Sp

YLHH

a

ARRAA

KLOLPSA

1

aai(mac

ea2occs

h

h0

Carbohydrate Polymers 173 (2017) 215–222

Contents lists available at ScienceDirect

Carbohydrate Polymers

j ourna l ho me page: www.elsev ier .com/ locate /carbpol

tructure features and in vitro hypoglycemic activities ofolysaccharides from different species of Maidong

ajun Gong, Jie Zhang, Fei Gao, Jiewen Zhou, Zhinan Xiang, Chenggao Zhou,uosheng Wan ∗, Jiachun Chen ∗

ubei Key Laboratory of Natural Medicinal Chemistry and Resource Evaluation, College of Pharmacy, Huazhong University of Science and Technology,angkong Road 13, Wuhan 430030, China

r t i c l e i n f o

rticle history:eceived 10 January 2017eceived in revised form 5 May 2017ccepted 24 May 2017vailable online 27 May 2017

eywords:iriope spicata

a b s t r a c t

Structures and in vitro hypoglycemic activities of polysaccharides from different species of Maidong werestudied. The primary structures of polysaccharides were elucidated on the basis of GC, GC–MS, infrared,NMR and periodate oxidation-Smith degradation. Liriope spicata polysaccharide (LSP), Ophiopogon japon-icus polysaccharide (OJP) and Liriope muscari polysaccharide (LMP) were composed of ˇ-fructose and�-glucose. The average molecular weights of LSP, OJP and LMP were 4742, 4925 and 4138 Da with poly-dispersity indexes of 1.1, 1.2 and 1.1, respectively. The backbones of polysaccharides were formed byFruf-(2→, →2)-Fruf-(6→, →6)-Glcp-(1→ and →1, 2)-Fruf-(6→ with a molar ratio of 5.0:18.2:1.0:5.3

phiopogon japonicusiriope muscariolysaccharidetructure elucidationnti-diabetic effect

(LSP), 6.8:15.8:1.0:5.8 (OJP), 8.3:12.3:1.0:3.9 (LMP), respectively. The RT-PCR and western blot analysisindicated that LSP, LMP and OJP increased the expression of PI3K, AKT, InsR, PPAR� and decreased theexpression of PTP1B in mRNA level and protein level in IR HepG2 cells. Furthermore, glucose consump-tion was increased after treated with polysaccharides. These results revealed that LSP, OJP and LMP hadpotential anti-diabetic effects.

© 2017 Elsevier Ltd. All rights reserved.

. Introduction

Maidong is the dried tuberous root of Ophiopogon japonicus, andnother two species, Liriope spicata and Liriope muscari are also useds substitutes in prescriptions. These three species all belong to Lil-aceae family and have been used as traditional Chinese medicineTCM) for thousands of years. Maidong are used to nourish Yin,

oisten the lungs, clean the heart-fire, relief the drought of mouthnd tongue (Xiaoke syndrome), and to treat vexation, insomnia andough (Pharmacopoeia of China, 2015).

Pharmacological studies have confirmed that Maidong has theffects of neuro-protection, ameliorated diabetic nephropathy,nti-cancer, immunomodulatory (Chen et al., 2009a; Kim et al.,012; Lu et al., 2013; Park et al., 2015; Wang et al., 2013). More-ver, phytochemical investigations have shown that the main

hemical compositions of Maidong are steroidal saponins, polysac-harides and flavonoids. Among these main compositions, steroidalaponins can promote anti-oxidative protection of the cardiovascu-

∗ Corresponding authors.E-mail addresses: wanlesheng1@163.com (L. Wan),

omespringchen@mail.hust.edu.cn (J. Chen).

ttp://dx.doi.org/10.1016/j.carbpol.2017.05.076144-8617/© 2017 Elsevier Ltd. All rights reserved.

lar system (Zhang et al., 2015) and exhibit neuritogenic activity (Quet al., 2011); polysaccharides have anti-diabetic and preventive dia-betic nephropathy activities (Chen et al., 2009b); flavonoids displayobvious anti-inflammation and antioxidant activities (Hung et al.,2010; Lin, Zhu, Qi, Qin, & Yu, 2010).

Polysaccharides, the main composition of Maidong with anextraction rate up to 35% (Huang, Xiu, & Liu, 2012; Xiao, Wang,Gan, & Chen, 2014; Xu et al., 2005), have attracted great attentionfrom the society of carbohydrate polymers. Due to the disor-dered sources and complicated structures, the structure propertiesof polysaccharides originated from these three species have notbeen completely settled. Taking OJP as an example, the molecularweight of OJP was inconsistent, ranging from 4020 to 48651 Da;monosaccharide compositions were ambiguous as well, such asglucopyranose: glucofuranose (7:3); arabinose: glucose: galactose(1:16:8); fructose: glucose (30:1); fructans. Backbone connectionsof OJP have been reported were Fruf-(2-1) with branch of Fruf-(2–6)Fruf-(2→; �-(1–4) glucosidic linkages (Chen et al., 2011; Lin et al.,2010; Shi, Li, Wang, & Feng, 2015; She & Shi, 2003; Wang, Sun,Zhang, Chen, & Liu, 2012; Wu et al., 2006). Similar inconsistencies

are also present in LSP (Chen et al., 2009b; Han, Cheng, Li, & Chen,2007) and LMP (Huang et al., 2012; Wang et al., 2011). Meanwhile,few published articles are available referring to compare the details

2 te Poly

orrshdioit(aa

2

2

OCAPVNPpfHwc

2

rwwMbuT

2

t2rtcDooTst

2

HU(m3

16 Y. Gong et al. / Carbohydra

f structural characteristics, the mechanism of improve insulinesistance and the hypoglycemic activities of Maidong polysaccha-ides from different species. In order to solve these issues, in presenttudy, polysaccharides from three species of traditional authenticerbs Maidong were prepared under the same experimental con-itions. The structural elucidations of these polysaccharides were

nvestigated by chemical analysis and NMR spectroscopy. More-ver, their effects on glucose consumption and the expression ofnsulin receptor (InsR), phosphatidylinositol-3-kinase (PI3K), pro-ein kinase B (Akt), peroxisome proliferators-activated receptors �PPAR�), protein tyrosine phosphatase 1B (PTP1B) in mRNA levelnd proteins level in insulin resistance HepG2 cells in vitro werelso investigated.

. Materials and methods

.1. Materials and reagents

Liriope spicata were purchased from Hubei province, China.phiopogon japonicus were purchased from Sichuan province,hina. Liriope muscari were purchased from Fujian province, China.ll the plants were identified by Professor Jiachun Chen, College ofharmacy, Huazhong University of Science and Technology (HUST).oucher specimens were deposited in Hubei Key Laboratory ofatural Medicinal Chemistry and Resource Evaluation, College ofharmacy, HUST. Monosaccharide standards, acetonitrile, insulin,apain were brought from Sigma (USA); dextran standards wererom Pharmacia (USA) and DMEM, double antibody were fromyClone (USA). Antibodies for PI3K, AKT, P-Akt, InsR, PPAR�, PTP1Bere obtained from Sungene Biotech Co., Ltd (Tianjin, China). Other

hemicals were of analytical grade.

.2. Extraction and purification of LSP, OJP, LMP

The crude polysaccharides were extracted from the tuberousoots as described (Xiao et al., 2014). Briefly, 100 g Maidong powderere extracted 3 times with water at 100 ◦C for 30 min 3% papainas used to hydrolyze proteins. The extracts were ultra-filtrated byini-pellicon system (MwCO 1000). The retention fluid was eluted

y distilled water on DEAE-52 cellulose columns (Whatman, UK)ntil phenol sulfuric acid reaction showed colorless (Yu et al., 2017).hree purified polysaccharides were obtained after vacuum dried.

.3. Determination of total polysaccharides content

Anthrone sulphuric acid colorimetry was used to determinehe polysaccharides content of LSP, LMP and OJP (Feng et al.,016). Fructose was used as the reference and the standard curveange was 3.4–17 �g/mL. Absorbance of polysaccharides solu-ion (1 mg/mL) was measured at 625 nm. Total polysaccharideontent = C × D × F/W × 100% (C = fructose concentration of sample,

= dilution ratio of sample, F = conversion factor 0.91, W = weightf samples). Different sugar has different impact on the resultsf polysaccharides content by anthrone-sulfuric acid method.herefore, using the conversion factor method to correct the mea-urement results is necessary, and the conversion coefficient wasested as demonstrated (Wu, Hu, Huang, & Jiang, 2013)

.4. Determination of molecular weight

The molecular weights of LSP, LMP and OJP were determined byPGPC with differential refractive index detector (G1362A, Agilent,

SA) and G4000 PWxl column (7.8 mm × 300 mm, TOSOH, Japan)

Ghoneim, Hassan, Mahmoud, & Asker, 2016; Xu et al., 2005). Theobile phase was distilled water at a flow rate of 0.6 mL/min at

5 ◦C. The molecular weights of polysaccharides were calculated

mers 173 (2017) 215–222

according to a calibration curve obtained with dextran standards(0.18, 3, 6, 7 and 10 KDa).

2.5. Infrared spectrum analysis of LSP, LMP, OJP

The functional groups of LSP, LMP and OJP were characterizedusing a Fourier transform infrared spectrometer (Bruker, Germany).2 mg of samples were ground with 100–200 mg KBr powder andthen pressed into pellets for infrared spectra measurement in afrequency range of 4000–500 cm−1 (HYPERION TM 2000).

2.6. Analysis of monosaccharide compositions

The monosaccharide compositions of LSP, LMP and OJP weredetermined by GC (Lopes et al., 2015). 10 mg samples werehydrolyzed with TFA (0.1 M) at 70 ◦C for 2 h, then derivations of thehydrolyzed polysaccharides and monosaccharide standards werecarried out using the trimethylsilylation reagents. After that, thetrimethylsilylated derivatives were loaded into GC (Agilent 7820A,USA) with HP-5 capillary column (30 m × 0.32 mm × 0.25 �m). GCanalysis condition: injection temperature 260 ◦C, detector temper-ature 300 ◦C, column temperature programmed from 150 to 185 ◦C(holding for 2 min) at 7 ◦C/min, then increasing to 190 ◦C (holdingfor 2 min) at 1 ◦C/min, and finally to 250 ◦C (holding for 1 min) at10 ◦C/min. Nitrogen was used as the carrier gas.

2.7. Methylation analysis

Methylation reactions were conducted 3 times until the dis-appearance of the hydroxyl group absorption in infrared spectraaccording to the methods (Yuan et al., 2016) with slight modifi-cation. Briefly, 10 mg samples were dissolved in DMSO (3 mL) andthen methylated with NaOH (20 mg) and methyl iodide (3.0 mL).After ultrasonicated for 60 min in darkness (KH-5200, 40 kHz, Kun-shan, China), the reaction was terminated with water (2.0 mL).Methylated polysaccharides were extracted with chloroform anddried on a rotary evaporator (RE52AA, Shanghai, China). Then thefully methylated products were hydrolyzed with 2 mL 90% (V/V)formic acid and 0.1 M TFA at 70 ◦C for 2 h, successively. The uronicacid in dried hydrolyzed samples were reduced with excess sodiumborohydride (NaBH4, 30 mg), and further acetylated by acetic anhy-dride (1 mL) in pyridine (1 mL) at 100 ◦C for 1 h. The dried acetylatedderivatives were dissolved in 300 �L chloroform and filtered beforeloaded into GC–MS (7890A/5975C, Agilent, USA) with HP-5 capil-lary column. The temperature program was set as follows: 50 ◦C for3 min, 50 ◦C/min to 200 ◦C for 3 min, 20 ◦C/min to 300 ◦C for 6 min;injection temperature: 250 ◦C; the ion source temperature of massspectrometer was 230 ◦C.

2.8. Periodate oxidation-Smith degradation reaction

Periodate oxidation-Smith degradation reaction was used todetect the substitution position of the sugar ring (Wang, Zeng,& Luo, 2016). 20 mg samples were dissolved in sodium perio-date solution (15 nmol/L, 20 mL) and reacted at 4 ◦C in darknessuntil the value of absorbance at 223 nm was stable. Consump-tion of sodium periodate was calculated according to the sodiumperiodate-sodium iodate standard curve. NaOH solution (0.1 M)was used to calculate the yield of formic acid produced. Remain-ing solutions were dialyzed (spectrum, USA) for 48 h by distilledwater. After concentration, retention fluid was reduced by NaBH4

(30 mg) for 10 h. Furthermore, after dialyzed for 24 h, the reten-tion fluid was vacuum dried and hydrolyzed by TFA (0.1 M, 5 mL) at70 ◦C for 2 h. The acetylation and sample treatments were the sameas methylation analysis in Section 2.7 before loaded into GC–MS.

te Poly

Ta

2

sc6npsw

2r

cHhtwti1rT9iw4w

tncpcat

F6

Y. Gong et al. / Carbohydra

he temperature program was set rising from 150 ◦C up to 200 ◦Ct 2.5 ◦C/min, with injector temperature at 250 ◦C.

.9. Nuclear magnetic resonance (NMR) spectroscopy

50 mg samples were dissolved in 0.5 mL D2O for NMR analy-is. Tetramethylsilane was used as internal standard for spectraalibration. Spectra were recorded by a spectrometer (Bruker DRX-00, Rheinstetten, Germany) at 25 ◦C. The 2D NMR, included heterouclear single quantum coherence (HSQC), hetero nuclear multi-le bond coherence (HMBC) and nuclear over hauser enhancementpectroscopy (NOESY), were performed using standard Bruker soft-are TopSpin.

.10. Effect of LSP, LMP and OJP on glucose metabolism in insulinesistant model HepG2 cells and 3T3-L1 cells

In order to explore the effect of LSP, LMP and OJP on glu-ose metabolism, experiments were performed on insulin-resistantepG2 cells and 3T3-L1 cells. HepG2 cells were incubated byigh concentration of insulin (1.0 × 10−6 mol/L) for 48 h based onhe method reported (Chen et al., 2016). 3T3-L1 preadipocytesere cultured in DMEM supplemented with 10% FBS for 48 h, and

hen transferred by a differentiation-induction medium (contain-ng 0.5 mmol/L IBMX, 5 �mol/L DEX, 5 mg/L insulin in DMEM with0% FBS). 48 h later, the differentiation-induction medium waseplaced with DMEM containing only insulin for another 2 days.hen cells were cultured in growth medium without insulin until0% of the cells displayed an adipocyte phenotype. To induce

nsulin resistance, fully differentiated adipocytes were culturedith growth medium supplemented with 1 �mol/L DXM 24 h, and

8 h respectively (Shen et al., 2012). The insulin resistance modelsere used for the following studies.

Each polysaccharide was dissolved in DMSO at the concentra-ion of 400 mg/mL and diluted to the required concentration withutrient fluid. Positive control was metformin (2 mg/mL); standardontrol was standard solution; blank control was distilled water;

olysaccharide groups were 400, 200, 100 �g/mL. The glucose con-entrations of supernatant were measured with glucose assay kitccording to the instruction. MTT method was used to detect cyto-oxicity (Moran-Santibanez et al., 2016).

ig. 1. GC spectrum of monosaccharide references, LSP, LMP and OJP. (A). monosaccharide-galactose, 7-glucose. (B)–(D) Gas chromatography of LSP, LMP, OJP, 1.2.3-fructose, 4-glu

mers 173 (2017) 215–222 217

2.11. Quantitative real-time polymerase chain reaction (RT-PCR)analysis

The mRNA expression level of PI3K, AKT, InsR, PPAR�, PTP1Bwas determined by RT-PCR (Nacher-Vazque et al., 2015). TotalRNA from HepG2 cells was extracted using TRIzol Reagent (Invitro-gen) according to the manufacturer’s instructions. RNA was reversetranscribed to cDNA using the RevertAid First Strand cDNA Syn-thesis kit (Fermentas). RT-PCR analysis was performed using theSYBR Green Kit (KAPA, USA) on the ABI stepone plus software.Primer pairs (Table S4) were designed with Primer 5.0 software.Fold changes were calculated using the 2−��Ct method.

2.12. Western blot

HepG2 cells were incubated with compounds for 24 h, followedby lysis in RIPA lysis buffer. Protein concentration was determinedby Bradford assay and proteins were separated by 10% SDS-PAGEand transferred to PVDF membranes (Millipore, USA). The proteinswere probed with primary antibodies against AKT, PTP1B, P-AKT(ser 473), PI3K, PPAR�, InsR. The blots were then incubated withsecondary antibody; reactive protein was detected by ECL chemi-luminescence system (Zhang et al., 2016).

2.13. Statistical analysis

SPSS 19 software was used for statistical analysis. The measure-ment data were expressed as (x ± s), and the experimental datawere statistically analyzed by ANOVA, *P < 0.05 was statisticallysignificant.

3. Results and discussion

3.1. Extraction, purification and total polysaccharides content ofLSP, OJP and LMP

The extraction rate of LSP, LMP and OJP was 53.2 ± 0.78%,54.7 ± 0.76% and 55.2 ± 1.86%, respectively, which were higher thanthe yields of other extraction methods (Huang et al., 2012; Xu et al.,

2005). The purification was proved to be simple, convenient andquick (Figs. S1, S3). Anthrone sulfuric acid method was used to mea-sure the content of polysaccharides with fructose as the reference(Fig. S2). Results demonstrated that the average content of LSP, LMP

references. 1-arabinose, 2-rhamnose, 3-xylose, 4-mannose+fructose, 5.6-fructose,cose.

218 Y. Gong et al. / Carbohydrate Polymers 173 (2017) 215–222

Table 1The methylation analysis of LSP, LMP and OJP.

Retention time(min) Methylated sugars Molar ratiosa Substituted sugar unit Major mass fragments (m/z)

LSP LMP OJP

8.31 2,5-Ac2-1,3,4,6-Me4-d-mannitol 5.0 8.3 6.8 Fru-(2→ 43,71,87,101,129,145,1619.48 2,5,6-Ac3-1,3,4-Me3-d-mannitol 18.2 12.3 15.8 →2)-Fru-(6→ 43,71,87,101,129,145,161,1899.77 2,3,4-Me -d-Glucitol 1.0 1.0 1.0 →6)-Glc-(1→ 43,71,87,101,117,131,161,189, 233

.9 5.8 →1,2)-Fru-(6→ 43,87,101,129,145,189

ah

3

at1

3

tfrpfrrcp22

3

3

elrcfastaGtc

3a

sp2(pswhw

3

10.53 1,2,5,6-Ac4-3,4-Me2-d-mannitol 5.3 3

a Molar ratio was calculated from peak areas.

nd OJP was 92.7 ± 3.5%, 90.1 ± 4.7% and 91.1 ± 5.1%, respectively,igher than previous reports (Tang, Li, Liu, & Chen, 2003).

.2. Measurement of molecular weight

Fitted curves (Fig. S3) were calculated by GPC software. Theverage molecular weight of the LSP, LMP and OJP were estimatedo be 4742, 4138 and 4925 Da with polydispersity indexes of 1.1,.2 and 1.1, respectively.

.3. Monosaccharide composition

As seen in Fig. 1, LSP, LMP and OJP after trimethylsilylation reac-ion showed obvious peaks with retention time corresponding toructose (peak 1–3 in Fig. 1 BCD) and glucose (peak 4 in Fig. 1 BCD),evealing that only glucose and fructose were present in these threeolysaccharides. Judging from the peak area, AUC ratio betweenructose and glucose were 28:1 (LSP), 24:1 (LMP) and 29:1 (OJP),espectively. These results demonstrated that the monosaccharideatio of the three polysaccharides were different, and the monosac-haride composition of LSP, LMP and OJP were also different fromrevious findings (Chen et al., 2011; Chen et al., 2009b; Han et al.,007; Huang et al., 2012; Lin et al., 2010; Shi et al., 2015; She & Shi,003; Wang et al., 2011).

.4. Structural characterization of polysaccharides

.4.1. Results of methylation analysisPeaks at ∼1716 cm−1 in infra-red spectra (Fig. S4) proved the

xistence of uronic acid in the structures (Nep et al., 2016), whiched to a reduction before acetylation of hydrolyzed polysaccha-ides. Four types of glycosidic linkages were detected in total ionhromatography (Fig. S5, Table 1). Alditol acetate derivatives wereormed after methylated, hydrolyzed and acetylated, successively,s analyzed by GC–MS (Liu, Shang, Yang, Wu, & Zhao, 2017). Theum of all the same productions was their real content. It was clearhat →2)-Fruf-(6→ was the main linkage, and the hydroxyl groupst C-3, C-4 of fructose have not been substituted. Moreover, −1)-lc-(6- was also found in trace amounts. These findings indicated

hat glucose exists in the backbone, with fructose being the mainonstituent in LSP, LMP and OJP.

.4.2. Results of periodate oxidation-Smith degradation reactionnalysis

In this study, UV absorbance of LSP, LMP and OJP aqueousolution at 223 nm was getting lower and lower during sodiumeriodate consumption, and finally came to a stable value after01 h. One molar of sugar residues consumes 1.28 (LSP), 1.25LMP), 1.24 (OJP) molar sodium periodate, respectively. Oxidationroducts were titrated with NaOH (0.1 mol/L), while almost no con-

umption of NaOH was observed. These results revealed that thereas little amount of acid produced, meaning that no adjacent threeydroxyl groups existed in the polysaccharides, and →2)-Fruf-(6→as the main glycosidic bond.

Fig. 2. The FT-infrared spectra of polysaccharides. (a). LSP infrared spectrum (b).LMP infrared spectrum (c). OJP infrared spectrum.

After dialysis, reduction and acetylation reaction, GC–MS anal-ysis was carried out. Three kinds of polysaccharides showed thesame GC–MS behaviors (Fig. S6). The major production of acetyla-tion was glycerol triacetate (at 2.727 min), indicating that oxidationtook place between the 3, 4-dihydroxy groups in the sugars. Fur-thermore, the hydroxyl group at the 2, 6-positions mainly formedglycoside bond, consistent with previous methylation analysis.

3.4.3. FT- infrared analysisThe FT-infrared spectra (Fig. 2) of three polysaccharides were

almost identical. Broad stretching intense characteristic peak wasshown at 3400 cm−1 for the −OH, while a weak C H stretchingband was observed ranged from 2933 to 2935 cm−1.The peaks at1642 cm−1 were from the bending vibration absorption of −OH.Absorption peaks ranged from 1460 to 1200 cm−1 were the variableangle vibrations of C H. Usually, pyran glycosides have three strongpeaks ranged from 1100 to 1010 cm−1 while furan glycosides haveonly two absorption peaks. Peaks ranged from 3600 to 3200 cm−1

and 1665–1635 cm−1 revealed that these compounds were carbo-hydrates. Peaks ranged from 930 to 932 cm−1 and 817–820 cm−1

indicated that polysaccharides were fructofuranose with ˇ-bondtypes. Moreover, and peaks at 1715 cm−1 without absorption fur-ther explained that the fructan had no uronic acid (Du, Lu, Cheng,& Chen, 2016; Kokoulin et al., 2016; Ustyuzhanina et al., 2016).

3.4.4. NMR analysisThe NMR spectra of LSP were shown in Fig. S7. In 1H NMR spec-

trum, chemical shift of the anomeric proton signal at more than ıH5 ppm is for the �-type, while less than 5 ppm is for the ˇ-type (Nep

et al., 2016). LSP has only one weak signal at ıH 5.38 ppm, indicat-ing that the signal may be derived from the �-d-Glc of the chain,since fructose is a ketose without proton at C-2. The rest absorptionpeaks between ıH 4.0–4.3 ppm were from protons at C-2 to C-6.

Y. Gong et al. / Carbohydrate Polymers 173 (2017) 215–222 219

re of

rwbbwW1coc11ıCcrncftCts

clG

repeat units in the polysaccharide backbone.Combined with all the experimental analysis, including GC,

methylation, periodate oxidation-Smith degradation, infrared,HPGPC, and 1D and 2D NMR spectra, the possible structure of LSP

Fig. 3. Possible structures of polysaccharides. (Sa). Structure of LSP (Sb). Structu

In 13C NMR spectrum, anomeric carbons signals were in theange of ıC 98–110 ppm, the chemical shift of C-2, C-3, C-4 thatithout substitution were usually from 69 to 77 ppm, and would

e down-field to ıC 77–85 ppm if hydrogen atoms on them hadeen substituted; C-6 was in the range of ıC 60–64 ppm, andould be down-field to ıC 64–70 ppm with substitution (Fontana,eintraub, & Widmalm, 2015). The absence of signals from 120 to

75 ppm suggested that there were no hexuronic acid, acetyl aminoompounds and carboxyl. A series of signals presented in the rangef ıC 103–104 ppm revealed that the polysaccharide was mainlyonstituted of ˇ-fructose. The peaks at ıC 103.9, 103.8, 103.6, 103.1,03.0 and 92.1 ppm were anomeric carbon chemical shift signals, ıC03.8, 103.1 and 103.6 ppm were C-2 signals of −2-Fruf-6- residue;C 103.9 ppm was the C-2 signal of Fru-(2→ residue; ıC 103.0 was-2 signal of-1, 2, 6- linked fructose. In general, �-type anomericarbon signals are ranged from ıC 97–101 ppm, ˇ-type signals areanged from ıC 103–106 ppm (Nep et al., 2016). The absence of sig-als ranged from ıC 97–101 ppm indicated that LSP were mainlyonstituted of ˇ-fructose. The weak signal at ıC 92.10 ppm mightrom �-d-Glc at the end of the chain. ıC 80.2, 81.0, 81.1 ppm werehe resonance signal of C-5; ıC 69–77 ppm were the signals of C-3,-4 without substitution; ıC 59.9–72.4 ppm were speculated to behe carbon signal from CH O ; ıC 63.08 ppm was the resonanceignal of C-6 at the end of the chain.

HSQC spectrum indicated the anomeric carbon of �-d-Glc had a

hemical shift of ıC 92.10 ppm. HMBC (Fig. S7d) revealed the corre-ations between the anomeric carbons and hydrogens in the chain.lcH-1/AC-2 correlation peak was weak, which suggested that glu-

LMP. (Sc). Structure of OJP. The “n” indicates the number of the repeating units.

cose is at the end of the sugar chain and forms glycosides with sugarA. DC-2/CH-6, BC-2/AH-6, CC-2/BH-6, EC-2/DH-1 and other corre-lation peaks intensity were strong, suggesting that they were the

Fig. 4. Effects of LSP, LMP and OJP on the protein expression of AKT, PTP1B, InsR,PPAR�, P-AKT and PI3K by western blot analysis. Results are representative of 3 inde-pendent experiments. �-actin was used as a loading reference. IR: Insulin resistancemodel, Met: metformin group, others were different concentrate polysaccharidesgroups (400, 200, 100 �g/mL).

220 Y. Gong et al. / Carbohydrate Polymers 173 (2017) 215–222

Fig. 5. Effect of LSP, LMP, OJP on mRNA and protein expression (n = 3), glucose consumption (n = 5, mmol/L) in IR cells induced by insulin. (A) the mRNA levels of PI3K (B) them ) the m( lls. Dar chari

wa

tfn

RNA levels of AKT. (C) the mRNA levels of InsR. (D).the mRNA levels of PPAR�. (EG) Glucose consumption in IR HepG2 cells. (H) Glucose consumption in IR 3T3-L1 ceesistance model, Met: metformin group, others were different concentrate polysac

as shown in Fig. 3Sa. Chemical shifts assignments for the 13C NMRnd 1H NMR spectra of LSP were shown in Table S1.

The NMR spectra of LMP were shown in Fig. S8. In 1H NMR spec-rum, only a single peak appeared at ıH 5.43 ppm, which camerom the anomeric proton of �-d-Glc. Six anomeric carbon sig-als at ıC 104.0, 103.2, 103.8, 103.1, 103.7 and 92.1 ppm appeared

RNA levels of PTP1B. (F) the ratio of phosphorylated AKT/total AKT protein levels.ta are shown as mean ± SD. Vertical lines represent standard deviations. IR: Insulin

des groups (400, 200, 100 �g/mL). *P < 0.05, **P < 0.01 compared with IR control.

in 13C NMR spectrum, indicating that five fructoses and one glu-cose were existed in one unit, and fructose were mainly presented

as ˇ-configuration. ıC 80.21, 81.05, 81.16, 81.17, 80.98 ppm wereattributed to the resonance signal of C-5; ıC 69.18–77 ppm wereC-3, C-4 signal without substitution; ıC 59.92–72.40 ppm werespeculated to be the signal of CH O ; ıC 62.24, 62.14 ppm were

te Poly

tiNp

ywetS

3p

i(iisettat(p

(RcripiricdAroc

4

aiMMtprwdLmstdcroO

Y. Gong et al. / Carbohydra

he resonance signal of C-6. The 2D NMR analyses of LMP weren the same as that of LSP. Chemical shift assignments for the 13CMR and 1H NMR spectra of LMP were showed in Table S2, and itsossible structure was showed in Fig. 3Sb.

The NMR spectra of OJP were shown in Fig. S9. Detailed anal-sis methods referred to LSP. Six anomeric carbon signals of OJPere detected, suggesting that five fructoses and one glucose were

xisted in one polysaccharide unit. Chemical shift assignments forhe 13C NMR and 1H NMR spectrum of OJP were showed in Table3, and the possible structure was displayed in Fig. 3Sc.

.5. Effect of LSP, LMP, OJP on glucose consumption, mRNA androtein expression level in IR cells

InsR, PI3K, AKT and PPAR� play very important role in mediat-ng insulin metabolic and closely correlative with insulin resistanceLiu et al., 2013). PTP1B can dephosphorylate the InsR as well asnsulin receptor substrate (IRS), it’s a key negative regulator of thensulin signaling pathways, and its increased activity and expres-ion are implicated in the pathogenesis of insulin resistance (Liut al., 2015). The results of MTT test showed that there was no cyto-oxicity at all concentration of LSP, LMP and OJP. Compared withhe IR control, metformin showed the best effect on glucose uptakend LSP, LMP, OJP significantly increased the glucose consump-ion at the concentration of 100, 200 and 400 mg/mL, respectivelyFig. 5G and H). Additionally, there was no significance among threeolysaccharides.

The mRNA expression of PI3K (Fig. 5A), AKT (Fig. 5B), InsRFig. 5C), PPAR� (Fig. 5D) and PTP1B (Fig. 5E) were evaluated byT-PCR. The RT-PCR amplification plots (Fig. S10) and meltingurves (Fig. S11–S16) were shown in supplementary material. Theesults showed that the expression of PI3K, AKT, InsR and PPAR�n mRNA level increased in all of the polysaccharides groups andositive control. Meanwhile, the mRNA expression level of PTP1B

n these groups decreased (p < 0.05). Western blot analysis (Fig. 4)evealed that the protein expression of InsR, PI3K, AKT and PPAR�n IR HepG2 cells were increased after treated with different con-entrations of LSP, LMP and OJP. The polysaccharides produced aown-regulation of the protein expression of PTP1B. Meantime, P-KT/total-AKT ratios were augmented (Fig. 5F) (p < 0.05). Theseesults showed that the LSP, LMP and OJP have positive effectsn PI3K/Akt signaling pathway and improve the glucose uptakeapacity of HepG2 cells and 3T3-L1 cells.

. Conclusion

Polysaccharides from Maidong were extracted with hot waternd purified using a series of procedures in this study. The compar-sons of the structures and in vitro hypoglycemic activity of three

aidong polysaccharides were performed for the first time. Threeaidong polysaccharides were all composed of glucose and fruc-

ose, with ˇ-d-Fructofuranose being the main component. Theyossessed similar molecular weights, backbones, and same hexoseesidues. However, the proportion of monosaccharide compositionas different among three polysaccharides, and also the results areifferent from previous reports. Hypoglycemic tests revealed thatSP, LMP and OJP exhibited satisfying effects on insulin resistanceodel of HepG2 cells and 3T3-L1 cells. RT-PCR and western blot

tudies suggested that LSP, LMP and OJP can mitigate insulin resis-ant in HepG2 cells through PI3K/AKT pathway, and the activityifferences among these three polysaccharides were not signifi-

ant, this might be related to their structural similarity. Presentesults provided preliminary chemical basis to support that Liri-pe spicata and Liriope muscari can be used as the substitutes ofphiopogon japonicus in clinical practice. However, the detailed

mers 173 (2017) 215–222 221

hypoglycemic mechanisms of these polysaccharides in vivo stillrequire further clarification.

Conflict of interest

The authors declare that they have no conflicts of interest withthis paper submitted.

Acknowledgements

This work was supported financially by National Natural ScienceFoundation of China [No. 30873373 and 81373917] and NationalScience and Technology Major Project of China (No. 2011ZX09102-004-02).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.carbpol.2017.05.076.

Referencess

Chen, X. H., Bai, X., Liu, Y. H., Tian, L. Y., Zhou, J. Q., Chen, J. C., et al. (2009).Anti-diabetic effects of water extract and crude polysaccharides from tuberousroot of Liriope spicata var. prolifera in mice. Journal of Ethnopharmacology, 122,205–209.

Chen, X., Liu, Y., Bai, X., Wen, L., Fang, J., Chen, J. C., et al. (2009). Hypoglycemicpolysaccharides from the tuberous root of Liriope spicata. Journal of NaturalProducts, 72, 1988–1992.

Chen, X., Jin, J., Tang, J., Wang, Z., Wang, J., Jin, L., et al. (2011). Extraction,purification: Characterization and hypoglycemic activity of a polysaccharideisolated from the root of Ophiopogon japonicus. Carbohydrate Polymers, 83,749–754.

Chen, Y., Wang, S., Tian, S. T., Hu, X., Xu, J., Yang, G. Z., et al. (2016).12b-hydroxy-des-D-garcigerin A enhances glucose metabolism ininsulin-resistant HepG2 cells via the IRS-1/PI3-K/Akt cell signaling pathway.Journal of Asian Natural Products Research, 18, 1091–1100.

Du, D., Lu, Y., Cheng, Z., & Chen, D. (2016). Structure characterization of two novelpolysaccharides isolated from the spikes of Prunella vulgaris and theiranticomplement activities. Journal of Ethnopharmacology, 193, 345–353.

Feng, H. B., Fan, J., Song, Z. H., Du, X. G., Chen, Y., Wang, J. S., et al. (2016).Characterization and immunoenhancement activities of Eucommia ulmoidespolysaccharides. Carbohydrate Polymers, 136, 803–811.

Fontana, C., Weintraub, A., & Widmalm, G. (2015). Structural elucidation of theO-Antigen polysaccharide from escherichia coli O181. Chemistry Open, 4,47–55.

Ghoneim, M. A., Hassan, A. I., Mahmoud, M. G., & Asker, M. S. (2016). Effect ofpolysaccharide from Bacillus subtilis sp. on cardiovascular diseases andatherogenic indices in diabetic rats. BMC Complementary and AlternativeMedicine, 16(112)

Han, F., Cheng, L., Li, L., & Chen, Y. (2007). Isolation, purification andmonosaccharide composition of polysaccharide from Liriope spicata. ChineseTraditional and Herbal Drugs, 38, 30–31.

Huang, R., Xiu, H., & Liu, Y. (2012). Studies on the purification and quantity analysismethods of Liriope muscari (Decne.) Baily polysaccharides. StraitPharmaceutical Journal, 24, 43–45.

Hung, T. M., Thu, C. V., Dat, N. T., Ryoo, S. W., Lee, J. H., Kim, J. C., et al. (2010).Homoisoflavonoid derivatives from the roots of Ophiopogon japonicus and theirin vitro anti-inflammation activity. Bioorganic & Medicinal Chemistry Letters, 20,2412–2416.

Kim, H. K., Lee, J. Y., Han, H. S., Kim, Y. J., Kim, H. J., Kim, Y. S., et al. (2012).Immunomodulatory effects of Liriope platyphylla water extract onlipopolysaccharide- activated mouse macrophage. Nutrients, 4, 1887–1897.

Kokoulin, M. S., Kuzmich, A. S., Kalinovsky, A. I., Tomshich, S. V., Romanenko, L. A.,Mikhailov, V. V., et al. (2016). Structure and anticancer activity of sulfatedO-polysaccharide from marine bacterium Cobetia litoralis KMM 3880(T).Carbohydrate Polymers, 154, 55–61.

Lin, X., Wang, S., Jiang, Y., Wang, Z. J., Sun, G. L., Xu, D. S., et al. (2010). Poly(ethyleneglycol)-radix Ophiopogonis polysaccharide conjugates: Preparation,characterization, pharmacokinetics and in vitro bioactivity. European Journal ofPharmaceutics and Biopharmaceutics, 76, 230–237.

Lin, Y., Zhu, D., Qi, J., Qin, M., & Yu, B. (2010). Characterization of homoisoflavonoidsin different cultivation regions of Ophiopogon japonicus and related antioxidant

activity. Journal of Pharmaceutical and Biomedical Analysis, 52, 757–762.

Liu, Y. H., Wan, L. S., Xiao, Z. Q., Wang, J. J., Wang, Y. L., & Chen, J. C. (2013).Antidiabetic activity of polysaccharides from tuberous root of Liriope spicatavar. prolifera in KKAy mice. Evidence-Based Complementary and AlternativeMedicine, 2013, 1–11.

2 te Poly

L

L

L

L

M

N

N

N

P

Q

S

S

S

T

and Experimental Therapeutics, 352, 166–174.Zhang, H. H., Luo, X. P., Ke, J. J., Duan, Y. Q., He, Y. Q., Zhang, D., et al. (2016).

Procyanidins from Castanea mollissima Bl. shell, induces autophagy followingapoptosis associated with PI3K/AKT/mTOR inhibition in HepG2 cells.Biomedicine & Pharmacotherapy, 81, 15–24.

22 Y. Gong et al. / Carbohydra

iu, Z. Q., Liu, T., Chen, C., Li, M. Y., Wang, Z. Y., Chen, R. S., et al. (2015).Fumosorinone, a novel PTP1B inhibitor: Activates insulin signaling ininsulin-resistance HepG2 cells and shows anti-diabetic effect in diabetic KKAymice. Toxicology and Applied Pharmacology, 285, 61–70.

iu, J., Shang, F. N., Yang, Z. M., Wu, M. Y., & Zhao, J. H. (2017). Structural analysis ofa homogeneous polysaccharide from Achatina fulica. International Journal ofBiological Macromolecules, 98, 786–792.

opes, S. M. S., Krausova, G., Rada, V., Goncalves, J. E., Goncalves Regina, A. C., & deOliveira, A. J. B. (2015). Isolation and characterization of inulin with a highdegree of polymerization from roots of Stevia rebaudiana (Bert.) Bertoni.Carbohydrate Research, 411, 15–21.

u, H. J., Tzeng, T. F., Hsu, J. C., Kuo, S. H., Chang, C. H., Huang, S. Y., et al. (2013). Anaqueous-ethanol extract of Liriope spicata var. prolifera ameliorates diabeticnephropathy through suppression of renal inflammation. Evidence-BasedComplementary and Alternative Medicine, 2013, 1–10.

oran-Santibanez, K., Cruz-Suarez, L. E., Ricque-Marie, D., Robledo, D.,Freile-Pelegrin, Y., Pena-Hernandez, M. A., et al. (2016). Synergistic effects ofsulfated polysaccharides from Mexican seaweeds against measles virus.BioMed Research International, 2016, 1–11.

acher-Vazque, M., Ballesteros, N., Canales, A., Rodriguez, S. J. S., Perez-Prieto, S. I.,Prieto, A., et al. (2015). Dextrans produced by lactic acid bacteria exhibitantiviral and immunomodulatory activity against salmonid viruses.Carbohydrate Polymers, 124, 292–301.

ational Pharmacopoeia Committee. (2015). . Pharmacopoeia of People’s Republic ofChina (1st vol) Beijing: China Medical Science Press [26, 155, 210, 295, 307].

ep, E. I., Carnachan, S. M., Ngwuluka, N. C., Kontogiorgos, V., Morris, G. A., Sims, I.M., et al. (2016). Structural characterisation and rheological properties of apolysaccharide from sesame leaves (Sesamum radiatum Schumach. & Thonn.).Carbohydrate Polymers, 152, 541–547.

ark, H. R., Lee, H., Park, H., Jeon, J. W., Cho, W. K., & Ma, J. Y. (2015).Neuroprotective effects of Liriope platyphylla extract against hydrogenperoxide-induced cytotoxicity in human neuroblastoma SH-SY5Y cells. BMCComplementary and Alternative Medicine, 15, 171.

u, Y., Zhang, Y., Pei, L., Wang, Y., Gao, L., Huang, Q., et al. (2011). New neuritogenicsteroidal saponin from Ophiopogon japonicus (Thunb.) Ker-Gawl. Bioscience,Biotechnology, and Biochemistry, 75, 1201–1204.

he, G. M., & Shi, J. P. (2003). Structural features of two neutral polysaccharidesMd-1: Md-2 from Ophiopogon japonocus. Journal of Chinese Medicinal Materials,26, 100–101.

hen, F. X., Gu, X. M., Pan, W., Li, W. P., Li, W., Ye, J., et al. (2012). Over-expression ofAQP7 contributes to improve insulin resistance in adipocytes. Experimental CellResearch, 318, 2377–2384.

hi, L. L., Li, Y., Wang, Y., & Feng, Y. (2015). MDG-1, an Ophiopogon polysaccharide:

regulate gut microbiota in high-fat diet-induced obese C57BL/6 mice.International Journal of Biological Macromolecules, 81, 576–583.

ang, L. Q., Li, C., Liu, S., & Chen, L. M. (2003). Determination of pachyman in radixophiogonis content by anthrone-sulphuric acid colrimetry. Anhui Medical andPharmaceutical Journal, 7, 39.

mers 173 (2017) 215–222

Ustyuzhanina, N. E., Bilan, M. I., Dmitrenok, A. S., Tsvetkova, E. A., Shashkov, A. S.,Stonik, V. A., et al. (2016). Structural characterization of fucosylatedchondroitin sulfates from sea cucumbers Apostichopus japonicus andActinopyga mauritiana. Carbohydrate Polymers, 153, 399–405.

Wang, G., Lin, R., Nie, L., Xie, W., & He, F. (2011). Liriope muscari polysaccharide andits extraction method. China patent, CN101935359A, 2011-01-05.

Wang, X. M., Sun, R. G., Zhang, J., Chen, Y. Y., & Liu, N. N. (2012). Structure andantioxidant activity of polysaccharide POJ-U1a extracted by ultrasound fromOphiopogon japonicus. Fitoterapia, 83, 1576–1584.

Wang, H. C., Wu, C. C., Cheng, T. S., Kuo, C. Y., Tsai, Y. C., Chiang, S. Y., et al. (2013).Active constituents from Liriope platyphylla root against Cancer Growthin vitro. Evidence-based Complementary and Alternative Medicine, 2013, 1–10.

Wang, Z., Zeng, Y., & Luo, D. (2016). Structure elucidation of a non-branched andentangled heteropolysaccharide from Tremella sanguinea Peng and itsantioxidant activity. Carbohydrate Polymers, 152, 33–40.

Wu, X., Dai, H., Huang, L., Gao, X., Tsim, K. W., & Tu, P. (2006). A fructan, from radixophiopogonis: Stimulates the proliferation of cultured lymphocytes: structuraland functional analyses. Journal of Natural Products, 69, 1257–1260.

Wu, G. H., Hu, T., Huang, Z. L., & Jiang, J. G. (2013). Characterization of water andalkali-soluble polysaccharides from Pleurotustuber-regium sclerotia.Carbohydrate Polymers, 96, 284–290.

Xiao, Z. Q., Wang, Y. L., Gan, S. R., & Chen, J. C. (2014). Polysaccharides from LiriopesRadix ameliorates hyperglycemia via various potential mechanisms in diabeticrats. Journal of the Science of Food and Agriculture, 94, 975–982.

Xu, D. S., Feng, Y., Lin, X., Deng, H. L., Fang, J. N., & Dong, Q. (2005). Isolation:Purification and structural analysis of a polysaccharide MDG-1 fromOphiopogon japonicus. Acta Pharmaceutica Sinica, 40, 636–639.

Yu, X. H., Liu, Y., Wu, X. L., Liu, L. Z., Fu, W., & Song, D. D. (2017). Isolation,purification: Characterization and immunostimulatory activity ofpolysaccharides derived from American ginseng. Carbohydrate Polymers, 156,9–18.

Yuan, Y., Wang, Y. B., Jiang, Y., Prasad, K. N., Yang, J., Qu, H., et al. (2016). Structureidentification of a polysaccharide purified from Lycium barbarium fruit.International Journal of Biological Macromolecules, 82, 696–701.

Zhang, Y. Y., Meng, C., Zhang, X. M., Yuan, C. H., Wen, M. D., Chen, Z., et al. (2015).Ophiopogonin D attenuates doxorubicin-induced autophagic cell death byrelieving mitochondrial damage in vitro and in vivo. Journal of Pharmacology