presence of other functional groups that enhance activity.
Despite the small number of lignans possessing a 6-acyloxy
group among the set of molecules investigated, the QSAR
model revealed the marked potentiation of activity by a 6-
acyloxy group in the final three-parameter model. The
importance of acylation at C-6 was not obvious directly from
the percentages of doxorubicin accumulation in a one-
parameter QSAR model, because all three lignans carrying a
6-acyloxy group also have a 7-hydroxy group, which decreases
the activity. The positive effect of esterification at C-6 shown in
the present study corresponds to a previous observation, in
The QSAR data indicated that a 1,2,3-trimethoxy moiety is a
key component for effective and potent P-gp inhibition. The
eight most active lignans found contain this functionality on the
aromatic A ring (Table 1). Even a small change in substitution
of this crucial structural component decreases activity. This may
be seen from the superior activity of (−)-2, containing a 1,2,3-
trimethoxy substituent (doxorubicin accumulation 277%), in
comparison to (−)-4 (accumulation 152%), with a 1-methoxy-
2,3-methylenedioxy moiety. Another example showing the
importance of the 1,2,3-trimethoxy moiety is the clear
difference in potentiation of doxorubicin accumulation between
1 (accumulation 283%) and 9 (accumulation 187%). The most
active, 1, is again substituted with a 1,2,3-trimethoxy moiety,
whereas 9 contains a 1-methoxy-2,3-methylenedioxy function-
ality.
The lignan 8 differs from the two enantiomers of 2 in the
absence of methyl groups at C-3 and C-11. The negative
contribution of hydroxy groups at these two positions was
evident from the comparisons of (−)-2 versus (−)-8 and (+)-2
versus (+)-8 (Table 1). This indicates that at least a methoxy
group attached to C-3 is essential for the potent inhibition of P-
gp.
Despite the importance of a 1,2,3-trimethoxy moiety
attached to the aromatic A ring, substitution of the B ring
with a 12,13,14-trimethoxy moiety had no significant effect on
accumulation according to the one-parameter QSAR model
D
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 6. Viability of the cells treated for 48 h by doxorubicin (DOXO) at concentrations of 0.02, 0.06, and 0.2 μM in the presence or absence of the
lignans or verapamil (V) at a concentration of 25 μM. The data show means SD of at least three independent experiments. The viability of cells
after treatment with doxorubicin alone was compared with viability after treatment with doxorubicin in combination with the test lignans or
verapamil (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).
which synthetic 6-acyl derivatives of lignans inhibited P-gp in
Kb-V1 multidrug resistant human cervix carcinoma cells.19
The ester moiety of cyclooctadiene lignans may be attached
to C-6 from either of the two stereochemically distinct sides of
the cyclooctene ring. The isomeric lignans 3 and 6 differ only in
the orientation at C-6 and C-7, and therefore they can be used
for a direct comparison. Compound 3 was slightly more active
than compound 6 (234% vs 201% doxorubicin accumulation).
Effect of Absolute Configuration. The interaction
between P-gp and chiral ligands is an interesting field of
research. Only a few reports have been made on the
comparison of the P-gp inhibitory activity of enantiomers.
For example, enantiomers of verapamil, omeprazole, and
propranolol displayed equal affinities to P-gp, suggesting that
their recognition of P-gp is nonchiral.20−22 On the other hand,
stereoselective P-gp inhibition by mefloquine was observed in
rat cells but not in human cells. The (+)-stereoisomer of
mefloquine was up to 8-fold more effective than its antipode in
increasing cellular accumulation of [3H]vinblastine in GPNT
(rat brain capillary endothelial) cells, while in Caco-2 (human
colon carcinoma) cells, both enantiomers were equally
effective.23 It was also found that mouse P-gp could distinguish
between two enantiomeric cyclic peptides.24
Dibenzocyclooctadiene lignans are chiral compounds con-
taining at least two asymmetric carbons at C-7 and C-8.
Moreover, they exhibit axial chirality, which arises from
restricted free rotation around the single bond connecting the
two aromatic rings. These compounds have been isolated
mostly from Schisandra species in the optically active form.
However, some of these, such as ( )-4, have been isolated as a
racemate.25 In a previous study, (+)-2 and ( )-4 were
identified as the most potent inhibitors of multidrug-
resistance-associated protein (MRP1) among nine lignans
evaluated. Since the common structural feature of both active
lignans is a M-biaryl configuration and the structurally similar
lignan, 1, possessing an opposite P-biaryl configuration was less
active, it was suggested that the absolute configuration of
lignans may play an important role in the modulation of ABC
transporter activity.14
The present results based on a one-parameter QSAR model
showed that lignans possessing a P-biaryl configuration are
more active than their counterparts with an M-biaryl
configuration, but this difference was not significant statistically
(Table 1). As shown above, the biaryl configuration was a
component in a statistically significant three-parameter model,
but the final three-parameter model, which included a 6-acyloxy
group as a parameter instead of biaryl configuration, was more
accurate. Due to the preparation of three pairs of enantiomers,
the importance of absolute configuration was evaluated directly
by using a t test for each pair of enantiomers. There is no doubt
that (+)-4 increased the accumulation of doxorubicin
substantially more effectively than (−)-4 (p < 0.0001), but
conversely, (−)-8 exceeded the effect of (+)-8 (p < 0.01). The
difference between the effects of (+)-2 and (−)-2 was only
slightly (ca. 10%) in favor of (−)-2.
The significant difference in activity between the enantiomers
of 4 may be explained by the fact that only (+)-4, but not
(−)-4, contains the activity enhancing the trimethoxy moiety
attached to the aromatic A ring.
Dibenzocyclooctadiene Lignans Overcome the Re-
sistance to Doxorubicin in HL60/MDR Cells. The present
results demonstrated clearly that lignans increased the
accumulation of doxorubicin in resistant cells. To determine
whether the lignans increase the sensitivity of HL60/MDR cells
to doxorubicin, cytotoxicity tests were performed.
The viability of cells treated for 48 h with three different
concentrations of doxorubicin (0.02, 0.06, and 0.2 μM) was
determined in the presence or absence of the lignan 1 or (+)-2
and verapamil (used as a positive control) at a concentration of
25 μM using the propidium iodide exclusion assay. The lignans
or doxorubicin alone exhibited a toxicity up to 10% dead cells,
while a combination of lignans with doxorubicin at all
concentrations led to an increased proportion of dead cells
on comparing to doxorubicin alone (Figure 6). The present
results indicate that the test lignans restored the cytotoxic
action of doxorubicin in resistant cells.
The effects of three enantiomeric lignan pairs and the active
lignan 1 (25 μM) in combination with three concentrations of
doxorubicin (0.02, 0.2, and 0.6 μM) on cell proliferation were
analyzed using the WST assay. Table 3 shows the viability of
the cells treated 48 h with doxorubicin alone or doxorubicin in
combination with lignans or verapamil related to an untreated
control. Lignans 1, (+)-2, (−)-2, and ( )-4 significantly
increased the toxicity of doxorubicin at a concentration of 0.2
μM, while all lignans with the exception of (−)-4 and (+)-8
were effective in combination with doxorubicin at a
Using chiral HPLC, it was found that all lignans isolated in
the present study are pure enantiomers except for racemic
( )-4 and 8, in which (−)-8 predominates. To investigate the
effect of absolute configuration of lignans on P-gp inhibition
more comprehensively, three pairs of enantiomers (2, 4, and 8)
were prepared, as described in the Experimental Section and
the Supporting Information.
E
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
increase of cells arrested in the G2/M phase (16%, Figure 7C)
was observed. Neither (+)-2 (12.1%, Figure 7D) nor 1 (8.8%,
Figure 7E) alone had any effect on the cell cycle. Both lignans
in combination with a lower concentration of doxorubicin
increased the proportion of G2/M cells significantly: (+)-2 with
doxorubicin by 67.9% (Figure 7F) and 1 with doxorubicin by
36.9% (Figure 7G). In addition to the cell cycle profile, also the
sub-G1 peak corresponding to apoptotic and/or necrotic cells
was observed upon treatment with doxorubicin at a
concentration of 0.2 μM and in cells treated with a combination
of lignans and doxorubicin at a concentration 0.02 μM (Figure
7B, F, and G). Since the cell cycle arrest in the G2/M phase is
usually a cellular response to doxorubicin treatment, it can be
assumed that these lignans potentiate the effect of doxorubicin
in doxorubicin-resistant HL60/MDR cells.
Table 3. Percentage of Viability of HL60/MDR Cells on
Treatment with Doxorubicin and/or Doxorubicin in
Combination with Selected Lignans
a
doxorubicin (μM)
compound
0
0.02
0.2
0.6
control (DMSO)
96
2
109
9
83 12
42 11*
52 4*
48 1*
41 13*
66 41
91 14
39 13*
43 2**
39 2*
36 19*
38 14*
81 14
1
128 28
92 11
144 23
88 13
102 26
121 12
108 30
106 12
(+)-2
(−)-2
( )-4
(+)-4
(−)-4
(+)-8
(−)-8
verapamil
117
0
94 27
106
85 10
78
93 12
8
100 16
97
94 17
52
29 10**
7
100
9
73
30 8*
5***
2
8
73 34
78 12
8
4
In the present study, a preliminary structure−activity
relationship of dibenzocyclooctadiene lignans as P-gp inhibitors
has been determined. The three main structural characteristics
involved in P-gp inhibition were identified: a 1,2,3-trimethoxy
moiety, a 6-acyloxy group, and the absence of a 7-hydroxy
group. The differences in activity between the enantiomers of 4
and the structural requirement for a trimethoxy moiety located
only on the A ring, but not on the B ring, have provided
evidence that the interaction between an active lignan of this
type and P-gp is stereospecific (Figure 8). The potentiation of
the doxorubicin effect on HL60/MDR cells was shown in
cytotoxicity assays and cell cycle analysis. Since none of the
lignans tested met all three structural requirements for potent
P-gp inhibition, the present results may stimulate the rational
design of more efficient molecules. Among the lignans
investigated, 1 and (+)-2 were found to be the most promising
P-gp modulators due to their strong increases of the
doxorubicin accumulation and potentiation of the cytotoxicity
shown for doxorubicin. These lead compounds might be
improved further by substitution at carbon C-6 and by
modification of the aromatic B ring.
a
Data are expressed as percentages relative to untreated cells and are
means SD of two independent experiments performed in triplicate.
The statistical analysis compared samples of selected lignans (or
verapamil) in combination with doxorubicin to doxorubicin only at the
same concentration (*p < 0.05, **p < 0.01, ***p < 0.001).
concentration of 0.6 μM. Verapamil enhanced the effect of
doxorubicin at concentrations of 0.2 and 0.6 μM.
The above-mentioned results demonstrated the strong
potentiation of doxorubicin cytotoxicity by most test lignans
and confirmed the results obtained in accumulation studies, in
which the lignans (+)-4 and (−)-8 were more potent than their
enantiomers.
A G2/M cell cycle arrest is a typical consequence of
doxorubicin treatment.14,26 Therefore, the ability of 1 or (+)-2
(25 μM) was tested in combination with doxorubicin at a
concentration of 0.02 μM to increase the proportion of
resistant HL60/MDR cells in the G2/M phase.
Doxorubicin at a concentration of 0.2 μM greatly increased
the percentage of cells in the G2/M phase (52%, versus 11% in
DMSO only treated cells, Figure 7A,B), while at a 10 times
lower concentration of doxorubicin (0.02 μM), only a weak
Figure 7. Cell cycle analysis. Lignan 1 or (+)-2 (25 μM) in combination with doxorubicin at a concentration of 0.02 μM increased the proportion of
G2/M significantly. The sub-G1 peak that could display apoptotic and/or necrotic cells is apparent (B, F, G: a). Control (A), doxorubicin 0.2 μM
(B), doxorubicin 0.02 μM (C), (+)-2 (D), 1 (E), (+)-2 + doxorubicin 0.02 μM (F), 1 + doxorubicin 0.02 μM (G).
F
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
A stock solution of each lignan (1 mg/50 μL) was prepared in
dimethyl sulfoxide (DMSO; Sigma-Aldrich Corp.) and stored at −20
°C. In biological experiments, the lignans were used in a concentration
range of 0.025−75 μM.
Cell Lines and Cultivation Conditions. Two cancer cell lines
were used: HL60 (human promyelotic leukemia) obtained from the
European Collection of Animal Culture (ECAC, Salisbury, U.K.) and
its multidrug-resistant subline HL60/MDR derived from parental
HL60 cells after treatment with doxorubicin, which was obtained from
Prof. B. Sarkadi (Semmelweis University Medicine, National Blood
Centre, Membrane Research Group, Hungarian Academy of Science,
Budapest, Hungary). The cells were grown in RPMI 1640
supplemented with 2 mM glutamine, 10% fetal calf serum, 100 IU/
mL penicillin, and 100 μg/mL streptomycin (PAA Laboratories,
Austria). Cells were incubated at 37 °C under 5% CO2 in a high-
humidity atmosphere and subcultured three times a week.
RNA Extraction and RT-PCR. RT-PCR was carried out to analyze
the mRNA levels of MRP1, MRP2, and P-gp (MDR1) as described
previously.14 Briefly, total RNA was extracted using the RNAeasy Mini
kit (Quiagen) according to the manufacturer’s instructions. PCR was
performed using gene-specific primers for MRP1 (forward: 5′-
GGTGCTTCCCACGGAGG-3′; reverse: 5′-TCAACCACAA-
AACTGCAGCC-3′), MRP2 (forward: 5′-GACATCAGAAA-
T A G A G A C C - 3 ′ ; r e v e r s e 5 ′ - C T A C T C C A T C A A -
TGATAATCTGACC-3′), and MDR1 (forward: 5′-AGTGGTTCAG-
GTGGCTCT-3′; reverse: 5′-TTCTGTCTTGGGCTTGTG-3′). The
β-actin gene was coamplified and served as a control (forward: 5′-
TTCCAGCCTTCCTTCCTGGG-3′; reverse: 5′-TTGCGCTCAG-
GAGGAGCAAT-3′). All primers were synthesized by Invitrogen.
Expected product sizes were 182 bp for MRP1, 341 bp for MDR1, 322
bp for MRP2, and 250 bp for β-actin. PCR amplification was
performed on a GeneAmp PCR System 9700 (Applied Biosystems).
PCR products were run on 2% agarose gels containing 0.6 μg/mL of
ethidium bromide and photographed under a TCX20 UV-
transiluminator (Vilber Lourmat, France).
Immunodetection of MDR1 Protein. The presence of MDR1
protein at the surface of HL60/MDR cells was studied using flow
cytometry analysis of living cells stained with mouse monoclonal anti
P-gp antibody conjugated with PE (IOTest) CD243 (P-gp; Beckman
Coulter). As an isotype control mouse PE/IgG1 (IOTest, Beckman
Coulter) was used. An excitation wavelength of 488 nm and an
emission wavelength of 575 nm (FL2) were used. A minimum of 10
000 cells were counted for each sample.
Drug Accumulation Assay. Since doxorubicin is a fluorescent
substance, its content in cells can be measured directly by flow
cytometry. For the drug accumulation assays HL60/MDR cells at a
concentration of 2 × 106 per mL grown in 150 μL of medium in 96-
well plates were preincubated in medium containing either a test
compound (25 μM) or verapamil (25 μM) for 15 min. Then
doxorubicin (10 μM) was added, and the cells were incubated for an
additional 60 min. The cells were collected and washed with ice-cold
PBS. The intracellular doxorubicin content was determined using a
Cytomics FC 500 flow cytometer (Beckman Coulter, Inc., CA, USA)
at an excitation wavelength of 488 nm and an emission wavelength of
575 nm (FL2). A minimum of 10 000 cells were counted for each
sample. Results were expressed as a percentage ratio of the mean
fluorescence of doxorubicin in the presence of a modulator to the
mean fluorescence of doxorubicin without a modulator.
Figure 8. Summary of the structural features of dibenzocyclooctadiene
lignans influencing the inhibition of P-gp in HL60/MDR cells. The
plus-circles specify the positive contribution of structural elements to
P-gp inhibition, and the minus-circles show the negative impact on P-
gp inhibition.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotations were
measured on a JASCO P-2000 polarimeter in a 1 dm tube filled
with chloroform solution at the D line of sodium. CD spectra were
recorded using a JASCO J-815 spectropolarimeter in methanol loaded
into a cuvette with a path length of 10 mm. 1H and 13C NMR spectra
were recorded on a Bruker AVANCE 300 MHz instrument or a
Bruker AVANCE III 500 MHz spectrometer equipped with 5 mm
Prodigy cryoprobe or a Bruker AVANCE III 700 MHz instrument
equipped with 5 mm TXO cryoprobe. All samples were dissolved and
measured in CDCl3 at 298 K. Chemical shifts are reported in ppm
relative to tetramethylsilane. The compounds were identified on the
basis of 1H, 13C, 13C APT, 1H−1H COSY, 1H−13C HSQC, and
1H−13C HMBC experiments. High-resolution mass spectra were
recorded on a MicrOTOF-QII (Bruker, Germany) mass spectrometer,
operated in a positive electrospray ionization mode.
Plant Material. The seeds of Schisandra chinensis were collected in
the surroundings of Vladivostok (Russia) and were imported by
Herbaton Ltd. (Klcov, Slovakia). The seeds were identified by Ing.
̌
Pavel Musil (head of the Centre of Medicinal Plant of Masaryk
University), and a voucher specimen (No. 728-12) was deposited in
the Department of Biochemistry, Faculty of Medicine, Masaryk
University, Brno. The seeds were air-dried for a week at ambient
temperature and then powdered with a mechanical grinder.
Extraction and Isolation. The lignans 1, (+)-2, 3, ( )-4, 5, 8, 9,
10, and 11 were isolated from seeds of S. chinensis as described in the
Supporting Information. The isolation of 6 and 7 is described in a
previous paper.27 The lignans were identified by using H and 13C
1
NMR, ESIMS, and CD experiments and by comparison of
experimental data with those described in the literature.25,28−38
Characterization of compounds is given in the Supporting
Information. The structure of 1 was also verified using X-ray
crystallographic analysis.39
Preparation of (+)-4, (−)-4, (+)-8, (−)-8, and (−)-2. Racemic
( )-4 (18 mg) was separated into the corresponding enantiomers on a
short semipreparative Chiralcel OD column (50 × 10 mm i.d., Daicel)
with a mobile phase containing 0.125% (v/v) propane-2-ol in n-
hexane. A sample injection was 1−2 mg of ( )-4. As a result of
semipreparative chiral chromatography, 7.8 mg of (+)-4 and 5.2 mg of
(−)-4 were obtained. Compound 8 (20 mg) was resolved into
enantiomers by repeating the separation on the Chiralcel OD column
(50 × 10 mm i.d., sample injection 1 mg) using a mixture of n-
hexane−propane-2-ol (99:1) as a mobile phase. After evaporation of
the solvent, 2.0 mg of (+)-8 and 15.7 mg of (−)-8 were obtained. The
lignan (−)-2 (4.7 mg) was prepared by methylation of (−)-8 (6.7 mg)
with methyl iodide (0.1 mL) and K2CO3 (9.5 mg) in acetone under
reflux.
Determination of Antiproliferative Activity. Cell Viability
Assay. The cell viability assay was based on the exclusion of propidium
iodide (PI; Sigma-Aldrich, St. Louis, MO, USA) by the intact viable
cells. The cells were plated in 12-well tissue culture test plates (Orange
Scientific, Braine-I’Alleud, Belgium) at 7 × 104 cells/mL of medium
and treated with test compounds at a concentration of 25 μM in
combination with doxorubicin at concentrations of 0.02, 0.06, and 0.2
μM.
After 48 h of incubation PI was added, and immediately after
addition, the percentage of dead (PI-positive) cells was detected using
a Cytomics FC 500 flow cytometry system (Beckman Coulter, Inc.)
with channel FL3 (emission at 620 nm). A total number of 10 000
G
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
cells were analyzed for each sample. The viability of cells after
treatment with doxorubicin alone was compared with the viability after
treatment with doxorubicin in combination with the test compounds
or verapamil (positive control).
AUTHOR INFORMATION
■
Corresponding Author
*Tel: +420549496985. Fax: 549491327. E-mail: ipokorna@
med.muni.cz.
WST Assay. Antiproliferative activity was assessed using a WST
assay on 96-well plates (Nunc A/S, Rockilde, Denmark). The assay is
based on the reduction of WST-1 (2-(4-iodophenyl)-3-(4-nitro-
phenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, Na salt) by viable
cells. The amount of the formazan dye generated by the activity of
dehydrogenases within cells is therefore directly proportional to the
number of living cells. All cells were seeded at a density of 5 × 104 cells
per mL (100 μL/per well) and incubated at 37 °C in a medium
containing doxorubicin, modulators (test compounds or verapamil), or
doxorubicin with modulators for 48 h. After incubation 10 μL of
working WST-1 solution [1 mM WST (Serva, Heidelberg, Germany)]
and 0.2 mM MMPM [(1-methoxy-5-methylphenazinium methyl
sulfate; Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA)]
were added to 100 μL of fresh medium in each well, and cells were
incubated for a further 4 h under normal cell culture conditions. At the
end of the incubation periods the optical density was read at 450 nm
using a DTX 880 multimode detector (Beckman Coulter, Inc.). Each
concentration of each of the test compounds was examined in four
replicate wells. The experiment was repeated at least three times.
Cell Cycle Analysis. Approximately 5 × 105 cells/mL (HL60/
MDR) were treated with doxorubicin at a concentration of 0.2 μM
and/or 0.02 μM either alone or in combination with 1 and/or (+)-2 at
a concentration of 25 μM and cultivated under standard conditions for
48 h. Cells were harvested into ice-cold PBS and fixed with 70%
ethanol for 30 min at room temperature. After washing in PBS, the
cells were incubated with RNase A (17 μg/mL) at 37 °C for 30 min
and then stained with PI (3 μg/mL) for 10 min (in darkness, at room
temperature).
Present Address
∥Public Health Institute Ostrava, Partyzan
́
ske
́
Nam
́
es
̌
ti 7, 702
́
00 Ostrava, Czech Republic.
Author Contributions
▽J. Slanina and G. Pac
́ ́
hnikova contributed equally to this
work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors thank Prof. B. Sarkadi (Semmelweis University
Medicine, National Blood Centre, Membrane Ressearch Group,
Hungarian Academy of Science, Budapest, Hungary) for
providing HL60/MDR cells and Mrs. Z. Prokopova
́
and I.
́
Komarkova for laboratory assistance. O. Julinek, Institute of
́
́
Chemical Technology, Prague, Czech Republic, is acknowl-
edged for optical rotation measurements. This work was
supported by Masaryk University Projects of Specific Research
MUNI/A/0938/2013 and MUNI/A/0954/2013.
REFERENCES
■
(1) Filipits, M. Drug Discovery Today 2004, 1, 229−234.
(2) Yu, X. Q.; Xue, C. C.; Wang, G.; Zhou, S. F. Curr. Drug Metab.
2007, 8, 787−802.
The cell cycle profile was analyzed using a Cytomics FC 500 flow
cytometer using the FL3 channel (emission at 620 nm), and 10 000−
20 000 events were acquired. Data from each sample were saved as
separate flow cytometry standard (FCS) files using CXP analysis
software (Beckman Coulter, Inc.) and were analyzed for cell cycle
phases using Multicycle AV for Windows software (Phoenix Flow
System, San Diego, CA, USA). DNA content analysis included the
determination of the percentage of G1, G2/M, and S phases and sub-
G1 fractions.
(3) Ozben, T. FEBS Lett. 2006, 580, 2903−2909.
(4) Wu, C. P.; Ohnuma, S.; Ambudkar, S. V. Curr. Pharm. Biotechnol.
2011, 12, 609−620.
(5) Li, L.; Lu, Q. H.; Shen, Y. W.; Hu, X. Biochem. Pharmacol. 2006,
71, 584−595.
(6) Li, L.; Pan, Q. R.; Sun, M.; Lu, Q. H.; Hu, X. Life Sci. 2007, 80,
741−748.
(7) Sun, M.; Xu, X. L.; Lu, Q. H.; Pan, Q. R.; Hu, X. Cancer Lett.
2007, 246, 300−307.
Statistical Analysis. All statistical analysis was performed with
Statistica version 10 (StatSoft software). The data for doxorubicin
accumulation were from at least four independent experiments
performed in triplicate. Statistical comparison of doxorubicin
accumulation induced by the test compound versus the control
group and the verapamil group was done using one-way analysis of
variance (ANOVA) followed by Dunnett’s test. Stepwise multiple
regression analysis was used to find the relation between the
percentage of P-gp inhibition and the descriptors in a QSAR model.
The least-squares regression analysis was used to determine the
correlation between the measured data and calculated percentage of
doxorubicin accumulation. A paired Student’s t test was applied to
compare the effects of enantiomers or pairs of test lignans differing
only in one structural parameter.
(8) Wan, C. K.; Zhu, G. Y.; Shen, X. L.; Chattopadhyay, A.; Dey, S.;
Fong, W. F. Biochem. Pharmacol. 2006, 72, 824−837.
(9) Huang, M.; Jin, J.; Sun, H.; Liu, G. T. Cancer Chemother.
Pharmacol. 2008, 62, 1015−1026.
(10) Slanina, J.; Glatz, Z. J. Chromatogr. B 2004, 812, 215−229.
(11) Hancke, J. L.; Burgos, R. A.; Ahumada, F. Fitoterapia 1999, 70,
451−471.
(12) Opletal, L.; Sovova,
357−371.
(13) Pan, Q. R.; Lu, Q. H.; Zhang, K.; Hu, X. Cancer Chemother.
Pharmacol. 2006, 58, 99−106.
́ ́ ́
H.; Bartlova, M. J. Chromatogr. B 2004, 812,
(14) Slaninova,
Vitro 2009, 23, 1047−1054.
́ ̌ ́ ́
I.; Brezinova, L.; Koubíkova, L.; Slanina, J. Toxicol. in
(15) Chiu, P. Y.; Leung, H. Y.; Poon, M. K. T.; Mak, D. H. F.; Ko, K.
M. Mol. Cell. Biochem. 2006, 289, 185−191.
Cytotoxicity data were evaluated from at least three independent
experiments performed in triplicate (PI-exclusion assay) or two
independent experiments performed in quadruplicate (WST assay).
The results were analyzed using Student’s t test.
(16) Chiu, P. Y.; Tang, M. H.; Mak, D. H. F.; Poon, M. K. T.; Ko, K.
M. Free Radical Biol. Med. 2003, 35, 368−380.
(17) Ip, S. P.; Ko, K. M. Biochem. Pharmacol. 1996, 52, 1687−1693.
(18) Yoo, H. H.; Lee, M.; Lee, M. W.; Lim, S. Y.; Shin, Y.; Kim, D. H.
Planta Med. 2007, 73, 444−450.
ASSOCIATED CONTENT
(19) Schobert, R.; Kern, W.; Milius, W.; Ackermann, T.; Zoldakova,
M. Tetrahedron Lett. 2008, 49, 3359−3362.
■
S
* Supporting Information
(20) Gruber, A.; Peterson, C.; Reiyenstein, P. Int. J. Cancer 1988, 41,
224−226.
The isolation procedure for the lignans investigated and their
characteristic details, including circular dichroism and NMR
spectra, are available free of charge via the Internet at http://
pubs.acs.org.
(21) Loetchutinat, C.; Heywang, C.; Priebe, W.; Garnier-Suillerot, A.
Biochem. Pharmacol. 2001, 62, 561−567.
(22) Pajeva, I. K.; Wiese, M. J. Med. Chem. 2002, 45, 5671−5686.
H
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(23) Pham, Y. T.; Regina, A.; Farinotti, R.; Couraud, P. O.; Wainer, I.
W.; Roux, F.; Gimenez, F. Biochim. Biophys. Acta 2000, 1524, 212−
219.
(24) Aller, S. G.; Yu, J.; Ward, A.; Weng, Y.; Chittlaboina, S.; Zhuo,
R.; Harrell, P. M.; Trinh, Y. T.; Zhang, Q.; Urbatsch, I. L.; Chang, G.
Science 2009, 323, 1718−1722.
(25) Ikeya, Y.; Taguchi, H.; Yosioka, I. Chem. Pharm. Bull. 1982, 30,
132−139.
(26) Savatier, J.; Rharass, T.; Canal, Ch.; Gbankoto, A.; Vigo, J.;
Salmon, J. M.; Ribou, A. C. Leuk. Res. 2012, 36, 791−798.
̌
̌
(27) Smejkal, K.; Slapetova,
S.; Innocenti, G.; Vanco
Dvorska, M.; Slanina, J.; Kramar
Planta Med. 2010, 76, 1672−1677.
́
T.; Krmencí
, J.; Casarin, E.; Carrara, M.; Kalvarova,
va, E.; Julínek, O.; Urbanova,
̌
k, P.; Babula, P.; Dall’Acqua,
̌
́
́
K.;
M.
́
́
o
̌
(28) Ikeya, Y.; Taguchi, H.; Yosioka, I. Chem. Pharm. Bull. 1978, 26,
328−331.
(29) Ikeya, Y.; Taguchi, H.; Yosioka, I.; Kobayashi, H. Chem. Pharm.
Bull. 1979, 27, 1383−1394.
(30) Ikeya, Y.; Taguchi, H.; Yosioka, I.; Iitaka, Y.; Kobayashi, H.
Chem. Pharm. Bull. 1979, 27, 1395−1401.
(31) Ikeya, Y.; Taguchi, H.; Yosioka, I.; Kobayashi, H. Chem. Pharm.
Bull. 1979, 27, 1583−1588.
(32) Ikeya, Y.; Taguchi, H.; Yosioka, I.; Kobayashi, H. Chem. Pharm.
Bull. 1979, 27, 2695−2709.
(33) Ikeya, Y.; Taguchi, H.; Yosioka, I.; Kobayashi, H. Chem. Pharm.
Bull. 1980, 28, 3357−3361.
(34) Ikeya, Y.; Taguchi, H.; Yosioka, I. Chem. Pharm. Bull. 1982, 30,
3207−3211.
(35) Ikeya, Y.; Sugama, K.; Minoru, O.; Mitsuhashi, H. Phytochemistry
1991, 30, 975−980.
(36) Hu, D.; Wang, X.; Cao, Y.; Liu, Z.; Han, N.; Yin, J. J. Tradit.
Med. 2009, 4, 14−18.
(37) Gnabre, J.; Unlu, I.; Chang, T. S.; Lisseck, P.; Bourne, B.;
Scolnic, R.; Jacobsen, N. E.; Bates, R.; Huang, R. C. J. Chromatogr. B
2010, 878, 2693−2700.
(38) Ikeya, Y.; Taguchi, H.; Sasaki, H.; Nakajima, K.; Yosioka, I.
Chem. Pharm. Bull. 1980, 28, 2414−2421.
(39) Marek, J.; Slanina, J. Acta Crystallogr., Sect. C: Cryst. Struct.
Commun. 1998, 54, 1548−1550.
I
dx.doi.org/10.1021/np500521v | J. Nat. Prod. XXXX, XXX, XXX−XXX