Polysaccharide from Seeds of Plantago asiatica L. Affects LipidMetabolism and Colon Microbiota of MouseJie-Lun Hu,†,§ Shao-Ping Nie,*,† Qi-Meng Wu,† Chang Li,† Zhi-Hong Fu,† Joshua Gong,†,§

Steve W. Cui,†,§ and Ming-Yong Xie†

†State Key Laboratory of Food Science and Technology, Nanchang University, Nanchang 330047, China§Guelph Food Research Center, Agriculture and Agri-Food Canada, 93 Stone Road West, Guelph, Ontario N1G 5C9, Canada

*S Supporting Information

ABSTRACT: Polysaccharide from the seeds of Plantago asiatica L. was given via oral administration to mice (0.4 g/kg bodyweight, 30 days) to observe its effects on mouse nutrient metabolism and colon microbiota. It was found the polysaccharideintake could lower the apparent absorption of lipid. Total triglyceride, cholesterol, and atherogenic index in blood serum withtotal lipid and cholesterol levels in liver of polysaccharide group mice were all significantly lower than those of the control group(p < 0.05). Furthermore, the effect of the polysaccharide intake on mouse colon bacterial communities was investigated. Micefrom the polysaccharide group showed a higher colon bacterial diversity than the control group. Bacteroides sp., Eubacterium sp.,butyrate-producing bacteria Butyrivibrio sp., and probiotics Bifidobacterium bifidum, Lactobacillus fermentum, and Lactobacillusreuteri in mouse colon were all increased after polysaccharide intake. These indicated that the intake of polysaccharide from P.asiatica L. could be beneficial for lipid metabolism and colon microbiota.

KEYWORDS: Plantago asiatica L., polysaccharide, lipid metabolism, colon microbiota

■ INTRODUCTION

In recent years, the influence of polysaccharides on majornutrient (carbohydrate, protein, and lipid) metabolism has beenof particular interest because major nutrient metabolism isrelated to many chronic diseases.1,2 In addition, the effects ofpolysaccharide on colon microbiota were also paid muchattention because of its benefits for the colon health, which is ofmore and more concern nowadays.3 The gut microbiota incolon contact with its host takes part in a great number ofmetabolic processes.4 The supply of fermentable carbohydrateis an important factor limiting the growth of bacteria in colonand, as in other environments, bacteria that can transform theavailable substrates most rapidly will occur in the greatestnumbers.5

Polysaccharides can be extracted from many kinds of herbalor plant materials. Some Plantago plants, such as Plantago afraL., Plantago psyllium L., Plantago ovata Forsk. (isabgul),Plantago indica L., and Plantago major L., are often used intraditional medicine throughout the world because the solublefibers in their seeds are able to improve some intestinalfunctions.6 Our research group has recently isolated a pure andhomogeneous polysaccharide from the seeds of Plantagoasiatica L. with a molecular mass of 1894 kDa.7 Thepolysaccharide was found to be a highly branched heteroxylan,which consisted of a β-1,4-linked Xylp backbone with sidechains attached to O-2 or O-3. The side chains consisted of β-T-linked Xylp, α-T-linked Araf, α-T-linked GlcAp, β-Xylp-(1→3)-α-Araf, and α-Araf-(1→3)-β-Xylp, etc.8 In addition, ourrecent studies have also shown that this polysaccharide mayinduce maturation of murine dendrite cells, have antioxidantactivity in vitro, promote mouse defecation, increase short-chain fatty acid (SCFA) production in mouse colon and in vitro

fermentation, and have some effects on intestinal function invitro.9−14 However, the influence of the polysaccharide from P.asiatica L. on nutrient metabolism and colon microbiota inmouse has not been studied.In this study, the effect of the polysaccharide from P. asiatica

L. on mouse nutrient metabolism was evaluated by determiningthe apparent absorption of major nutrients, blood serumingredients, and liver lipid contents. Additionally, the effect ofthe polysaccharide on mouse colonic bacterial communities wasinvestigated. Amplicons of the V3 variable regions of bacterial16S rDNA were analyzed by denaturing gradient gel electro-phoresis (DGGE), cloning, and sequencing. Furthermore, therelationship between lipid metabolism difference and bacterialcommunity change of mouse after polysaccharide intake wasalso discussed.

■ MATERIALS AND METHODSMaterials and Animals. The seeds of P. asiatica L. were purchased

from Ji’an, Jiangxi, China, and dried in the sun before use. All reagentsused were of analytical grade and purchased from Shanghai Chemicalsand Reagents Co. (Shanghai, China).

Kunming mice, weighing 20.0 ± 2.0 g [grade II, certificate SCXK(gan) 2006-0001], were purchased from Jiangxi College of TraditionalChinese Medicine, Jiangxi Province, China. All animals used in thisstudy were cared for in accordance with the Guidelines for the Care andUse of Laboratory Animals published by the U.S. National Institutes ofHealth (NIH Publication 85-23, 1996), and all procedures were

Received: September 12, 2013Revised: December 4, 2013Accepted: December 16, 2013

Article

pubs.acs.org/JAFC

© XXXX American Chemical Society A dx.doi.org/10.1021/jf4040942 | J. Agric. Food Chem. XXXX, XXX, XXX−XXX

approved by the Animal Care Review Committee (Animal applicationapproval 0064257), Nanchang University, China.Polysaccharide Preparation. Polysaccharide from P. asiatica L.

seeds was prepared using our published method.7 Briefly, the seeds ofP. asiatica L. (50 g) were defatted with 1 L of ethanol at roomtemperature for 24 h under stirring and then extracted with 500 mL ofdoubly distilled water at 100 °C for 2 h. The residue was re-extracted.The combined aqueous extract (a highly viscous gel) was centrifuged(4800g, 10 min) and prefiltered through a cotton cloth bag. After that,the filtrate was concentrated by a rotary evaporator under reducedpressure at 55 °C to yield P. asiatica L. water extract. The filtrate wasmixed with 1.5 g/L papain and heated in water at 60 °C for 2 h. Theresulting aqueous solution was extensively dialyzed against doublydistilled water for 72 h and precipitated by adding 4 volumes ofanhydrous ethanol at 4 °C for 12 h. After centrifugation, theprecipitate was washed with anhydrous ethanol, dissolved in water, andlyophilized to yield the polysaccharide.Animal Experiment Design. Male 6-week-old Kunming mice

(20.0 ± 2.0 g) were individually housed in stainless steel cages in aroom with controlled temperature (25 ± 0.5 °C), relative humidity(50 ± 5%), and a 12 h/12 h light/dark cycle. All mice were fed thesame amount of basal diet (Table 1), which was prepared according to

the published formula,15 and water was provided ad libitum. All micewere randomly divided into two groups: (1, Polysaccharide group)Mice were orally administered polysaccharide at the dose of 0.4 g/kgbody weight of mice. The dose for the polysaccharide administrationwas selected according to our previous research.11,12 (2, Controlgroup) Mice were given distilled water of the same volume as themean volume of the polysaccharide groups. Each group had 12 mice,and mice were housed individually. All of the mice were given oraladministration of polysaccharide or water at about 9:00 a.m. every dayfor 30 days. In addition, we determined the body weight of eachmouse every day before gavage and changed the volume of thepolysaccharide to ensure that the mice were given polysaccharide atthe dose of 0.4 g/kg body weight by gavage every day. Throughout theexperiment, the animals’ general health status and body mass wereobserved twice daily.At the end of the designated experiment period, feces of the mice

were collected individually and used to measure the apparentabsorption of major nutrients (carbohydrate, protein, and fat). Bloodwas drawn from the eyes of the mice, and the serum was isolated bybrief centrifugation of the whole blood. After that, the mice weresacrificed, and the liver and colon (with contents) were obtained.Then the colons were aseptically removed immediately and placed onan ice-cold plate and longitudinally opened, and the colon contentswere collected. The feces, liver, and colon content were then kept at−20 °C for further analysis.

Measurement of the Apparent Absorption of MajorNutrients (Carbohydrate, Protein, and Lipid). The feeds andfreeze-dried feces were crushed and measured for their carbohydrate,crude protein, and crude lipid.16 The carbohydrate content wasdetermined colorimetrically: To an aliquot of the feed (200 mg) orfeces (200 mg) was added 10 mL of a 0.2 M acetate buffer (pH 4.8)and 5 mL of a 0.12% glucoamylase solution (Biochemical Co.,Shanghai, China). After the addition of distilled water to give 50 mL ofthe solution, a few drops of toluene were added, and the mixture wasincubated at 37 °C for 24 h. After hydrolysis, the mixture was filtered,and an aliquot of the solution was used for determining the reducingsugar according to the method of Somoyi-Nelson, 0.9 being adoptedas the conversion factor to carbohydrate. The crude protein contentwas determined by titration: An aliquot of the feed (200 mg) or feces(200 mg) was weighed and degraded by H2SO4. The nitrogen contentwas determined using the Kjeldahl method with a Kjeltec TM 8400instrument (Foss, Denmark), 6.25 being adopted as the conversionfactor to protein. The crude lipid content was determined gravimetri-cally according to the Soxhlet method, using an aliquot of the feed(200 mg) or feces (200 mg). The apparent absorption of majornutrients (carbohydrate, protein, and lipid) was calculated by using thefollowing equation:

= − ×

apparent nutrient absorption (%)

[(intake excretion)/intake] 100 (1)

Measurement of Blood Serum Ingredients and Liver LipidContents. Total serum protein, albumin, albumin/globulin ratio,glucose, urea nitrogen, creatinine, triglyceride, total cholesterol, high-density lipoprotein (HDL)-cholesterol, low-density lipoprotein(LDL)-cholesterol, atherogenic index, glutamic oxalacetic trans-aminase activity (GOT), and glutamic pyruvic transaminase activity(GPT) were determined by using an Automatic Biochemical Analyzer(AU2700, Olympus, Hamburg, Germany). Freeze-dried livers wereground, and the lipids were extracted with chloroform/methanol (2:1,v/v).17 The total lipid content was quantified gravimetrically, and thetotal cholesterol and triglycerides were determined by using theAutomatic Biochemical Analyzer, whereas phospholipid contents weredetermined colorimetrically with a determiner assay kit (JinghaoMedix Co., Beijing, China).

Colon Content DNA Extraction and PCR Amplification. Todescribe bacterial diversity in the mouse colon tract, total DNA wasisolated from colon content samples and used as a template for PCRamplification. Amplicons of the V3 regions of bacterial 16S rDNAwere analyzed using the DGGE analysis method. Sample DNAextraction (200 mg of frozen colon content for every mouse) wasperformed by using a QIAamp DNA Stool Mini Kit (QIAGEN, Inc.,Shanghai, China), according to the manufacturer’s instruction. ForDGGE analysis, PCR amplifications with V3 region primers against the16S rDNA were performed in a 25 μL (total volume) mixturecontaining 1.2 μL of forward primer with a GC-clamp [primer F338-GC, sequence (5′−3′), CGCCCGCCGCGCGCGGCGGGCGGG-GCGGGGGCACGGGGGGACTCCTACGGGAGGCAGCAG] and1.2 μL pf reverse primer [primer R518, sequence (5′−3′), ATTACC-GCGGCTGCTGG], 0.5μL of deoxynucleoside triphosphate at aconcentration of 10 mM, 2.5 μL of PCR reaction buffer, 0.2 μL ofDNA polymerase (5 U/μL), 1.0 μL of DMSO (4%, v/v), 1 μL ofDNA template with an appropriate amount, and 17.4 μL of sterilewater.18 PCR amplification was performed with a Biosci PCR systemas follows: an initial denaturation of 94 °C for 4 min, followed by 30cycles of 94 °C for 0.5 min, 56 °C for 0.5 min, and 72 °C for 1 min,and a final elongation at 72 °C for 10 min. Aliquots of 5 μL wereanalyzed by electrophoresis on an agarose gel (1%) to check the sizeand amounts of the amplicons.

DGGE Profiling. DGGE was performed with the Bio-RadDCodeTM system (Bio-Rad Ltd., Hercules, CA, USA). PCR productswere loaded onto 1 mm thick 8% (w/v) polyacrylamide (acrylamide/bisacrylamide = 37.5:1) gels containing a 35−65% linear denaturinggradient. TAE buffer (1×, 40 mM Tris-acetate, 1 mM Na-EDTA, pH8.0) was used as the electrophoresis buffer. The electrophoresis was

Table 1. Composition of the Basal Diet for Mice

component g/kg dry weight (DW) basis

corn starch 454soybean meal 240sucrose 100corn oil 60cellulose 50gelatinized starch 10vitamin mixturea 10mineral mixtureb 76

aContaining, in g/kg (DW basis): 0.25 (250000 IU) vitamin A; 0.002(65000 IU) vitamin D3; 2 (2000 IU) vitamin E; 0.41 menadinoe; 0.07folic acid; 1.25 niacin; 0.5 calcium pantothenate; 0.2 riboflavin; 0.25thiamin; 0.5 pyridoxine; 0.003 cyanocobalamine; 0.025 biotin; 70.0cholin chloride. bContaining, in g/kg (DW basis): 69.6 CaCO3; 313.1C12H10Ca3·4H2O; 114.6 CaHPO4·2H2O; 222.2 K2HPO4·3H2O; 126.7KCl; 78.3 NaCl; 38.9 MgCO3; 35.8 MgCO3; 0.204 MnSO4·H2O;0.042 KI; 0.515NaF; 0.091 AlNH4(SO4)2·12H2O.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4040942 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXB

initiated by prerunning for 5 min at a voltage of 220 V andsubsequently run at a fixed voltage of 85 V for 16 h at 60 °C. The gelwas stained with AgNO3 and developed after completion ofelectrophoresis. The obtained DGGE patterns were subsequentlynormalized and analyzed with Bio-Rad Quantity One 4.4.0 software.During the processing, the different lanes were defined, backgroundwas subtracted, differences in the intensity of the lanes werecompensated during normalization, and the correlation matrix wascalculated. Clustering was done with Pearson correlation and theunweighted pair group mean average (UPGMA) method.Identification of DGGE DNA Bands. The bands of interest were

excised from the gel using a sterile blade and incubated overnight at 4°C in Tris−EDTA buffer (pH 8.0) for future sequencing. PCRamplifications of the DNA fragments from the excised gel were carriedout according to the same protocols described above with thecorresponding primers [forward primer F338, sequence (5′−3′),ACTCCTACGGGAGGCAGCAG; reverse primer R518, sequence(5−3′), ATTACCGCGGCTGCTGG] and 1 μL aliquot of the gelelution.18 PCR products were excised from a 1.0% agarose gel andpurified with a DNA Gel Extraction Kit (Tiangen, China). Isolation ofDNA bands from the DGGE gel, cloning of excised bands, andsequence analysis were performed as described by Gong et al.19 Thesequences were submitted to GenBank and the RDP database todetermine their most related bacterium. The DNA sequences areavailable in the GenBank database with accession numbers.Statistical Analysis. All of the experiments were done in triplicate.

Statistical analysis was carried out using SPSS (version 16.0, Chicago,IL, USA). The results were expressed as the mean ± standarddeviation and compared using the Tukey test at a 5% confidence level.DGGE banding patterns in each group were digitalized by QuantityOne software (Bio-Rad). The intensity and relative position of eachDGGE band were determined manually with background subtraction.The intensity of each band was expressed as percentage of theintegrated intensity of the entire lane. The matrix of intensity andrelative position was analyzed by principal component analysis (PCA)using SPSS software.

■ RESULTS AND DISCUSSION

General Health Status of Mice. Throughout the experi-ment period, no noticeable behavioral or activity changes wereobserved in the mice, and no treatment-related illness or deathoccurred. The growth of mice appeared normal throughout theexperiment period, and no mouse experienced diarrhea orconstipation. There was no observable difference in theanimals’ body mass and hair luster between the polysacchar-ide-treated group and control group.Effect of Polysaccharide on Absorption of Major

Nutrient of Mice. The apparent absorption of major nutrients(carbohydrate, protein, and lipid) is summarized in Table 2.Although the values for nutrient absorption were all nearlyquantitative, there was a significant difference (p < 0.05)between the control group (94.8 ± 0.2%) and polysaccharide

group (87.4 ± 0.3%) for lipid apparent absorption. Thisobservation suggested that the polysaccharide from P. asiaticaL. may have had some influence on lowering the absorption oflipid, which was consistent with our previous finding that thepolysaccharide from P. asiatica L. could inhibit some pancreaticlipase activity and reduce cholesterol available for absorption.13

The polysaccharide from P. asiatica L., which was found topromote mouse defection in our previous study, might alsohave some effect on promoting cholesterol and lipidexcretion.11 In addition, the polysaccharide was found toincrease the propionic acid production in mouse colon, whichmight influence the lipid metabolism.12 Furthermore, thepolysaccharide from P. asiatica L. was found to bind bile acidin vitro,13 which may also affect the excretion and metabolismof cholesterol and lipid. Other studies showed that some otherpolysaccharides could enhance the excretion of lipids in feces,20

which was similar to our results.Effect of Polysaccharide on the Characteristics of

Blood Serum and Liver Lipid Levels. The characteristics ofthe blood serum and the liver lipid levels in mice of thepolysaccharide group or control group are given in Tables 3

and 4, respectively. There was in general no marked differencein serum chemistry, apart from total triglyceride, totalcholesterol, and atherogenic index (Table 3). Total triglyceride,total cholesterol, and atherogenic index in blood serum tendedto be lower in the group of mice given oral administration ofpolysaccharide, with the difference between the control group

Table 2. Effect of Polysaccharide from P. asiatica L. onApparent Absorption of Carbohydrate, Protein, and Lipid inMicea

apparent absorption (%)

control group polysaccharide group

carbohydrate 99.9 ± 0.0 a 99.6 ± 0.3 alipid 94.8 ± 0.2 a 87.4 ± 0.3 bprotein 99.9 ± 0.1 a 99.9 ± 0.1 a

aEach value is the mean ± standard deviation (n = 12); means in thesame row not sharing a common letter are significantly different (p <0.05).

Table 3. Effect of Polysaccharide from P. asiatica L. onCharacteristic Blood Serum Chemistry of Micea

control group polysaccharide group

glucose (mmol/L) 6.72 ± 0.51 a 6.56 ± 0.39 atotal protein (g/L) 55.3 ± 1.1 a 54.0 ± 1.1 aalbumin (g/L) 31.3 ± 0.9 a 31.6 ± 0.5 aalbumin/globulin 1.3 ± 0.1 a 1.4 ± 0.1 aurea nitrogen (mmol/L) 9.5 ± 0.6 a 9.4 ± 0.5 acreatinine (μmol/L) 33 ± 2 a 35 ± 3 atriglyceride (mmol/L) 1.90 ± 0.11 a 1.25 ± 0.12 btotal cholesterol (mmol/L) 3.27 ± 0.26 a 2.65 ± 0.28 bHDL-cholesterol (mmol/L) 2.62 ± 0.11 a 2.34 ± 0.14 aLDL-cholesterol (mmol/L) 0.58 ± 0.14 a 0.32 ± 0.09 aatherogenic indexb 0.22 ± 0.01 a 0.17 ± 0.01 bGOT (U/L)c 193 ± 21 a 192 ± 29 aGPT (U/L)d 132 ± 19 a 134 ± 21 a

aEach value is the mean ± standard deviation (n = 12); means in thesame row not sharing a common letter are significantly different (p <0.05). bAtherogenic index = (LDL-cholesterol)/(HDL-cholesterol).cGOT, glutamic oxalacetic transaminase activity. dGPT, glutamicpyruvic transaminase activity.

Table 4. Effect of Polysaccharide from P. asiatica L. on LiverLipid Levels in Micea

control group polysaccharide group

cholesterol (μmol/g of liver) 6.4 ± 0.4 a 5.5 ± 0.2 btriglyceride (μmol/g of liver) 33.5 ± 1.2 a 32.9 ± 0.9 aphospholipid (μmol/g of liver) 35.4 ± 2.2 a 35.9 ± 1.8 atotal lipid (mg/g liver) 49.6 ± 2.2 a 44.6 ± 2.2 b

aEach value is the mean ± standard deviation (n = 12); means in thesame row not sharing a common letter are significantly different (p <0.05).

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4040942 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXC

and the polysaccharide group being significant (p < 0.05).Studies have shown that an elevation of serum lipids, especiallycholesterol, is one of the most important risk factors leading tocoronary heart disease.21 Thus, the cholesterol-lowering effectof the polysaccharide from P. asiatica L. was beneficial forhealth.A difference in liver lipid levels was apparent for cholesterol

and total lipid (Table 4). The liver cholesterol and total lipidlevel in mice given oral administration of polysaccharide wereboth significantly lower (p < 0.05) than that of the controlgroup. Thus, the polysaccharide from P. asiatica L. appears toplay a role in lowering the liver cholesterol level. Some otherstudies also indicated that polysaccharides present in some foodmaterials could reduce the cholesterol in serum and lipid inliver,22 which was similar to our results.In contrast to liver cholesterol level, the liver triglyceride level

in mice was not significantly decreased after polysaccharideadministration for 30 days (Table 4, p > 0.05). We haveobserved that the polysaccharide intake could significantlylower the serum triglyceride level of mice (Table 3, p < 0.05).The influence of the polysaccharide on serum triglyceride levelmay be faster than its effect on liver triglyceride level becausethe triglyceride was accumulated in liver and the reducing effectby polysaccharide intake may need more time. However, forcholesterol, the polysaccharide could significantly reduce both

the serum and liver cholesterol levels (Tables 3 and 4, p <0.05). Although cholesterol was also accumulated in the liver,the liver cholesterol level was significantly reduced by thepolysaccharide intake. This may be because except for thepolysaccharide itself, the organic acids, particularly propionate,that are produced by colonic fermentation could also influencethe liver cholesterol metabolism.23 The polysaccharide from P.asiatica L. was found to increase the propionic acid productionof mouse colon in our previous study.12 Thus, the effect of thepolysaccharide on lowering the liver cholesterol level may bestronger than its effect on lowering the liver triglyceride level.

Effect of Polysaccharide on Mice Colon Microbiota.Diversity indices for the DGGE bacterial profiles (Figure 1A)generated from the 16S rDNA V3 region are shown in Table 5.The richness of the DGGE banding profiles for thepolysaccharide group (30.80 ± 3.96) was significantly higherthan for the control group (24.60 ± 3.05, p < 0.05) after 30days of administration. In addition, a significant increase wasalso observed in the Shannon−Wiener index and maximumShannon−Wiener index for the polysaccharide group relative tothe control group (p < 0.05). There was no significantdifference for the evenness between the polysaccharide groupand control group (p > 0.05).The effect of polysaccharide on the bacterial communities

after 30 days of oral administration was investigated by PCR-

Figure 1. Microbiota analysis for the colon content in mice from polysaccharide group and control group: (A) denaturing gradient gelelectrophoresis (DGGE) of V3 regions of bacterial 16S rDNA amplifications; (B) UPGMA cluster analysis of DGGE bacterial profiles; (C) PCAscores plot of DGGE bacterial profiles. PC1, PC2, and PC3 represent principal factors and their probability (%) of influence on the bacterial profiles.C1−C5, colon content samples individually from five mice randomly chosen from the control group after 30 days of administration; P1−P5, coloncontent samples individually from five mice randomly chosen from the polysaccharide group after 30 days of administration.

Table 5. Diversity Indices Calculated from the DGGE Banding Profiles Generated from the V3 Region of Bacterial 16S rDNA

treatment richness (S) H′a H′maxb evenness (E)

control group 24.60 ± 3.05 ac 3.02 ± 0.13 a 3.10 ± 0.12 a 0.97 ± 0.01 apolysaccharide group 30.80 ± 3.96 b 3.34 ± 0.14 b 3.42 ± 0.13 b 0.98 ± 0.01 a

aH′, Shannon−Wiener index. bH′max, maximum Shannon−Wiener index. cValues are mean of three determinations ± standard deviation. Differentletters indicate significant difference (p < 0.05) from each other in the same column.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4040942 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXD

DGGE bacterial profiling and identification of putative bacterialspecies through sequence analysis of recovered DGGE DNAbands. To compare the predominant bacterial communitiesbetween different treatments, PCR amplification of the V3region of the 16S rRNA gene was carried out withcorresponding primers. DGGE banding patterns of the V3region of colon content of each group are shown in Figure 1A.The influence of polysaccharide intake on bacterial communitystructure was investigated. DGGE image analysis showed thatthe bacterial community structure from the colon contentsamples of the polysaccharide group was similar (Figure 1A)and also grouped together in the UPGMA cluster analysis(Figure 1B). In addition, it could also be seen that the bacterialcommunity structure of colon content from polysaccharidegroup mice was low in similarity with the control group, whichmeans the influence of polysaccharide intake appeared to be amajor factor affecting the community. Furthermore, the PCAscores plot (Figure 1C) was also used to compare thedifferences between banding patterns of different treatments.In the PCA scores plot, “sit together” demonstrates similaritybetween samples. As shown in Figure 1C, the banding patternof the control group was different from that of thepolysaccharide group. Because banding patterns were relatedto the bacterial communities, the above results suggest that thepredominant bacterial communities in the colon content ofpolysaccharide group mice were different from those of thecontrol group mice.Several numerically dominant bacterial species in the colon

content for the polysaccharide group corresponding to thepresence of some intense DGGE DNA bands, which weredifferent from the control group, were observed (Figure 1A).To further clarify the differences, bands that showed differentpercentages of density among groups were sequenced fordetermining their putative identity of bacteria. The putativeidentity of bacteria is presented in Table 6, which may suggesttheir possible role in the utilization of the polysaccharide. Asshown in Figure 1A and Table 6, 14 interesting excised bandswere identified.Although bands 1, 3−6, and 8−12 have the same migration

in different lanes, higher peak density of bands can be observedin the polysaccharide-treated group than in the control group(Figure 1A). Sequencing results showed that their most related

bacteria (bands 9−12) were Bacteroides sp. Bacteroidetes candegrade polysaccharides and metabolic intermediates. Theywere also able to produce SCFAs such as acetate, lactate,propionate, and others.24 Band 1 was closest to butyrate-producing bacteria and affiliated with the family of Butyrivibrio.A few species-related bands, 5, 6, and 8, had a closerrelationship with probiotics Bif idobacterium bif idum, Lactoba-cillus fermentum, and Lactobacillus reuteri, respectively. Theseputative butyrate-producing bacteria or probiotics wereobserved to increase after polysaccharide intake in this study,similar to these bacteria reported to utilize fibers previously.25

Bands 3 and 4 were sequenced and identified as belonging toEubacterium sp., which were reported to be very activefermentation bacteria for fiber.26

Bands 2, 7, 13, and 14 were sequenced as well (Figure 1A),which had higher peak density in the control group than in thepolysaccharide-treated groups. The sequencing results showedthat their most related bacteria were Clostridiales sp. andHelicobacter sp., which are the common commensal bacteria inthe colon and not good for health in a large amount.18 Thesesequencing results indicated that polysaccharide treatmentcould decrease Clostridiales sp. and Helicobacter sp. in themouse colon microbiota.On the basis of the banding patterns (Figure 1A) and

sequencing results (Table 6), it can be inferred that orallyadministered polysaccharide could substantially affect themouse colon microbiota. In addition, some other papers haveindicated that the predominant increased species in colonmicrobiota belong to the genera Bacteroides, Eubacterium, andBif idobacterium after the supply of fermentable carbohydrates.Many other bacteria also occur in high numbers, includingLactobacilli and various anaerobic Gram-positive cocci.27 Theirresults were similar to ours.It was reported that there were some relationships between

lipid metabolism difference and bacterial community change.The intestinal microbiota develops an important biochemicalactivity within the host body by providing additional metabolicfunctions. Specific components of the commensal microbiotaalso regulate serum lipids and cholesterol by taking part inmetabolism.28 Some intestinal bacteria could influence lipidmetabolism. Butyrivibrio proteoclasticus, a butyrate-producingruminal bacterium, could have some effects on cholesterol andlipid utilization.29 Lactic acid bacteria and Bif idobacteria couldremove cholesterol or carry out cholesterol assimilation. Forexample, Lactobacillus fermentum was reported to reduce serumlipids and cholesterol in subjects.30 Low lipid level wasassociated with higher Bif idobacterium numbers in cecal contentin mice.31 Furthermore, Eubacterium rectale, Bacteroidesthetaiotaomicrometer, and Bacteroides vulgatus were also foundto be cholesterol-reducing bacteria.32 Thus, the Bacteroides sp.,Eubacterium sp., Butyrivibrio proteoclasticus, Bifidobacteriumbifidum, Lactobacillus fermentum, and Lactobacillus reuteri,which were found increased in mouse colon after the intakeof polysaccharide P. asiatica L., may have some effect on themouse lipid metabolism in this study.In addition, some probiotics-related metabolites might also

exert beneficial effects on cholesterol and lipid metabolism.Propionic acid was reported to affect the liver cholesterolmetabolism.23 The polysaccharide from P. asiatica L. was foundto increase the propionic acid production in mice colon in aprevious study,12 which might influence the lipid metabolism.Bile acid was found to be closely related to cholesterol and lipidmetabolism; thus, the polysaccharide from P. asiatica L., which

Table 6. Band Identification for Searching the BacterialIdentity in Denaturing Gradient Gel Electrophoresis(DGGE) Using Bacterial V3 Primers

strain closest relative similarity (%) GenBank no.

1 Butyrivibrio proteoclasticus 96 U37378.12 Clostridiales sp. 97 AB781636.13 Eubacterium rectale 100 AB626630.14 Eubacterium sp. 98 AY169428.15 Bif idobacterium bif idum 95 GU936674.16 Lactobacillus fermentum 98 JQ669802.17 Helicobacter pylori 96 JX455160.18 Lactobacillus reuteri 100 HQ615667.19 Bacteroides salanitronis 100 AB253731.110 Ruminococcus torques 99 NR_036777.111 Bacteroides thetaiotaomicrometer 99 NR_074277.112 Bacteroides vulgatus 95 HQ012024.113 Clostridium clarif lavum 98 NR_041235.114 Clostridium caenicola 99 NR_041311.1

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4040942 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXE

was found to bind bile acid in vitro,13 might also affectcholesterol and lipid metabolism.Our results showed the effect of polysaccharide from P.

asiatica L. on mouse lipid metabolism and colon microbiota forthe first time. The intake of polysaccharide from P. asiatica L.could lower the apparent absorption of lipid and decrease theserum and liver lipid level in mice. Mice from thepolysaccharide group showed a higher colon bacterial diversitythan the control group. Bacteroides sp., Eubacterium sp.,butyrate-producing bacteria Butyrivibrio sp., and probioticsBif idobacterium bif idum, Lactobacillus fermentum, and Lactoba-cillus reuteri in mouse colon were all increased after mice weregiven oral administration of the polysaccharide. Thesesuggested that the intake of polysaccharide from P. asiatica L.may be beneficial for lipid metabolism and colon microbiota.The changed colon bacteria may play a role in utilization of thepolysaccharide and mouse lipid metabolism, which was alsoaffected by polysaccharide intake.

■ ASSOCIATED CONTENT*S Supporting InformationStructure of the polysaccharide from the seeds of P. asiatica L.This material is available free of charge via the Internet athttp://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*(S.-P.N.) Phone/fax: +86-791-88304452. E-mail: spnie@ncu.edu.cn.FundingThis study is financially supported by the Key Program of theNational Natural Science Foundation of China (31130041), theNational Key Technology R&D Program of China(2012BAD33B06), the National Natural Science Foundationof China (20802032 and 21062012), the Program for NewCentury Excellent Talents in University (NCET-12-0749), andResearch Projects of the State Key Laboratory of Food Scienceand Technology (SKLF-ZZA-201301 and SKLF-KF-201202),which are gratefully acknowledged.NotesThe authors declare no competing financial interest.

■ ABBREVIATIONS USEDHDL, high-density lipoprotein; LDL, low-density lipoprotein;DGGE, denaturing gradient gel electrophoresis; PCA, principalcomponent analysis; UPGMA, unweighted pair group meanaverage; SCFA, short-chain fatty acid

■ REFERENCES(1) Neyrinck, A. M.; Possemiers, S.; Verstraete, W.; De Backer, F.;Cani, P. D.; Delzenne, N. M. J. Nutr. Biochem. 2012, 23, 51−59.(2) Gropper, S. S.; Smith, J. L. In Advanced Nutrition and HumanMetabolism, 1st ed.; Wadsworth Publishing: New York, 2012.(3) Xu, X.; Xu, P.; Ma, C.; Tang, J.; Zhang, X. Biotechnol. Adv. 2012,31, 318−337.(4) Topping, D. L.; Clifton, P. M. Physiol. Rev. 2001, 81, 1031−1064.(5) Cummings, J. H.; Macfarlane, G. T. J. Appl. Microbiol. 1991, 70,443−459.(6) Mahady, G. B.; Fong, H. S.; Farnsworth, N. R. In WHOMonographs on Selected Medicinal Plants, 1st ed.; World HealthOrganisation: Geneva, Switzerland, 1999.(7) Yin, J. Y.; Nie, S. P.; Zhou, C.; Wan, Y.; Xie, M. Y. J. Sci. FoodAgric. 2010, 90, 210−217.

(8) Yin, J.; Lin, H.; Li, J.; Wang, Y.; Cui, S. W.; Nie, S.; Xie, M.Carbohydr. Polym. 2012, 87, 2416−2424.(9) Huang, D. H.; Xie, M. Y.; Yin, J. Y.; Nie, S. P.; Tang, Y. F.; Zhou,C. J. Ethnopharmacol. 2009, 124, 493−498.(10) Tang, Y.; Huang, D.; Yin, J.; Zhou, C.; Xie, X.; Xie, M. Food Sci.2007, 28, 517−520.(11) Wu, G.; Tian, Y.; Xie, M.; Li, C. Food Sci. 2007, 28, 514−516.(12) Hu, J. L.; Nie, S. P.; Min, F. F.; Xie, M. Y. J. Agric. Food Chem.2012, 60, 11525−11532.(13) Hu, J. L.; Nie, S. P.; Li, C.; Xie, M. Y. Int. J. Biol. Macromol.2013, 54, 264−269.(14) Hu, J. L.; Nie, S. P.; Li, C.; Fu, Z. H.; Xie, M. Y. J. Agric. FoodChem. 2013, 61, 6092−6101.(15) Adachi, T.; Ono, Y.; Koh, K. B.; Takashima, K.; Tainaka, H.;Matsuno, Y.; Nakagawa, S.; Todaka, E.; Sakurai, K.; Fukata, H. FoodChem. Toxicol. 2004, 42, 445−452.(16) Takahashi, T.; Maeda, H. Biosci., Biotechnol., Biochem. 1999, 63,1340−1345.(17) Folch, J.; Lees, M.; Sloane-Stanley, G. J. Biol. Chem. 1957, 226,497−509.(18) Muyzer, G.; De Waal, E. C.; Uitterlinden, A. G. Appl. Environ.Microbiol. 1993, 59, 695−700.(19) Gong, J.; Yu, H.; Liu, T.; Gill, J.; Chambers, J.; Wheatcroft, R.;Sabour, P. J. Appl. Microbiol. 2008, 104, 1372−1382.(20) Dall’Agnol, R.; Lino von Poser, G. J. Ethnopharmacol. 2000, 71,337−341.(21) Kannel, W. B.; Castelli, W. P.; Gordon, T. Ann. Intern. Med.1979, 90, 969−975.(22) Gallaher, C. M.; Munion, J.; Hesslink, J. R.; Wise, J.; Gallaher, D.D. J. Nutr. 2000, 130, 2753−2759.(23) Nishimura, N.; Nishikawa, H.; Kiriyama, S. J. Nutr. 1993, 123,1260−1269.(24) Duncan, S.; Louis, P.; Flint, H. Lett. Appl. Microbiol. 2007, 44,343−350.(25) Crittenden, R.; Karppinen, S. J. Sci. Food Agric. 2002, 82, 781−789.(26) Varel, V. H.; Yen, J. T. J. Anim. Sci. 1997, 75, 2715−2722.(27) Finegold, S. M.; Attebery, H. R.; Sutter, V. L. Am. J. Clin. Nutr.1974, 27, 1456−1469.(28) Gill, S. R.; Pop, M.; DeBoy, R. T.; Eckburg, P. B.; Turnbaugh, P.J.; Samuel, B. S.; Gordon, J. I.; Relman, D. A.; Fraser-Liggett, C. M.;Nelson, K. E. Science 2006, 312, 1355−1359.(29) Moon, C. D.; Pacheco, D. M. Int. J. Syst. Evol. Microbiol. 2008,58, 2041−2045.(30) Simons, L. A.; Amansec, S. G. Nutr. Metab. Cardiovasc. 2006, 16,531−535.(31) Cani, P. D.; Amar, J.; Iglesias, M. A.; Poggi, M.; Knauf, C.;Bastelica, D.; Neyrinck, A. M.; Fava, F.; Tuohy, K. M.; Chabo, C.Diabetes 2007, 56, 1761−1772.(32) Sadzikowski, M.; Sperry, J.; Wilkins, T. Appl. Environ. Microbiol.1977, 34, 355−362.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf4040942 | J. Agric. Food Chem. XXXX, XXX, XXX−XXXF