Polyoxypregnane steroids with an open-chain sugar moiety from Marsdenia tenacissima and their chemoresistance reversal activity

Article abstract of DOI:10.1016/j.phytochem.2016.03.006

A polyoxypregnane aglycone, 12β-O-acetyl-11α-O-isobutyryltenacigenin B, and four polyoxypregnane glycosides with a pachybionic acid ester moiety, 12β-O-acetyl-3-O-(6-deoxy-3-O-methyl-β-d-allopyranosyl-(1→4)-β-d-oleandronyl)-11α-O-isobutyryltenacigenin B, 12β-O-acetyl-3-O-(6-deoxy-3-O-methyl-β-d-allopyranosyl-(1→4)-β-d-oleandronyl)-11α-O-tigloyltenacigenin B, 12β-O-acetyl-3-O-(6-deoxy-3-O-methyl-β-d-allopyranosyl-(1→4)-β-d-oleandronyl)-11α-O-2-methylbutyryltenacigenin B, and 12β-O-acetyl-3-O-(β-d-glucopyranosyl-(1→4)-6-deoxy-3-O-methyl-β-d-allopyranosyl-(1→4)-d-oleandronyl)-11α-O-tigloyltenacigenin B, were isolated from the canes of Marsdenia tenacissima, together with a disaccharide derivative. Their structures were elucidated by extensive spectroscopic analysis, and the absolute configurations were further determined by X-ray crystallographic analysis. With the exception of the disaccharide derivative, all five compounds are unusual naturally occurring polyoxypregnane glycosides bearing an open-chain sugar moiety. Two of these exhibit a wide spectrum of chemoresistance reversal activity, and potential mechanisms were studied accordingly.

Full text of DOI:10.1016/j.phytochem.2016.03.006

Phytochemistry 126 (2016) 47–58
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Polyoxypregnane steroids with an open-chain sugar moiety from
Marsdenia tenacissima and their chemoresistance reversal activity
Sheng Yao a,b,1, Kenneth Kin-Wah To b,c,1, Liang Ma a,b, Chun Yin b,d, Chunping Tang a,b, Stella Chai b,d
,
Chang-Qiang Ke a,b, Ge Lin b,d, , Yang Ye a,b,e,
⇑
⇑
a State Key Laboratory of Drug Research and Natural Products Chemistry Department, Shanghai Institute of Materia Medica, Chinese Academy of Sciences,
555 Zu-Chong-Zhi Road, Zhangjiang Hi-Tech Park, Shanghai 201203, China
b Joint Research Laboratory of Promoting Globalization of Traditional Chinese Medicines between Shanghai Institute of Materia Medica,
Chinese Academy of Sciences and The Chinese University of Hong Kong, China
c School of Pharmacy, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong, China
d School of Biomedical Sciences, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong, China
e School of Life Science and Technology, ShanghaiTech University, Shanghai, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 June 2015
Received in revised form 29 February 2016
Accepted 8 March 2016
Available online 14 March 2016
A polyoxypregnane aglycone, 12b-O-acetyl-11
a
-O-isobutyryltenacigenin B, and four polyoxypregnane glyco-
sides with a pachybionic acid ester moiety, 12b-O-acetyl-3-O-(6-deoxy-3-O-methyl-b-
D
-allopyranosyl-(1?4)-
-O-isobutyryltenacigenin B, 12b-O-acetyl-3-O-(6-deoxy-3-O-methyl-b- -allopyranosyl-
-O-tigloyltenacigenin B, 12b-O-acetyl-3-O-(6-deoxy-3-O-methyl-b- -allopyra-
-O-2-methylbutyryltenacigenin B, and 12b-O-acetyl-3-O-(b- -glucopyra-
nosyl-(1?4)-6-deoxy-3-O-methyl-b- -allopyranosyl-(1?4)- -oleandronyl)-11 -O-tigloyltenacigenin B, were
b-D
-oleandronyl)-11
a
D
(1?4)-b-
D
-oleandronyl)-11
a
D
nosyl-(1?4)-b-
D
-oleandronyl)-11
a
D
D
D
a
Keywords:
Marsdenia tenacissima
Asclepiadaceae
C21 steroids
Open-chain sugars
Absolute configuration
X-ray crystallography diffraction
Chemoresistance reversal activity
Structure elucidation
isolated from the canes of Marsdenia tenacissima, together with a disaccharide derivative. Their structures
were elucidated by extensive spectroscopic analysis, and the absolute configurations were further deter-
mined by X-ray crystallographic analysis. With the exception of the disaccharide derivative, all five com-
pounds are unusual naturally occurring polyoxypregnane glycosides bearing an open-chain sugar moiety.
Two of these exhibit a wide spectrum of chemoresistance reversal activity, and potential mechanisms
were studied accordingly.
Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction
17b-tenacigenin B (Deng et al., 2005b) and the diester derivatives
of tenacigenin B (Luo et al., 1993); and marstenacigenins A and B
Marsdenia tenacissima (Roxb.) Wight et Arn., a plant belonging
to the family Asclepiadaceae, is widely distributed from tropical
to subtropical Asia. Its rhizomes and roots have long been used
in China for the treatment of asthma, cancer, tracheitis, tonsillitis,
pharyngitis, cystitis, and pneumonia (Jiangsu New College of
Medicine, 1977). Previous chemical investigations resulted in the
identification of more than 15 polyoxypregnane genins from
M. tenacissima, classified into three types: tenacigenins A, B, and
C (Yang et al., 1981) and their analogues (Deng et al., 2005a);
(Qiu et al., 1996). On the basis of the above aglycones, over 20 gly-
cosides of tenacigenin B have been reported: from M. tenacissima
tenacissosides A–N (Chen et al., 1999; Liu et al., 2008; Miyakawa
et al., 1986; Wang et al., 2006; Xing et al., 2004) and marsdeno-
sides A–K (Deng et al., 2005a, 2005c); seven glycosides of 17b-
tenacigenin B named marsdenosides L and M (Huang et al., 2009)
and tenacissimosides E–I (Yao et al., 2014), as well as four marste-
nacigenin glycosides named marstenacissides A–D (Xia et al.,
2011). In addition, cissogenin (another type of compound) and its
analogues, have been reported from seeds of the plant growing
in India (Singhal et al., 1980a,b). Biological evaluations of extracts
and isolated compounds from M. tenacissima also have antiasth-
matic (Zhou et al., 1980), anticancer (Xue et al., 2012; Ye et al.,
2014), and multidrug resistance (MDR) reversal (Han et al., 2012;
Hu et al., 2008) activities. Furthermore, a new polyoxypregnane
aglycone from this plant was able to circumvent P-glycoprotein
(P-gp)-mediated MDR, as well as potentiating the activity of
erlotinib and gefitinib in epidermal growth factor receptor tyrosine
⇑
Corresponding authors at: Joint Research Laboratory of Promoting Globalization
of Traditional Chinese Medicines between Shanghai Institute of Materia Medica,
Chinese Academy of Sciences and The Chinese University of Hong Kong, China (G.
Lin). State Key Laboratory of Drug Research and Natural Products Chemistry
Department, Shanghai Institute of Materia Medica, Chinese Academy of Sciences,
555 Zu-Chong-Zhi Road, Zhangjiang Hi-Tech Park, Shanghai 201203, China (Y. Ye).
E-mail addresses: linge@cuhk.edu.hk (G. Lin), yye@mail.shcnc.ac.cn (Y. Ye).
1
Contributed equally to this work.
http://dx.doi.org/10.1016/j.phytochem.2016.03.006
0031-9422/Ó 2016 Elsevier Ltd. All rights reserved.

48
S. Yao et al. / Phytochemistry 126 (2016) 47–58
R1 R2
kinase inhibitor (EGFR TKI)-resistant non-small-cell lung cancer
cells (Yao et al., 2014).
O
21
1
H
H
iBu
20
17
1a
2
mBu
O
O
18
13
S1 iBu
S1 Tig
S1 mBu
S2 Tig
In a continuation of efforts in the search for MDR-reversal
agents from natural sources, the EtOAc fraction of the EtOH extract
prepared from the stems of M. tenacissima was investigated. A
polyoxypregnane aglycone (1) and four new C21 steroids with an
open-chain sugar moiety (2–5), together with a known disaccha-
ride derivative (6), were afforded. Here, the isolation and structural
characterization of these new compounds, their chemoresistance
reversal activity, and mechanistic studies are presented.
R2O
3
12
11
19
16
4
14
5
1
4
9
6
15
8
7
2
3
10
Tenacissoside A S3 Tig
O
5
R1O
HO
O
O
O
O
O
O
O
OH
O
Tig
iBu
6
O
OH
O
O
S1
=
HO
O
OH
O
2. Results and discussion
O
mBu
OH
OH
O
O
2.1. Structure elucidation
O
O
O
S2
=
=
HO
HO
O
OH
O
OH
O
O
A 95% EtOH extract of the air-dried stems of M. tenacissima
(8.0 kg) was suspended in water and extracted with EtOAc. The
EtOAc fraction was dissolved in MeOH, and partitioned into a
MeOH and a petroleum ether extract. After repeated column chro-
matography over silica gel, Sephadex LH-20, and MCI gel, and
preparative HPLC, compounds 1–6 were obtained (Fig. 1).
OH
O
O
O
O
S3
HO
HO
O
OH
OH
Fig. 1. Structures of the isolated compounds.
Compound 1 was obtained as a colorless, amorphous powder.
Its HR-ESI-MS peak at m/z 499.2642 ([M+Na]+, calculated for
C
Table 1
1H NMR spectroscopic data of the aglycone moieties of compounds 1–5 (d in ppm, J in
27H40O7Na 499.2672) suggested a molecular formula of C27H40O7
Hz).
with eight degrees of unsaturation. Its IR spectrum showed absorp-
tion bands for hydroxy groups at 3435 cm–1 and carbonyl groups at
1736 and 1707 cm–1. The 1H NMR spectrum (Table 1) displayed
proton signals for three methyls at dH 1.06 (s), 1.08 (s), and 2.18
(s), and two oxygenated methines at dH 4.98 (d, J = 10.1 Hz) and
5.35 (t, J = 10.1 Hz). The proton signals at dH 1.96 (s), and dH 1.09
(6H, d, J = 2.7 Hz) and 2.38 (m), along with carbon resonances at
dC 170.6 (s) and 20.8 (q), and at dC 176.0 (s), 34.5 (d), 18.9 (q),
and 18.4 (q), indicated the presence of acetyl and isobutyryl ester
groups, respectively. By comparison with polyoxygenated preg-
nane steroids from the title plant, the NMR data (Tables 1 and 2)
Position 1a,c
2b,c
3a,c
4a,c
5a,d
1a
1b
2a
2b
3
4a
4b
5
6a
6b
7a
7b
9
1.24, m
1.30, m
1.51, m
1.41, m
1.72, m
4.68, m
1.43, m
1.66, m
1.43, m
1.42, m
1.60, m
1.23, m
1.87, m
2.02, d
(10.1)
1.22, m
1.52, m
1.40, m
1.66, m
4.66, m
1.40, m
1.66, m
1.40, m
1.38, m
1.52, m
1.24, m
1.86, m
2.00, d
(10.3)
1.27, m
1.50, m
1.41, m
1.70, m
4.66, m
1.41, m
1.67, m
1.41, m
1.36, m
1.52, m
1.22, m
1.87, m
2.00, d
(10.1)
1.33, m
1.54, m
1.46, m
1.67, m
4.68, m
1.42, m
1.72, m
1.52, m
1.44, m
1.58, m
1.27, m
1.90, m
2.02, d
(10.2)
1.48, m
1.37, m
1.69, m
3.58, m
1.33, m
1.63, m
1.35, m
1.40, m
1.57, m
1.24, m
1.87, m
1.98, d
(10.1)
of 1 almost resembled those of 12b-O-acetyl-11a-O-2-methylbu-
tyryltenacigenin B (1a) (Hu et al., 2008). The only difference found
was that the NMR data assigned to a 2-methylbutyryl group in 1a
was replaced by those assigned to an isobutyryl unit. The
long-range correlation of H-11 (dH 5.35)/C-10 (dC 176.0) inferred
that the isobutyryl ester group was located at C-11. Therefore,
11
12
5.35, t
(10.1)
4.98, d
(10.1)
5.34, t
(10.1)
4.98, d
(10.1)
5.38, t
(10.3)
4.90, d
(10.3)
5.33, t
(10.1)
4.96, d
(10.1)
5.36, t
(10.2)
5.01, d
(10.2)
15a
15b
16a
16b
17
1.56, m
1.98, m
1.58, m
2.15, m
2.92, d
(7.6)
1.55, m
1.98, m
1.61, m
2.16, m
2.92, d
(7.4)
1.06, s
1.06, s
2.19, s
iBu
2.37, m
1.07, d
(2.7)
1.07, d
(2.7)
1.58, m
1.98, m
1.60, m
2.14, m
2.90, d
(7.3)
1.07, s
1.04, s
2.19, s
Tig
1.57, m
1.95, m
1.59, m
2.14, m
2.90, d
(7.4)
1.04, s
1.04, s
2.18, s
mBu
1.52, m
1.98, m
1.66, m
2.21, m
3.00, d
(7.3)
1.09, s
1.05, s
2.13, s
Tig
compound 1 was determined to be 12b-O-acetyl-11a-O-isobu-
tyryltenacigenin B.
Compound 2 had a molecular formula of C41H64O15 containing
10 degrees of unsaturation, as deduced by HR-ESI-MS. IR absorp-
tion bands at 3450, 1738, and 1707 cm–1 indicated the presence
of hydroxy and carbonyl groups. The 1H NMR spectrum of com-
pound 2 (Tables 1 and 3) displayed signals for three singlet methyl
groups at dH 1.06 (6H, s) and 2.19 (s), two doublet methyls at dH
1.07 (6H, d, J = 2.7 Hz), a multiplet methine at dH 2.37, and two oxy-
genated methines at dH 4.98 (d, J = 10.1 Hz) and 5.34 (t, J = 10.1 Hz).
The anomeric proton signal at dH 4.62 (d, J = 8.0 Hz) and the corre-
sponding carbon resonance at dC 100.4 are indicative of a sugar
moiety with a b-linkage. Extensive analyses established that the
NMR data of the aglycone of 2 closely resembled those of com-
pound 1, except for significant differences in the chemical shifts
of C-2, C-3, and C-4. This suggested a linkage of sugar moieties to
C-3 in 2. Interpretation of the H-H COSY and HSQC spectra of com-
pound 2 revealed a further two partial moieties –C(60)H3–C(50)H–C
(40)H–C(30)H–C(20)H2– and –C(600)H3–C(500)H–C(400)H–C(300)H–C(200)
H–C(100)H–. In the HMBC spectrum of 2, the long-range correlations
from the methoxy signal at dH 3.41 (s) to the oxygenated methine
carbon at dC 78.3 (C-30), and from the methoxy signal at dH 3.65 (s)
to the oxygenated methine carbon at dC 80.6 (C-300) indicated that
the two methoxy groups were attached to C-30 and C-300,
18
19
21
C-11
20
1.08, s
1.06, s
2.18, s
iBu
2.38, m
1.09, d
(2.7)
2.15, m
1.30, m
30
6.72, q
(6.9)
1.72, d
(6.9)
6.80, q
(7.6)
1.76, d
(7.6)
40
50
1.09, d
(2.7)
0.86, t
(7.2)
1.01, d
(7.2)
Ac
1.94, s
1.72, s
1.77, s
C-12
200
Ac
1.96, s
Ac
1.96, s
Ac
1.85, s
Ac
1.82, s
a
b
c
Measured at 300 MHz.
Measured at 400 MHz.
In CDCl3.
d
In CD3OD.
respectively. The HMBC correlation from H-30 (dH 4.00) to the car-
bonyl at dC 171.2 suggested the presence of a hexanyl ester group.
The HMBC correlations from H-500 (dH 3.52 to C-100 (dC 100.4), and

S. Yao et al. / Phytochemistry 126 (2016) 47–58
49
the ROESY correlations of H-100/OMe-300 and H-200/OMe-300 allowed
the sugar unit 6-deoxy-3-O-methyl-b-allopyranosyl, which was
then placed at C-40 of the hexanyl ester group inferred from the
HMBC correlation of H-100/C-40 (dC 80.8). Consequently, the sub-
stituent attached to C-3 was identified as 4-O-(6-deoxy-3-O-
methyl-b-allopyranosyl)-5-hydroxy-3-methoxyhexanoyl.
To determine the stereochemistry of the side-chain, compound
2 was treated with 1 N aqueous NaOH solution in THF (Scheme 1)
and the corresponding sodium carboxylate was released into the
water phase. After acidification, d-lactone (6), the hydrolysate of
the side-chain, was afforded. Its overall structure was identified
on the basis of 1D NMR and ESI-MS data. The absolute configura-
tion of 6 was determined by an X-ray crystallographic analysis
Table 2
13C NMR spectroscopic data of the aglycone moieties of compounds 1–5 (d in ppm).
Position
1a,c
2b,c
3a,c
4a,c
5a,d
1
2
3
4
5
6
7
8
37.5, t
31.1, t
70.4, d
38.3, t
44.0, d
26.6, t
31.7, t
66.8, s
51.0, d
38.9, s
68.5, d
75.1, d
45.7, s
71.3, s
26.5, t
24.9, t
60.0, d
16.7, q
12.7, q
210.6, s
29.8, q
iBu
37.2, t
27.1, t
73.2, d
34.1, t
43.8, d
26.5, t
31.6, t
66.7, s
50.9, d
38.9, s
68.5, d
75.1, d
45.8, s
71.4, s
26.5, t
24.9, t
60.1, d
16.7, q
12.7, q
210.7, s
30.0, q
iBu
36.7, t
27.0, t
73.1, d
33.9, t
43.7, d
26.3, t
31.4, t
66.4, s
50.8, d
38.5, s
68.4, d
74.8, d
45.5, s
71.2, s
26.3, t
24.7, t
59.6, d
16.4, q
12.4, q
210.6, s
29.8, q
Tig
37.1, t
27.0, t
73.1, d
34.0, t
43.7, d
26.4, t
31.5, t
66.6, s
50.8, d
38.8, s
68.3, d
75.0, d
45.7, s
71.3, s
26.4, t
24.8, t
60.0, d
16.7, q
12.6, q
210.6, s
29.6, q
mBu
38.7, t
28.8, t
79.3, d
35.8, t
45.6, d
28.2, t
33.3, t
68.6, s
53.1, d
40.5, s
70.6, d
76.4, d
47.5, s
73.5, s
28.2, t
26.4, t
61.3, d
17.4, q
13.6, q
213.6, s
30.8, q
Tig
9
10
11
12
13
14
15
16
17
18
19
20
21
C-11
10
experiment using Mo K
derivative (6a) (Scheme 2, Fig. 2). Compound 6 was finally deduced
to be 6-deoxy-3-O-methyl-b- -allopyranosyl-(1?4)- -oleandronic
a radiation of its bis(4-bromobenzoyl)
D
D
acid d-lactone, identical to pachybionic acid d-lactone, the disac-
charide lactone previously isolated from M. tenacissima. It has been
reported that the stereochemistry of sodium cymaronate is identi-
cal to that of cymaronic acid d-lactone (Yoshikawa et al., 2000).
Therefore, the absolute configurations in the side-chain of 2 were
determined and the full structure was established as being 12b-
176.0, s
34.5, d
18.4, q
18.9, q
176.1, s
34.5, d
18.4, q
18.9, q
167.0, s
128.3, s
138.0, d
14.2, q
11.7, q
Ac
175.6, s
41.2, d
26.1, t
11.7, q
15.2, q
Ac
169.2, s
130.2, s
140.4, d
15.0, q
12.6, q
Ac
20
O-acetyl-3-O-(6-deoxy-3-O-methyl-b-D-allopyranosyl-(1?4)-D-ole-
30
andronyl)-11 -O-isobutyryltenacigenin B.
a
40
The molecular formulae of compounds 3 and 4 were deter-
mined by HR-ESI-MS to be C42H64O15 and C42H66O15, respectively.
The NMR data of compounds 3 and 4 (Tables 1–3) indicated that
both structures were quite similar to that of compound 2, except
that a tigloyl and a 2-methylbutyryl moiety were present in com-
pounds 3 and 4, respectively, rather than an isobutyryl group as in
2. The location of these acyl groups in compounds 3 and 4 was
determined from the HMBC correlations from dH 5.38 (H-11 of
the aglycone) to dC 167.0 (C@O of the tigloyl unit) for 3 and from
50
C-12
100
200
Ac
Ac
170.6, s
20.8, q
170.6, s
20.8, q
170.4, s
20.2, q
170.6, s
20.8, q
172.7, s
21.0, q
a
b
c
Measured at 100 MHz.
Measured at 125 MHz.
In CDCl3.
d
In CD3OD.
Table 3
1H and 13C NMR spectroscopic data of the disaccharide moieties of compounds 2–5 (d in ppm, J in Hz).
2a
3b
4b
5c
dH
dC
dH
dC
dH
dC
dH
dC
Ole
1
2a
2b
3
4
5
6
7
171.2, s
35.0, t
171.0, s
34.9, t
171.2, s
35.0, t
172.7, s
37.6, t
2.54, dd (16.6, 6.5)
2.78, dd (16.6, 5.8)
4.00, m
3.65, m
3.98, m
2.51, dd (16.5, 6.8)
2.74, dd (16.5, 5.4)
3.96, m
3.62, m
3.94, m
2.52, dd (16.6, 6.5)
2.77, dd (16.6, 5.8)
3.99, m
3.63, m
3.96, m
2.53, dd (16.0, 8.5)
2.73, dd (17.1, 5.1)
4.02, m
3.51, dd (6.6, 2.6)
3.88, m
78.3, d
80.8, d
67.6, d
19.4, q
58.0, q
78.1, d
80.8, d
67.4, d
19.1, q
57.8, q
78.1, d
80.9, d
67.6, d
19.3, q
58.0, q
171.2, s
74.8, d
84.8, d
68.8, d
20.6, q
1.26, d (6.4)
3.41, s
1.23, d (6.5)
3.39, s
1.24, d (6.5)
3.39, s
1.27, d (5.4)
3.38, s
Allo
1
2
3
4
5
6
7
4.62, d (8.0)
3.51, m
3.77, t (2.9)
3.20, m
3.52, m
1.23, d (6.2)
3.65, s
100.4, d
72.6, d
80.6, d
72.7, d
71.7, d
17.7, q
62.1, q
4.60, d (8.0)
3.48, m
3.74, t (3.0)
3.18, dd (9.7, 3.4)
3.50, m
100.3, d
72.3, d
80.5, d
72.5, d
70.9, d
17.5, q
61.9, q
4.60, d (8.1)
3.48, m
3.74, t (3.0)
3.17, m
3.50, m
1.20, d (6.2)
3.63, s
100.5, d
72.5, d
80.7, d
72.6, d
70.9, d
17.6, q
62.1, q
4.59, d (7.9)
3.35, m
3.94, m
3.32, m
3.77, m
103.1, d
73.5, d
83.4, d
84.3, d
70.5, d
18.6, q
62.5, q
1.20, d (6.3)
3.65, s
1.25, d (6.1)
3.57, s
Glu
1
2
3
4
5
6a
6b
4.34, d (7.6)
3.16, m
3.34, m
3.28, m
3.23, m
106.7, d
76.0, d
78.4, d
78.5, d
71.3, d
63.5, t
3.64, dd (11.9, 5.7)
3.88, m
a
b
c
300 MHz for 1H, 125 MHz for 13C, in CDCl3.
300 MHz for 1H, 100 MHz for 13C, in CDCl3.
300 MHz for 1H, 100 MHz for 13C, in CD3OD.

50
S. Yao et al. / Phytochemistry 126 (2016) 47–58
OAc
O
O
OAc
O
O
RO
RO
O
O
1. 1N NaOH, THF, rt
2. 1N HCl, 30
O
HO
O
+
O
OH
O
C
°
OH
O
O
HO
O
HO
O
OH
O
O
2 R=iBu
1
6
R=iBu
3a R=Tig
1a
3
4
R=Tig
R=mBu
R=mBu
Scheme 1. Hydrolysis of compounds 2–4.
O
O
O
O
O
O
O
O
O
O
O
Br
O
O
O
O
O
O
HO
O
Cl
O
O
MeO
O
Br
O
OH
O
Br
O
DMAP, CHCl , 35
C
O
°
3
O
Br
Br
6
6a
6a
Scheme 2. 4-Bromobenzoylation of compound 6.
Compound 5 had a molecular formula of C48H74O20 with 12
degrees of unsaturation, as determined by HR-ESI-MS. The IR spec-
trum exhibited absorption bands for hydroxy (3427 cm–1) and car-
bonyl (1736 and 1707 cm–1) groups. Besides three methyl group
signals at dH 1.05 (s), 1.09 (s), and 2.13 (s), and two oxygenated
methine resonances at dH 5.01 (d, J = 10.2 Hz) and 5.36 (t,
J = 10.2 Hz), the 1H NMR spectrum of 5 displayed an additional
methyl resonances at dH 1.82 (s) and a set of signals at dH 1.77
(s), 1.76 (d, J = 7.6 Hz), and 6.80 (q, J = 7.6 Hz). Interpretation of
HMBC correlations of these signals supported the existence of an
acetyl and a tigloyl unit. Two anomeric proton signals at dH 4.59
(d, J = 7.9 Hz) and 4.34 (d, J = 7.6 Hz), together with the correspond-
ing carbon resonances at dC 103.1 and 106.7, indicated the pres-
ence of two sugar moieties with b-linkages. Upon cellulase
hydrolysis, compound 5 afforded a sugar unit and a steroidal resi-
due. The steroidal residue was deduced to be identical to com-
pound 3 on the basis of the same retention times on HPLC,
similar NMR data and specific rotation values. The sugar product
was identified as
ple on HPLC, as well as from the specific rotation. The location of
the b-
-glucose, connected to C-400 of the 6-deoxy-3-O-methyl-b-
D-glucose by comparison with an authentic sam-
D
D
-allopyranosyl fragment in the side-chain, was determined from
the HMBC correlation from H-1000 (dH 4.34) of the glucose unit to
C-400 (dC 84.3) of the allose moiety. Consequently, the structure of
compound 5 was established as 12b-O-acetyl-3-O-(b-
nosyl-(1?4)-6-deoxy-3-O-methyl-b- -allopyranosyl-(1?4)-
andronyl)-11 -O-tigloyltenacigenin B.
D
-glucopyra-
D
D-ole-
a
2.2. Biological properties of isolated compounds with an open-chain
sugar structure versus that with a closed-ring sugar moiety
Fig. 2. Perspective ORTEP drawing of compound 6a.
2.2.1. Reversal of chemoresistance due to different mechanisms
Compounds 3–5 and tenacissoside A were tested for their rever-
sal of chemoresistance due to different mechanisms, using four
pairs of parental and drug-resistant cell lines with defined overex-
pression of the major MDR transporters (P-gp, MRP1, ABCG2, or
MRP2). In all of these cell lines, there was no effect on cell prolifer-
ation for each compound tested alone at concentrations below
dH 5.33 (H-11 of the aglycone) to dC 175.6 (C@O of the 2-methylbu-
tyryl group) for 4. Thus, the aglycones of 3 and 4 were ascertained
to be 12b-O-acetyl-11
a
-O-tigloyltenacigenin B (3a) and 12b-O-
(1a), respectively.
acetyl-11 -O-2-methylbutyryltenacigenin
a
B
The side-chain of compounds 3 and 4 was identical to that of com-
pound 2, which was confirmed by NMR analysis and the specific
optical rotations of their hydrolysate products, after the treatment
as described for compound 2. The location of the side-chain for
both was deduced to be at C-3 by the HMBC correlations from H-
3 (dH 4.66) to C-20 (dC 34.9) of the oleandronyl unit in 3, and from
H-3 (dH 4.66) to C-20 (dC 35.0) in 4. Therefore, the structures of 3
and 4 were established as shown in Fig. 1.
50
25, or 50
l
M. Therefore, the compounds were evaluated at 0.4, 2, 10,
M when combined with other anticancer drugs in the
l
MDR reversal assay. As illustrated in Table 4, all three compounds
(3–5) with an open-chain sugar structure were found to reverse P-
gp-mediated MDR at concentrations as low as 0.4
lM. Compounds
3 and 4 (used at 2–50 M) were also found to remarkably sensitize
l

S. Yao et al. / Phytochemistry 126 (2016) 47–58
51
Table 4
Table 5
Reversal of P-gp-mediated MDR by compounds 3–5, but not tenacissoside A.
Reversal of MRP1-mediated MDR by compounds 3 and 4, but not compound 5 and
tenacissoside A.
IC50 s.d. (
l
M) (fold-resistance)a
IC50 s.d. (
MCF-7
l
M) (fold-resistance)a
SW620
SW620 Ad300 (P-gp
overexpressing)
MCF-7 VP (MRP1
overexpressing)
Doxorubicin alone
+ PSC833 (0.4 M)
0.17 0.05
0.10 0.06
(1.0)
(0.6)
28.22 0.52
0.22 0.03⁄⁄
21.30 0.90⁄
16.95 1.96⁄⁄
4.95 0.39⁄⁄
(166)
(1.3)
l
Doxorubicin alone
+ MK571 (40 M)
0.22 0.06
0.22 0.07
(1.0)
(1.0)
2.48 0.33
0.18 0.04⁄
1.75 0.16⁄
1.14 0.14⁄
(11)
(0.8)
l
+ Compound 3 (0.4
+ Compound 3 (2
l
M)
M)
M)
0.03 0.05
0.20 0.08
0.15 0.05
(0.2)
(1.2)
(0.9)
(125)
(100)
(29)
l
+ Compound 3 (2
l
M)
0.22 0.06
0.22 0.09
(1.0)
(1.0)
(8.0)
(5.2)
+ Compound 3 (10
l
+ Compound 3 (10
l
M)
+ Compound 4 (0.4
l
M)
M)
M)
0.07 0.05
0.09 0.06
0.10 0.08
(0.4)
(0.5)
(0.6)
3.93 0.06⁄⁄
3.37 0.06⁄⁄
0.67 0.07⁄⁄
20.46 1.76⁄
15.95 2.15⁄
3.49 0.35⁄
(23)
(20)
(3.9)
+ Compound 4 (2
+ Compound 4 (10
l
l
M)
M)
0.90 0.09
0.15 0.16
(4.1)
(0.7)
1.03 0.06⁄⁄
1.05 0.03⁄⁄
(4.7)
(4.8)
+ Compound 4 (2
l
+ Compound 4 (10
l
+ Compound 5 (10
+ Compound 5 (50
l
l
M)
M)
0.18 0.02
0.21 0.11
(0.8)
(1.0)
2.31 0.41
2.26 0.34
(11)
(10)
+ Compound 5 (0.4
l
M)
M)
M)
0.09 0.07
0.22 0.06
0.19 0.09
(0.5)
(1.3)
(1.1)
(120)
(94)
(21)
+ Compound 5 (2
l
+ Tenacissoside A (10
+ Tenacissoside A (50
l
l
M)
M)
0.11 0.04
0.25 0.13
(0.5)
(1.1)
2.58 0.48
2.36 0.41
(12)
(11)
+ Compound 5 (10
l
+ Tenacissoside A (10
+ Tenacissoside A (50
l
l
M)
M)
0.27 0.22
0.21 0.11
(1.6)
(1.2)
27.56 2.15
22.41 1.76
(162)
(132)
*p < 0.05, **p < 0.01, compared with doxorubicin alone in MRP1-overexpressing
resistant MCF-7 VP cells.
*p < 0.05, **p < 0.01, compared with doxorubicin alone in P-gp-overexpressing
Fold-resistance was calculated by the IC50 under the designated treatment
a
resistant SW620 Ad300 cells.
condition/IC50 of doxorubicin alone in parental MCF-7 cells.
a
Fold-resistance was calculated by the IC50 under the designated treatment
condition/IC50 of doxorubicin alone in parental SW620 cells.
Table 6
Reversal of (A) ABCG2- and (B) MRP2-mediated MDR by compounds 3 and 4, but not
compound 5 and tenacissoside A.
resistant cancer cells due to MRP1, ABCG2, or MRP2 overexpres-
sion, whereas compound 5 (with an open-chain sugar structure)
and tenacissoside A (with a closed-ring sugar structure) were
devoid of this effect (Tables 5 and 6). Since platinum (Pt)-based
anticancer drugs are widely used in combination chemotherapy,
but their efficacy is adversely affected by drug resistance, the pos-
sible reversal of Pt drug resistance by these compounds was also
(A)
IC50 s.d. (
l
M) (fold-resistance)a
MCF-7
MCF-7 FLV1000
(ABCG2
overexpressing)
Mitoxantrone alone
0.0052 0.0021 (1.0)
1.81 0.32 (348)
+ FTC (10
lM)
0.0047 0.0024 (0.9) 0.029 0.005⁄ (5.6)
tested. Compounds 3 and 4 (at concentrations as low as 10 lM),
+ Compound 3 (2
+ Compound 3 (10
+ Compound 3 (20
l
l
l
M)
M)
M)
0.0047 0.0019 (0.9)
0.0049 0.0023 (0.9)
0.0032 0.0022 (0.6)
1.74 0.42 (335)
0.53 0.06⁄ (102)
0.34 0.09⁄ (65)
0.51 0.10⁄ (98)
0.30 0.15⁄ (58)
with an open-chain sugar structure, were also found to reverse
Pt resistance in a cisplatin-selected resistant esophageal cancer cell
line HKESC-1 cisR and an MRP2-overexpressed HEK293 cell line
(Table 7). The effect was specific because cisplatin cytotoxicity
was not affected in the parental HKESC-1 and pcDNA3 (backbone
vector)-transfected HEK293 cells (Table 7). Compound 5 and
tenacissoside A did not demonstrate any Pt resistance reversal
activity. The chemoresistance reversal activity of the tested com-
pounds is summarized in Table 8. The mechanisms for reversal of
Pt resistance by these compounds were further evaluated
(Section 2.3).
+ Compound 4 (2
+ Compound 4 (10
l
l
M)
M)
0.0108 0.0647 (2.1)
0.0086 0.0982 (1.7)
+ Compound 5 (10
+ Compound 5 (50
l
l
M)
M)
0.0123 0.0041 (2.4)
0.0086 0.0054 (1.7)
1.69 0.32 (325)
1.59 0.45 (306)
+ Tenacissoside A (10
+ Tenacissoside A (50
l
l
M) 0.0052 0.0729 (1.0)
M) 0.0086 0.0523 (1.7)
1.54 0.67 (296)
1.93 0.41 (371)
aFold-resistance was calculated by the IC50 under the designated treatment
condition/IC50 of mitoxantrone alone in parental MCF-7 cells.
*p < 0.05, compared with mitoxantrone alone in ABCG2-overexpressing
resistant MCF-7 FLV1000 cells.
(B)
IC50 s.d. (
HEK293 pcDNA3
l
M) (fold-resistance)a
HEK293 MRP2 stably
transfected
2.3. Mechanisms for Pt resistance circumvention by compounds 3 and 4
Etoposide
+ Benzbromarone (50
0.465 0.042
0.469 0.033
(1.0)
(1.0)
4.445 0.679
1.370 0.131⁄
(9.6)
(2.9)
2.3.1. Compounds 3 and 4 inhibit MRP2 efflux activity and enhance
total cellular accumulation of a Pt anticancer drug
lM)
+ Compound 3 (10
+ Compound 3 (20
l
l
M)
M)
0.507 0.077
0.522 0.061
(1.1)
(1.1)
4.444 0.595
(9.6)
(4.8)
MRP2 is a known efflux transporter for Pt-based anticancer
drugs and its overexpression has been reported to cause Pt resis-
tance (Taniquchi et al., 1996). The inhibition of MRP2 transport
activity, as indicated by the retention of an MRP2-specific fluores-
cent probe substrate (CFDA), was assessed as the difference in
2.244 0.395⁄⁄
+ Compound 4 (10
+ Compound 4 (20
l
l
M)
M)
0.481 0.069
0.428 0.060
(1.0)
(0.9)
3.433 0.557⁄⁄
2.329 0.431⁄⁄
(7.4)
(5.0)
+ Compound 5 (50
l
M)
0.448 0.086
0.448 0.082
(1.0)
(1.0)
4.498 0.690
4.511 0.749
(9.7)
(9.7)
mean fluorescence intensity (
of the tested compounds. The specific and potent MRP2 inhibitor
benzbromarone (50 M) was used as a positive control for compar-
DMFI) in the presence and absence
+ Tenacissoside A (50
l
M)
aFold-resistance = IC50 of etoposide in various combinations/IC50 of etoposide
alone in HEK293 pcDNA3 cells.
l
*p < 0.05, **p < 0.01, compared with etoposide alone in MRP2-stably
transfected cells.
ison. Both compounds 3 and 4 were found to increase the retention
of CFDA fluorescence in the two MRP2-overexpressing cell lines,
indicating inhibition of MRP2 transport activity (Table 9). The
MRP2 inhibitory effect of 3 and 4 was specific because
DMFI was
negligible in HEK293 pcDNA3 and parental HKESC-1 cells with
minimum MRP2 expression (Table 9).
The cellular accumulation of cisplatin, in the absence and pres-
ence of tested compounds, was also assessed. Consistent with the
fact that MRP2 is an efflux pump for Pt-based anticancer drugs, cis-
be significantly lower in the MRP2-overexpressing HEK293 MRP2
and HKESC-1 cisR cells than in the corresponding parental
HEK293 pcDNA3 and HKESC-1 cells (Fig. 3). While co-administra-
tion of the tested compounds with cisplatin had no significant
effect on the total cellular Pt accumulation in the parental cells,
the inclusion of either compound 3 or 4 was found to enhance Pt
platin accumulation after incubation for 4 h (200 lM) was found to

52
S. Yao et al. / Phytochemistry 126 (2016) 47–58
Table 7
Table 9
Circumvention of cisplatin resistance, by compounds 3 and 4, in (A) MRP2-stably
transfected HEK293 cells and (B) cisplatin-selected resistant human esophageal
cancer HKESC-1 cisR cells.
Flow cytometric analysis of MRP2 transport activity in MRP2-stably transfected
HEK293 cells and cisplatin-resistant HKESC-1 cisR cells.
(A)
D
MFIa
(A)
IC50 s.d. (l
M) (fold-resistance)a
pcDNA3 stably transfected MRP2 stably transfected
HEK293 pcDNA3
HEK293 MRP2 stably
transfected
HEK293
–
HEK293
–
CFDA only
Cisplatin
+Benzbromarone (50
1.940 0.537
2.017 0.286
(1.0)
(1.0)
22.55 2.18
5.78 0.88⁄⁄
17.05 1.62⁄⁄
12.86 0.35⁄
(12)
(3.0)
CFDA
l
M)
+ Benzbromaroneb
78 23
65 31
4380 510*
232 98*
+Compound 3 (10
+Compound 3 (20
l
l
M)
M)
1.897 0.761
2.020 0.524
(1.0)
(1.0)
(8.8)
(6.6)
+ Compound 3
(50
+Compound 4
(50 M)
+Compound 5
(50 M)
+Tenacissoside A
(50 M)
lM)
+Compound 4 (10
+Compound 4 (20
l
l
M)
M)
2.127 0.643
2.177 0.601
(1.1)
(1.1)
14.46 0.99⁄⁄
12.77 0.81⁄
(7.5)
(6.6)
69 19
73 30
55 32
466 34*
l
+Compound 5 (50
l
M)
1.863 0.782
1.793 0.563
(1.0)
(0.9)
23.94 3.23
24.30 2.37
(12)
(13)
26
7
+Tenacissoside A (50
lM)
l
aFold-resistance = IC50 of cisplatin in various combinations/IC50 of cisplatin
alone in HEK293 pcDNA3 cells.
45 29
l
*p < 0.05, **p < 0.01, compared with cisplatin alone in MRP2-stably transfected
cells.
(B)
D
MFIa
HKESC-1 Parental
–
HKESC-1 cisR
–
(B)
IC50 s.d. (
HKESC-1 parental
l
M) (fold-resistance)a
CFDA only
HKESC-1 cisR
CFDA
Cisplatin alone
+Benzbromarone (50
0.567 0.102
0.507 0.032
(1.0)
(0.9)
4.907 0.337
2.013 0.052⁄
3.064 0.135⁄
2.406 0.141⁄
2.533 0.080⁄
1.946 0.098⁄
(8.7)
(3.6)
+ Benzbromaroneb
15
22
19
5
9
3
1060 76⁄
204 39⁄
305 93⁄
l
M)
+ Compound 3 (50
+ Compound 4 (50
+ Compound 5 (50
l
l
l
M)
+Compound 3 (10
+Compound 3 (20
l
l
M)
M)
0.615 0.089
0.527 0.062
(1.1)
(0.9)
(5.4)
(4.2)
M)
M)
+Compound 4 (10
+Compound 4 (20
l
l
M)
M)
0.520 0.091
0.594 0.061
(0.9)
(1.0)
(4.5)
(3.4)
26 13
15
14
23
6
8
+ Tenacissoside A (50
l
M)
9
+ Compound 5 (50
l
M)
0.590 0.056
0.530 0.015
(1.0)
(0.9)
5.225 0.224
4.796 0.184
(9.2)
(8.5)
a
D
MFI = difference in intracellular fluorescence intensity in the presence and
+ Tenacissoside A (50
l
M)
absence of the tested inhibitor.
Benzbromarone (50 lM), a potent and specific MRP2 inhibitor, was used as a
positive control.
aFold-resistance = IC50 of cisplatin in various combinations/IC50 of cisplatin
alone in HKESC-1 parental cells.
b
*p < 0.01, compared with cisplatin alone in HKESC-1 parental cells.
*
Data are reported as mean s.d. from three independent experiments; p < 0.05,
different from CFDA alone.
Table 8
2.3.3. Compounds 3 and 4 restore cell cycle arrest caused by cisplatin
in resistant cells
Summary of chemoresistance reversal activity of compounds 3–5 and tenacissoside A.
Circumvention of different mechanisms of drug
resistance
Cisplatin is known to mediate its cytotoxic effect through arrest-
ing cancer cells at the G2/M phase. Consistent with the potent cyto-
toxic effect of cisplatin in the sensitive parental HKESC-1 cells,
cisplatin was found to cause remarkable G2/M arrest (Fig. 5). How-
ever, the resistant HKESC-1 cisR cells did not exhibit appreciable cell
cycle arrest upon treatment with cisplatin, which is believed to
cause the observed cisplatin resistance. Interestingly, the concomi-
tant treatment of compounds 3 or 4 with cisplatin was found to par-
tially restore the G2/M arrest in the resistant HKESC-1 cisR cells
(Fig. 5). In contrast, compound 5 and tenacissoside A did not impose
any significant changes on the cell cycle distribution in the resistant
cells. It is noteworthy that none of the tested compounds alone
induced cell cycle arrest (data not shown).
Steroids with
P-gp
ABCG2
MRP1
MRP2
Pt resistance
Open-chain sugar
Compound 3
Compound 4
Compound 5
p
p
p
p
p
p
p
p
p
p
p
ꢀ
ꢀ
ꢀ
ꢀ
Closed-ring sugar
Tenacissoside A
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
accumulation in a concentration-dependent manner in the MRP2-
stably transfected HEK293 MRP2 cells and cisplatin-resistant
HKESC-1 cisR cells (Fig. 3). In contrast, compound 5 and tenacis-
soside A had no effect on Pt accumulation (Fig. 3).
3. Conclusions
2.3.2. Pt resistance reversal is associated with upregulation of p21 by
compounds 3 and 4
Compounds 3–5 represent the first examples of naturally occur-
ring polyoxypregnane glycosides bearing an open-chain sugar moi-
ety. It is assumed that there might be two possible pathways for
the formation of this type of compound. Taking compound 3 as
an example (Scheme 3), its formation may involve enzymatic oxi-
dation of the first anomeric carbon of its precursor (see path I), or
probably is associated with nucleophilic attack of compound 6 by
1a, which leads to lactone cleavage and generation of the pachy-
bionic acid ester moiety at the same time (see path II). The disac-
charide lactone 6, pachybionic acid d-lactone, has been obtained
for the first time from nature, although it has been synthesized
previously from pachybiose (Abisch et al., 1959; Jaeggi et al.,
1967). Since 1977, there have been several publications outlining
the discovery of oligosaccharides, in which the first sugar unit is
The DNA repair gene (ERCC1) and cell cycle regulatory gene
(p21), widely reported to manifest the Pt resistance phenotype,
were also evaluated in the presence of the tested compounds.
Compared with the parental HKESC-1 cells, higher ERCC1
(2.28 0.23-fold) but lower p21 (0.15 0.01-fold) expression was
noted in the cisplatin-resistant HEKSC-1 cisR cells (Fig. 4), which
presumably contribute to cisplatin resistance. In parental HKESC-
1 cells, the tested compounds did not change ERCC1 and p21 levels.
However, in the cisplatin-resistant HKESC-1 cisR cells, compounds
3 and 4 (but not compound 5 or tenacissoside A) caused a signifi-
cant increase in p21 expression (Fig. 4, lower panel, p < 0.05)
whereas they had no effect on the ERCC1 level (Fig. 4, upper panel).

S. Yao et al. / Phytochemistry 126 (2016) 47–58
53
Fig. 4. Quantitative real-time PCR analysis of two Pt resistance genes (ERCC1 and
p21) in parental HKESC-1 and cisplatin-selected drug-resistant HKESC-1 cisR cells.
Cells were treated with cisplatin (10 lM) or the various tested compounds (25 lM)
for 24 h before harvest for PCR analysis. Data are presented as mean s.d. from
three independent experiments. *p < 0.05, **p < 0.01, versus no treatment in
resistant cells.
Fig. 3. Total cellular accumulation of cisplatin, in the presence or absence of tested
compounds, in parental and Pt-resistant sublines. Parental (HEK293 pcDNA3 or
HKESC-1) and cisplatin-resistant (HEK293 MRP2 stably transfected or HKESC-1
tenacissoside A with a closed-ring sugar structure was devoid of
this chemoresistance reversal effect. The mechanisms for Pt resis-
tance reversal were further evaluated in detail. Overexpression of
the MRP2 efflux transporter is known to cause Pt resistance by
actively pumping out glutathione S–Pt conjugates. The data pre-
sented indicate that compounds 3 and 4 inhibit MRP2 functional
activity (Table 9) and enhance total cellular accumulation of Pt
drugs (Fig. 3). It should be noted that compounds 3 and 4 do not
alter the MRP2 expression level (data not shown) but work by
inhibiting MRP2 functional activity. Furthermore, compounds 3
and 4 were also shown to upregulate p21 expression in cisplatin-
resistant HKESC-1 cisR cells (Fig. 4, lower panel) to partially restore
the G2/M cell cycle arrest mediated by cisplatin treatment (Fig. 5).
Due to the wide spectrum of chemoresistance reversal by com-
pounds 3 and 4 and their multifunctional activities in circumvent-
ing drug resistance, more studies are warranted to further develop
these compounds as useful adjuvants to enhance the anticancer
activity of combination chemotherapy.
cisR) cells were incubated with cisplatin (200
lM), without or with 25 lM (+) or
50 M (++) of the designated compounds 3–5 and tenacissoside A (T.A), for 4 h.
l
Total cellular accumulation was assessed by atomic absorption analysis after
harvesting the cells. The protein content of the samples was measured for
normalization purposes. Data are presented as mean s.d. from three independent
measurements. *p < 0.05, **p < 0.01, versus cisplatin alone in resistant cells.
oxidized to a lactone, from steroid-rich plants (Feng et al., 2008;
Kawanishi et al., 1977; Wang et al., 2010; Yoshikawa et al.,
2000). These reports support the idea that the hemi-acetal could
be oxidized enzymatically to a lactone in plants. However, when
the glycosidic bond was already formed, the oxidation could not
be achieved, suggesting that path I is not the likely pathway for
the formation of compound 3.
The effects of compounds 3–5 with an open-chain sugar moiety
on the circumvention of chemoresistance were investigated.
Tenacissoside A, another compound having a highly similar struc-
ture but with a closed-ring sugar moiety, was used for comparison.
Of the tested compounds, compounds 3 and 4 were apparently
able to circumvent MDR caused by overexpression of the four
major MDR transporters (P-gp, MRP1, ABCG2, and MRP2). Pt-based
anticancer drugs are indispensible in most combination
chemotherapeutic regimens. However, resistance to Pt drugs,
which is caused by multifactorial mechanisms, represents a major
hurdle to successful chemotherapy. Importantly, compounds 3 and
4. Experimental
4.1. General experimental procedures
Optical rotations were measured on a Perkin–Elmer M341
polarimeter, whereas IR spectra were obtained on a Perkin–Elmer
577 spectrophotometer using KBr disks. NMR spectra were
4
were also found to reverse Pt resistance. In contrast,

54
S. Yao et al. / Phytochemistry 126 (2016) 47–58
Fig. 5. Cell cycle analysis in parental and cisplatin-resistant cells after cisplatin treatment in the presence or absence of tested compounds. Upper panel: Parental HKESC-1
and cisplatin-resistant HKESC-1 cisR cells were exposed to cisplatin (10 M) alone, or its combination with various tested compounds (25 M), for 24 h before harvest for cell
l
l
cycle analysis. A representative set of data from three independent experiments is shown. Lower panel: Summary of cell cycle analysis results from three independent
experiments. Data are presented in histograms as mean s.d.; *p < 0.01, versus no treatment in cisplatin-resistant HKESC-1 cisR cells. The various tested compounds alone did
not cause any notable change to the cell cycle distribution (data not shown).
O
OAc
O
O
O
O
OAc
O
O
O
O
O
O
O
I
HO
O
OH
Oxidation
O
OH
O
O
O
HO
O
OH
O
O
O
OAc
O
O
3
O
II
HO
1a
O
O
O
HO
O
O
OH
O
6
Scheme 3. Plausible pathways for the formation of compound 3.

S. Yao et al. / Phytochemistry 126 (2016) 47–58
55
recorded on Varian Mercury-plus 300, Varian Mercury-plus 400,
4.3.1. 12b-O-Acetyl-11a-O-isobutyryltenacigenin B (1)
20
Bruker Avance III 400, Bruker Avance III 500, and Varian Inova
600 NMR spectrometers. Chemical shift (d) values are given in
ppm with TMS as internal standard, and coupling constants (J)
are in Hz. LR-ESI-MS were recorded on a Waters 3100 Mass Detec-
tor and HR-ESI-MS were recorded on a Waters Q-TOF UltimaTM Glo-
bal Mass Spectrometer. Silica gel was used for flash
chromatography and was obtained from Qingdao Marine Chemical
White powder; [
a
]
D
+3.0 (c 0.4, MeOH); IR (KBr) mmax 3435,
2972, 2931, 2864, 1736, 1707, 1632, 1470, 1387, 1365, 1246,
1190, 1157, 1030, 935 cm–1; For 1H NMR; 13C NMR spectroscopic
data, see Tables 1 and 2; ESI-MS m/z 499.1 [M+Na]+; HR-ESI-MS
m/z 499.2642 ([M+Na]+, calculated for C27H40O7Na, 499.2672).
4.3.2. 12b-O-Acetyl-3-O-(6-deoxy-3-O-methyl-b-
(1?4)- -oleandronyl)-11 -O-isobutyryltenacigenin B (2)
White powder; [
2974, 2935, 1738, 1707, 1649, 1452, 1367, 1248, 1163, 1082,
D-allopyranosyl-
Industrials. MCI gel CHP20P (75–150
l
m, Mitsubishi Chemical
D
a
22
Industries, Japan) and Sephadex LH-20 (Pharmacia Biotech AB,
Uppsala, Sweden) were used for column chromatography (CC).
TLC was carried out on precoated silica gel GF254 plates (Yantai
Chemical Industrials, China) and the TLC spots were viewed at
254 nm and visualized with 5% H2SO4 in EtOH containing
10 mg/ml vanillin. Analytical HPLC was performed on a Waters
2695 instrument with a 2998 PAD (photodiode array detector)
and coupled with an Alltech ELSD 2424 detector [SunfireTM RP-C18
a]
+3.3 (c 0.4, CHCl3); IR (KBr) mmax 3450,
D
1032, 920, 754 cm–1 1H NMR see Tables 1 and 3; 13C NMR see
;
Tables 2 and 3; ESI-MS m/z 819.5 [M+Na]+, 796.9 [M+H]+, 637.0
[M+HꢁAllo]+, 548.9 [M+HꢁAlloꢁisobutyric acid]+, 459.0 [M
+HꢁAlloꢁOle acid]+, 370.9 [M+HꢁAlloꢁOle acidꢁisobutyric acid]+,
310.9 [M+HꢁAlloꢁOle acidꢁisobutyric acidꢁCH3COOH]+; HR-ESI-
MS m/z 819.4136 ([M+Na]+, calculated for C41H64O15Na, 819.4143).
column (100 ꢀ 4.6 mm, 3.5
lm); CH3CN:H2O eluant system; flow
1 ml/min]. X-ray crystallographic analysis was carried out on a
4.3.3. 12b-O-Acetyl-3-O-(6-deoxy-3-O-methyl-b-
(1?4)- -oleandronyl)-11 -O-tigloyltenacigenin B (3)
White powder; [
2974, 2935, 1738, 1707, 1649, 1452, 1369, 1269, 1165, 1130,
1082, 1030, 920, 754 cm–1 1H NMR see Tables 1 and 3; 13C NMR
D-allopyranosyl-
Bruker Smart Apex CCD diffractometer with graphite-monochro-
D
a
23
mated Mo K
a
radiation, k = 0.71073 Å. Tenacissoside A was isolated
a]
+0.4 (c 0.4, CHCl3); IR (KBr) mmax 3446,
D
by our group from the same plant.
;
4.2. Plant material
see Tables 2 and 3; ESI-MS m/z 831.5 [M+Na]+, 809.7 [M+H]+,
649.6 [M+HꢁAllo]+, 549.4 [M+HꢁAlloꢁ2-methylbut-2-enoic
acid]+, 471.4 [M+HꢁAlloꢁOle acid]+, 371.2 [M+HꢁAlloꢁOle
acidꢁ2-methylbut-2-enoic acid]+, 311.2 [M+HꢁAlloꢁOle acidꢁ2-
methylbut-2-enoic acidꢁCH3COOH]+; HR-ESI-MS m/z 831.4127
([M+Na]+, calculated for C42H64O15Na, 831.4143).
Stems of M. tenacissima were collected in Simao County in
Yunnan Province, China and identified by Professor Jingui Shen,
from the Shanghai Institute of Materia Medica. A voucher (No.
20091024) has been deposited at the Herbarium of the Shanghai
Institute of Materia Medica, Chinese Academy of Sciences.
4.3.4. 12b-O-Acetyl-3-O-(6-deoxy-3-O-methyl-b-
(1?4)- -oleandronyl)-11 -O-2-methylbutyryltenacigenin B (4)
White powder; [
2972, 2937, 1736, 1707, 1454, 1367, 1248, 1167, 1082, 1030,
918, 733 cm–1 1H NMR see Tables 1 and 3; 13C NMR see Tables 2
D-allopyranosyl-
4.3. Extraction and isolation
D
a
23
a]
+4.9 (c 0.4, CHCl3); IR (KBr) mmax 3446,
D
Dried and powdered canes of M. tenacissima (8 kg) were
extracted with EtOH–H2O (3 ꢀ 35 l, 95:5 v/v, 4 days each) at room
temperature. After evaporation of the solvent, the obtained residue
was suspended in H2O, and then extracted with EtOAc. The EtOAc
extract (540 g) was dissolved in MeOH (5 l) and extracted with pet-
roleum ether (3 ꢀ 2 l), affording MeOH and petroleum ether
extracts. The MeOH extract (480 g) was subjected to silica gel CC
eluted with EtOAc, EtOAc–MeOH (from 10:1 to 1:1), and MeOH
to yield six fractions (Frs. 1–6). Fr. 2 (20 g) was applied to a silica
gel column eluting with CH2Cl2–MeOH (from 50:1 to 10:1) to give
five subfractions, Frs. 2a–2e. Fr. 2c (8 g) was separated by MCI gel
CC eluting with MeOH–H2O (from 3:7 to 10:0) to give six subfrac-
tions, Frs. 2c1–2c6. Fr. 2c1 (400 mg) was purified over Sephadex
LH-20 (MeOH) to yield 6 (65 mg). Fr. 2c3 (530 mg) was firstly puri-
fied over Sephadex LH-20 (MeOH) and then by preparative HPLC
;
and 3; ESI-MS m/z 833.4 [M+Na]+, 811.6 [M+H]+, 651.6 [M
+HꢁAllo]+, 549.4 [M+HꢁAlloꢁ2-methylbutyric acid]+, 473.4 [M
+HꢁAlloꢁOle acid]+, 371.4 [M+HꢁAlloꢁOle acidꢁ2-methylbutyric
acid]+, 311.2 [M+HꢁAlloꢁOle acidꢁ2-methylbutyric acidꢁ
CH3COOH]+; HR-ESI-MS m/z 833.4271 ([M+Na]+, calculated for
C42H66O15Na, 833.4299).
4.3.5. 12b-O-Acetyl-3-O-(b-
methyl-b- -allopyranosyl-(1?4)-
tigloyltenacigenin B (5)
D
-glucopyranosyl-(1?4)-6-deoxy-3-O-
D
D
-oleandronyl)-11 -O-
a
20
White powder; [
2933, 1736, 1707, 1649, 1369, 1269, 1163, 1080, 1032, 922,
563 cm–1 1H NMR see Tables 1 and 3; 13C NMR see Tables 2 and
a]
ꢁ1 (c 0.24, MeOH); IR (KBr) mmax 3427,
D
;
[SunfireTM RP-C18 column (150 ꢀ 30 mm, 5
lm); CH3CN:H2O (from
3; ESI-MS m/z 993.4 [M+Na]+, 971.7 [M+H]+, 809.5 [M+HꢁGlu]+,
649.4 [M+HꢁGluꢁAllo]+, 549.3 [M+HꢁGluꢁAlloꢁ2-methylbut-2-
enoic acid]+, 471.3 [M+HꢁGluꢁAlloꢁOle acid]+, 371.3 [M
+HꢁGluꢁOle acidꢁ2-methylbut-2-enoic acid]+, 311.3 [M
+HꢁGluꢁOle acidꢁ2-methylbut-2-enoic acidꢁCH3COOH]+; HR-
ESI-MS m/z 993.4666 ([M+Na]+, calculated for C48H74O20Na,
993.4651).
40:60 to 60:40); flow 25 ml/min], giving compounds 2 (10 mg), 3
(60 mg), and 4 (200 mg). Fr. 3 (100 g) was subjected to silica gel
CC eluting with CH2Cl2–MeOH (from 20:1 to 5:1) to give eight sub-
fractions, Frs. 3a–3h. Fr. 3b (15 g) was separated by MCI gel CC
eluting with MeOH–H2O (from 2:8 to 10:0) to give seven subfrac-
tions, Frs. 3b1–3b7. Compound 5 (10 mg) was obtained from Fr.
3b2 (300 mg) by preparative HPLC [SunfireTM RP-C18 column
(150 ꢀ 30 mm, 5
lm); CH3CN:H2O (from 20:80 to 50:50); flow
4.4. Hydrolysis of compounds 2–4
25 ml/min]. The petroleum ether extract (60 g) was subjected to
silica gel CC eluting successively with petroleum ether–EtOAc
(from 5:1 to 0:1) and MeOH to yield four fractions (Frs. 1–4). Fr.
3 (10 g) was separated by CC over MCI gel eluting with MeOH–
H2O (from 5:5 to 10:0) to give six subfractions, Frs. 3a–3f. Fr. 3c
(340 mg) was firstly purified over Sephadex LH-20 (MeOH) and
then separated by preparative HPLC [SunfireTM RP-C18 column
Compound 2 (5.0 mg, 6.3
l
mol) was dissolved in THF (1 ml), to
which 1 N aq. NaOH (25 l, 25
l
lmol) was added. The reaction mix-
ture was stirred at ambient temperature for 2 h, then diluted with
H2O (1 ml) and extracted with EtOAc (3 ꢀ 4 ml). The combined
organic layers were washed with brine (3 ml), dried over Na2SO4,
and concentrated in vacuo to give 1 (2.7 mg, 5.6 lmol, 90% yield).
(150 ꢀ 19 mm, 5
l
m); CH3CN:H2O (from 30:70 to 55:45); flow
The aqueous phase was acidified with 1 N aq. HCl soln. to pH 1–
2, and stirred at 30 °C overnight, then extracted with EtOAc
15 ml/min] to give compound 1 (13 mg).

56
S. Yao et al. / Phytochemistry 126 (2016) 47–58
(3 ꢀ 5 ml). The combined organic layers were washed with brine
(5 ml), dried over Na2SO4, and concentrated in vacuo. The remain-
ing residue was purified over Sephadex LH-20 eluting with MeOH
4.7. Single crystal X-ray crystallography of 6a
Colorless needle, C28H30Br2O10, Mr 686.34, orthorhombic, crys-
tal size 0.100 ꢀ 0.120 ꢀ 0.180 mm, space group P2(1), a = 7.6121
(5) Å, b = 17.2576(10) Å, c = 21.9680(13) Å, V = 2885.9(3) Å3, Z = 4,
Dcalcd = 1.580 Mg/m3, F(000) = 1392, reflections collected 12,668,
to give 6 (1.8 mg, 5.6 lmol, 90% yield). The same hydrolysis proce-
dure was applied to compounds 3 (5 mg) and 4 (5 mg), and the
aqueous layer of their hydrolysates was acidified and then
extracted with EtOAc. The resulting organic phase was dried and
concentrated, and the residue was identified to contain pachy-
bionic acid d-lactone, as determined by comparing HPLC and
reflections unique 5054 (Rint = 0.0220), final R indices for I > 2
r(I),
R1 = 0.0210, wR2 = 0.0475, indices for all data R1 = 0.0259,
R
wR2 = 0.0657, completeness to 2h (55.00) 100.0%, maximum
transmission 0.7456, minimum transmission 0.6526, absolute
structure parameter 0.003(7). The structure was solved by direct
methods using the program SHELXS-97. The refinement method
was full-matrix least-squares on F2, and goodness-of-fit on F2
was 1.053. The X-ray diffraction data have also been deposited at
the Cambridge Crystallographic Data Centre (CCDC 1403538).
NMR data, as well as specific optical rotations.
22
Pachybionic acid d-lactone (6): White powder; C14H24O8; [a]
D
+56.4 (c 0.58, CHCl3); 1H NMR (600 MHz, CDCl3): d 4.70 (1H, d,
J = 8.0 Hz, Allo-H-1), 4.25 (1H, dq, J = 12.9, 6.4 Hz, Ole-H-5), 3.97
(1H, dd, J = 6.2, 3.8 Hz, Ole-H-3), 3.79 (1H, t, J = 3.1 Hz, Allo-H-3),
3.68 (3H, s, Allo-H-OMe), 3.63 (1H, dd, J = 7.5, 2.3 Hz, Ole-H-4),
3.60 (1H, m, Allo-H-5), 3.47 (1H, ddd, J = 8.1, 5.2, 3.1 Hz, Allo-
H-2), 3.39 (3H, s, Ole-H-OMe), 3.25 (1H, dq, J = 9.8, 3.3 Hz,
Allo-H-4), 2.75 (2H, d, J = 3.8 Hz, Ole-H-2), 2.35 (1H, d, J = 5.2 Hz,
Allo-H-2-OH), 2.18 (1H, d, J = 10.6 Hz, Allo-H-4-OH), 1.50 (3H, d,
J = 6.4 Hz, Ole-H-6), 1.28 (3H, d, J = 6.2 Hz, Allo-H-6); 13C NMR
(125 MHz, CDCl3): d 170.2 (Ole-C-1), 30.0 (Ole-C-2), 78.1 (Ole-
C-3), 81.1 (Ole-C-4), 76.0 (Ole-C-5), 19.1 (Ole-C-6), 57.0
(Ole-OMe), 102.3 (Allo-C-1), 72.7 (Allo-C-2), 80.8 (Allo-C-3), 73.0
(Allo-C-4), 71.1 (Allo-C-5), 17.9 (Allo-C-6), 62.5 (Allo-OMe);
ESI-MS m/z 321.1 [M+H]+.
4.8. Cell culture
Pairs of parental and drug-resistant cell sublines with overex-
pression of the three major MDR transporters (P-gp, ABCG2, and
MRP1) were generous gifts obtained from Dr. Susan Bates
(National Cancer Institute, Bethesda, MD, USA), including human
colon cancer SW620/its P-gp-overexpressing SW620 Ad300 sub-
line (Lai et al., 1991), human breast cancer MCF-7/its ABCG2-
overexpressing MCF-7 FLV1000 subline (Robey et al., 2001), and
MCF-7/its MRP1-overexpressing MCF-7 VP-16 subline (Schneider
et al., 1994). They have been fully characterized and were found
to overexpress only the designated MDR transporter. The resistant
cells were allowed to grow in a drug-free culture medium for more
than 2 weeks before assays. The resistance phenotype was stable
for at least 3 months in the drug-free medium. The human embry-
onic kidney cell line HEK293 and its pcDNA3.1- or MRP2-stably
transfected cell lines were also used. The expression vector of
MRP2 for stable transfection was purchased from GeneCopoeia
(Rockville, MD, USA). The HKESC-1 human esophageal squamous
cell carcinoma cell line was kindly provided by Prof. G. Srivastava
(Department of Pathology, The University of Hong Kong, Hong
Kong, China). Its cisplatin-resistant subline HKESC-1 cisR was
developed by growing the parental HKESC-1 cells in a gradually
increasing concentration of cisplatin over a period of 5 months
(To et al., 2012). It was found to overexpress both ABCG2 and
MRP2. The stably transfected HEK293 cells were maintained in
DMEM medium supplemented with 2 mg/ml G418. The other cell
lines were grown in DMEM (MCF-7, HKESC-1, and their resistant
sublines) or RPMI-1640 medium (SW620 and its resistant subline)
supplemented with 10% fetal bovine serum, 100 units/ml strepto-
mycin sulfate, and 100 units/ml penicillin G sulfate, and incubated
at 37 °C in 5% CO2.
4.5. Cellulase treatment of compound 5
Compound 5 (6 mg, 6.2 lmol) was treated with cellulase
(20 mg) (Sigma–Aldrich, 481 C1184) in a 0.1 M HOAc–NaOAc (pH
5.0) buffer soln. (1 ml) at 37 °C for 7 days. The reaction mixture
was then extracted with CHCl3. The organic phase was dried with
Na2SO4 and concentrated. The residue (4 mg) was identical to com-
pound 3, determined by comparing their HPLC and NMR data, as
well as specific optical rotations. The aqueous layer was filtered
and the filtrate was subjected to a ProElut PXC column (DIKMA,
1 g/20 ml) eluting with 0.1 M HOAc soln. (10 ml); the aqueous
solution was evaporated to afford
D-glucose (1 mg, 88% yield),
which was identified by HPLC analysis [Waters 2695 LC with a
Waters 2424 ELSD detector and Waters Atlantis Hilic silica column
(4.6 ꢀ 150 mm, 5
lm); H2O–CH3CN 3:97 (each containing 0.1% tri-
fluoro acetic acid); flow 1.0 ml/min; temp. 30 °C; Rt(Glu-std) = 8.23
min; Rt(Glu-syn) = 8.31 min; ELSD gas pressure 28 psi; drift tube
temp. 60 °C; nebulizer mode: heating 60%, gain 135, data rate 2
20
pps] and the specific optical rotation of [
a
]
+52.3 (c 0.10, H2O)
D
20
20
[D
-glucose: [
a]
+52.7 (c 0.16, H2O);
L
-glucose: [
a
]
D
ꢁ56.6 (c
D
0.11, H2O)].
4.6. 4-Bromobenzoylation of pachybionic acid d-lactone (6)
4.9. Growth inhibition assay
Compound 6 (8.0 mg, 25.0
then DMAP (3.0 mg, 24.6 mol) and 4-bromobenzoyl chloride
(21.9 mg, 100.0 mol) were added. The reaction mixture was stir-
red at 35 °C for 2 h, then diluted with H2O (5 ml) and extracted
with CHCl3 (3 ꢀ 5 ml). The combined organic layers were washed
with 1 N HCl (10 ml), satd. NaHCO3 (10 ml), and brine (10 ml),
dried over Na2SO4, and concentrated in vacuo. The remaining resi-
due was purified over Sephadex LH-20 (CHCl3–MeOH 1:1) to give
lmol) was dissolved in CHCl3 (3 ml),
l
Growth inhibitory effects of various anticancer drugs, with or
without the concomitant treatment of tested compounds, on the
cell lines were evaluated by the sulforhodamine B assay (Skehan
et al., 1990). Cells were seeded into 96-well tissue culture plates
l
in 100 ll at a plating density of 3000–5000 cells/well, and allowed
to incubate overnight. The cells were then treated with various
anticancer drugs at a range of concentrations in the presence or
absence of a fixed concentration of the designated compounds
6a (15.0 mg, 22.0 lmol, 88% yield): White powder; C28H30O10Br2;
1H NMR (400 MHz, CDCl3): d 7.91 (4H, dd, J = 8.6, 2.1 Hz), 7.61
(4H, d, J = 8.6 Hz), 5.14 (1H, d, J = 8.2 Hz), 5.04 (1H, dd, J = 8.2,
2.8 Hz), 4.85 (1H, dd, J = 9.9, 2.6 Hz), 4.23 (1H, dq, J = 12.4,
6.2 Hz), 4.13 (1H, t, J = 2.7 Hz), 4.05 (1H, dd, J = 8.4, 6.3 Hz), 4.01
(1H, m), 3.57 (1H, d, J = 7.5 Hz), 3.44 (3H, s), 3.39 (3H, s), 2.75
(1H, dd, J = 15.9, 2.1 Hz), 2.68 (1H, dd, J = 15.9, 4.2 Hz), 1.28 (3H,
d, J = 6.2 Hz), 1.26 (3H, d, J = 8.4 Hz); ESI-MS m/z 707 [M+Na]+.
(0.4, 2, 10, 25, or 50 lM), and allowed to incubate at 37 °C in 5%
CO2 for 72 h. Each drug concentration was tested in quadruplicate
and controls were tested in replicates of eight. Fold-resistance was
calculated by dividing the IC50 of the anticancer drug with or with-
out the tested compounds in the resistant cells by the IC50 of the
anticancer drug alone in the parental cells. Each experiment was
performed independently at least three times. The Student’s t test

S. Yao et al. / Phytochemistry 126 (2016) 47–58
57
was carried out to evaluate the differences between IC50 values,
with p < 0.05 being considered significant.
primers used were as follows: MRP2 (forward) 50-ACA-
GAGGCTGGTGGCAACC-30 (reverse) 50-ACCATTACCTTGTCACTGTC-
CATGA-30; ERCC1 (forward) 50-GAGCTGGCTAAGATGTGTAT-30
(reverse) 50-AGGCCAGATCTTCTCTTGAT-30; and p21 (forward) 50-
GAGCGATGGAACTTCGACTTT-30 (reverse) 50-GGGCTTCCTCTTGGA-
GAAGAT-30. PCRs were performed at 95 °C for 5 min, followed by
45 cycles of 95 °C for 10 s and 60 °C for 10 s. The fluorescence sig-
nal was acquired at the end of the elongation step of every PCR
cycle (72 °C for 10 s) to monitor the progress of PCR amplification.
The comparative Ct method was used to calculate the relative gene
expression values between resistant and parental cells (or treated
and untreated cells).
4.10. Flow cytometric analysis of MRP2 activity using a fluorescent
probe substrate
A flow cytometry based assay was employed to study the inhi-
bition of MRP2 by tested compounds, as described previously with
minor modifications (To et al., 2011). Briefly, MRP2-stably trans-
fected HEK293 cells or MRP2-overexpressing HKESC-1 cisR cells
were incubated for 30 min in phenol red free, complete medium
with an MRP2-specific fluorescent probe substrate [0.2 lM car-
boxyfluorescein diacetate (CFDA)], with or without different con-
centrations of tested compounds. Then, the cells were washed
twice with ice-cold PBS and incubated in CFDA-free medium at
37 °C for 1 h continuing with the tested compounds to generate
the inhibitor/efflux histogram, or without the tested compounds
to generate the efflux histogram. Cells were finally washed twice
with ice-cold PBS and placed on ice in the dark until analysis by
flow cytometry. Benzbromarone, an MRP2-specific inhibitor, was
used as control for comparison. The inhibited efflux was deter-
4.13. Cell cycle analysis
The standard propidium iodide staining method was adopted
for cell cycle analysis (Pozarowski and Darzynkiewicz, 2004).
Briefly, after the designated treatment, cells were harvested in
PBS and fixed in 70% ethanol overnight. Fixed cells were washed
once in PBS and then treated with 10
30 min. Afterwards, 50 g/ml propidium iodide was added to the
cells which were allowed to incubate at room temperature in the
dark for at least 30 min. The DNA profile of samples was then ana-
lyzed on an LSRFortessa cell analyzer (BD Biosciences). Cell cycle
progression was then analyzed by fitting the data with ModFit LT
software (Verity Software House, Topsham, ME, USA).
lg/ml RNase A at 37 °C for
l
mined as the difference in mean fluorescence intensity (DMFI)
between the inhibitor/efflux and efflux histograms. To determine
significant difference between intracellular fluorescence values,
the Student’s t test was performed, with p < 0.05 being considered
statistically significant. Samples were analyzed on an LSRFortessa
cell analyzer (BD Biosciences, San Jose, CA, USA). CFDA fluorescence
was detected with a 488-nm argon laser and a 530-nm bandpass
filter. 10,000 events were collected for the samples. Cell debris
was eliminated by gating on forward versus side scatter and dead
cells were excluded by propidium iodide staining. All assays were
performed in at least three independent experiments.
Acknowledgements
The authors thank Professor Xuefeng Mei from the Shanghai
Institute of Materia Medica for the X-ray diffraction experiment.
Financial support from National Science & Technology Major
Project ‘‘Key New Drug Creation and Manufacturing Program”
(No. 2012ZX09301001-001), National Natural Science Funds
of China (No. 81302657), Ministry of Science and Technology
(2010DFA30980), Chinese Academy of Sciences (KSZD-EW-Z-004-
01), Youth Innovation Promotion Association CAS, Shanghai Com-
mission of Science and Technology (11DZ1970700, 12JC1410300),
Health and Medical Research Fund (HHSRF Project No.
08090481) from the Food and Health Bureau, HKSAR, Direct Grant
(2008.1.079) from The Chinese University of Hong Kong, One-off
Funding for Joint Lab/Research Collaboration from CUHK and Seed
Fund for Joint Establishments from CUHK School of Biomedical
Science, CUHK are gratefully acknowledged.
4.11. Cellular platinum accumulation assay
Cells (HKESC-1, HKESC-1 cisR, HEK293 pcDNA3, and HEK293
MRP2) were incubated with 200 lM cisplatin, with or without
the tested compounds, at 37 °C for 4 h for the Pt accumulation
assay, as described previously with minor modifications (To
et al., 2004). After incubation, cells were washed twice with ice-
cold PBS and collected by a scraper. The cell pellets obtained were
resuspended in lysis buffer (25 mM Tris (pH 7.5), 300 mM NaCl,
and 1% Triton X-100), and subjected to digestion in 70% nitric acid
at 80 °C for 4 h. Total cellular Pt content was measured by atomic
absorption spectroscopy (wavelength: 265.9 nm, apparatus: Per-
kin–Elmer Analyst 100 atomic absorption spectrometer with gra-
phite cell atomizer HGA800). Protein concentration was
evaluated by the standard Bradford method for normalization. All
experiments were repeated three times.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.phytochem.2016.
03.006.
4.12. Reverse transcription and quantitative real-time PCR
The effect of the tested compounds on expression of the MDR
transporter genes and a few Pt resistance genes was determined
by reverse transcription and real-time PCR (To et al., 2011). Total
RNA was isolated from cells after treatment with the tested com-
pounds using TRI Reagent (Molecular Research Center, Cincinnati,
References
Abisch, E., Tamm, C., Reichstein, T., 1959. Die Glykoside der Wurzeln von
Pachycarpus lineolatus (DECNE.) BULLOCK (oder P. schweinfurthii (N. E. BR.)
BULLOCK) Glykoside und Aglykone, 201. Mitteilung. Helv. Chim. Acta 42, 1014–
1052.
Chen, J.J., Zhang, Z.X., Zhou, J., 1999. New C21 steroidal glycosides from Marsdenia
tenacissima. Acta Bot. Yunnanica 21, 369–377.
Deng, J., Liao, Z.X., Chen, D.F., 2005a. Marsdenosides A–H, polyoxypregnane
glycosides from Marsdenia tenacissima. Phytochemistry 66, 1040–1051.
Deng, J., Liao, Z.X., Chen, D.F., 2005b. Two new C21 steroids from Marsdenia
tenacissima. Chin. Chem. Lett. 16, 487–490.
Deng, J., Liao, Z.X., Chen, D.F., 2005c. Three new polyoxypregnane glycosides from
Marsdenia tenacissima. Helv. Chim. Acta 88, 2675–2682.
OH, USA). RNA (1 lg) was reverse transcribed using the Prime-
Script First Strand cDNA Synthesis Kit (TaKaRa-Bio, Dalian, China).
Quantitative real-time PCR was performed to determine the rela-
tive expression level of genes of interest using the KAPA SYBR FAST
qPCR Kit (KapaBiosystems, Woburn, MA, USA) in a LightCycler 480
Instrument I (Roche Applied Science, Indianapolis, IN, USA). The
human GAPDH transcript level was measured in parallel for nor-
malization. The specific primers used for MDR1, ABCG2, and
MRP1 have been described previously (To et al., 2012). The other
Feng, J.Q., Zhang, R.J., Zhou, Y., Chen, Z.H., Tang, W., Liu, Q.F., Zuo, J.P., Zhao, W.M.,
2008. Immunosuppressive pregnane glycosides from Periploca sepium and
Periploca forrestii. Phytochemistry 69, 2716–2723.

58
S. Yao et al. / Phytochemistry 126 (2016) 47–58
Han, S.Y., Zhao, M.B., Zhuang, G.B., Li, P.P., 2012. Marsdenia tenacissima extract
restored gefitinib sensitivity in resistant non-small cell lung cancer cells. Lung
Cancer 75, 30–37.
Skehan, P., Storneng, R., Scudiero, D., Monks, A., McMahon, J., Vistica, D., Warren, J.T.,
Bokesch, H., Kenney, S., Boyd, M.R., 1990. New colorimetric cytotoxicity assay
for anticancer-drug screening. J. Natl. Cancer Inst. 82, 1107–1112.
Taniquchi, K., Wada, M., Kohno, K., Nakamura, T., Kawabe, T., Kawakami, M.,
Kagotani, K., Okumura, K., Akiyama, S., Kuwano, M., 1996. A human canalicular
multispecific organic anion transporter (cMOAT) gene is overexpressed in
cisplatin-resistant human cancer cell lines with decreased drug accumulation.
Cancer Res. 56, 4124–4129.
Hu, Y.J., Shen, X.L., Lu, H.L., Zhang, Y.H., Huang, X.A., Fu, L.C., Fong, W.F., 2008.
Tenacigenin
B
derivatives reverse P-glycoprotein-mediated multidrug
resistance in HepG2/Dox cells. J. Nat. Prod. 71, 1049–1051.
Huang, X.D., Liu, T., Wang, S., 2009. Two new polyoxypregnane glycosides from
Marsdenia tenacissima. Helv. Chim. Acta 92, 2111–2117.
Jaeggi, K.A., Weiss, E., Wehrli, W., Reichstein, T., 1967. Die Glykoside der Wurzeln
To, K.K., Wang, X., Yu, C.W., Ho, Y.P., Au-Yeung, S.C., 2004. Protein phosphatase 2A
inhibition and circumvention of cisplatin cross-resistance by novel TCM-
platinum anticancer agents containing demethylcantharidin. Bioorg. Med.
Chem. 12, 4565–4573.
To, K.K., Robey, R., Zhan, Z., Bangiolo, L., Bates, S.E., 2011. Upregulation of ABCG2 by
rodidepsin via the aryl hydrocarbon receptor pathway. Mol. Cancer Res. 9, 516–
527.
To, K.K., Yu, L., Liu, S., Fu, J., Cho, C.H., 2012. Constitutive AhR activation leads to
concomitant ABCG2-mediated multidrug resistance in cisplatin-resistant
esophageal carcinoma cells. Mol. Carcinog. 51, 449–464.
Wang, S., Lai, Y.H., Tian, B., Yang, L., 2006. Two new C21 steroidal glycosides from
Marsdenia tenacissima (ROXB.) WIGHT et ARN. Chem. Pharm. Bull. 54, 696–698.
Wang, L., Yin, Z.Q., Wang, Y., Zhang, X.Q., Li, Y.L., Ye, W.C., 2010. Perisesaccharides
A–E, new oligosaccharides from the root barks of Periploca sepium. Planta Med.
76, 909–915.
von Gongronema taylorii (SCHLTR.
& RENDLE) BULLOCK. Glykoside und
Aglykone, 292. Mitteilung. Helv. Chim. Acta 50, 1201–1229.
Jiangsu New College of Medicine, 1977. A Dictionary of Traditional Chinese Drugs.
Shanghai Science and Technology Press, Shanghai, p. 1976.
Kawanishi, S., Kasai, R., Sakuma, S., Shoji, J., 1977. Constituents of Chinese crude
drug ‘‘Wujiapi.” VVIII. On the structures of new oligosaccharides C1, D2, F1 and
F2 of Bei-Wujiapi. Chem. Pharm. Bull. 25, 2055–2060.
Lai, G.M., Chen, Y.N., Mickley, L.A., Fojo, A.T., Bates, S.E., 1991. P-glycoprotein
expression and schedule dependence of adriamycin cytotoxicity in human
colon carcinoma cell lines. Int. J. Cancer 49, 696–703.
Liu, J., Yu, Z.B., Ye, Y.H., Zhou, Y.W., 2008. A new C21 steroid glycoside from
Marsdenia tenacissima. Chin. Chem. Lett. 19, 444–446.
Luo, S.Q., Lin, L.Z., Cordell, G.A., Xue, L., Johnson, M.E., 1993. Polyoxypregnanes from
Marsdenia tenacissima. Phytochemistry 34, 1615–1620.
Miyakawa, S., Yamaura, K., Hayashi, K., Kaneko, K., Mitsuhashi, H., 1986. Five
glycosides from the Chinese drug ‘‘tong-guang-san”: the stems of Marsdenia
tenacissima. Phytochemistry 25, 2861–2865.
Xia, Z.-H., Mao, S.-L., Lao, A.-N., Uzawa, J., Yoshida, S., Fujimoto, Y., 2011. Five new
pregnane glycosides from the stems of Marsdenia tenacissima. J. Asian Nat. Prod.
Res. 13, 477–485.
Pozarowski, P., Darzynkiewicz, Z., 2004. Analysis of cell cycle by flow cytometry
methods. Mol. Biol. 281, 301–311.
Qiu, S.X., Luo, S.Q., Lin, L.Z., Cordell, G.A., 1996. Further polyoxypregnanes from
Marsdenia tenacissima. Phytochemistry 41, 1385–1388.
Robey, R.W., Medina-Perez, W.Y., Nishiyama, K., Lahusen, T., Miyake, K., Litman, T.,
Senderowicz, A.M., Ross, D.D., Bates, S.E., 2001. Overexpression of the ATP-
binding cassette half-transporter, ABCG2 (Mxr/BCrp/ABCP1), in flavopiridol-
resistant human breast cancer cells. Clin. Cancer Res. 7, 145–152.
Schneider, E., Horton, J.K., Yang, C.H., Nakagawa, M., Cowan, K.H., 1994. Multidrug
resistance-associated protein gene overexpression and reduced drug sensitivity
of topoisomerase II in a human breast carcinoma MCF7 cell line selected for
etoposide resistance. Cancer Res. 54, 152–158.
Xing, W.X., Cheng, B., Mi, H.M., Yang, G.J., Wu, Y.T., 2004. Two new C21 steroidal
glycosides from Marsdenia tenacissima. Acta Pharm. Sin. 39, 272–275.
Xue, H.L., Huang, X.D., He, D., Lin, S.J., Wang, S., Niu, T., 2012. Effects of Marsdenia
tenacissima extract on proliferation and apoptosis of hematologic neoplasm cell
line cells. J. Sichuan Univ. (Med. Sci. Ed.) 43, 147–149.
Yang, R.Z., Yang, C.R., Zhou, J., 1981. The structures of tenacigenin A, B, and C. Acta
Bot. Yunnanica 3, 271–278.
Yao, S., To, K.K.W., Wang, Y.Z., Yin, C., Tang, C.P., Chai, S., Ke, C.Q., Lin, G., Ye, Y., 2014.
Polyoxypregnane steroids from the stems of Marsdenia tenacissima. J. Nat. Prod.
77, 2044–2053.
Ye, B.G., Yang, J.R., Li, J., Niu, T., Wang, S., 2014. In vitro and in vivo antitumor
activities of tenacissoside C from Marsdenia tenacissima. Planta Med. 80, 29–38.
Yoshikawa, K., Nishino, H., Arihara, S.G.B., Chang, H.C., Wang, J.D., 2000.
Oligosaccharides from Hoya carnosa. J. Nat. Prod. 63, 146–148.
Zhou, J., Yang, C.R., Yang, R.Z., 1980. The structure of tenacigenin A. J. Integr. Plant
Biol. 22, 67–74.
Singhal, S., Khare, M.P., Khare, A., 1980a. Cissogenin,
a pregnane genin from
Marsdenia tenacissima. Phytochemistry 19, 2427–2430.
Singhal, S., Khare, M.P., Khare, A., 1980b. Tenasogenin, a pregnane ester from
Marsdenia tenacissima. Phytochemistry 19, 2431–2433.