Structural features and anti-gastric cancer activity of polysaccharides from stem, root, leaf and flower of cultivated Dendrobium huoshanense

Article abstract of DOI:10.1016/j.ijbiomac.2019.12.041

The structure features and anti-gastric cancer activities in vitro of stem, root, leaf and flower polysaccharides from cultivated Dendrobium huoshanense were investigated systematically. Stem polysaccharide (cDHPS) was composed of →4)-β-D-Glcp-(1→, →4)-β-

Full text of DOI:10.1016/j.ijbiomac.2019.12.041

Journal Pre-proof
Structural features and anti-gastric cancer activity of
polysaccharides from stem, root, leaf and flower of cultivated
Dendrobium huoshanense
Bing Liu, Zhen-Zi Shang, Qiang-Ming Li, Xue-Qiang Zha, De-
Ling Wu, Nian-Jun Yu, Lan Han, Dai-Yin Peng, Jian-Ping Luo
PII:
S0141-8130(19)37923-1
DOI:
https://doi.org/10.1016/j.ijbiomac.2019.12.041
BIOMAC 14068
Reference:
To appear in:
International Journal of Biological Macromolecules
Received date:
Revised date:
Accepted date:
30 September 2019
2 December 2019
5 December 2019
Please cite this article as: B. Liu, Z.-Z. Shang, Q.-M. Li, et al., Structural features and anti-
gastric cancer activity of polysaccharides from stem, root, leaf and flower of cultivated
Dendrobium huoshanense, International Journal of Biological Macromolecules(2018),
https://doi.org/10.1016/j.ijbiomac.2019.12.041
This is a PDF file of an article that has undergone enhancements after acceptance, such
as the addition of a cover page and metadata, and formatting for readability, but it is
not yet the definitive version of record. This version will undergo additional copyediting,
typesetting and review before it is published in its final form, but we are providing this
version to give early visibility of the article. 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.
©
2018 Published by Elsevier.

Journal Pre-proof
Title:
Structural features and anti-gastric cancer activity of polysaccharides from stem,
root, leaf and flower of cultivated Dendrobium huoshanense
Authors:
a,b
a,b
a,b
a,b
c
c
Bing Liu , Zhen-Zi Shang , Qiang-Ming Li , Xue-Qiang Zha , De-Ling Wu , Nian-Jun Yu ,
a,b
c
a,b,
Lan Han , Dai-Yin Peng , Jian-Ping Luo *
Affiliations:
a
School of Food and Biological Engineering, Hefei University of Technology, Hefei 230009,
China
b
Engineering Research Center of Bio-process, Ministry of Education, Hefei University of
Technology, Hefei 230009, China
c
School of Pharmacy, Anhui University of Chinese Medicine, Hefei, 230012, China
*
Corresponding author:
Prof. Jian-Ping Luo, School of Food and Biological Engineering, Hefei University of Technology,
Hefei 230009, People’s Republic of China. E-mail: jianpingluo@hfut.edu.cn.
1

Journal Pre-proof
Graphical abstract:
2

Journal Pre-proof
Abstract:
The structure features and anti-gastric cancer activities in vitro of stem, root, leaf and flower
polysaccharides from cultivated Dendrobium huoshanense were investigated systematically. Stem
polysaccharide (cDHPS) was composed of →4)-β-D-Glcp-(1→, →4)-β-D-Manp-(1→,
5
→
4)-3-O-acetyl-β-D-Manp-(1→ with the molecular weight of 2.59×10 Da; root polysaccharide
(
cDHPR) was composed of →3,5)-α-L-Araf-(1→, →4)-β-D-Glcp-(1→, →4)-β-D-Manp-(1→,
→
4,6)-β-D-Manp-(1→, →6)-α-D-Galp-(1→ and terminal β-L-Araf with the molecular weight of
4
1
.41×10 Da; leaf polysaccharide (cDHPL) was composed of →4)-β-D-Glcp-(1→,
→
4)-β-D-Manp-(1→, →4)-3-O-acetyl-β-D-Manp-(1→, →3,6)-β-D-Manp-(1→ and terminal
5
α-D-Galp with the molecular weight of 2.09×10 Da; and flower polysaccharide (cDHPF) was
composed of →4)-β-D-Glcp-(1→, →4)-β-D-Manp-(1→, →3,6)-β-D-Manp-(1→ and terminal
5
α-D-Galp with the molecular weight of 4.78×10 Da. Among these four polysaccharides, cDHPS
showed the best anti-gastric cancer activity evidenced by the inhibited growth and c-myc
expression as well as the enhanced apoptosis and p53 expression of murine forestomach
carcinoma (MFC) cells, suggesting their difference in anti-gastric cancer activity should be
contributed to their difference in structure features.
Keywords: Dendrobium huoshanense polysaccharide; Structural feature; Anti-gastric cancer
activity.
3

Journal Pre-proof
1
Introduction
Dendrobium huoshanense C. Z. Tang et S. J. Cheng (traditionally called as ‘Mihu’) is a
perennial plant of Orchidaceae family and has been used in the traditional Chinese medicine over
000 years [1]. D. huoshanense contains diverse chemical constituents, mainly including
2
polysaccharides, alkaloids, aromatic compounds and sesquiterpenoids, and possesses diverse
biological functions, mainly including immunomodulatory, antioxidative, hepatoprotective,
anti-cataract, neuroprotective and anti-tumor activities [2]. Since the water extract of D.
huoshanense is traditionally consumed as a tonic with the potential of regulating gastrointestinal
function, the structures and functions of the water-soluble active constituents, especially
polysaccharides that account for more than 70% in water extract, have attracted much attention.
In the past decade, the tissue-cultured D. huoshanense at different growth stages including
protocorm-like bodies and micropropagated plantlets, which were grown on Murashige and
Skoog medium under the controlled conditions in the sterile plant tissue rooms, were employed as
materials for the isolation, purification, structural identification and bioactivity assessment of
polysaccharides due to the limitation of wild resource, which results from slow growth, habitat
destruction, over-collection and commercial trade [1,3-7].
In view of the fact that the tissue-cultured materials cannot be used for medicinal purpose or
health care product development, great efforts have made to improve artificial cultivation
technology of D. huoshanense since 1960 [8]. Recently, D. huoshanense has been successfully
cultivated in its natural habitat, and the stem, leaf and flower of cultivated D. huoshanense as well
as their water extracts have been approved as medicinal and edible materials. However, very little
research has been conducted up to date on the structure and activity of polysaccharides from
4

Journal Pre-proof
cultivated D. huoshanense. It is not clear whether the structural feature of polysaccharides from
artificially cultivated D. huoshanense is different from that of tissue-cultured D. huoshanense due
to the difference in growth patterns and environments.
We previously reported the monosaccharide compositions and glycosidic linkage types of
total polysaccharides from the cultivated D. huoshanense stems [9]. In the present study, the
homogeneous polysaccharides from stem, root, leaf and flower of the cultivated D. huoshanense
were isolated and purified, and their structure features were systematically investigated by
HPGPC, FT-IR, HPLC, GC-MS and NMR techniques. Moreover, the anti-gastric cancer activity
of these four polysaccharides was compared according to the regulated proliferation, apoptosis
and gene expression of gastric cancer cells.
5

Journal Pre-proof
2
2
Materials and methods
.1 Materials and reagents
The cultivated D. huoshanense plants were collected at 2016 from Huoshan (Anhui Province
of China), which were grown from seedlings of seeds under artificial-sheltered forests. Murine
forestomach carcinoma (MFC) cell line was purchased from Shanghai Institutes for Biological
Sciences, Chinese Academy of Sciences. DEAE-Cellulose DE-52, Sephadex G-50 and Sephacryl
S-400 were purchased from GE Healthcare (Uppsala, Sweden). Dimethyl sulphoxide (DMSO),
methyl iodide (CH I) and deuterium oxide (D O), monosaccharide standards including
3
2
D-mannose (D-Man), L-rhamnose (L-Rha), D-glucose (D-Glc), D-galactose (D-Gal) and
L-arabinose (L-Ara), as well as dextran series standards including 5000, 25000, 80000, 150000,
4
20000 and 670000 Da were purchased from Sigma-Aldrich (St Louis, Missouri, USA).
Trifluoroacetic acid (TFA), sodium borodeuteride (NaBD ) and sodium periodate were obtained
4
from Aladdin Industrial Corporation (Shanghai, China). Dulbecco’s modified eagle medium/high
glucose (DMEM-H) was purchased from Hyclone (Logan, Utah, USA) and fetal bovine serum
(FBS) was purchased from Clark Bioscience (Richmond, Virginia, USA). CCK-8 reagent was
provided by Beyotime Institute of Biotechnology (Shanghai, China). The Annexin V-FITC
apoptosis assay kit was provided by BD Biosciences (San Jose, California, USA). TRI®zol was
provided by Invitrogen (Carlsbad, California, USA). RevertAid First Strand cDNA Synthesis Kit
TM
was provided by Thermo Fisher Scientific (Waltham, Massachusetts, USA) and the iTaq
@
Universal SYBR Green Master Mix kit was provided by Bio-Rad Laboratories (Hercules,
California, USA). All primers were obtained from Sangon Biotech Co., Ltd (Shanghai, China).
The 5-fluorouracil (5-FU) was purchased from Shanghai EKEAR Bio-tech Co., Ltd (Shanghai,
China). All other reagents were from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).
6

Journal Pre-proof
2
.2 Isolation and purification of polysaccharides
The stems, roots, leaves and flowers from whole plants of cultivated D. huoshanense were
separated firstly and then dried to constant weight in oven at 60 °C. Fifty grams of the materials
were crushed into powder and were pre-extracted with 2 L 95 % ethanol at 25 ºC for 24 h.
Subsequently, the residue was extracted three times with 2 L distilled water at 70 ºC, each for 1 h.
All of the extract was filtered and centrifuged at 5000 g for 20 min at 4 ºC. The supernatant was
concentrated by vacuum rotary evaporator under reduced pressure at 60 ºC and precipitated with
ethanol at a final concentration of 80 % (v/v) allowing to stand overnight at 25 ºC. The α-amylase
(46 U/mL) of 50 μL was added into the precipitate which was dissolved in 1 L distilled water to
hydrolyze starch at 60 ºC for 1 h. The hydrolyzed mixture was collected by centrifuging at 5000 g
for 20 min at 4 ºC. The supernatant, after concentrated by vacuum rotary evaporator under
reduced pressure at 60 ºC, was mixed with ethanol at a final concentration of 80 % (v/v) to obtain
the precipitate at 25 ºC for 24 h. The resulting precipitate was dissolved in distilled water and
lyophilized at −50 °C below 20 Pa for 48 h to yield the crude polysaccharides. Dissolved crude
polysaccharide (10 % w/v) in distilled water was fractionated on DEAE-Cellulose DE-52
anion-exchange column (1.6 cm × 60 cm) through stepwise elution with pure water and NaCl
solutions of 0.1-0.5 M at a flow rate of 5 mL/min. Fractions from water elution were further
purified through gel permeation chromatography on a Sephadex G-50 column (1.6 cm × 80 cm)
or Sephacryl S-400 column (1.6 cm × 80 cm) with pure water elution. After concentrated by
vacuum rotary evaporator under reduced pressure at 60 ºC, dialyzed at 25 °C for 24 h (molecular
weight cut off of 3500 Da) and lyophilized at −50 °C below 20 Pa for 48 h, the polysaccharides
including cDHPS from stems, cDHPR from roots, cDHPL from leaves and cDHPF from flowers
7

Journal Pre-proof
of cultivated D. huoshanense were obtained.
.3 Measurement of physicochemical properties
The carbohydrate content was determined by phenol-sulfuric acid method [10]. The O-acetyl
2
group content was determined by hydroxylamine colorimetry [11]. The aqueous solutions (2
mg/mL) of polysaccharides were scanned on a TU-1901 UV-Vis spectrophotometer (Persee,
Beijing, China) within the wavelength range of 190 ~ 400 nm.
2
.4 HPGPC and FT-IR analysis
High performance gel permeation chromatography (HPGPC) was used to determine the
homogeneity and the relative molecular weight (Mw) of polysaccharide. Agilent 1260 Infinity
system equipped with the TSK G5000PWXL (7.8 mm × 300 mm) column under the circumstance
of 35 C was used to determine the retention times of samples with refractive index detector
(RID). The ultrapure water was used as mobile phase with a flow rate of 0.5 mL/min. The
retention time of dextran series standards of 5000, 25000, 80000, 150000, 420000 and 670000 Da
was measured under the same conditions. Then the molecular weight standard curve was fitted
using Agilent GPC data analysis in ChemStation B.04.03 and the Mw of cDHPS, cDHPR,
cDHPL and cDHPF was calculated according to the standard curve.
For FT-IR analysis, the dried polysaccharide samples were mixed with KBr powder and
pressed into flake, and then scanned on Nicolet 6700 spectrometer (Thermo Nicolet, Madison, WI,
-1
USA) to record FT-IR spectrum ranging from 4000 to 400 cm .
2
.5 Monosaccharide composition analysis
Five milligrams of dried samples were dissolved in 3 mL of 2 M trifluoroacetic acid (TFA)
and 1 mL of glucosamine hydrochloride (0.96 mg/mL) was added as an internal standard. The
8

Journal Pre-proof
mixed solution was hydrolyzed at 110 ºC for 4 h. After the solution was cooled to room
temperature, the reaction mixture was evaporated at 60 ºC under reduced pressure to remove TFA
and the residue was dissolved in 4 mL of ultrapure water. Then, 400 μL of the sample solution
was added with 400 μL of 1-phenyl-3-methyl-5-pyrazalone (PMP) methanol solution (0.5 M) and
4
00 μL of NaOH solution (0.3 M), and was permitted to react in water bath at 70 C for 100 min.
Subsequently, the sample solution was washed three times with chloroform following the addition
of HCl solution (0.3 M) with 500 μL, and the resulting aqueous layer was collected by
centrifugation at 500 g at 4 C for 15 min and filtration with 0.22 μm membrane to obtain the
PMP-labeled samples for HPLC analysis on Agilent 1260 Infinity system equipped with a UV
detector. The mobile phase was acetonitrile-0.02 M ammonium acetate solution (2 : 8, v/v) with a
flow rate of 1 mL/min and the Agilent XDB-C18 column (250 mm × 4.6 mm) was used at the
circumstance of 25 ºC [12].
2
.6 Periodate oxidation and Smith degradation
The dried samples (20 mg) were dissolved in 10 mL double distilled water and reacted with
3
0 mM NaIO (10 mL) under 4 °C in the dark. Ten microliters of the reaction solution were
4
monitored every 6 h with the absorption at 223 nm until a constant value and the un-reacted
NaIO was quenched with ethylene glycol. Then the production of formic acid was determined by
4
titration with 0.1 M NaOH (phenolphthalein as an indicator). Further, the freshly prepared NaBH4
solution (1 M) was added into reaction solution to reduce polyaldehyde and the resulting polyol
was desalted by dialysis bag of molecular weight cutoff of 3500 Da followed by lyophilization at
−
50 °C below 20 Pa for 48 h. The lyophilizate was dissolved in 4 M trifluoroacetic acid (TFA),
hydrolyzed at 110 °C for 4 h and acetylated with pyridine and acetic anhydride. The acetylated
9

Journal Pre-proof
products were extracted by dissolving them with an appropriate amount of dichloromethane and
the dichloromethane layer was analyzed by GC-MS following by dehydration with anhydrous
sodium sulfate and filtration with 0.22 μm organic membrane [13]. Agilent
GC7890A-MSD5975C instrument equipped with a HP-5 capillary column (30 m × 0.32 mm
×
0.25 μm) was used. The inlet temperature was 280 °C and the procedure of column temperature
was set as follows: first 150 °C for 1 min, then rise to 200 °C at 10 °C/min, finally increase to
2
1
2
70 °C at 4 °C/min. The flow rate of helium was 1 mL/min and temperature of ion source was
50 °C.
.7 Methylation analysis
The methylation method was carried out according to the modified Ciucanu method [14].
The dried sample (20 mg) was dissolved in 5 mL DMSO, followed by the addition of 200 mg
NaOH powder which was fully ground in liquid nitrogen and the sonication at the frequency of
9
0 KHz for 50 min at 35 °C. Then, the mixture in ice bath was treated with methyl iodide in the
dark. The complete methylation was detected by FT-IR according to the disappearance of
hydroxyl groups. All the above procedures were performed under a nitrogen flow condition. Then
the dried methylated products were depolymerized with formic acid at 100 °C for 3 h followed by
a further acid hydrolysis with 4 M TFA at 110 °C for 4 h. Lastly, the partially methylated sugar
residues were reduced with NaBD , acetylated with pyridine and acetic anhydride, and analyzed
4
by GC-MS as described in Section 2.6.
2
.8 NMR analysis
One hundred milligrams of the samples were exchanged with deuterium oxide (D O) by
2
lyophilizing for three times. The lyophilizate was dissolved in 99.9 % D O containing sodium
2
10

Journal Pre-proof
4
,4-dimethyl-4-silapentan-sulfonate (DSS) and transferred to the NMR tube (5 mm, Wilmad-Lab
1
Glass, Buena, NJ, USA) after filtration of 0.22 μm membrane. All NMR spectra including H
13
1
1
(
298 K and 328 K), C, H- H COSY, HSQC, HMBC, TOCSY, ROESY were recorded on a
Varian VNMRS 600 MHz superconducting nuclear magnetic resonance spectrometer with a 5
1
mm OneNMR probe (Agilent Technologies Inc., USA). The operating frequencies of H NMR
13
and C NMR were 599.81 MHz and 150.84 MHz. The 1D NMR spectra were recorded using
s2pul pulse sequence and the 2D NMR spectra were acquired using gCOSY, zTOCSY, HSQCAD,
gHMBCAD and ROESYAD pulse sequences, respectively.
2
.9 Assessment of cell viability
The MFC cells were seeded in 96 well plates at a density of 5×10 cells/well. After 24 h
3
adhesion, the cells were incubated with cDHPS, cDHPR, cDHPL or cDHPF at the concentrations
of 0.025, 0.05, 0.1, 0.2, 0.4, 0.8, 1.0, 1.5, 2.0 and 2.5 mg/mL, respectively. Medium and 5-Fu (0.1
mg/mL) were served as the normal control (NC) and positive control (PC), respectively. After
incubation for 24 h at 37 °C in a humidified incubator with 5 % CO , CCK-8 reagent (10 μL/well)
2
was added into each well and the plates were further incubated for 2 h. Subsequently, the optical
density (OD) values were measured at 450 nm on the Multiskan FC microplate reader (Thermo
Fisher Scientific, Finland). The cell viability was calculated as follows: the cell viability (%) =
(
1–ODexperiment/ODblank control) × 100 %. All assays were separately repeated three times and the half
maximal inhibitory concentration (IC ) was also calculated.
50
2
.10 Detection of cell apoptosis
The MFC cells were seeded at a concentration of 1×10 cells/well into a 6-cell plate and
incubated for 12 h at 37 °C in a humidified incubator with 5 % CO before the addition of cDHPS,
6
2
11

Journal Pre-proof
cDHPR, cDHPL or cDHPF at the concentration of 0.025, 0.05, 0.1, 0.2, 0.4, 0.8, 1.0 and 1.5
mg/mL for 24 h, respectively. Medium and 5-Fu (0.1 mg/mL) were served as the normal control
(
NC) and positive control (PC), respectively. The cells were harvested by centrifuging at 2000 g
for 10 min at 4 °C after digested with EDTA-free trypsin. Then the cells were washed with
pre-cooled phosphate buffer (0.01 M, pH 7.4) and resuspended in 300 μL 1 × binding buffer
provided by the apoptosis assay kit. Annexin V-FITC (5 μL) and 5 μL of propidiom iodide (PI)
from assay kit were added to the cell solution and incubated for 30 min in the dark at room
@
temperature. The sample was detected by MoFlo XDP flow cytometry (Beckman coulter, USA)
and the percentages of early and late apoptotic rates were quantified by FlowJo software 7.6.
2
.11 Quantitative RT-PCR analyses
The MFC cells were seeded into a 24-well plate at a density of 1×10 cells/mL. As described
above, MFC cells were treated with cDHPS, cDHPR, cDHPL or cDHPF at the concentration of
.025, 0.05, 0.1, 0.2, 0.4, 0.8, 1.0 and 1.5 mg/mL for 24 h incubation at 37 °C in a humidified
6
0
incubator with 5 % CO , respectively. Medium and 5-Fu (0.1 mg/mL) were served as the normal
2
control (NC) and positive control (PC), respectively. The TRI®zol reagent was employed to
extract total RNA, which was reverse-transcribed using RevertAid First Strand cDNA Synthesis
Kit. The mixture was incubated in PCR system (Applied Biosystems GeneAmp 2400, Thermo
Fisher, USA) at 42 °C for 60 min, followed by incubating at 70 °C for 5 min. PCR reaction was
performed on a iQ5 Real Time PCR system (BIO-RAD, USA) with a reaction volume of 20 μL
TM
@
containing 10 μL 2 × iTaq Universal SYBR Green Master Mix, 0.6 μL forward primer, 0.6 μL
reverse primer, 1 μL cDNA and 7.8 μL RNase-Free ddH O. Primers were designed as follows:
2
5
’-TGTGGAGTATTTGGATGACA-3’ (Forward) and 5’-GAACATGAGTTTTTTATGGC-3’
12

Journal Pre-proof
(Reverse) for the p53 gene; 5’-GAACAAGAAGATGAGGAAGA-3’ (Forward) and
5
5
’-AGTTTGTGTTTCAACTGTTC-3’
(Reverse)
for
the
c-myc
gene;
’-GAGACCTTCAACACCCAGCC-3’ (Forward) and 5’- GCGGGGCATCGGAACCGCTCA-3’
(
Reverse) for the β-actin gene (use as an internal control). The thermal cycling procedures were
set as follows: initial denaturation at 94 °C for 1 min followed by 40 cycles of denaturation at
5 °C for 20 s, annealing at 60 °C for 1 min, extension at 72 °C for 2 min, and a final extension
was at 72 °C for 6 min.
9
2
.12 Statistical analysis
All experiment results were expressed as the mean ± SE. All data were analyzed by one-way
analysis of variance (ANOVA) and multiple comparisons were performed by SPSS 23.0 software.
13

Journal Pre-proof
3
3
Results
.1 Physicochemical properties of stem, root, leaf and flower polysaccharides
After water extraction, ethanol precipitation and α-amylase digestion as well as purification
on anion-exchange and gel permeation chromatography, polysaccharides including cDHPS from
stems, cDHPR from roots, cDHPL from leaves and cDHPF from flowers were obtained with the
yield of 13.14±1.44 %, 0.81±0.46 %, 3.68±0.83 % and 2.77±0.37 %, respectively. The
carbohydrate content of these polysaccharides was over 95 %, with O-acetyl group contents
ranging from 0.39 % to 2.78% (Table 1). The UV scanning (Supplemental Figure 1) suggested
that these polysaccharides were free of proteins and nucleic acids. In the HPGPC spectra (Fig. 1A,
2
A, 3A and 4A), cDHPS, cDHPR, cDHPL and cDHPF all exhibited a symmetrical peak,
2
indicating their homogeneity. According to standard curve (y = - 0.3122 x + 10.39, R = 0.9976),
5
the relative molecular weights of cDHPS, cDHPR, cDHPL and cDHPF were 2.59×10 Da,
4
5
5
1
.41×10 Da, 2.09×10 Da and 4.78×10 Da, respectively.
FT-IR spectra (Fig. 1B, 2B, 3B and 4B) demonstrated that cDHPS, cDHPR, cDHPL and
cDHPF had the characteristic peaks of polysaccharides. The broad intense peak at around 3700
-1
-1
-1
cm —3100 cm was attributed to the stretching vibration of -OH. The peaks at around 2930 cm
-1
and 2880 cm were from the stretching vibration of C-H. Characteristic absorptions for the C-O
vibration of C-O-C and C-O-H in the pyranose rings of polysaccharides were occurred between
-1
-1
-1
-1
1
250 cm and 950 cm . The peaks between 1400 cm and 1200 cm were from the symmetric
-1
C-H bending vibration, the peak at 1740 cm from the stretching vibration of C=O and the peak
-1
-1
at 1245 cm from the stretching vibration of C-O. In addition, the peak at 895 cm for
β-glycosidic bond in all polysaccharides could be found in FT-IR spectra.
14

Journal Pre-proof
3
.2 Structural characterization of stem, root, leaf and flower polysaccharides
.2.1 cDHPS
3
cDHPS was mainly composed of D-Man and D-Glc in a molar ratio of 3.04 : 1.00
(
Supplemental Figure 2). The analysis of periodate oxidation found that 1.77 mol NaIO was
4
consumed without the production of formic acid, indicating the existence of (1→2), (1→2, 6) or
1→4), (1→4,6)-linked glycosidic bonds. The oxidized products after reduced, hydrolyzed and
(
acetylated, were subjected to GC-MS analysis (Supplemental Figure 3), and the presence of
ethylene glycol and erythritol confirmed the existence of (1→4) or (1→4,6)-linked glycosidic
bonds in cDHPS. Methylation analysis found two peaks in the TIC profile of cDHPS
(
Supplemental Figure 4 and Supplemental Table 1), which were identified as 2,3,6-Me -Manp
3
and 2,3,6-Me -Glcp with a molar ratio of 2.88 : 1.00. These results suggested that cDHPS
3
contained two glycosidic linkages, i.e. 1,4-linked Manp and 1,4-linked Glcp.
1
In order to avoid the interference of deuterium oxide (D O) solvent peak, the analysis of H
2
NMR spectrum was carried out at temperature of 298 K (Fig. 1C) and 328 K (Fig. 1D),
respectively, and three anomeric protons were found at δ 4.52 ppm, δ 4.75 ppm and δ 4.77 ppm,
which were named as residue A , B and C .
S
S
S
The residue A should be β-configuration on the basis of its anomeric proton signal at δ 4.52
S
3
ppm with a large coupling constant value of J (~7.42 Hz), which was supported by its anomeric
1
,2
13
carbon signal at δ 105.49 in C NMR spectrum (Fig. 1E) and the existence of a cross signal at δ
4
.52/105.49 ppm in HSQC spectrum (Fig. 1F). According to the chemical shift of H-1, the cross
signal at δ 4.52/3.33 ppm (H-1/H-2) and δ 3.33/3.62 ppm (H-2/H-3) could be discerned in the
COSY spectrum (Fig. 1G). Although H-3 (δ 3.63 ppm) and H-4 (δ 3.62 ppm) were very close to
15

Journal Pre-proof
each other and almost close to the diagonal, H-4/H-5 (δ 3.63/3.56 ppm) could be easily found
1
based on the characteristics of H-5. In the TOCSY spectrum (Supplemental Figure 5A), the H
chemical shifts were further confirmed by the spatial relationship among protons inside the
residue with the cross-signals of H-1/H-2 (δ 4.52/3.33 ppm), H-1/H-3 (δ 4.52/3.62 ppm), H-1/H-4
(
δ 4.52/3.63 ppm), H-1/H-6a (δ 4.52/3.91 ppm) and H-1/H-6b (δ 4.52/3.82 ppm). In the HSQC
spectrum (Fig. 1F), cross-peaks for H-2/C-2 (δ 3.33/75.49 ppm), H-3/C-3 (δ 3.62/76.77 ppm),
H-4/C-4 (δ 3.63/80.24 ppm) and H-5/C-5(δ 3.56/77.72 ppm) could be found while the peak for
H-6/C-6 was overlapped with other residue signals. All chemical shifts of the corresponding
13
carbons were confirmed in the C NMR spectrum. Thus, the residue A was established as
S
→
4)-β-D-Glcp-(1→ as comparing above analysis with literature’s data [15], and the detail
information was listed in Table 2.
As for the residue B , its anomeric proton at δ 4.75 ppm in the H NMR spectrum (Fig. 1D)
1
S
1
3
and anomeric carbon at δ 102.87 ppm in the C NMR spectrum (Fig. 1E), which was evidenced
by the dominant coupling signal at δ 4.75/102.87 ppm in the HSQC spectrum (Fig. 1F), suggested
it was β-configuration. In the COSY spectrum (Fig. 1G), proton chemical shifts of H-1/H-2 (δ
4
.75/4.11 ppm), H-2/H-3 (δ 4.11/3.72 ppm) and H-4/H-5 (δ 3.75/3.50 ppm) were further assigned
although the chemical shifts of H-3 and H-4 were very close to each other as found for residue AS.
Based on the proton chemical shifts, the correlations between H-1/C-1 (δ 4.75/75.49 ppm),
H-2/C-2 (δ 4.11/72.61 ppm), H-3/C-3 (δ 3.72/74.82 ppm), H-4/C-4 (δ 3.78/79.18 ppm) and
H-5/C-5 (δ 3.50/77.77 ppm) could be found in the HSQC spectrum (Fig. 1F), where the
cross-peak of H-6/C-6 was overlapped with other peaks. Comparing above analysis with
literature’s data [15], it could be assumed that the residue B should be →4)-β-D-Manp-(1→, and
S
16

Journal Pre-proof
all signals of H and C resonances were shown in Table 2.
As the residue B , the residue C also existed as the β-linkage, which was supported by the
S
S
1
13
anomeric proton at δ 4.77 ppm in H NMR (Fig. 1D), the anomeric carbon at δ 101.99 ppm in C
NMR (Fig. 1E) and their coupling signal at δ 4.77/101.99 ppm in HSQC (Fig. 1F). Further, the
signals for other H and C, which were shown in Table 2, were assigned according to COSY
spectrum (Fig. 1G) and HSQC spectrum (Fig. 1F). Compared with the residue B , the signal for
S
H-3 of the residue C shifted to the lower field, while its C-2 and C-4 signals moved toward the
S
higher field. These changes may be attributed to the presence of O-acetyl group, which was
1
revealed both by H NMR spectrum, where three hydrogens on the methyl group in O-acetyl
1
3
group occurred around δ 2.19 ppm, δ 2.18 ppm and δ 2.15 ppm (Fig. 1C), and by C NMR
spectrum, where the resonances of carbonyl carbons and methyl carbons in O-acetyl group
occurred at δ 174.17 ppm and δ 23.04 ppm (Fig. 1E). Considering the methylation analysis of
glycosidic linkages, the residue C could be assigned as →4)-3-O-acetyl-β-D-Manp-(1→.
S
In the upper left region of the HMBC spectrum (Fig. 1H), it was easy to find the
inter-residual cross peaks between A H-1/ A C-4 (δ 4.75/80.24 ppm), A H-1/ C C-4 (δ 4.52/
S
S
S
S
7
5.70 ppm), and C H-1/ A C-4 (δ 4.77/79.18 ppm). Thus, the possible repeat unit of cDHPS
S S
could be inferred as follows:
3
.2.2 cDHPR
cDHPR consisted of D-Man, D-Glc, D-Gal and L-Ara in a molar ratio of 2.38 : 1.00 : 8.49 :
17

Journal Pre-proof
5
.23 (Supplemental Figure 2). Further, the results of periodate oxidation and Smith degradation
experiment (Supplemental Figure 3) and methylation analysis (Supplemental Figure 4 and
Supplemental Table 1) implied that the glycosidic linkages of cDHPR included 1-linked-Araf,
1
,4-linked Manp, 1,4-linked Glcp, 1,6-linked Galp, 1,4,6-linked Manp and 1,3,5-linked Araf in a
molar ratio of 3.55 : 1.74 : 1.52 : 8.37 : 1.00 : 2.41.
1
In the H NMR spectra at 298 K (Fig. 2C) and at 328 K (Fig. 2D), the signals for anomeric
protons of cDHPR were found at δ 5.24 ppm, δ 4.49 ppm, δ 4.73 ppm, δ 5.02 ppm, δ 4.57 ppm
and δ 5.08 ppm, which were named as residue A , B , C , D , E and F , respectively. Among
R
R
R
R
R
R
these residues, B and C were basically the same as the residue A and B of cDHPS, and were
R
R
S
S
respectively assigned as →4)-β-D-Glcp-(1→ and →4)-β-D-Manp-(1→ according to the
information from 1D and 2D NMR (Fig. 2 and Table 2).
For the residue A , the anomeric proton at δ 5.24 ppm (Fig. 2C and 2D), the anomeric
R
carbon at δ 111.82 ppm (Fig. 2E) and their coupling signal at δ 5.24/111.82 ppm (Fig. 2F)
indicated that it presented as α-configuration. In COSY (Fig. 2G), the cross signals for H-1/H-2,
H-2/H-3, H-3/H-4 and H-4/H-5 were found at δ 5.24/3.97 ppm, δ 3.97/3.96 ppm, δ 3.96/4.15 ppm
and δ 4.15/3.80 ppm with the assistance of TOCSY spectrum (Supplemental Figure 5C).
Accordingly, the cross signals for H-2/C-2, H-3/C-3, H-4/C-4 and H-5/C-5 were found at δ
8
4.53/3.97 ppm, δ 85.94 /3.96 ppm, δ 82.98 /4.15 ppm and δ 68.51 /3.80 ppm in HSQC. Thus, the
residue A could be attributed to →3,5)-α-L-Araf-(1→ based on the methylation analysis and
R
literature data [16].
For the residue D , the anomeric region signal at δ 5.02 ppm (Fig. 2C and 2DB), δ 101.69
R
1
13
ppm (Fig. 2E) and their coupling signal at δ 5.02/101.69 (Fig. 2F) were found in H NMR, C
18

Journal Pre-proof
NMR and HSQC, respectively. Furthermore, signals for H-2 to H-5 and C-2 to C-5 were found in
sequence in COSY (Fig. 2G) and HSQC (Fig. 2F), respectively. Consulting methylation analysis
and literature data [17], the residue D could be designated as β-L-Araf-(1→.
R
The residue E existed as β-configuration, which was inferred based on chemical shifts of
R
anomeric proton at δ 4.57 ppm (Fig. 2C and 2D), anomeric carbon at δ 100.21 ppm (Fig. 2E) and
their coupling-signal at δ 4.57 ppm/100.21 ppm (Fig. 2F). In COSY (Fig. 2G), the H-1/H-2 (δ
4
.57 ppm/3.96 ppm), H-2/H-3 (δ 3.96 ppm/4.08 ppm), H-3/H-4 (δ 4.08 ppm/3.77 ppm), H-4/H-5
δ 3.77 ppm/3.78 ppm) and H-5/H-6 (δ 3.78 ppm/3.95 ppm) were found. This assignment was
supported by the correlations between δ 3.96 ppm/70.98 ppm (H-2/C-2), δ 4.08 ppm/70.35 ppm
H-3/C-3), δ 3.77 ppm/78.22 ppm (H-4/C-4), δ 3.78 ppm/69.91 ppm (H-5/C-5) and δ 3.78
(
(
pm/68.36 ppm (H-6/C-6) in HSQC (Fig. 2F). Compared with the residue C , the chemical shift of
R
C-6 (δ 68.36 ppm) in the residue E moved to the lower field, suggesting the occurrence of
R
glycosylation [18]. Thus, the residue E should be →4,6)-β-D-Manp-(1→.
R
Residue F had anomeric proton at δ 5.08 ppm (Fig. 2C and 2D) and anomeric carbon at δ
R
9
8.73 ppm (Fig. 2E), as well as their cross-signal at δ 5.08/98.73 ppm in HSQC (Fig. 2F),
indicating it was α-configuration residue. Starting from H-1, the signals of H-2 (δ 3.99 ppm), H-3
δ 3.69 ppm), H-4 (δ 4.06 ppm), H-5 (δ 3.92 ppm), H-6a (δ 3.66 ppm) and H-6b (δ 3.76 ppm)
(
were found in COSY (Fig. 2G) and TOCSY (Supplemental Figure 5C). Correspondingly, the
chemical shifts from C-2 to C-6 (δ 70.68 ppm, δ 69.13 ppm, δ 68.99 ppm, δ 71.90 ppm and δ
6
6.22 ppm) were found in HSQC. Referring to the literature data[19], the residue F could be
R
assigned as →6)-α-D-Galp-(1→, where the downfield shift of C-6 (δ 66.22 ppm) signal
suggested the formation of glycosidic bond.
19

Journal Pre-proof
The linkage sequence of glycosyl residues in cDHPR was further derived according to the
cross-peaks in HMBC (Fig. 2H). These cross-peaks included δ 5.24/68.99 ppm, δ 4.49/66.22 ppm,
δ 4.73/68.51 ppm, δ 4.57/80.98 ppm, δ 5.02/68.36 ppm, δ 5.02/68.51 ppm, δ 5.02/85.94 ppm, δ
5
.08/66.22 ppm, δ 5.08/78.22 ppm and δ 5.08/79.58 ppm, indicating the linkages of A H-1 / F
R R
C-4, B H-1/ F C-6, C H-1/ A C-5, E H-1 / B C-4, D H-1/E C-6, D H-1/A C-5, D
R
R
R
R
R
R
R
R
R
R
R
H-1/A C-3, F H-1/F C-6, F H-1/E C-4 and F H-1/C C-4 in cDHPR, which were evidenced
R
R
R
R
R
R
R
by the ROESY experiment (Supplemental Figure 5D). Accordingly, the structure of cDHPR could
be proposed to be composed of →3,5)-α-L-Araf-(1→, →4)-β-D-Glcp-(1→,
4)-β-D-Manp-(1→, →4,6)-β-D-Manp-(1→ and →6)-α-D-Galp-(1→ as the backbone, along
→
with the branches consisting of β-L-Araf-(1→. The possible repeat unit of cDHPR was shown as
follows:
3
.2.3 cDHPL
cDHPL comprised a large proportion of D-Man and a trace of D-Glc and D-Gal in a molar
ratio of 19.15 : 1.32 : 1.00 (Supplemental Figure 2). Methylation analysis (Supplemental Figure 4
and Supplemental Table 1) suggested the glycosidic linkage types of cDHPL included
1
-linked-Galp, 1,4-linked Manp, 1,4-linked Glcp and 1,3,6-linked Man in a molar ratio of 1.00 :
20

Journal Pre-proof
1
7.89 : 1.33 : 1.58, which was supported by the periodate oxidation and Smith degradation
experiment (Supplemental Figure 3).
1
The H NMR experiment at 298 K (Fig. 3C) and at 328 K (Fig. 3D) showed the anomeric
protons of cDHPL were at δ 4.48 ppm, δ 4.73 ppm, δ 4.82 ppm, δ 5.17 ppm and δ 4.76 ppm,
which were marked as residue A , B , C , D and E . Correspondingly, the anomeric carbons of
L
L
L
L
L
these residues were found at δ 105.04 ppm, δ 102.69 ppm, δ 101.13 ppm, δ 96.51 ppm and δ
13
9
6.35 ppm were obtained in the C NMR experiment (Fig. 3E). Among these residues, the
identification of residue A , B and C was similar to that of cDHPS, and they were
L
L
L
→
4)-β-D-Glcp-(1→, →4)-β-D-Manp-(1→ and →4)-3-O-acetyl-β-D-Manp-(1→, respectively, and
all signals of H and C resonances were shown in Table 2.
For the residue D , anomeric proton at δ 5.17 ppm (Fig. 3C and 3D), anomeric carbon at δ
L
9
6.51 ppm (Fig. 3E) and their coupling signal at δ 5.17/96.51 ppm (Fig. 3F) indicated it was
α-configuration residue. According to H-1 (δ 5.17 ppm), the chemical shifts from H-2 to H-6 can
13
be found in COSY (Fig. 3G) and TOCSY (Supplemental Figure 5E). Subsequently, the
C
chemical shifts were determined from the cross-peaks of H and C in HSQC. Thus, the residue DL
was identified as α-D-Galp-(1→ based on its unique signal of C-2 (δ 67.95 ppm) and C-6 (δ 61.10
ppm) [20].
The residue E existed as β-configuration according to the anomeric proton at δ 4.76 ppm
L
(Fig. 3D), the anomeric carbon at δ 96.35 ppm (Fig. 3E) and their coupling signal at δ 4.76/96.35
ppm in HSQC (Fig. 3F). The correlation signals of different protons as well as cross signals
between protons and carbons could be found in COSY (Fig. 3G) and HSQC (Fig. 3F). The
downfield shifts of C-3 (δ 77.53 ppm) and C-6 (δ 69.56 ppm) with reference to the pertinent data
21

Journal Pre-proof
[
21] indicated the occurrence of replacement, suggesting that the residue EL was
→
3,6)-β-D-Manp-(1→.
The linkage sequence of different residues in cDHPL was inferred from the correlations
between C and H in HMBC (Fig. 3H). The cross-peaks between A H-1 and E C-3 at δ
L
L
4
.48/77.53 ppm, B H-1 and A C-4 at δ 4.73/79.83 ppm , B H-1 and C C-4 at δ 4.73/75.83
L L L L
ppm , B H-1 and B C-4 at δ 4.73/79.26 ppm, C H-1 and B C-4 at δ 4.82/79.26 ppm, E H-1
L
L
L
L
L
and B C-4 at δ 4.76/79.26 ppm indicated that the backbone of cDHPL could be composed of
L
→
→
4)-β-D-Glcp-(1→,
→4)-β-D-Manp-(1→,
→4)-3-O-acetyl-β-D-Manp-(1→
and
3,6)-β-D-Manp-(1→. The cross-peak between D H-1 and E C-6 at δ 5.17/69.56 ppm
L
L
indicated that the residue D as the branch was attached to the C-6 of residue E . These results
L
L
were confirmed by the ROESY experiment (Supplemental Figure 5F), and the possible repeat
unit of cDHPL was presented as follows:
3
.2.4 cDHPF
cDHPF was composed of D-Man, D-Glc and D-Gal with molar ratio of 9.68 : 3.26 : 1.00
Supplemental Figure 2). Mthylation analysis (Supplemental Figure 4 and Supplemental Table 1),
(
with the assistance of periodate oxidation and Smith degradation experiment (Supplemental
Figure 3), indicated that the glycosidic linkage types of cDHPF included 1-linked-Araf,
1
,4-linked Manp, 1,4-linked Glcp, and 1,3,6-linked Manp in a molar ratio of 1.00 : 9.18 : 2.65 :
22

Journal Pre-proof
1
.12.
1
13
According to H NMR (Fig. 4C and 4D) and C NMR (Fig. 4E) experiments, four residues,
i.e. A , B , C and D , were found, and their assignment process were consistent with the residue
F
F
F
F
A of cDHPS, the residue B of cDHPS, the residue D of cDHPL and theresidue E of cDHPL,
S
S
L
L
respectively. Based on the signals of all H and C (Table 2), which inferred from 2D NMR (Fig.
4
F-H and Supplemental Figure 5G-H), A , B , C and DF were →4)-β-D-Glcp-(1→,
F F F
→
4)-β-D-Manp-(1→, α-D-Galp-(1→ and →3,6)-β-D-Manp-(1→, respectively. The inter-residual
cross-peaks, i.e. A H-1/D C-3 (δ 4.49/77.25 ppm), B H-1/A C-4 (δ 4.74/80.72 ppm), B
F
F
F
F
F
H-1/B C-4 (δ 4.74/80.22 ppm), C H-1/D C-6 (δ 5.17/69.06 ppm) and D H-1/B C-4 (δ
F
F
F
F
F
4
.90/80.22 ppm), were easy to be found in HMBC (Fig. 4H) and confirmed by the ROESY
experiment (Supplemental Figure 5H). Accordingly, the structure of cDHPF could be composed
of →4)-β-D-Glcp-(1→, →4)-β-D-Manp-(1→ and →3,6)-β-D-Manp-(1→ as the backbone with
the branch consisting of α-D-Galp-(1→, and the possible repeat unit was proposed as follows:
3
.3 Effects of different polysaccharides on MFC cell viability
The treatment with cDHPS, cDHPR, cDHPL and cDHPF in a certain range of concentrations
could inhibited the proliferation of gastric cancer MFC cells, and their concentrations showing
the highest inhibition effect were 0.2 mg/mL for cDHPS, 0.8 mg/mL for cDHPL, 1.0 mg/mL for
cDHPF and 2.0 mg/mL for cDHPR, respectively (Fig. 5). Obviously, among these four
23

Journal Pre-proof
polysaccharides, cDHPS had the best inhibitory effect on the proliferation of MFC cells with an
IC of 0.327 mg/mL followed by cDHPL with an IC of 1.524 mg/ mL. cDHPF and cDHPR
5
0
50
only displayed weak inhibitory effects, and their IC values (3.588 mg/mL and 4.651 mg/mL,
50
respectively) were significantly higher than those of cDHPS and cDHPL.
3
.4 Effects of different polysaccharides on the MFC cell apoptosis
MFC cell apoptosis was analyzed with Annexin V-FITC & PI assay by flow cytometry, and
the total percentage of apoptosis both at early stage and late stage was presented in upper right
area and lower right area of Fig. 6A-F, respectively. These data showed that cDHPS could
significantly induce cell apoptosis in the tested concentration range, and the total apoptosis rate
rose from 16 % at 25 μg/mL to 25.05 % at 100 μg/mL, which was significantly higher than that of
untreated MFC cells (2.12 ± 0.48 %) and was very near to that of positive group (28.45 ± 0.98%).
Compared to cDHPS, cDHPL and cDHPF induced a lower apoptosis of MFC cells although the
percentage of total apoptotic cells was higher than that of untreated MFC cells. The highest
apoptosis for cDHPL occurred at 800 μg/mL and for cDHPF at 1.0 mg/mL. cDHPR only showed
a tiny effect on the apoptosis of MFC cells, and no significant difference was found between
cDHPR treatment at all tested concentrations and the normal control group. These results
suggested that the cDHPS had the best ability to induce gastric cancer cell apoptosis as compared
to cDHPR, cDHPL and cDHPF (p＜0.01) (Fig. 6G).
3
.5 Effects of different polysaccharides on the gene expression of p53 and c-myc in MFC
cells
qRT-PCR analysis showed that cDHPS at all tested concentrations could significantly
promote the gene expression of wild-type tumor suppressor gene p53 (Fig. 7A) and inhibit the
24

Journal Pre-proof
expression of proto-oncogene c-myc (Fig. 7B) in MFC cells, which was similar to that of the
positive control even if the concentration of cDHPS was lowered to 0.025 mg/mL. Compared to
cDHPS, cDHPL enhanced the expression of p53 (p<0.05) and decreased the expression of c-myc
(p<0.05) at the concentration of more than 0.4 mg/mL, and cDHPF displayed markedly influence
on the gene expression of p53 and c-myc at the concentration of more than 0.8 mg/mL. However,
cDHPR had no significant effect on p53 and c-myc gene expression at all tested concentration as
compared to the normal control. Obviously, cDHPS showed the strongest effect on up-regulated
expression of p53 gene and down-regulated expression of c-myc gene.
25

Journal Pre-proof
4
Discussion
In the current study, four homogeneous polysaccharides, named as cDHPS, cDHPR, cDHPL
and cDHPF, were isolated and purified from stem, root, leaf and flower of artificially cultivated D.
huoshanense, respectively. Based on the analysis of HPLC, FT-IR, GC-MS and NMR, these four
polysaccharides, although they had the similarity in monosaccharide compositions of D-Man and
D-Glc as well as glycosidic linkages of →4)-β-D-Manp-(1→ and →4)-β-D-Glcp-(1→, were
substantive different in structural features. Firstly, the relative molecular weights of cDHPS,
cDHPR, cDHPL and cDHPF were different from each other. Secondly, these four polysaccharides
were composed of different monosaccharides with different molar ratios. Besides D-Man and
D-Glc, cDHPR contained D-Gal and L-Ara, while cDHPL and cDHPF contained D-Gal. Thirdly,
these four polysaccharides were different in glycosidic linkage types and sequences. cDHPS was
a linear O-acetylated polysaccharide consisting of →4)-β-D-Manp-(1→ and →4)-β-D-Glcp-(1→
and acetyl groups were attached to the O-3 position of →4)-β-D-Manp-(1→. Compared to
cDHPS, other polysaccharides (cDHPR, cDHPL and cDHPF) were all branched polysaccharides.
For cDHPL and cDHPF, although they both were composed of a backbone consisting of
→
4)-β-D-Manp-(1→, →4)-β-D-Glcp-(1→ and →3,6)-β-D-Manp-(1→, and a branch of
α-D-Galp-(1→ that was attached to C-6 of →3,6)-β-D-Manp-(1→ in their repeat unit, the
backbone of cDHPL contained acetyl groups that were attached to the O-3 position of
→
→
4)-β-D-Manp-(1→. For cDHPR, its backbone contained →4)-β-D-Manp-(1→,
4)-β-D-Glcp-(1→, →6)-α-D-Galp-(1→, →4,6)-β-D-Manp-(1→ and →3,5)-α-L-Araf-(1→, and
the branches were composed of β-L-Araf-(1→ that was attached to C-3 of →3,5)-α-L-Araf-(1→,
C-5 of →3,5)-α-L-Araf-(1→and C-6 of →4,6)-β-D-Manp-(1→, respectively.
26

Journal Pre-proof
Previously, five polysaccharide fractions were isolated from the tissue-cultured D.
huoshanense including protocorm-like bodies and regenerated plantlets, two from water-eluted
fraction of crude polysaccharides on DEAE cellulose anion-exchange column [3,4], two from
NaCl-eluted fraction of crude polysaccharides on DEAE cellulose anion-exchange column [5,6],
and one from water-eluted fraction of crude polysaccharides on polyamide column [7]. As
compared to the polysaccharides from stem, root, leaf and flower of cultivated D. huoshanense,
these five polysaccharides showed obviously different structure features. Hsieh et al. reported the
structure of an immunoregulatory polysaccharide from the mixed mucilage from stem and leaf of
a D. huoshanense hybrid, which is from the hybridization of D. huoshanense and D. tosaense
[
22]. Although this polysaccharide was composed of →4)-β-D-Manp-(1→ and
4)-β-D-Glcp-(1→, its structure including the molar ratio of Man and Glc, relative molecular
→
weight, the sequence of glycosyl resides and the site of acetyl groups was quite different from
that of stem polysaccharide of cultivated D. huoshanense. These results indicated that growth
pattern, environment, developmental stage and genetic background remarkably influenced the
structural features of D. huoshanense polysaccharides.
It has been generally accepted that the biological activity of polysaccharide depends on its
structural features. In the present study, based on changes in viability and apoptosis as well as p53
and c-myc gene expression of gastric cancer cells after polysaccharide treatment, cDHPS showed
the best anti-gastric cancer activity, followed by cDHPL and cDHPF, and the anti-gastric cancer
activity of cDHPR was the weakest. Undoubtedly, the different structure features of
polysaccharides from stem, root, leaf and flower of D. huoshanense led to different anti-gastric
cancer activity. Since the different parts of D. huoshanense have been widely used for medicinal
27

Journal Pre-proof
and edible purpose at present, our results on the different structure features and anti-gastric cancer
activity of stem, root, leaf and flower polysaccharides suggested that the different parts of
cultivated D. huoshanense should be treated differently in future applications.
28

Journal Pre-proof
5
Conclusion
The homogeneous stem, root, leaf and flower polysaccharides from cultivated D.
huoshanense were isolated and purified, and the structural identification exhibited that there were
significant differences in relative molecular weights, monosaccharide compositions, glycosidic
linkage types and sequences as well as acetylated sites between these four polysaccharides. Stem
polysaccharide (cDHPS) was composed of →4)-β-D-Glcp-(1→, →4)-β-D-Manp-(1→,
5
→
4)-3-O-acetyl-β-D-Manp-(1→ with the relative molecular weight of 2.59×10 Da, root
polysaccharide (cDHPR) was composed of →3,5)-α-L-Araf-(1→, →4)-β-D-Glcp-(1→,
→
4)-β-D-Manp-(1→, →4,6)-β-D-Manp-(1→, →6)-α-D-Galp-(1→ and terminal β-L-Araf with
4
the relative molecular weight of 1.41×10 Da, leaf polysaccharide (cDHPL) was composed of
→
→
4)-β-D-Glcp-(1→,
→4)-β-D-Manp-(1→,
→4)-3-O-acetyl-β-D-Manp-(1→,
5
3,6)-β-D-Manp-(1→ and terminal α-D-Galp with relative molecular weight of 2.09×10 Da,
and flower polysaccharide (cDHPF) was composed of →4)-β-D-Glcp-(1→, →4)-β-D-Manp-(1→,
→
3,6)-β-D-Manp-(1→ and terminal α-D-Galp-(1→ with the relative molecular weight of
5
4
.78×10 Da. Among these four polysaccharides, cDHPS had the best anti-gastric cancer activity,
followed by cDHPL, cDHPF and cDHPR. Further work is needed to investigate their anti-gastric
cancer activity in vivo as well as their structure-activity relationship in order to develop functional
products with suitable polysaccharides from D. huoshanense.
29