Discovery of cardio-protective constituents of Gualou Xiebai Decoction, a classical traditional Chinese medicinal formula

Article abstract of DOI:10.1016/j.phymed.2018.04.047

Backgroud: Finding effective compounds of TCMs has always been the basis for achieving marker-based quality control which is currently most widely used quality control strategy. Gualou Xiebai Decoction (GLXB), a classical TCM formula, is recorded and proven as a therapy for curing coronary heart disease but the effective constituents are unidentified and the substantial basis of the therapeutic effects is not clear. Purpose: The present research is an investigation on the chemistry of this formula aiming at finding and precisely identifying effective compounds. Study design and methods: This research started with screening for effective fractions of GLXB by rat myocardial infarction model and H9c2 cell hypoxia/reoxygenation model, then compounds in effective fractions were isolated and identified by phytochemical and spectroscopic methods. The cardio-protective activities of the compounds were tested in vitro and one of the effective compounds was taken as example to investigate the mechanisms. Results: The water-insoluble parts of GLXB were identified as effective parts in both in vitro and in vivo experiments. Systematic isolation of compounds in the effective fractions resulted in the isolation of 34 compounds including 7 new compounds, whereas 8 compounds were effective in protecting H9c2 cells against hypoxia/reoxygenation injury. One of the effective compounds, macrostemonoside P (MP) possibly exerted its effect by activating RISK pathway and attenuating apoptosis. Conclusion: An array of effective constituents of GLXB were discovered, and discovery of these compounds contributed to elucidating the substantial basis for the therapeutic effects of this formula, and provides fundaments for establishing Q-markers for further reliable quality control of GLXB.

Full text of DOI:10.1016/j.phymed.2018.04.047

Accepted Manuscript
Discovery of cardio-protective constituents of Gualou Xiebai
Decoction, a classical traditional Chinese medicinal formula
Weiyang Zhang , Yang Yu , Lulu Yan , Chong Li , Jingyan Han ,
Zifei Qin , Yi Dai , Zhihong Yao , Hua Zhou , Xinsheng Yao
PII:
DOI:
Reference:
S0944-7113(18)30147-8
10.1016/j.phymed.2018.04.047
PHYMED 52486
To appear in:
Phytomedicine
Received date:
Revised date:
Accepted date:
14 December 2017
19 March 2018
16 April 2018
Please cite this article as: Weiyang Zhang , Yang Yu , Lulu Yan , Chong Li , Jingyan Han ,
Zifei Qin , Yi Dai , Zhihong Yao , Hua Zhou , Xinsheng Yao , Discovery of cardio-protective con-
stituents of Gualou Xiebai Decoction, a classical traditional Chinese medicinal formula, Phytomedicine
(
2018), doi: 10.1016/j.phymed.2018.04.047
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service
to our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and
all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Discovery of cardio-protective constituents of Gualou Xiebai Decoction, a
classical traditional Chinese medicinal formula
a,1
b,1
a,c
a
c,d
b
Weiyang Zhang , Yang Yu , Lulu Yan , Chong Li , Jingyan Han , Zifei Qin , Yi
b
b,e
a,*
a,b,*
Dai , Zhihong Yao , Hua Zhou , Xinsheng Yao
a
Faculty of Chinese Medicine, Macau University of Science and Technology and State
Key Laboratory of Quality Research in Chinese Medicine (Macau University of
Science and Technology), Taipa, Macau, PR China
b
College of Pharmacy, Jinan University, Guangzhou 510632, PR China
c
Tasly Microcirculation Research Center, Peking University Health Science Center,
Beijing 100191, PR China
d
Department of Integration of Chinese and Western Medicine, School of Basic
Medical Sciences, Peking University, Beijing 100191, PR China
e
State key Laboratory of Drug Research, Shanghai Institute of Materia Medica,
Chinese Academy of Sciences, Shanghai 201203, P.R. China
*
Corresponding authors
Prof. Hua Zhou, Faculty of Chinese Medicine, Macau University of Science and
Technology and State Key Laboratory of Quality Research in Chinese Medicine
(Macau University of Science and Technology), Taipa, Macau, PR China, Tel:
+853-88972458 Fax: +853-28825886.
E-mail address: hzhou@must.edu.mo
Prof. Xinsheng Yao, Faculty of Chinese Medicine, Macau University of Science and
Technology and State Key Laboratory of Quality Research in Chinese Medicine
(
Macau University of Science and Technology), Taipa, Macau, PR China; College of
Pharmacy, Jinan University, 601 Huangpu Avenue, Guangzhou 510632, PR China. Tel.
86-20-85225849. Fax: +86-20-85221559,
E-mail address: tyaoxs@jnu.edu.cn
+
1

ACCEPTED MANUSCRIPT
1
Both authors contributed equally to this research.
2

ACCEPTED MANUSCRIPT
ABSTRACT
Backgroud: Finding effective compounds of TCMs has always been the basis for
achieving marker-based quality control which is currently most widely used quality
control strategy. Gualou Xiebai Decoction (GLXB), a classical TCM formula, is
recorded and proven as a therapy for curing coronary heart disease but the effective
constituents are unidentified and the substantial basis of the therapeutic effects is not
clear.
Purpose: The present research is an investigation on the chemistry of this formula
aiming at finding and precisely identifying effective compounds.
Study Design and Methods: This research started with screening for effective
fractions of GLXB by rat myocardial infarction model and H9c2 cell
hypoxia/reoxygenation model, then compounds in effective fractions were isolated
and identified by phytochemical and spectroscopic methods. The cardio-protective
activities of the compounds were tested in vitro and one of the effective compounds
was taken as example to investigate the mechanisms.
Results: The water-insoluble parts of GLXB were identified as effective parts in both
in vitro and in vivo experiments. Systematic isolation of compounds in the effective
fractions resulted in the isolation of 34 compounds including 7 new compounds,
whereas 8 compounds were effective in protecting H9c2 cells against
hypoxia/reoxygenation injury. One of the effective compounds, macrostemonoside P
(MP) possibly exerted its effect by activating RISK pathway and attenuating
apoptosis.
Conclusion: An array of effective constituents of GLXB were discovered, and
discovery of these compounds contributed to elucidating the substantial basis for the
therapeutic effects of this formula, and provides fundaments for establishing
Q-markers for further reliable quality control of GLXB.
Keywords
Traditional Chinese medicine formula, Gualou Xiebai Decoction, Allii Macrostemonis
Bulbus, Trichosanthis Fructus, myocardial infarction, H9c2 cells
3

ACCEPTED MANUSCRIPT
Abbreviations
AAR, areas-at-risk; Akt, protein kinase B; ANOVA, analysis of variance; Bax,
Bcl-2-associated X protein; Bcl-2, B-cell lymphoma 2; CHD, coronary heart disease;
DEPT, distortionless enhancement by polarization transfer; DMEM, Dulbecco's
modified Eagle medium; DMSO, dimethylsulfoxide; eNOS, endothelial nitric oxide
synthase; Erk1/2, extracellular signal–regulated kinases 1/2; FBS, fetal bovine serum;
FITC/PI,
fluorescein
isothiocyanate
/
propidium
iodide;
GAPDH,
glyceraldehyde-3-phosphate dehydrogenase; GLXB or GX, Gualou Xiebai Decoction;
1
1
H- H COSY, proton chemical shift correlation spectroscopy; H/R
1
hypoxia/reoxygenation; HMBC, H detected heteronuclear multiple bond correlation;
HPLC, high-performance liquid chromatography; HRESIMS, high-resolution
electro-spray ionization mass spectroscopy; HRP, horseradish peroxidase; HSQC,
heteronuclear singular quantum correlation; IA, infarct areas; IR, infrared
spectroscopy, infrared spectrum; IRI, ischemia-reperfusion injury; KRB,
Krebs-Ringer bicarbonate; LAD, left anterior descending branch; LV, left ventricle;
MP, macrostemonoside P; MTT, methyl thiazolyl tetrazolium; NMR, nuclear
magnetic resonance; NOESY, nuclear Overhauser effect correlation spectroscopy; OD,
optical density; ODS, octadecyl silane; PI3K, phosphoinositide 3-kinase; Q-marker,
quality marker; RISK pathway, reperfusion injury salvaging kinase pathway;
SDS-HCl, sodium dodecyl sulfate-hydrochloric acid; SDS-PAGE, sodium dodecyl
sulfate polyacrylamide gel electrophoresis; TCM, traditional Chinese medicine(s);
TMS, tetramethyl silane; TTC, 2,3,5-triphenyl-2H-tetrazolium chloride;
UHPLC-TOF-MS, ultra-high performance liquid chromatography-time-of-flight-mass
spectroscopy; UV/vis, ultraviolet-visible spectroscopy.
4

ACCEPTED MANUSCRIPT
Introduction
Gualou Xiebai Decoction (GLXB), the alcoholic decoction of Trichosanthis
Fructus and Allii Macrostemonis Bulbus, is one of the classical formulas originally
recorded in “Jin Kui Yao Lue”. The indication of GLXB is depicted as “the symptoms
of chest impediment and heart pain”, which corresponds with myocardial ischemia
and coronary heart disease (CHD) in Western medicine (Zhu, 1999). CHD is the
leading cause of death globally (WHO, 2007).
Great efforts and progress have been made on the pharmacological and chemical
aspects of GLXB. Previously as depicted (Lei et al., 2013), the combination of T.
kirilowii and A. macrostemon can improve the lesions of acute myocardium ischemia
rats, similar results were afforded by several investigations. Ding et al investigated the
preventive effect of GLXB against myocardial fibrosis (Ding et al., 2013) and their
work also involved the screening for active parts of the formula and a preliminary
chemical profiling (Ding, 2014). A study led by our team partially unveiled the
chemical constituents of GLXB (He et al., 2002). Nevertheless, there has always been
a pitfall elucidating the substantial basis for the therapeutic effects of this formula.
Currently, effective constituents of GLXB are not sufficiently understood, and no
quality control system has been proposed yet.
Establishing reliable quality control system for Chinese medicine formula is
essential and needs chemical markers that can correctly reflect the safety and
effectiveness of traditional Chinese medicines (TCMs) and can be qualitatively and
quantitatively determined. These markers are usually bioactive chemical constituents
of the formula. (Liu et al., 2017) The quality control system is questionable if quality
control markers are not associated with the efficacy and/or safety. Therefore, a
concept, “Q-marker”, was proposed to define such quality control markers (Liu et al.,
2
016).
In order to find and precisely identify effectiveness-related compounds as
substantial basis for future establishment of Q-markers of GLXB, the present research
involved screening of effective fractions of GLXB for preventing I/R-induced
myocardial infarction, isolating compounds by phytochemical approaches and
5

ACCEPTED MANUSCRIPT
identifying isolated compounds by spectroscopic methods. Then the effects of these
compounds on hypoxia/reoxygenation (H/R) -induced cell death were investigated to
find effective constituents.
Materials and methods
Plant materials
The samples of T. Pericarpium and A. Macrostemonis Bulbus were purchased
from Handan Chinese Herbal Medicinal Materials Company (Handan, Hebei, China)
and identified by Prof. Guang-xiong Zhou of Jinan University. Voucher specimens
were deposited in the Institute of Traditional Chinese Medicine & Natural Products,
Jinan University (No. JNU-TK-201503 and No. JNU-AM-201503), Guangzhou,
China.
Extraction and isolation
Dry T. pericarpium and A. Macrostemonis Bulbus were extracted by heat-reflux
in 60% ethanol at a ratio of 1.5:1 (15 kg:10 kg). The extract (GX, a. 5 kg) was yielded
by removing the solvents in vacuo, then loaded on an HP-20 macroporous resin
column (ϕ 40×160 cm, Mitsubishi Chemical, Tokyo, Japan) and eluted with water,
3
0% ethanol and 95% ethanol successively, three fractions were yielded respectively,
namely GX-1 (a. 4 kg, 80%), GX-2 (a. 780 g, 15.6%) and GX-3 (a. 240 g, 4.8%).
UHPLC-TOF-MS chromatographs of GX, GX-1, GX-2 and GX-3 were established
(Fig. S54). UHPLC-TOF-MS analysis was finished on an Agilent 6230 system with
TM
an Acquity
BEH-C18 column (ϕ10×100 mm, 1.7 μm). The procedures of
systematic isolation and purification of compounds from GX-2 and GX-3 were
depicted in the Supplementary Information. The purities of all compounds tested by
1
H-NMR and HPLC were > 90%.
Spectroscopic determination
All NMR spectra were acquired on a Bruker AV 600 (Bruker, Co. Ltd, Bremen,
German) using pyridine-d or DMSO-d as solvents and tetramethyl silane (TMS) as
5
6
internal reference. High-resolution electro-spray ionization mass spectroscopies
6

ACCEPTED MANUSCRIPT
(HRESIMS) were acquired on a Waters Synapt G2 mass spectrometer (Waters,
Manchester, UK). Infrared spectra (IR) were acquired on a JASCO FT/IR-480 plus
spectrometer (JASCO International Co. Ltd, Tokyo, Japan). The spectroscopic data
were listed in the Tables 1 and 2 or described in Supplementary Information. The
absolute configuration of saccharide moieties were determined by o-tolyl
isothiocyanate derivatization and HPLC following a previously described method (Ni
et al., 2017).
Rat myocardial ischemia reperfusion
Adult male Sprague-Dawley rats (240-260 g body weight) were kept in cages at
room temperature (22 ± 2 °C) and constant humidity (40 ± 5%) with regular 12 h
day-night switch and constant fresh air. The animals were provided free access to
water and food unless fasted. All procedures were in accordance to the guidelines of
Commission of Animal Research of Peking University and approved by Peking
University Institutional Animal Care and Use Committee (IACUC).
GX, GX-1, GX-2 and GX-3 were administrated at 4 g/kg, 3.06, 0.34 and 0.13
g/kg body weight by intragastric administration, respectively. The doses were
calculated from the origin records in Jin Kui Yao Lue and confirmed by
pre-experiments (data not shown, calculation and conversion of doses please see in
the Supplementary Information). A course of 30 min ischemia and 90 min reperfusion
was induced by left anterior descending branch (LAD) ligation following a reported
protocol (Lin et al., 2013).
After 90 min reperfusion, LAD was ligated and 2 ml of 4% Evans blue in normal
saline was injected through femoral vein then the heart was harvested immediately
and cut into 1 mm cross sections. The cut sections were incubated in 0.375%
2
,3,5-triphenyl-2H-tetrazolium chloride (TTC) phosphate buffer solution for 15 min
at 37 °C. The images of the sections were recorded with a digital camera (DS-5M-U1,
Nikon, Nanjing, China) connected to a stereoscope. Infarct areas were manifested by
pale white color, ischemic but viable tissues (areas at-risk, AAR) were stained in red,
and non-infarction areas were blue. AAR/LV (%) and IA/AAR (%) were calculated
7

ACCEPTED MANUSCRIPT
and the average ratios of 5 slices were used to indicate the degree of myocardial
infarction.
H9c2 cell hypoxia/reoxygenation assay
H9c2 myoblast cells (ATCC, Maryland, US) were cultured in Dulbecco’s
modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS), 100 U/ml
penicillin and 100 µg/ml streptomycin (Gibco Invitrogen Corp., California, US) at
3
8
7 °C under normoxic condition (95% air / 5% CO ). When the confluence reached
2
0-90%, the cells were harvested by trypsinization (0.25% trypsin EDTA solution,
Gibco Invitrogen Corp., California, US) and seeded in 96-well plates at 5000
cells/well. All tested compounds were dissolved in DMSO at 100 mM as stock
solution. Thirty six h after seeding, the cells were pretreated by the compounds at 0.5,
5
and 50 μM for 1 h, then the cells were washed once with KRB (Krebs-Ringer
Bicarbonate) culture buffer pre-balanced in nitrogen overnight and then given the
compounds at the same concentrations (diluted with pre-balanced KRB culture buffer).
The cells were deposited in a hypoxia chamber infused with nitrogen at 37 °C for 3 h.
After hypoxia, the cells were given full DMEM and moved back to normoxic
environment for 3 h’ culture. For each well, 100 μl 0.5 mg/ml MTT (Sigma Aldrich
Chemical Co., Missouri, US) dissolved in DMEM was added at the end of
reoxygenation. After incubation at 37 °C for 4 h, 100 μl 10% sodium dodecyl
sulfate-hydrochloric acid (SDS-HCl) was added in each well to dissolve the yielded
formazan. Optical densities (OD) was determined after the formazan was fully
dissolved (570 Abs / 610 Ref). Survival rates were calculated based on OD values and
protection rates versus H/R group were calculated from survival rates.
Flow cytometry
Cells were stained with Annexin-V FITC/PI (fluorescein isothiocyanate/
propidium iodide) apoptosis detection kit (BD Biosciences, California, US). H9c2
cells were seeded in a ϕ 3.5 cm culture dish. After overnight culture, the cells were
treated by corresponding compounds and H/R following the same procedure in the
9
6-well plate assay. The cells were harvested by trypsinization and then treated with
8

ACCEPTED MANUSCRIPT
the following protocol provided by the manufacturer. Briefly, after washing the
harvested cells twice with ice-cold PBS, the cells were resuspended with the binding
6
buffer at 1×10 cells/ml, and 100 μl of the suspension was collected and added with
5
μl FITC labeled Annexin-V solution and 5 μl PI solution. After incubating the cell
suspension at room temperature in the dark for 15 min, the flow cytometry analysis
TM
was performed on a FACSAria III Flow Cytometer (FACS Vantage-BD Sciences,
Beckon Dickinson Company, California, US).
Western blotting
H9c2 cells were seeded in a ϕ 3.5 cm culture dish. After overnight culture, the
cells were treated by the compound and H/R following the same procedure in the
9
6-well plate assay. The cells were washed twice with ice-cold PBS and lysed with
RIPA buffer (Thermo Scientific, California, US). After a very short vortex, the lysed
cells were kept in ice bath for 20 min. Then the lysates were centrifuged at 14,000 g
for 15 min, the supernatant were collected and the contents of protein in the lysate
were determined with Bio-Rad protein assay reagent (Bio-Rad, California, US).
Samples equivalent to 60 μg protein were loaded to 10% sodium dodecyl sulfate
polyacrylamide gel electrophoresis (SDS-PAGE). After running, the proteins were
electrotransferred to nitrocellulose filter membrane (PALL Corp., New York, US) and
then incubated with 5% BSA solution at room temperature for 2 h. Primary antibodies
of Bcl-2 (1:200, mouse, all antibodies were from Cell Signalling Technologies,
Massachusetts, US), Bax (1:200, rabbit), β-actin (1:500, mouse), GAPDH (1:200,
rabbit), phosphor-Erk 1/2 (1:200, rabbit), phosphor-Akt (1:200, rabbit),
phosphor-PI3K (1:200, rabbit) and phosphor-eNOS (1:200, rabbit) were added to the
membranes correspondingly and incubated at 4 °C overnight. Afterward, HRP-linked
anti-rabbit or anti-mouse secondary antibodies were applied for 1.5 h at room
temperature and HRP substrate solution was added to visualize the proteins. Photos
were recorded on an Amersham Imager 600 System (G.E., Massachusetts, US) and
the densities of bands were quantified by Image-J software.
Statistical analysis
9

ACCEPTED MANUSCRIPT
*
One-way ANOVA with Tukey’s multiple comparison test, significant levels: P <
**
***
*
#
0
.05, P < 0.01, P < 0.001, α = 0.05, vs model group, vs normal control.
Results
Active fractions of GLXB
An in vivo ischemia-reperfusion (I/R) model in rats was employed to examine
the effect of GLXB and different factions of GLXB.. As shown in Fig. 1A and 1B,
pretreatment of GX at 4 g/kg body weight can significantly reduce the infarction areas
(P < 0.05). When administrated at equivalent doses according to their proportions in
the total extract (GX-1 3.06 g/kg, GX-2 0.34 g/kg and GX-3 0.13, m/m), GX-2 can
significantly reduce the infarction size (P < 0.05), while GX-3 showed some effects
but weaker. The ischemic area of each group was similar.
Then the screening was conducted on in vitro model of H9c2 myoblast cell
protection against H/R as well. The results indicated that the total extract (GX) and
two water-insoluble fractions (GX-2 and GX-3) significantly increased cell survival
rates in the MTT test (Fig. 1C). Taking both in vivo and in vitro results into
consideration, GX, GX-2 and GX-3 were cardio-protective.
Structure identification of new compounds
Compound 1 was obtained as white gum, and its molecular formula was assigned
+
as C H O according to the HRESIMS (m/z 627.3739 [M-OCH ] , C H O calcd
34
58 12
3
33 55 11
for 627.3744). The IR spectrum of 1 showed absorption bands for hydroxyl (3398
-1
-1
1
cm ) and ketal (1079 cm ) groups. H-NMR spectrum (Table 1) showed four methyl
groups at δ 0.82 (3H, s, H -18), 1.04 (3H, d, 5.1 Hz, H -27), 1.17 (3H, d, 5.6 Hz,
H
3
3
H -21) and 1.37 (3H, s, H -19), a methoxyl group at δ 3.26 (3H, s, 22-OCH ), and an
3
3
H
3
1
3
anomeric proton signal at δ 4.84 (1H, d, 7.7 Hz, Glc-1ʹ). In the C-NMR spectrum, a
H
total of 34 carbon signals were observed. Except for the signals for a methoxyl and
glucosyl groups, the remained 27 carbon resonances resolved as four methyls, 8
methylenes, 12 methines and three quaternary carbons, were characteristic of a C-27
steroidal aglycon moiety. Detailed comparison of 1D-NMR data of 1 with those of its
analogue macrostemonoside M indicated that 1 was very likely a methoxylated
10

ACCEPTED MANUSCRIPT
derivative of the latter (Chen et al., 2006).
1
1
Data assignment for 1 was conducted based on HSQC, HMBC and H- H COSY
spectra. H -19 (δ 1.37) showed HMBC correlations with C-1, C-5, C-9 and C-10,
3
H
combined with sequential COSY correlations H-1/H-2/H-3/H -4/H-5/H -6/H -7/H-8,
2
2
2
indicating a A/B ring partial structure with four hydroxyls substituted at C-1,2,3,6. As
other HMBC correlations of H -18/C-12, C-13, C-14, C-17 and H -21/C-17, C-20,
3
3
C-22, together with COSY correlations of H -11/H -12 and H-15/H -16/H-17/H-8, the
2
2
2
assignment was established for C-E rings substructure. In addition, a spin system of
H -23/H -24/H-25/H -26 (H -27) and HMBC correlation of H-23/C-22 suggested the
2
2
2
3
presence of a saturated C-22 side chain with an oxygenated group at C-26. Finally, the
methoxyl and glucosyl groups were located at C-22 and C-26 as indicated by key
HMBC correlations of 22-OCH /C-22 and H -26/C-1ʹ, respectively. The aglycone (1a)
3
2
and monosaccharide were obtained after acid hydrolysis of 1. Derivatization and
HPLC experiments were used to identify the monosaccharide as a D-glucose.
Furthermore, the β-configuration of anomeric proton was determined by a large
coupling constant (J = 7.7 Hz). The relative configuration of A and B rings was
determined as the same as macrostemonoside M on the basis of NOESY correlations
of H -19/H-5/H-6, H-2/H-9, H-1/H -11, H-1/H-2, H-2/H-3 and H-6/H-8. The 25S
3
α
configuration of 1 was determined by the chemical shift deviation between H-26 and
a
H-26 . Geminal proton resonances for H -26 are more resolved in 25S isomers
b
2
compared with 25R compounds, which was summarized as a criterion by Agrawal
Agrawal, 2005). In 1, the difference (δ – δ ) was about 0.56 ppm near to the
(
a
b
empirical deviation range of 25S-configuration (δ – δ > 0.57 for 25S and δ – δ <
a
b
a
b
-
1
0
2
2
.48 for 25R). Meanwhile, peak strength 916 > 898 cm provided consistent result of
5S-configuration.
Thus,
1
was
identified
as
22-methoxyl
5S-furostan-1β,2β,3β,6α,26-pentol 26-O-β-D-glucopyranoside.
Compound 2 was isolated as white gum. HRESIMS gave a quasi-molecular peak
+
of [M+Na] at m/z 683.3613 (C H O Na calcd for 683.3619). IR absorptions of
33
56 13
-1
-1
hydroxyl groups at 3396 cm and ketal C-O bond at 1077 cm were observed. The
D-NMR and MS data indicated that 2 and 1 were very similar but differed at the
1
11

ACCEPTED MANUSCRIPT
downfield shift of C-5 and absence of a methoxyl group. These were further
supported by HMBC correlations of H -19/C-1, C-5, C-9 and C-10, indicating a
3
hydroxyl group at C-5. The absence of a methoxyl group signal indicated that C-22
was attached by a hydroxyl group. Similarly, the relative configurations of A and B
rings in 2 were suggested to be the same as in 1. The 25R configuration could be
determined by the chemical shift deviation of H-26 and H-26 , whereas the IR
a
b
-1
spectra of the aglycon yielded a same result as 25R (917 < 899 cm ). Therefore,
compound was identified as 25R-furostan-1β,2β,3β,5β,6α,26-hexol
6-O-β-D-glucopyranoside.
Also obtained as white gum, 3 had the same molecular formula of C H O as
2
2
3
3
56 13
+
compound 2 by HRESIMS ([M+Na] m/z 683.3612, C H O Na calcd for 683.3619).
3
3
56 13
The NMR data of 3 showed high similarity to those of 2 except for slight differences
of the proton chemical shifts of H -26 and H -27, indicating a possible configurational
2
3
change at C-25. In 3, the chemical shift deviation between H-26 and H-26 is 0.57,
a
b
suggesting
a
25S configuration. In conclusion,
3
was identified as
2
5S-furostan-1β,2β,3β,5β,6α,26-hexol 26-O-β-D-glucopyranoside.
Compound 4 was isolated as white powder. HRESIMS showed a
+
quasi-molecular peak [M+H] at m/z 625.3593, indicative of a molecular formula of
C H O (calcd for 625.3595). The 1D-NMR data of 4 showed a close resemblance
33
52 11
to a known compound, macrostemonoside N (Chen et al., 2006), except for the
presence of an olefinic carbons at C-20 and C-22. Thus, it was reasonable to infer that
4
was a dehydrated derivative of macrostemonoside, which was verified by HMBC
correlations of H -21/C-20, C-22, and a terminal double bond resonance at C-25(27).
3
Thus, 4 was identified as 5β-furost-20(22), 25(27)-dien-1β,2β,3β,6α,26-pentol
2
6-O-β-D-glucopyranoside.
Compound 5 was isolated as white powder, the quasi-molecular peak at m/z
+
+
6
27.3742 ([M+H] ) was indicative of a molecular formula of C H O ([M+H]
3
3
54 11
calcd for 627.3739). Except for two less olefinic carbon signals, other NMR data of 5
1
were very similar to 4. The presence of H -27 methyl peak in H-NMR and HMBC
3
correlations of H -27/C-25, C-26 and H-25/C-23, C-24 indicated that C-25(27) was
3
12

ACCEPTED MANUSCRIPT
saturated. In conclusion, the structure of 5 was defined as 25S-furost-20(22)-en-1β,2β,
3
β,6α,26-pentol 26-O-β-D-glucopyranoside.
Compound 6 was white powder when isolated, the quasi-molecular ion peak at
+
m/z 937.5008 was indicative of a molecular formula of C H O ([M+H] calcd for
4
5
76 20
1
9
37.5008). In the H-NMR spectrum of 6, methyl signals were seen as δ 0.85 (3H, s,
H
H -18), 1.03 (3H, d, 6.5 Hz, H -27), 1.24 (3H, s, H -19) and 1.32 (3H, d, H -21),
3
3
3
3
1
3
while three anomeric protons were present at δ 4.81, 4.95 and 5.36. The C-NMR
H
spectrum of 6 showed 45 signals including 27 carbons of the aglycon and 18 signals
of three saccharide moieties. Acidic hydrolysates of 6 contained D-glucose and
D-galactose in a ratio of 2:1 as determined by HPLC after derivatization.
Similarity between 6 and macrostemonoside P (Chen et al., 2007) was revealed
by data assignment and comparison. Thereafter, differences were seen at the proton
chemical shifts of H -26 and H -27. As the chemical shift deviation between H-26
2
3
a
and H-26 was 0.60 and H-27 was observed at δ 1.03, 6 was defined as a 25S-isomer
b
H
of macrostemonoside P. Thus, compound 6 was identified as 25S-26-O-β-D-glucosyl
β-furostan-1β,3β,26-triol 3-O-β-D-glucosyl (1→2)-β-D-galactoside.
Compound 7 was isolated as white powder. HRESIMS showed an [M+H] peak
at m/z777.1872 (C H O calcd for 777.1878). As displayed in ultraviolet-visible
5
+
3
5
36 20
spectrum, typical absorption peaks of flavonoids were shown at 203, 267 and 345 nm.
1
H-NMR spectrum of 7 showed two sets of characteristic ABX aromatic protons [δH
7
.17 (1H, d, 8.7 Hz, H-5ʹ), 7.51 (1H, dd, 8.7, 2.2 Hz, H-6ʹ), and 7.58 (1H, d, 2.2 Hz,
H-2ʹ)] and [δ 6.66 (1H, d, 8.2 Hz, H-5ʹʹʹʹ), 7.14 (1H, dd, 8.2, 1.9 Hz, H-6ʹʹʹʹ) and 7.24
H
(1H, d, 1.9 Hz, H-2ʹʹʹʹ)], as well as a pair of meta-positioned protons on an aromatic
ring [δ 6.36 (1H, br s, H-8) and 6.17 (1H, br s, H-6)]. Two anomeric protons were
H
observed at [δ 5.54 (1H, d, 7.4 Hz, H-1ʹʹ) and 4.83 (1H, d, 7.3 Hz, H-1ʹʹʹ)], and a
H
1
3
methoxyl group was seen at δ 3.68 (3H, s). C-NMR spectrum of 7 showed 35
H
resonances in total including two secondary carbons, 18 tertiary carbons, 14
quaternary carbons, and a methoxyl group.
HMBC correlations of H-5ʹʹʹʹ/C-3ʹʹʹʹ, C-4ʹʹʹʹ, H-2ʹʹʹʹ/C-3ʹʹʹʹ, C-4ʹʹʹʹ and H-2ʹʹʹʹ,
H-6ʹʹʹʹ/C=O were indicative of a benzoyl group with two oxygenated substituents, one
13

ACCEPTED MANUSCRIPT
of which was a methoxyl group as verified by 4ʹʹʹʹ-OCH /C-4ʹʹʹʹ correlation.
3
The rest 27 carbons belonged to a flavonoid aglycon and two hexoses. Two
broadened singlets [δ 6.17 (1H, br s) and 6.36 (1H, br s)] were typical of H-6 and
H
H-8 of 5,7-dihydroxyl A ring of flavonoid and the other ABX coupling system were
arisen from the 3,4-disubstituted B ring. Chemical shifts of C-2 (δ 155.8) and C-3
C
(δC 133.9) were indicative of 3-OH, and the aglycon was defined as
3
,5,7,3′,4′-pentahydroxyl flavonoid, namely quercetin (Bendaikha et al., 2014; Xu et
al., 2009). Meanwhile, two hexosyls were both identified as β-D-glucosyls.
Linkages between the aglycon and the glucosyls were established based on
HMBC cross peaks of the anomeric protons and oxygenated carbons as H-1′′/C-3 and
H-1′′′/C-3′. Whereas HMBC correlations of H -6′′′/C=O indicated that the
2
4
-hydroxyl-3-methoxyl benzoyl was attached to C-6′′′ of 3′-O-glucosyl. Conclusively,
was identified as quercetin-3-O-β-D-glucosyl-3′-O-6′′-(4-hydroxyl-3-methoxyl
7
benzoyl)-β-D-glucoside.
Other compounds were known compounds identified based on their
spectroscopic
data,
including
2,3,4,9-tetrahydro-1H-pyrido
[3,4-b]
indole-3-carboxylic acid (8) (He et al., 2003), macrostemonoside M (9),
macrostemonoside N (10) (Chen et al., 2006), macrostemonoside G (11),
macrostemonoside
6-[(β-D-glucopyranosyl)oxy]-2β,22-dihydroxy-5β-furostan-25(27)-en-3β-yl
O-β-D-glucopyranosyl-(1→2)- β-D-galactopyranoside (13), macrostemonoside J (14)
Peng et al., 1994), macrostemonoside P (15), macrostemonoside O (16) (Chen et al.,
007), timosaponin BII (17) (Guo et al., 2015), 25R-timosaponin BII (18),
Q
(12)
(Chen
et
al.,
2007),
2
(
2
(
-
-
(
2α,3β,5α)-27-O-(β-D-glucopyranosyl)-2,22-dihydroxyl
furost-25-en-3-yl
β-D-glucopyranosyl- (1→2)-O-
[β-D-glucopyranosyl-(1→3)]
O-β-D-glucopyranosyl- (1→4)- β-D-galactopyranoside (19), macrostemonoside R
20) (Chen et al., 2007), syringaresinol 4'-O-β-D-glucopyranoside (21) (Lami et al.,
1
991b), galangin 3-O-β-D-glucoside (22) (Kaouadji et al., 1988), luteolin (23) (Park et
al., 2007), chrysoeriol 7-O-β-D-glucoside (24) (Fan et al., 2011b), quercetin
-O-L-riboside (25) (Kizu et al., 1995), luteoside (26) (Lami et al., 1991a),
3
14

ACCEPTED MANUSCRIPT
isoquercetin (27) (Jin et al., 2009), quercitrin (28) (Lee et al., 2017), linarin (29)
(
Zhang et al., 2012), rutin (30) (Li et al., 2008), quercetin (31) (Chang et al., 2000),
'-deoxy-5'-β-methylsulphinyl-adenosine (32) (Wang and Lee, 1997), calycosin (33)
Cui et al., 1993) and davidigenin (34) (Siddaiah et al., 2006).
5
(
Anti-H/R effect of the compounds on H9c2 cells
A total of 18 out of all compounds were tested in the H9c2 cell
hypoxia-reoxygenation assay. All tested compounds 1) were absorbed compounds or
structural analogues of absorbed compounds (Lin, et al, 2018), 2) were abundant
enough for finishing the tests (> 20 mg obtained). Protection rates of the compounds
were illustrated in Fig. 3 (mean ± SD). As the most potent candidate, rutin (30)
showed a protection rate of 63.7 ± 12.1% at 50 μM. While two other flavonoids, 25
and 27 were active as well (25.3 ± 7.2% and 42.1 ± 3.8%, respectively). Other
effective compounds, were all steroid saponins including 2, 6, 13, 14 and 15, among
which the effect of 2 was significant at 5 and 50 μM and 15 was significantly effective
at 0.5, 5 and 50 μM whereas others were significantly active at 50 μM only.
Effects of macrostemonoside P on H/R-induced apoptosis and RISK pathway
As the effects and mechanisms of steroidal saponins were seldom investigated,
macrostemonoside P (MP) was taken as an example to confirm the effects and
investigate the mechanisms as it was the most potent saponin in this test and relatively
abundant when isolated (27.7 mg). During IRI and H/R, apoptosis is an important cell
death pathway. Cell apoptosis was analyzed by flow cytometry after Annexin-V FITC
PI staining. As illustrated in Fig. 4A, the normal morphology or the cells collapsed
after H/R and MP-treated cells mostly maintained their normal shape. Flow
+
cytometric analysis indicated that H/R drastically increased FITC cell proportion
(mean 40.9%), whereas 25 μM MP-treated cells showed decreased apoptotic
proportions but the difference was not significant (29.5%), and 50 μM group showed
significant decrease (24.0%, P < 0.05). Meanwhile, ratio of Bax/Bcl-2 was
determined. Compared with the H/R group, MP groups showed a tendency of
decrease in Bax/Bcl-2 ratios and the difference was significant in 50 μM group (P <
15

ACCEPTED MANUSCRIPT
0
.01).
The interaction between MP and phosphorylation activation of key proteins of
RISK pathway was investigated. The data were illustrated as folds of normal group in
Fig. 5. MP-treatment significantly increased phosphorylation of PI3K and eNOS at
concentration of 50 μM (P < 0.05) and Akt at both 25 and 50 μM (P < 0.05) as shown
in Fig. 5. Erk1/2 were activated when the cells underwent H/R and MP further
promoted the phosphorylation of Erk1/2 to salvage the cells.
Discussion
In this work, effective compounds in the effective fractions of GLXB were
identified. The total extract and the less water-soluble fractions of GLXB were
effective against myocardial infarction induced by ischemia-reperfusion in vivo and
H/R-induced H9c2 cell death in vitro.
Investigation on the chemistry of active fractions yielded 34 compounds,
including 18 steroidal saponins, 13 flavonoids and three compounds of other types
(alkaloids and lignan). Steroidal saponin is the most important species of compounds
in A. macrostemon. Approximately 50 steroidal saponins have been discovered from A.
macrostemon, besides, A. macrostemon also contains sulfide volatile oil, lignans and
alkaloids (Yao et al., 2016). Major constituents of T. Fructus (containing peels and
seeds) are triterpenoid saponins (Akihisa et al., 1988; Akihisa et al., 1994), and
flavonoids (Li et al., 2014) etc. However, since sugar, pectin and oil in the fruits and
seeds can considerably interfere isolation and detection, meanwhile T. Pericarpium
injection is approved for curing angina pectoris (Yao et al., 2017), T. Pericarpium was
used instead of whole fruit in this research. The work of He (He, 2002) is the first
report on active compounds in GLXB in which anticoagulative fractions were
separated to yield 27 compounds and 12 of them were anticoagulative in vitro.
Antithrombotic effect is an aspect in the prevention of atherosclerosis and CHD, so
finding anticoagulative compounds in GLXB was a different view on this formula,
and yielded different compounds.
In the H9c2 cell assay, flavonoids were mostly active including 25, 27 and 30,
whereas other active components were steroid saponins including 2, 6, 13, 14 and 15.
16

ACCEPTED MANUSCRIPT
The results indicated these compounds were effective constituents of GLXB for
cardioprotective effects. During the occurrence of CHD, IRI and acute myocardial
infarction compromise myocardium (Baliga, 2001; Morel et al., 2012), oxidative
stress is one of the most important factors threatening cell survival (Frank et al., 2012).
Flavonoids are potent naturally occurring anti-oxidants (Rice-Evans et al., 1996), the
roles and mechanisms of flavonoids in alleviating IRI and countering oxidative stress
have long been recognized (Cotelle, 2001). Steroidal saponins in A. macrostemon
were highly diversified in structures, whereas their pharmacological effects were also
diversified including protection against renal IRI (Manivannan et al., 2015),
anti-coagulation, and cytotoxicity (Chen et al., 2009). Steroidal saponins from A.
macrostemon were discovered as cardio-protective substances involved in this
research, but they differed in cell-protective activity. Pitifully current results were not
sufficient for a structure-activity relationship discussion which will be a good
guidance for finding more efficacy-related substances in GLXB.
Moreover, this is the first to report the bioactivity and mechanism of
macrostemonoside P. Macrostemonoside P possibly alleviated H/R-induced apoptosis,
and the mechanism is associated with activating pro-survival protein kinases
including PI3K, Akt and Erk1/2 to further activate eNOS. These proteins were termed
reperfusion injury salvage kinase (RISK) pathway (Hausenloy and Yellon, 2004).
Reliable quality control system requires appropriate quality markers which 1) are
components of the herbal materials, 2) have definite chemical structures and can be
determined qualitatively and/or quantitatively, 3) are closely related to the efficacy
and/or safety of the TCM herbs/formulas (Yang et al., 2017a, b). In the current
research, as the effective constituents were obtained from effective fractions by
phytochemical approaches and identified by comprehensive spectroscopic methods,
the structures were fully elucidated including stereochemistry (Fig. 2). This approach
can provide precise structural information and materials for pharmacological assays
but not a full chemical profiling. Meanwhile, accurate evaluation of efficacy depends
on appropriate pharmacological models. In vivo rat myocardial infarction model was
employed in the screening for effective fractions and in vivo H9c2 cell H/R model
17

ACCEPTED MANUSCRIPT
was employed in testing single compounds. Both are recognized as pharmacological
models for the basic research of cardioprotection. Thus, as a linkage between
compounds and therapeutic effects was built based on current results, these effective
compounds are the substantial basis for the therapeutic effects of GLXB, at the same
time potential Q-markers. Nevertheless, contents, detectability and measurability
should be taken into consideration when establishing Q-markers based on substantial
basis.
As the present research is limited to cardio-protective effects, while the effects of
Chinese medicine formulae are mostly various and holistic, other aspects of this
formula cannot be reflected from current results, for example, potential
lipid-regulatory effects. Different roles of the component herbs have not been
reflected from the effective compounds, either.
In conclusion, GLXB extract can significantly reduce infarction areas in rats
undergone myocardial IRI, whereas water-insoluble parts of GLXB were the effective
cardioprotective parts as revealed by both in vivo and in vitro experiments. Thereafter,
isolation and structure elucidation of compounds from the active parts of GLXB
provided a vision to the effective constituents of the formula, which are important
fundaments for establishing quality markers. In vivo assay indicated that effective
constituents
for
cardioprotective
effects
are
as
follows:
2
2
2
5R-furostan-1β,2β,3β,5β,6α,26-hexol
26-O-β-D-glucopyranoside
(2),
5S-macrostemonoside
P
(6),
26-[(β-D-glucopyranosyl)oxy]-
O-β-D-glucopyranosyl-(1→2)-
β,22-dihydroxy-5β-furostan-25(27)-en-3β-yl
β-D-galactopyranoside (13), macrostemonoside J (14), macrostemonoside P (15),
quercetin 3-O-L-riboside (25), isoquercetrin (27) and rutin (30), and these compounds
are therefore potential Q-marker for establishment of reliable quality control system
of GLXB products.
Conflict of interest
The authors declare no conflict of interest.
18

ACCEPTED MANUSCRIPT
Acknowledgment
The authors thank all the supports from Macao Science and Technology
Development Fund (No. 035/2014/A1), the National Natural Science Foundation of
China (81773880) and the State Key Laboratory of Drug Research
(SIMM1705KF-17).
19

ACCEPTED MANUSCRIPT
References
Agrawal, P.K., 2005. Assigning stereodiversity of the 27-Me group of furostane-type steroidal saponins via NMR chemical shifts. Steroids 70, 715-724.
Akihisa, T., Tamura, T., Matsumoto, T., Eggleston, D.S., Kokke, W., Shimizu, N., 1988. Karounidiol [D: C-friedo-oleana-7, 9 (11)-diene-3α, 29-diol] and its
3
4
-O-benzoate: novel pentacyclic triterpenes from Trichosanthes kirilowii. X-Ray molecular structure of karounidiol diacetate. J. Chem. Soc., Perkin Trans. 1,
39-443.
Akihisa, T., Yasukawa, K., Kimura, Y., Takido, M., Kokke, W.C., Tamura, T., 1994. 7-Oxo-10α-Cucurbitadienol from the seeds of Trichosanthes kirilowii and
its anti-inflammatory effect. Phytochemistry 36, 153-157.
Baliga, R.R., 2001. Apoptosis in myocardial ischemia, infarction, and altered myocardial states. Cardiol. Clin. 19, 91-112.
Bendaikha, S., Gadaut, M., Harakat, D., Magid, A., 2014. Acylated flavonol glycosides from the flower of Elaeagnus angustifolia L. Phytochemistry 103,
129-136.
Chang, Y.C., Chang, F.R., Wu, Y.C., 2000. The constituents of Lindera glauca. J. Chin. Chem. Soc. 47, 373-380.
Chen, H.F., Wang, G.H., Luo, Q., Wang, N.L., Yao, X.S., 2009. Two new steroidal saponins from Allium macrostemon Bunge and their cytotoxity on different
cancer cell lines. Molecules 14, 2246-2253.
Chen, H.F., Wang, G.H., Wang, N.L., Yang, M.S., Wang, Z., Wang, X.W., Yao, X.S., 2007. New furostanol saponins from the bulbs of Allium macrostemon
Bunge and their cytotoxic activity. Die Pharmazie 62, 544-548.
Chen, H.F., Wang, N.L., Sun, H.L., Yang, B.F., Yao, X.S., 2006. Novel furostanol saponins from the bulbs of Allium macrostemon B. and their bioactivity on
i
[
Ca2+] increase induced by KCl. J. Asian Nat. Prod. Res. 8, 21-28.
Cotelle, N., 2001. Role of flavonoids in oxidative stress. Curr. Top. Med. Chem. 1, 569-590.
Cui, B., Nakamura, M., Kinjo, J., Nohara, T., 1993. Chemical Constituents of Astragali Semen. Chem. Pharm. Bull. (Tokyo) 41, 178-182.
Ding, Y.F., 2014. Establishment and Application of Myocardial Fibrosis Cell Model Based on Endothelial-Mesenchymal Transition and the Application of
Gualou Xiebai Baijiutang. Nanjing University of Chinese Medicine.
Ding, Y.F., Peng, Y.R., Li, J., Shen, H., Shen, M.Q., Fang, T.H., 2013. Gualou Xiebai Decoction prevents myocardial fibrosis by blocking TGF-beta/Smad
signalling. J. Pharm. Pharmacol. 65, 1373-1381.
Fan, X.M., Chen, G., Su, S.S., Lu, X., Pei, Y.H., 2011a. Isolation and identification of constituents from fruit of Trichosanthes kirilowii Maxim. Journal of
Shenyang Pharmaceutical University 12, 6.
Fan, X.M., Chen, G., Su, S.S., Lu, X., Pei, Y.H., 2011b. Isolation and identification of constituents from fruit of Trichosanthes kirilowii Maxim. Journal of
Shenyang Pharmaceutical University 12, 6-.
Frank, A., Bonney, M., Bonney, S., Weitzel, L., Koeppen, M., Eckle, T., 2012. Myocardial ischemia reperfusion injury: from basic science to clinical bedside,
Semin. Cardiothorac. Vasc. Anesth. SAGE Publications Sage CA: Los Angeles, CA, pp. 123-132.
Guo, J., Xu, C.H., Xue, R., Jiang, W.X., Wu, B., Huang, C.G., 2015. Cytotoxic activities of chemical constituents from rhizomes of Anemarrhena
asphodeloides and their analogues. Arch. Pharm. Res. 38, 598-603.
Hausenloy, D.J., Yellon, D.M., 2004. New directions for protecting the heart against ischaemia–reperfusion injury: targeting the Reperfusion Injury Salvage
Kinase (RISK)-pathway. Cardiovasc. Res. 61, 448-460.
He, X.J., 2002. The metabolites research of andrographolide in rats and the active constituents research of Gualou-Xiebai-Baijiu-Tang. Shenyang
Pharmaceutical University.
20

ACCEPTED MANUSCRIPT
He, X.J., Qiu, F., Shoyama, Y., Tanaka, H., Yao, X.S., 2002. Two New Steroidal Saponins from “Gualou-xiebai-baijiu-tang” Consisting of Fructus trichosanthis
and Bulbus allii macrostemi. Chem. Pharm. Bull. (Tokyo) 50, 653-655.
He, X.J., Qiu, F., Yao, X.S., 2003. The active constituents research of Gualou Xiebai Baijiutang (IV) nitrogen containing compounds and others. Research and
Exploration of Natural Products 15, 9-12.
Jin, H.Z., Chen, G., Li, X.F., Shen, Y.H., Yan, S.K., Zhang, L., Yang, M., Zhang, W.D., 2009. Flavonoids from Rhododendron decorum. Chem. Nat. Compd. 45,
8
5-86.
Kaouadji, M., Ravanel, P., Tissut, M., Creuzet, S., 1988. Novel methylated flavonols with unsubstituted B ring from Platanus acerifolia buds. J. Nat. Prod. 51,
53-356.
3
Kizu, H., Shimana, H., Tomimori, T., 1995. Studies on the constituents of Clematis species. VI. The constituents of Clematis stans Sieb. et Zucc. Chem. Pharm.
Bull. (Tokyo) 43, 2187-2194.
Lami, N., Kadota, S., Kikuchi, T., Momose, Y., 1991a. Constituents of the roots of Boerhaavia diffusa L. III. Identification of Ca2+ channel antagonistic
compound from the methanol extract. Chemical & pharmaceutical bulletin 39, 1551-1555.
2
+
Lami, N., Kadota, S., Kikuchi, T., Momose, Y., 1991b. Constituents of the roots of Boerhaavia diffusa L. III. Identification of Ca channel antagonistic
compound from the methanol extract. Chem. Pharm. Bull. (Tokyo) 39, 1551-1555.
Lee, H.E., Kim, J.A., Whang, W.K., 2017. Chemical Constituents of Smilax china L. Stems and Their Inhibitory Activities against Glycation, Aldose Reductase,
α-Glucosidase, and Lipase. Molecules 22, 451.
Lei, Y., Huang, F., Han, L.T., Dan, H.X., Yan, B.W., Li, M.M., 2013. Influence of Trichosanthes kirilowii Maxim and Allium macrostemon Bge. on MAPK
Signal Path against Rat Acute Myocardial Ischemia. China Pharmaceuticals 22, 11.
Li, A.F., Sun, A.L., Liu, R.M., Zhang, Y.Q., Cui, J.C., 2014. An efficient preparative procedure for main flavonoids from the peel of Trichosanthes kirilowii
Maxim. using polyamide resin followed by semi-preparative high performance liquid chromatography. J. Chromatogr. B 965, 150-157.
Li, Y.L., Li, J., Wang, N.L., Yao, X.S., 2008. Flavonoids and a new polyacetylene from Bidens parviflora Willd. Molecules 13, 1931-1941.
Lin, P., Qin, Z.F., Yao, Z.H., Wang, L., Zhang, W.Y., Yu, Y., Dai, Y., Zhou, H., Yao, X.S., Metabolites profile of Gualou Xiebai Baijiu decoction (a classical
traditional Chinese medicine prescription) in rats by ultra-performance liquid chromatography coupled with quadrupole time-of-flight tandem mass
spectrometry. Under Review of Journal of Chromatography B., Mar 2018.
Lin, S.Q., Wei, X.H., Huang, P., Liu, Y.Y., Zhao, N., Li, Q., Pan, C.S., Hu, B.H., Chang, X., Fan, J.Y., Yang, X.Y., Wang, C.S., Liu, H.N., Han, J.Y., 2013.
QiShenYiQi Pills® prevents cardiac ischemia–reperfusion injury via energy modulation. Int. J. Cardiol. 168, 967-974.
Liu, C.X., Chen, S.L., Xiao, X.H., Zhang, T.J., Hou, W.B., Liao, M.L., 2016. A new concept on quality marker of Chinese materia medica: quality control for
Chinese medicinal products. Chin. Tradit. Herbal Drugs, 1443-1457.
Liu, C.X., Cheng, Y.Y., Guo, D.A., Zhang, T.J., Li, Y.Z., Hou, W.B., Huang, L.Q., Xu, H.Y., 2017. A new concept on quality marker for quality assessment and
process control of Chinese medicines. Chinese Herbal Medicines 9, 3-13.
Manivannan, J., Shanthakumar, J., Silambarasan, T., Balamurugan, E., Raja, B., 2015. Diosgenin, a steroidal saponin, prevents hypertension, cardiac
remodeling and oxidative stress in adenine induced chronic renal failure rats. RSC Adv. 5, 19337-19344.
Morel, O., Perret, T., Delarche, N., Labèque, J.-N., Jouve, B., Elbaz, M., Piot, C., Ovize, M., 2012. Pharmacological approaches to reperfusion therapy.
Cardiovasc. Res. 94, 246-252.
Ni, Y., Li, L., Zhang, W.Y., Lu, D., Zang, C.X., Zhang, D., Yu, Y., Yao, X.S., 2017. Discovery and LC-MS characterization of new crocins in Gardeniae Fructus
and their neuroprotective potential. J. Agric. Food Chem. 65, 2936-2946.
21

ACCEPTED MANUSCRIPT
1
13
Park, Y.H., Moon, B.H., Lee, E.J., Lee, Y.S., Yoon, Y.D., Ahn, J.H., Lim, Y.H., 2007. H and C‐NMR data of hydroxyflavone derivatives. Magn. Reson.
Chem. 45, 674-679.
Peng, J.P., Yao, X.S., Okada, Y., Okuyama, T., 1994. Further studies on new furostanol saponins from the bulbs of Allium macrostemon. Chem. Pharm. Bull.
(
Tokyo) 42, 2180-2182.
Rice-Evans, C.A., Miller, N.J., Paganga, G., 1996. Structure-antioxidant activity relationships of flavonoids and phenolic acids. Free Radic. Biol. Med. 20,
33-956.
9
Siddaiah, V., Rao, C.V., Venkateswarlu, S., Subbaraju, G.V., 2006. A concise synthesis of polyhydroxydihydrochalcones and homoisoflavonoids. 37, 841-846.
Wang, P.H., Lee, S.S., 1997. Active chemical constituents from Sauropus androgynus. J. Chin. Chem. Soc. 44, 145-149.
WHO, 2007. WHO international standard terminologies on traditional medicine in the western pacific region. Manila: WHO Regional Office for the Western
Pacific.
Xu, W.H., Jacob, M.R., Agarwal, A.K., Clark, A.M., Liang, Z.S., Li, X.C., 2009. Flavonol glycosides from the native American plant Gaura longiflora.
Heterocycles 78, 2541-2548.
Yang, W.Z., Zhang, Y.B., Wu, W.Y., Huang, L.Q., Guo, D.A., Liu, C.X., 2017a. Approaches to establish Q-markers for the quality standards of traditional
Chinese medicines. Acta Pharmaceutica Sinica B 7, 439-446.
Yang, W.Z., Zhang, Y.B., Wu, W.Y., Huang, L.Q., Guo, D.A., Liu, C.X., 2017b. Approaches to establish Q-markers for the quality standards of traditional
Chinese medicines. Acta pharmaceutica Sinica. B 7, 439-446.
Yao, Z., Wei, X., Weiwei, L., Chongbai, X., Ning, G., 2017. Gualoupi (Pericarpium Trichosanthis) injection in combination with convention therapy for the
treatment of angina pectoris: a Meta-analysis. J. Tradit. Chin. Med. 37, 1-11.
Yao, Z.H., Qin, Z.F., Dai, Y., Yao, X.S., 2016. Phytochemistry and pharmacology of Allii Macrostemonis Bulbus, a traditional Chinese medicine. Chinese
journal of natural medicines 14, 481-498.
Zhang, W.K., Xu, J.K., Zhang, L., Du, G.H., 2012. Flavonoids from the bran of Avena sativa. Chinese Journal of Natural Medicines 10, 110-114.
Zhu, W.F., 1999. Standards for the diagnosis and therapy of diseases in Chinese internal medicine, Changsha, China.
22

ACCEPTED MANUSCRIPT
23

ACCEPTED MANUSCRIPT
Fig. 1. The total extract of GLXB can prevent myocardial infarction after IRI and two of the fractions (GX-2 and GX-3) were effective as well. The protective effects were also shown in in
vitro H/R model. (A) Evans blue and TTC stained tissues of each group after 30 min ischemia and 90 min reperfusion. White areas are infarction areas (IA), scarlet areas are area-at-risk (AAR).
GX was administrated at 4 g/kg and the fractions were administrated at equivalent doses according to their proportions in the total extract (GX-1 3.06 g/kg, GX-2 0.34 g/kg and GX-3 0.13,
*
#
m/m). (B) Calculated area ratio of each group, P < 0.05, vs Sham, P < 0.05 vs I/R. (C) Cell viability of H9c2 cells after treatment by 4 h hypoxia and 4 h reoxygenation and different extracts
*
**
and fractions, P < 0.001, vs H/R.
HO
HO
O
O
O
O
HO
OH
OH
O
R3
OH
OH
OH
OH
O
O
OH
O
OH
O
OH
O
COOH
NH
HO
HO
HO
O
HO
O
R2
OH
OH
O
OH
OH
OH
OH
R1
OH
HO
HO
HO
HO
N
H
OH
1
2
3
9
*
*
*
R1 = β-H, R2 = S1, R3 = OCH3,25R
R1 = β-OH, R2 = S1, R3 = OH, 25R
R1 = β-OH, R2 = S1, R3 = OH, 25S
R1 = β-H, R2 = S1, R3 = OH, 25R
S1
OH
H
8
S2
H
OH
4
OH
5*
*
HO
HO
O
HO
O
O
O
O
O
O
OH
OH
R4
HO
O
OH
OH
O
OH
OH
R1
O
O
OH
OH
OH
OH
R2
R3
R1
R2
OH
R4
HO
OH
R3
O
OH
OH
O
H
H
HO
HO
OH
6
1
*
R1 = β-OH, R2 = H, R3 = S2, R4 = S1, 25S
R1 = α-OH, R2 = β-OH, R3 = S2, R4 = S1, 25R
14 R1 = H, R2 = β-OH, R3 = S2, R4 = S1, 25R
15 R1 = β-OH, R2 = H, R3 = S2, R4 = S1, 25R
R1 = R2 = H, R3 = S2, R4 = S1, 25S
18 R1 = R2 = H, R3 = S2, R4 = S1, 25R
11 R1 = H, R2 = S2, R3 = S1, R4 = β-OH
13 R1 = β-OH, R2 = S2, R3 = S1, R4 = H
16 R1 = H, R2 = S2, R3 = S1, R4 = H
19 R1 = β-OH, R2 = S3, R3 = S1, R4 = H
OH
S3
2
H
OH
1
7
O
1
0
O
O
O
2
0
R1 = H, R2 = α-OH, R3 = S3, R4 = S1, 25S
OH
OH
R4
R5
OH
OH
R1
O
OCH3
OH
OH
OH
S4
R3
R2
O
O
NH2
O
OCH3
7
*
R1 = R2 = R4 = OH, R3 = S1, R5 = S5
R1 = R2 = OH, R3 = S1, R4 = R5 = H
R1 = R2 = R4 = R5 = OH, R3 = H
R1 = S1, R2 = R4 = OH, R3 = H, R5 = OCH3
R1 = R2 = R4 = R5 = OH, R3 = -O-β-Rib
R1 = S1, R2 = R4 = R5 = OH, R3 = H
R1 = R2 = R4 = R5 = OH, R3 = S1
R1 = R2 = R4 = R5 = OH, R3 = -O-α-L-Rham
R1 = S4, R2 = OH, R3 = R5 = H, R4 = OCH
R1 = R2 = R4 = R5 = OH, R3 = S4
H
N
N
H3CO
HO
H
O
O
O
22
23
24
25
26
27
28
29
30
31
H3CO
O
N
N
S
O
O
O
OH
OH
HO
OH OH
O
OH
OH
OCH3
OH
32
S5
OH
R1 = R2 = R3 = R4 = R5 = OH
21
HO
O
HO
OH
O
OCH3
OH O
OH
33
34
Fig. 2. Structures of compounds isolated from GLXB (starred numbers signify new compounds)
24

ACCEPTED MANUSCRIPT
Fig. 3. Protective effects of GLXB compounds against H/R-induced cell death; adenosine was used as positive control (PC) at 50 μM; data were presented in protection rate (mean ± SD %, n =
*
**
***
3
), P < 0.05, P < 0.01, P < 0.001.
25

ACCEPTED MANUSCRIPT
Fig. 4. MP alleviated H/R-induced apoptosis. (A) Microscopic photos and Annexin-V FITC/PI staining and flow cytometry: apoptotic percentages were illustrated (right, mean ± SD %, n = 3),
*
**
***
P < 0.05, P < 0.01, P < 0.001. (B) Representative Western blot photos of Bax and Bcl-2 in normal control, H/R, MP (25 and 50 μM); ratios of Bax/Bcl-2 were shown in folds of normal
*
**
***
*
#
control, (mean ± SD, n = 3), P < 0.05, P < 0.01, P < 0.001, vs H/R, vs normal control.
26

ACCEPTED MANUSCRIPT
Fig. 5. MP activated RISK-pathway. (A) Representative Western blot photo of each group of cells; (B-E) Phosphorylation activation of PI3K, Akt, Erk1/2 and eNOS were illustrated as folds of
*
**
***
*
#
normal control, (mean ± SD, n = 3), P < 0.05, P < 0.01, P < 0.001, vs H/R, vs normal control
graphic abstract
27

ACCEPTED MANUSCRIPT
1
13
5
Table 1. NMR data of 1-5 (600 MHz for H-NMR and 150 MHz for C-NMR in pyridine-d )
1
2
3
4
5
Positions
δ
C
δ
H
δ
C
δ
H
δ
C
δ
H
δ
C
δ
H
δ
C
δ
H
1
2
3
4
5
6
7
8
9
78.4
67.9
72.4
27.7
37.9
66.8
35.6
35.0
41.7
43.1
21.2
39.8
41.1
56.0
32.2
81.2
64.3
16.4
19.9
4.22
4.06
78.1
68.2
71.1
32.5
78.4
72.4
37.0
34.3
45.6
46.4
21.9
40.1
41.0
56.0
32.5
81.0
64.0
16.7
14.2
4.34, br s
4.16
78.2
68.3
71.2
32.5
78.5
72.4
37.0
34.5
45.6
46.4
21.9
40.1
41.1
56.0
32.5
81.0
63.9
16.7
14.2
4.34, br s
4.16
78.5
68.0
72.5
27.7
38.0
66.8
35.7
34.8
41.6
43.2
21.4
39.8
43.6
54.4
34.7
84.4
64.6
14.4
20.0
4.25
4.08
78.6
68.0
72.4
27.7
38.0
66.8
35.8
34.8
41.6
43.2
21.4
39.8
43.7
54.4
34.5
84.3
64.6
14.4
20.0
4.24
4.08
4.65
4.76
4.75
4.68
4.68
2.73, 2.06
2.87
2.99, br d (15.6) 2.37, br d (15.6)
3.00, br d (14.7), 2.38, dd (15.0, 2.2)
2.75, 2.11
2.89
2.76, 2.10
2.90
-
4.23
-
4.25
4.46
4.48
4.49
1.95, 1.48
1.77
2.03, 1.31
1.84
2.03, 1.33
1.86
1.95, 1.50
1.73
1.95, 1.49
1.74
1.38
1.31
1.34
1.38
1.38
1
0
-
-
-
-
-
11
1.40, 1.31
1.63, 1.06
-
1.48, 1.43
1.68, 1.07
-
1.50, 1.45
1.69, 1.08
-
1.44, 1.33
1.68, 1.14
-
1.45, 1.34
1.69, 1.14
-
1
1
1
1
1
1
1
1
2
3
4
5
6
7
8
9
1.19
1.16
1.17
1.00
1.00
2.05, 1.41
4.49
2.08, 1.47
4.96, dd (14.6, 7.2)
1.97
2.07, 1.47
4.97, dd (14.8, 7.3)
1.97
2.16, 1.49
4.82
2.16, 1.50
4.84
1.80
2.51
2.50
0.82, s
1.37, s
0.91, s
1.59, s
0.91, s
1.62, s
0.71, s
1.39, s
0.71, s
1.40, s
28

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
2
0
1
2
3
4
5
6
7
40.4
16.3
112.6
30.9
28.1
34.4
75.1
17.5
105.0
75.2
78.4
71.6
78.5
2.23
1.17, d (5.6)
-
40.7
16.5
110.7
37.3
28.5
34.1
75.3
17.5
105.0
75.4
78.6
71.8
78.7
2.25
1.33, d (6.6)
-
40.7
16.5
110.7
37.2
28.4
34.2
75.3
17.5
105.2
75.4
78.6
71.8
78.7
2.24
103.8
11.7
-
1.62, s
-
103.5
11.7
152.4
23.6
31.4
33.6
75.1
17.1
105.2
75.1
78.6
71.7
78.5
-
1.61, s
-
1.32, d (6.7)
-
151.7
24.6
1.93, 1.84
1.70, 1.44
1.87
2.06
2.03
2.49, 2.40
2.50
2.25, 2.17
1.85, 1.45
1.95
2.04, 1.68
1.93
2.07, 1.69
31.0
1.93
146.2
71.7
-
4.08, 3.52
1.04, d (5.1)
4.84, d (7.7)
4.04
3.95, 3.65
0.98, d (6.6)
4.83, d (7.7)
4.04
4.08, 3.49
4.62, 4.37
5.38, 5.07
4.92, d (7.7)
4.07
4.07, 3.49
1.04, d (6.7)
4.84, d (7.8)
4.02
1.04, d (6.6)
111.6
103.9
75.1
Glc-1′
4.83, d (7.7)
2
3
4
5
′
′
′
′
4.04
4.22
4.25
4.25
78.6
4.24
4.24
4.22
4.22
4.23
3.95
71.7
4.25
4.24
3.96
3.95
78.5
3.96
3.96
4
.56, dd (11.7, 2.3), 4.40, dd (11.7,
6
′
62.8
47.2
4.56, br d (11.8), 4.38, dd (11.8, 4.5)
3.26, s
62.9
4.56, br d (11.6), 4.40, dd (11.6, 5.2)
62.9
62.8
4.56, br d (11.5), 4.40, dd (12.6, 6.1)
62.8
4.56, br d (11.8), 4.39, dd (11.6, 5.0)
5.2)
2
2-OCH
3
29