2015Extraction optimizatin, characterization and antioxidant activity

Carbohydrate Polymers 128(2015)52–62Contents lists available at ScienceDirectCarbohydratePolymersj o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /c a r b p olExtraction optimization,characterization and antioxidant activity in vitro of polysaccharides from mulberry (Morus alba L.)leavesQingxia Yuan a ,Yufeng Xie a ,Wei Wang b ,Yuhua Yan a ,Hong Ye a ,Saqib Jabbar a ,c ,Xiaoxiong Zeng a ,∗aCollege of Food Science and Technology,Nanjing Agricultural University,Nanjing 210095,Jiangsu,ChinabCollege of Life Sciences and Laboratory Center of Life Sciences,Nanjing Agricultural University,Nanjing 210095,Jiangsu,China cInstitute of Food Science and Nutrition,University of Sargodha,Sargodha,Pakistana r t i c l ei n f oArticle history:Received 10February 2015Received in revised form 3April 2015Accepted 7April 2015Available online 21April 2015Keywords:Mulberry leaf Polysaccharide ExtractionCharacterization Antioxidanta b s t r a c tExtraction optimization,characterization and antioxidant activity in vitro of polysaccharides from mul-berry leaves (MLP)were investigated in the present study.The optimal extraction conditions with an extraction yield of 10.0±0.5%for MLP were determined as follows:extraction temperature 92◦C,extraction time 3.5h and ratio (v/w,mL/g)of extraction solvent (water)to raw material 34.Two purified fractions,MLP-3a and MLP-3b with molecular weights of 80.99and 3.64kDa,respectively,were obtained from crude MLP by chromatography of DEAE-Cellulose 52and Sephadex G-100.Fourier transform-infrared spectroscopy revealed that crude MLP,MLP-3a and MLP-3b were acidic polysaccha-rides.Furthermore,crude MLP and MLP-3a had more complicated monosaccharide compositions,while MLP-3b had a relatively higher content of uronic acid.Crude MLP,MLP-3a and MLP-3b exhibited potent Fe 2+chelating power and scavenging activities on 1,1-diphenyl-2-picrylhydrazyl,hydroxyl,superoxide and 2,2 -azinobis-(3-ethyl-benzothiazolin-6-sulfonic acid)radicals.The results suggested that MLP could be explored as natural antioxidant.©2015Elsevier Ltd.All rights reserved.1.IntroductionMulberry (Morus alba L.),a multipurpose agro-forestry plant that belongs to the family of Moraceae,is widely distributed in tropical,subtropical and temperate areas (Agarwal &Kanwar,2007).It is commonly used as a silkworm (Bombyx mori L.)diet,alternative medicine in China and Japan and a kind of tea due to its low toxicity and good therapeutic performance (Chung,Kim,Kim,&Kwon,2013;Jia,Tang,&Wu,1999;Nookabkaew,Rangkadilok,&Satayavivad,2006;Wang,Fang,Ma,&Zhang,2014;Zhong,Furne,&Levitt,2006).It has been reported that mulberry contains a lot of bioactive compounds including polysaccharides,1-deoxynojirimycin,moracin,chlorogenic acid,rutin,flavonol gly-cosides and anthocyanins,which are associated with its biological functions such as anti-obesity,anti-diabetes,anti-oxidation,anti-inflammation and anti-atherosclerosis (Harauma et al.,2007;Hunyadi,Martins,Hsieh,Seres,&Zupkó,2012;Katsube,Tsurunaga,Sugiyama,Furuno,&Yamasaki,2009;Kimura,Nakagawa,Kubota,∗Corresponding author.Tel.:+862584396791;fax:+862584396791.E-mail address:zengxx@ (X.Zeng).Kojima,&Goto,2007;Li et al.,2011;Peng et al.,2011;Yang,Wang,Wang,&Zhang,2012;Yatsunami,Ichida,&Onodera,2008;Zhang et al.,2014a ).Mulberry leaves polysaccharides (MLP),the main active compo-nents of mulberry leaves,have been attracted increasing attention as other herb polysaccharides,due to their multiple biological activities such as anti-diabetic,anti-tumor,anti-inflammatory and immunostimulatory effects (Li,Chen,Wang,Tian,&Zhang,2010;Li et al.,2011;Wang,Li,&Jiang,2010b;Yan,Wang,&Wu,2014;Yang,Zhao,Yang,&Ruan,2008;Yang et al.,2012;Zhang et al.,2010,2014a ).Besides,it has been reported that many polysaccharides including MLP have potential antioxidant activities (Scartezzini &Speroni,2000;Wang et al.,2013).However,the reports on the antioxidant activity of MLP are relatively insufficient (Samavati &Yarmand,2013;Wang et al.,2010b ).It is well known that the biological functions including antioxidant activity of polysaccha-rides are intimately related to their structure features such as chemical components,molecular weight,monosaccharide com-position and glycosidic linkage (You et al.,2013;Zeng,Zhang,&Jia,2014).Accordingly,a comprehensive study of purification,characterization and antioxidant activities of MLP will provide useful information on the relationship between antioxidant activ-ity and structure features.Furthermore,in order to study and/10.1016/j.carbpol.2015.04.0280144-8617/©2015Elsevier Ltd.All rights reserved.Q.Yuan et al./Carbohydrate Polymers128(2015)52–6253 explore MLP better,an efficient technique for the extraction of MLPis necessary.For polysaccharide extraction,hot water extractiontechnology is still the main and classic method due to its con-venience,low cost and high extraction yield(Passos&Coimbra,2013;Wei et al.,2010).However,the existing water extractionmethod for MLP needs long extraction time and several extractioncycles(Samavati&Yarmand,2013).Therefore,we report here theextraction,optimization,characterization and antioxidant activ-ity in vitro of MLP.First,response surface methodology(RSM)based on a Box–Behnken design(BBD)was applied to optimizethe extraction conditions,and the resulting extraction conditionswere used to prepare crude MLP through water extraction andethanol precipitation.The crude MLP was then purified by ion-exchange chromatography of DEAE-52cellulose and size exclusionchromatography of Sephadex G-100,and the crude MLP and itspurified fractions were further characterized via chemical analy-sis,high performance liquid chromatography(HPLC)and Fouriertransform-infrared(FT-IR)spectroscopy.Finally,the antioxidantactivities in vitro of crude MLP and its purified fractions were inves-tigated.To the best of our knowledge,it is thefirst report on theantioxidant activities of the purified fractions of MLP.2.Materials and methods2.1.Materials and chemicalsThe mulberry leaves were collected from the mulberry planta-tion in Bozhou(Anhui,China),washed with top water,air dried atroom temperature and ground intofine powder.DEAE-52cellu-lose,Sephadex G-100,mannose(Man),arabinose(Ara),galactose(Gal),galacturonic acid(GalA),glucose(Glc),glucuronic acid(GlcA),ribose(Rib),xylose(Xyl),3-methyl-1-phenyl-2-pyrazolin-5-one(PMP),nitroblue tetrazolium(NBT),1,1-diphenyl-2-picrylhydrazyl(DPPH),phenazine methosulphate(PMS),reduced nicotinamideadenine dinucleotide(NADH),ferrozine,[2,2 -azinobis-(3-ethyl-benzothiazolin-6-sulfonic acid)]diammonium salt(ABTS)and2,4,6-tris(2-pyridyl)-s-triazine(TPTZ)were purchased from SigmaChemical Co.(St.Louis,MO,USA).Rhamnose(Rha)and fucose(Fuc)were purchased from Aladdin Chemical Reagent Co.,Ltd.(Shanghai,China).All other chemicals used were of analytical grade.2.2.Extraction of polysaccharidesIn order to defat and remove most of the monosaccharides,oligosaccharides,pigments and other small molecules,the powderof mulberry leaves was treated with85%ethanol twice at roomtemperature for24h.The resulting residue was dried and usedfor next extraction.The dried pretreated sample was extractedwith designed extraction temperature,extraction time and ratioof extraction solvent(deionized water)to raw material.The result-ing solution wasfiltered,mixed with a triple volume of absoluteethanol and kept overnight.The precipitates were collected by cen-trifugation at4000rpm for15min,washed with absolute ethanoland acetone and dried,affording the crude MLP.The extractionyield was calculated according to the following formula:Extraction yield(%)=W1W0×100where W1and W0are the weights of crude MLP and pretreated sample,respectively.2.3.Experimental design of RSMEffects of extraction parameters including extraction tempera-ture,extraction time,extraction cycles and ratio of water to raw material on the yields of MLP were investigated by single-factor tests(data not shown).Accordingly,three major factors(extraction temperature,extraction time and ratio of water to raw material) were chosen and their proper ranges were determined based on the preliminary experimental results.Furthermore,a three-level, three-variable BBD was applied to determine the optimal levels of extraction variables including the extraction temperature(X1), extraction time(X2)and ratio of water to raw material(X3)for the extraction of MLP.For statistical calculation,the variables were coded according to the following equation:X i=i− 0iwhere X i is the coded value of independent variable, i is the actual value of the independent variable, 0is the actual value of the inde-pendent variable at the central point,and i is the step change of the variable.Table1shows the range of independent variables and their levels.The whole design consisted of17experimental runs,including 12factorial points and5axial points.The5axial points were used to allow for estimation of a pure error sum of squares.The exper-iments were carried out in random order,and the experimental data(Table1)werefitted to the following second-order polynomial modeY=ˇ0+3i=1ˇi X i+3i=1ˇi i X2i+2i=13j=i+1ˇi j X i X jwhere Y,extraction yield of MLP,is the predicted response;ˇ0,ˇi,ˇii andˇij are the regression coefficients for intercept,linear, quadratic and interaction terms,respectively;X i and X j are the independent variables(i/=j).2.4.Purification of crude MLPCrude MLP was dissolved in deionized water and loaded onto a DEAE-52cellulose column(2.6×50cm).Then,the column was stepwise eluted with0,0.1,0.3and0.5M sodium chloride(NaCl) solution at aflow rate of60mL/h.Three completely separated frac-tions,MLP-1,MLP-2and MLP-3,were collected by checking the absorbance at490nm by using the phenol-sulphuric acid method (Dubois,Gilles,Hamilton,Rebers,&Smith,1956).As the main frac-tion,MLP-3was loaded onto a column(1.6×100cm)of Sephadex G-100and the column was eluted with0.2M NaCl solution at a flow rate of15mL/h.The elution was checked as described above. As a result,two purified fractions were collected,dialyzed and lyophilized,affording MLP-3a and MLP-3b,respectively.2.5.Characterization of MLP2.5.1.Determination of contents of carbohydrate,protein,uronic acid,sulfuric radical and total polyphenolsThe contents of carbohydrate in crude MLP and its purified fractions were determined by the phenol-sulphuric acid method (Dubois et al.,1956)using glucose as the standard.The content of protein was determined according to the reported method (Bradford,1976)using bovine serum albumin as the standard.The content of uronic acid was determined according to the method of Blumenkrantz and Asboe-Hansen(1973)using galacturonic acid as the standard.The content of sulfate radical was determined accord-ing to the reported method(Doigson&Price,1962).The content of total polyphenols was estimated by the Folin–Ciocalteu colori-metric method(Li,Nie,Xie,&Li,2014)using gallic acid(GA)as the standard,and it was expressed as mg GA equivalent(GAE)per 100mg dry sample.54Q.Yuan et al./Carbohydrate Polymers128(2015)52–62Table1Box–Behnken design matrix and the response values for the yield of MLP.Run Temperature(◦C)Time(h)Ratio of water to rawmaterial(mL/g)Polysaccharide yield(%) X1Code X1X2Code X2X3Code X319512−13008.529002−120−18.239003030010.14900303009.85951304019.3685−13020−1 6.879002−14018.78900414019.59900303009.91085−1413007.711900303009.8129004120−18.41385−12−1300 6.9149513020−17.5159********.71685−1304017.317951413008.92.5.2.Determination of homogeneity and molecular weightThe homogeneity and molecular weight of MLP-3a and MLP-3b were determined by high-performance gel permeation chromatog-raphy(HPGPC)using an Agilent1200series(Agilent Technologies, Santa Clara,CA,USA)apparatus equipped with a Shodex OH-pak SB-804HQ column(8×300mm).The column was eluted with 0.2M NaCl solution at aflow rate of0.5mL/min.The HPLC system was pre-calibrated with D-series dextran standards(D-2,D-3,D-4, D-5,D-6,D-7and D-8).2.5.3.Analysis of monosaccharide compositionThe monosaccharide composition of MLP was analyzed accord-ing to the reported method(Dai et al.,2010)with some modifications.The polysaccharide solution(100␮L,5mg/mL)was hydrolyzed with100␮L4M trifluoroacetic acid(TFA)at120◦C for 2h.After hydrolysis,the excess TFA was removed by the addition of methanol and evaporated at reduced pressure.The hydrolysate was dissolved with deionized water(100␮L)and mixed with100␮L 0.6M NaOH.Then,100␮L of the resulting mixture was labeled with PMP by adding100␮L of0.5M methanol solution of PMP and incu-bating at70◦C for100min.After being cooled to room temperature, the resultant solution was neutralized by adding50␮L of0.3M HCl and evaporated to dryness.The residue was dissolved in1.0mL deionized water,and the excess PMP was leached with chloroform for three times.The aqueous layer wasfiltered through a0.45␮m membrane and analyzed by an Agilent1100HPLC system equipped with photodiode array detector.The chromatographic conditions were as follows:column,Eclipse Plus C18(4.6×250mm,5␮m, Agilent);column temperature,30◦C;mobile phase,a mixture of phosphate buffered saline(PBS,0.1M,pH6.7)and acetonitrile in a ratio of83:17(v/v);flow rate,1.0mL/min;detector wave-length,245nm.The injection volume was20␮L.In similar manner, the monosaccharide standards were PMP-labeled and analyzed by HPLC.2.5.4.FT-IR spectrometric analysisThe dried crude MLP and its purified fractions were mixed with spectroscopic-grade potassium bromide powder,ground and pressed into pellets for FT-IR measurement.FT-IR spectra were recorded with a Nicolet6700FT-IR spectrometer(Thermo Fisher Scientific Inc.,Waltham,MA,USA)at the frequency range of 4000–400cm−1.2.6.Assay of antioxidant activity in vitro of MLP2.6.1.Assay of DPPH radical scavenging activityThe DPPH radical scavenging activity was measured by the method of previous report(Shimada,Fujikawa,Yahara,& Nakamura,1992)with slight modifications.Briefly,MLP was dis-solved in deionized water to afford a series of concentrations (0.0625,0.125,0.25,0.5,1.0,2.0,3.0and4.0mg/mL).Then,50␮L of sample solution,25␮L of DPPH-ethanol solution(0.4mM)and 100␮L of deionized water were mixed in a96-well plate.The mix-ture was kept at room temperature in the dark for30min,and the absorbance(Abs)at517nm was measured by a microplate reader (BioTek Instruments Inc.,Winooski,VT,USA).Ascorbic acid was used as positive control.DPPH radical scavenging activity was cal-culated by the following formula:DPPH radical scavenging activity(%)=1−(Abs1−Abs2)Abs0×100where Abs0is the Abs of the control(deionized water instead of sample),Abs1is the Abs of the sample,and Abs2is the Abs of the sample under identical conditions as Abs1with ethanol instead of DPPH-ethanol solution.2.6.2.Assay of hydroxyl radical scavenging activityThe hydroxyl radical scavenging activity of MLP was determined according to the reported method(Liu,Luo,Ye,&Zeng,2012). Briefly,the reaction mixture contained50␮L ferrosin(0.75mM), 75␮L phosphate buffer(0.15M,pH7.4),50␮L FeSO4(0.75mM), 50␮L sample solution and50␮L H2O2(0.01%,w/v).After incuba-tion at37◦C for30min,the Abs at536nm was measured.Ascorbic acid was used as positive control.The scavenging activity was cal-culated using the following equation:Hydroxyl radical scavenging activity(%)=[(Abs2−Abs0)−(Abs1−Abs0)]×100where Abs0is the Abs of the control(deionized water instead of sample),Abs1is the Abs of the deionized water instead of H2O2 and sample,and Abs2is the Abs of the sample.2.6.3.Assay of superoxide radical scavenging activityThe scavenging ability on superoxide radical was measured by the previously described method(Bi et al.,2013;Robak&Q.Yuan et al./Carbohydrate Polymers128(2015)52–6255 Gryglewski,1988)with minor modifications.The reaction mix-ture contained50␮L of sample solution,50␮L of NBT solution(156␮M),50␮L of NADH solution(156␮M)and50␮L of PMS solu-tion(60␮M).The mixture was incubated at25◦C for5min,and theAbs at560nm was determined.Ascorbic acid was used as positivecontrol.The superoxide radical scavenging activity was calculatedaccording to the following formula: Superoxide radical scavenging activity(%)=1−(Abs1−Abs2)Abs0×100where Abs0is the Abs of the control(deionized water instead of sample),Abs1is the Abs of the sample,and Abs2is the Abs of the sample under identical conditions as Abs1with0.1M phosphate buffer instead of NBT solution.2.6.4.Assay of ABTS radical scavenging activityThe ABTS radical scavenging activity of MLP was measured by ABTS radical cation decolorization assay(Re et al.,1999;Xie,Hu, Wang,&Zeng,2014).Briefly,ABTS solution(7mM)was oxidized with4.95mM of potassium persulphate for12h in the dark at room temperature.The ABTS+solution was then diluted with PBS(0.2M, pH7.4)to an Abs of0.70±0.02at734nm,and200␮L of the result-ing ABTS+solution was mixed with20␮L of sample solution.The mixture was kept for6min at room temperature,and the Abs at 734nm was then measured.The ABTS radical scavenging effect was calculated by the following equation:ABTS radical scavenging activity(%)=[1−(Abs1−Abs2)]Abs0×100where Abs0is the Abs of the control(deionized water instead of sample),Abs1is the Abs of the sample,and Abs2is the Abs of the sample only(PBS instead of ABTS+solution).2.6.5.Assay of Fe2+chelating activityThe Fe2+chelating ability of MLP was measured according to the reported method(Decker&Welch,1990).Briefly,50␮L sample solution was mixed with2.5␮L of ferrous chloride(FeCl2)solu-tion(5.0mM),10␮L of ferrozine solution(5mM)and137␮L of deionized water.The mixture was incubated for10min at room temperature,and then the Abs at562nm was measured.EDTA was used as positive control.The Fe2+chelating activity was calculatedby the following equation:Fe2+chelating activity(%)=1−(Abs1−Abs2)Abs0×100where Abs0is the Abs of the control(deionized water instead of sample),Abs1is the Abs of the sample,and Abs2is the Abs of the sample only(deionized water instead of FeCl2solution).2.7.Statistical analysisThe Design-Expert software version8.0.6was used for the experimental design and data analysis of RSM.The results of the antioxidant assay are reported as mean±SD of three replicates. The data were statistically analyzed by one-way analysis of vari-ance(ANOVA)procedure with SPSS software version19.0(Chicago, IL,USA),followed by the Duncan test.P-value of less than0.05was regarded as significant.3.Results and discussion3.1.Optimization of extraction parameters3.1.1.Predicted model and statistical analysisThe experimental design along with the extraction yields is shown in Table1,and the data were analyzed by Design-Expert Table2ANOVA for response surface quadratic model of MLP extraction.Source Sum of squares df Mean square F-value P-valueModel18.929 2.1082.47<0.0001a X1 3.661 3.66143.49<0.0001a X20.6010.6023.510.0019a X3 1.921 1.9275.34<0.0001a X1X20.06810.068 2.650.1475b X1X30.4710.4718.400.0036a X2X30.08110.081 3.190.1175b X128.1918.19321.37<0.0001a X220.8010.8031.250.0008a X32 2.171 2.1785.02<0.0001a Residual0.1870.025Lack offit0.09630.032 1.550.3315b Pure error0.08240.021Cor total19.1016R2=0.9907,adjusted R2=0.9786.a5%significance level.b Not significant relative to the pure error.software.As a result,a second-order polynomial equation describ-ing the correlation between the MLP yield and the test variables was obtained by the following equation:Y=−454.60135+10.05870X1+4.80828X2−0.17728X3−0.02612X1X2+0.00683X1X3+0.01416X2X3−0.05583X21−0.43506X22−0.00718X23where Y represents the extraction yield of MLP,X1,X2and X3rep-resent extraction temperature(◦C),extraction time(h)and ratio (v/w,mL/g)of water to raw material,respectively.The ANOVA,lack-of-fit and the adequacy of the model are indi-cated in Table2.The model F-value of82.47implied that the model was significant,indicating that there was only a0.01%chance that the model F-value could occur due to noise.The determination coefficient(R2)and the adjusted determination coefficient(adj-R2) were0.9907and0.9786,respectively,which showed a good agree-ment between the experimental and the predicted values of the MLP yield with goodness-of-fit of the regression equation.The P-value is used as a tool to check the significance of each coefficient,which in turn may indicate the pattern of the interac-tions between the variables.The smaller the P-value is,the more significant the corresponding coefficient is.Table2shows that the linear coefficients(X1,X2and X3),quadratic term coefficients(X12, X22and X32)and cross product coefficients(X1X3)were significant on extraction yield of MLP due to P-value<0.05.The results also suggested that,among the independent variables,the extraction temperature was the most significant parameter affecting the yield of MLP followed by ratio of water to raw material and extraction time.3.1.2.Response surface plot and contour plotThe3D response surface and2D contour plots are provided as the graphical representations of a regression equation.They show the type of interactions between two tested variables and the rela-tionship between responses and experiment levels of each variable. In the present study,the response surface and contour plots(Fig.1) were obtained by using Design-Expert.Fig.1A and B shows the effects of extraction temperature(X1),extraction time(X2)and their reciprocal interaction on extraction yield when the ratio of water to raw material(X3)wasfixed at level0.The X1and X2 demonstrated quadratic effects on the extraction yield.The yield of MLP increased to the maximum value with increase of extrac-tion temperature but it then decreased slightly with the further increase of extraction temperature.The extraction time showed a similar effect on the yield of MLP.Generally,higher extraction56Q.Yuan et al./Carbohydrate Polymers 128(2015)52–62Fig.1.Response surface plots (A,C and E)and contour plots (B,D and F)showing the effects of variables (X 1,extraction temperature;X 2,extraction time;X 3,ratio of water to material)and their mutual effects on the extraction yield of MLP.temperature and longer extraction time can promote the disso-lution of polysaccharides from plant tissues.However,too high extraction temperature and too long extraction time may result in degradation of polysaccharides and hence decrease the polysac-charides yield (Samavati &Yarmand,2013).Fig.1C and D shows the quadratic effects of X 1and X 3on the extraction yield when X 2was fixed at level 0.The elliptical contour plot as shown in Fig.1Dindicated that the mutual interactions between the extraction tem-perature and ratio of water to raw material were significant.The extraction yield of MLP increased with the increases of extrac-tion temperature and ratio of water to raw material from 85to 93.27◦C and 20to 39.56mL/g,but it did not further increase when extraction temperature and ratio of water to raw material were over 93.27◦C and 39.56mL/g.The result is in agreement with thatQ.Yuan et al./Carbohydrate Polymers128(2015)52–6257of previous reported(Chen,Li,Li,Jin,&Lu,2015).Fig.1E and F shows the effects of X2and X3and their reciprocal interaction on the extraction yield when X1wasfixed at level0.The maxi-mum yield of MLP could be achieved when extraction time and ratio of water to raw material were in the range of2.78–3.95h and29.27–38.36mL/g,respectively.However,the extraction yield of MLP decreased when extraction time and ratio of water to raw material were over3.95h and38.36mL/g.This phenomenon could be explained that the polysaccharides will be completely dissolved in the extraction solvent and partial of polysaccharides may be hydrolyzed due to longer extraction time.The result is in agreement with that reported by Qiao et al.(2009).3.1.3.Optimization of extracting parameters and validation ofthe modelBy employing the Design-Expert software,the optimal condi-tions for MLP extraction were extraction temperature91.52◦C, extracting time3.53h,and ratio(v/w,mL/g)of water to raw mate-rial34.09.In the optimal conditions,the maximum predicted yield of MLP was10.1%.For operation convenience,the optimal param-eters were determined as follows:extraction temperature92◦C, extracting time3.5h,and ratio(v/w,mL/g)of water to raw mate-rial34.To ensure that the predicted result was not biased toward the practical value,experimental rechecking was performed using the deduced optimal conditions.A mean value of10.0±0.5%(n=3) was obtained from actual experiments,which demonstrated the validation of the RSM model and indicating that the model was ade-quate for the extraction of MLP.The optimal extraction conditions with high yield were more efficient than that reported(Samavati& Yarmand,2013).3.2.Purification and characterization of MLPCrude MLP was prepared using the optimal extraction condi-tions.Then,the crude MLP was separated by a DEAE-52cellulose chromatography column,affording three independent elution peaks of MLP-1,MLP-2and MLP-3(Fig.2A).The third fraction of MLP-3,the main fraction of crude MLP,was further purified with a Sephadex G-100gel permeation chromatography column,result-ing in two fractions of MLP-3a and MLP-3b(Fig.2B).3.2.1.Contents of carbohydrate,protein,uronic acid,sulfuricradical and total polyphenols in MLPTable3shows the contents of carbohydrate,protein,uronic acid, sulfuric radical and total polyphenols in crude MLP,MLP-3a and MLP-3b.The carbohydrate contents in crude MLP,MLP-3a and MLP-3b were52.09%,89.74%and37.20%,respectively.The contents of protein in crude MLP,MLP-3a and MLP-3b were2.16%,0.83%and 0.22%,respectively.Among all the polysaccharides tested,MLP-3b Fig.2.Stepwise elution curve of crude MLP on DEAE-cellulose column(A)and elution curve of polysaccharides fraction(MLP-3)on Sephadex G-100column(B).contained the highest contents of uronic acid.The contents of total polyphenols for MLP-3a and MLP-3b(0.17and0.16mg GAE/100mg,respectively)were much lower than that(1.93mg GAE/100mg)for crude MLP,indicating that most of the polyphenols in crude MLP was removed during the purification.3.2.2.Molecular weight of MLPThe homogeneity and molecular weights of MLP-3a and MLP-3b were analyzed by HPLC.Both MLP-3a and MLP-3b gave a single and symmetrical peak,indicating that they were homogeneous.Table3Preliminary characterization of crude MLP,MLP-3a and MLP-3b.Item Crude MLP MLP-3a MLP-3bCarbohydrate(%)52.09±1.1089.74±0.3137.20±0.59 Protein(%) 2.16±0.020.83±0.030.22±0.02 Uronic acid(%)32.45±0.92 6.53±0.1365.29±1.53 Sulfuric radical(%) 1.73±0.10 1.40±0.02 1.22±0.02 Total polyphenols(mg GAE/100mg) 1.93±0.020.17±0.010.16±0.01 Molecular weight(kDa)–80.99 3.64 Monosaccharide composition(mol)Man0.510.77Rha 5.13 4.53 1.57GlcA 2.230.810.20GalA 3.02 1.21 6.10Glc 2.13 3.47Gal 2.9512.55 1.27Ara 2.5511.140.89–:Not detected.58Q.Yuan et al./Carbohydrate Polymers128(2015)52–62Fig.3.FT-IR spectra of crude MLP(A),MLP-3a(B)and MLP-3b.The average molecular weights of MLP-3a and MLP-3b were estimated to be80.99kDa and3.64kDa(Table3),respectively, according to the calibration curve.The molecular weight of MLP-3b was lower than those of polysaccharides from mulberry leaves as reported,8kDa(Zhang et al.,2014b),15kDa(Xia et al.,2008Xia, Liu,Zhang,&Luo,2008),24.898and61.131kDa(Ying,Han,&Li, 2011),55.7062kDa(Jia,Zhang,Lan,Yang,&Sun,2013)and550kDa (Katayama,Takano,&Sugimura,2008),which indicating that MLP-3b was a novel polysaccharide from mulberry leaves.3.2.3.Monosaccharide composition of MLPThe results of monosaccharide compositions of crude MLP and its purified fractions(MLP-3a and MLP-3b)determined by HPLC are shown in Table3.It was found that both crude MLP and MLP-3a were composed of Man,Rha,GlcA,GalA,Glc,Gal,and Ara,but the molar ration was quite different.The molar ratio of Man,Rha,GlcA, GalA,Glc,Gal,and Ara for crude MLP was0.51:5.13:2.23:3.02: 2.13:2.95:2.55,while it was0.77:4.53:0.81:1.21:3.47:12.55: 11.14for MLP-3a.For MLP-3b,the molar ratio of Rha,GlcA,GalA,Gal and Ara was1.57:0.20:6.10:1.27:0.89,and Man and Glc were not detected.These results indicated that the crude MLP and its purified fractions were acidic polysaccharide.In addition,the monosaccha-ride composition and molar ratio of MLP were different from those of the earlier reports for mulberry leaf polysaccharides(Katayama et al.,2008;Xia et al.,2008).As reported,a pectic polysaccharide from mulberry leaves was comprised of Rha,Ara,Xyl,Glc,Gal and GalA in a molar ratio of5:4:1:2:6:38(Xia et al.,2008).Katayama et al.(2008)separated and characterized a main polysaccharide from mulberry leaves,which was composed of Rha,Gal,Glc,GalA and GlcA in a molar ratio of1:0.2:0.5:2.3:1.5.The difference might be related to the extraction,purification and the raw mate-rial(mulberry leaf).It has been reported that different results for compositional components of tea polysaccharides were given in various reports due to different tea raw material or purification process(Nie&Xie,2011).3.2.4.FT-IR spectrumThe FT-IR spectra of crude MLP and its purified fractions are shown in Fig.3.Broad and strong absorption bands around 3400cm−1for C H stretching vibrations and2939cm−1for C H stretching vibrations were observed in the FT-IR spectra of crude MLP and its purified fractions.The absorption peaks at1604.48cm−1and1421.3cm−1for crude MLP,1734.8cm−1 and1641.1cm−1for MLP-3a,1738.6cm−1and1610.6cm−1for MLP-3a were attributed to the stretching vibrations of ester car-bonyl groups(C O)and carboxylic groups(COO)(Li et al.,2013;Santhiya,Subramanian,&Natarajan,2002;Zhao,Yang,Yang,Jiang, &Zhang,2007),respectively,which indicated that crude MLP, MLP-3a and MLP-3b were acidic polysaccharides.The results are consistent with the analytical results of monosaccharide composi-tions for crude MLP,MLP-3a and MLP-3b as mentioned above.The absorbance peaks at1414.91cm−1for MLP-3a and1415.46cm−1 for MLP-3b were associated with the stretching of the pectin methyl ester group(OCH3),which suggested that some of the uronic acids in polysaccharides were esterified(Al-Sheraji et al.,2012).A char-acteristic peak at around894cm−1was found in crude MLP,MLP-3a and MLP-3b,indicating the existence of␤-glycosidic bonds in the three polysaccharides(Coimbra,Gonc¸alves,Barros,&Delgadillo, 2002;Yang et al.,2006).3.3.Antioxidant activity in vitro of MLP3.3.1.DPPH radical scavenging activityDPPH free radical is a stable free radical with a characteristic absorption at517nm,which will decrease significantly on expo-sure to proton-donating substance(Yamaguchi,Takamura,Matoba, &Terao,1998).Accordingly,it has been widely used to evaluate the antioxidant activity of natural antioxidants.In the present study, the scavenging ability of MLP on DPPH free radicals was examined and the results are shown in Fig.4A.The scavenging rates of MLP and Vc increased with the increase of sample concentration.At a concentration of4.0mg/mL,the DPPH free radical scavenging activ-ities for crude MLP,MLP-3a,MLP-3b and Vc were68.21%,44.96%, 60.17%and91.16%,pared with purified fractions, the crude MLP exhibited relative higher(P<0.05)DPPH radical scavenging activity,which might be partly attributed to its relative higher content of polyphenols(Table3).The scavenging abilities of MLP-3a and MLP-3b might mainly come from the polysaccha-rides due to their relatively low contents of total polyphenols.It has been reported that the antioxidant activity of polysaccharides was related to their molecular weight,uronic acid content,degree of sulfation,type of monosaccharide and glycosidic linkage(Liu et al., 2010;Melo,Feitosab,Freitasa,&de Paula,2002;Sun,Wang,Li,& Liu,2014;Wang,Chang,Stephen Inbaraj,&Chen,2010a;Zeng et al., 2014).MLP-3b displayed higher antioxidant activity than MLP-3a, which might be due to the lower molecular weight and higher con-tent of uronic acid of MLP-3b(Asker,Mahmoud,&Ibrahim,2007; Chen,Zhang,Qu,&Xie,2008).But the exact mechanism is unclear.3.3.2.Hydroxyl radical scavenging activityHydroxyl radical,which is well known as one of the most reac-tive free radical,can react with almost the biomacromolecules in living cells and induce severe damage to the adjacent biomolecules (Spencer et al.,1994).Therefore,the hydroxyl radical scavenging activities of MLP and its purified fractions were investigated.As shown in Fig.4B,all MLP samples and Vc exhibited scavenging activity on hydroxyl radicals in a dose-dependent manner.At a concentration of4.0mg/mL,the scavenging abilities of crude MLP, MLP-3a,MLP-3b and Vc were88.85%,57.89%,68.12%and92.44%, respectively.Notably,crude MLP showed similar(P>0.05)scaveng-ing activity as Vc.The results indicated that the crude MLP exhibited strong scavenging activity and its purified fractions had moderate scavenging activities.It has been reported that the hydroxyl scav-enging ability is related to the number of hydroxyl or amino groups in polysaccharide(Guo et al.,2005).This might be explained by the factor that crude MLP would provide more active hydroxyl groups than its purified fractions as crude MLP contained much higher content of polyphenols.3.3.3.Superoxide radical scavenging activitySuperoxide radical is a long lifetime radical that is generated by numerous biological and photochemical reactions(Banerjee,。