(Morella Rubra Sieb. Et Zucc.) Fruit Extracts and -Glucosidase Inhibitory

molecules

Article Puriﬁcation of Flavonoids from Chinese Bayberry (Morella rubra Sieb. et Zucc.) Fruit Extracts and α-Glucosidase Inhibitory Activities of Different Fractionations

Shuxia Yan 1,2, Xianan Zhang 1,2, Xin Wen 1,2, Qiang Lv 1,2, Changjie Xu 1,2, Chongde Sun 1,2 and Xian Li 1,2,*

1 Zhejiang Provincial Key Laboratory of Horticultural Plant Integrative Biology, Zhejiang University, Hangzhou 310058, China; [email protected] (S.Y.); [email protected] (X.Z.); [email protected] (X.W.); [email protected] (Q.L.); [email protected] (C.X.); [email protected] (C.S.) 2 Laboratory of Fruit Quality Biology, the State Agriculture Ministry Laboratory of Horticultural Plant Growth, Development and Quality Improvement, Zhejiang University, Hangzhou 310058, China * Correspondence: [email protected]; Tel.: +86-571-8898-2630; Fax: +86-571-8898-2224

Academic Editor: Nancy D. Turner Received: 6 July 2016; Accepted: 26 August 2016; Published: 31 August 2016

Abstract: Chinese bayberry (Morella rubra Sieb. et Zucc.) fruit have a diverse flavonoid composition responsible for the various medicinal activities, including anti-diabetes. In the present study, efficient simultaneous purification of four flavonoid glycosides, i.e., cyanidin-3-O-glucoside (1), myricetin-3-O-rhamnoside (2), quercetin-3-O-galactoside (3), quercetin-3-O-rhamnoside (4), from Chinese bayberry pulp was established by the combination of solid phase extract (SPE) by C18 Sep-Pak® cartridge column chromatography and semi-preparative HPLC (Prep-HPLC), which was followed by HPLC and LC-MS identification. The purified flavonoid glycosides, as well as different fractions of fruit extracts of six bayberry cultivars, were investigated for α-glucosidase inhibitory activities. The flavonol extracts (50% methanol elution fraction) of six cultivars showed strong α-glucosidase inhibitory activities (IC50 = 15.4–69.5 µg/mL), which were higher than that of positive control acarbose (IC50 = 383.2 µg/mL). Four purified compounds 1–4 exerted α-glucosidase inhibitory activities, with IC50 values of 1444.3 µg/mL, 418.8 µg/mL, 556.4 µg/mL, and 491.8 µg/mL, respectively. Such results may provide important evidence for the potential anti-diabetic activity of different cultivars of Chinese bayberry fruit and the possible bioactive compounds involved.

Keywords: Morella rubra fruit; α-glucosidase inhibitory activity; ﬂavonoid glycosides; puriﬁcation

1. Introduction Diabetes is an endocrine disorder characterized by elevated blood glucose levels resulting from the defects in insulin-secretory response or resistance to insulin action [1]. One of the therapeutic methods for reducing postprandial hyperglycemia is to delay the absorption of carbohydrates by inhibiting the α-glucosidase activity [2,3]. So far, acarbose, voglibose, and miglitol are frequently used drugs to control hyperglycemia after meals. However, side effects of these compounds, such as ﬂatulence, stomach ache, and diarrhea, are also apparent [4–6]. Natural plant products may provide good alternatives for new anti-diabetic compounds with high safety. Chinese bayberry (Morella rubra Sieb. et Zucc.) is an important subtropical fruit tree in the Myricaceae family and the fruit contains high nutritional and medicinal value [7–9]. Many bioactive compounds, such as cyanidin-3-O-glucoside (C3G) (1), myricetin-3-O-rhamnoside

Molecules 2016, 21, 1148; doi:10.3390/molecules21091148 www.mdpi.com/journal/molecules Molecules 2016, 21, 1148 2 of 12

Molecules 2016, 21, 1148 2 of 11 (M3R) (2), quercetin-3-O-galactoside (Q3Gal) (3), quercetin-3-O-rhamnoside (Q3R) (4), and asquercetin-3- cyanidin-3-O-glucosideO-glucoside (Q3G) (C3G) have (1), been myricetin-3- identifiedO-rhamnoside in bayberry [(M3R)10]. Previously, (2), quercetin-3- our resultsO-galactoside showed (Q3Gal)that the ( C3G-rich3), quercetin-3- bayberryO-rhamnoside fruit extract (Q3R) from (4), the and ‘Biqi’ quercetin-3- (BQ) cultivarO-glucoside significantly (Q3G) have decreased been identified blood inglucose bayberry levels [10]. in Previously, streptozotocin our results (STZ)-induced showed that diabetic the C3G-rich mice bayberry [11]. Recently, fruit extract bayberry from fruitthe ‘Biqi’ extracts (BQ) cultivarfrom nine significantly cultivars weredecreased evaluated blood forglucose the glucose levels in consumption streptozotocin activities (STZ)-induced in HepG2 diabetic cells. mice Results [11]. Recently,showed thatbayberry bayberry fruit extracts cultivars from contained nine cultivars higher were contents evaluated of C3G for the and glucose three quercetinconsumption glycosides activities in HepG2 cells. Results showed that bayberry cultivars contained higher contents of C3G and three (Q3G, Q3Gal, and Q3R) showed higher glucose consumption activities than those cultivars with lower quercetin glycosides (Q3G, Q3Gal, and Q3R) showed higher glucose consumption activities than those contents of C3G and quercetin glycosides [12]. However, further characterization and comprehensive cultivars with lower contents of C3G and quercetin glycosides [12]. However, further characterization and investigation of these important bioactive compounds from bayberry were limited by the availability comprehensive investigation of these important bioactive compounds from bayberry were limited by the of pure compounds and relative high quantities required by further studies. So far, few efficient availability of pure compounds and relative high quantities required by further studies. So far, few efficient separation methods have been established for simultaneous purification of these compounds with separation methods have been established for simultaneous purification of these compounds with high recoverieshigh recoveries and high and yields. high yields. TheThe present present study study was was designed designed to develop to develop an effi ancient efficient isolation isolation method method to purify to C3G purify and C3Gthree flavonolsand three (M3R, flavonols Q3Gal, (M3R, and Q3R) Q3Gal, (Figure and 1) Q3R) from (Figure BQ pulp1) at from the same BQ pulp time, at using the samea combination time, using of solid a ® phasecombination extraction of solid(SPE) phase by C18 extraction Sep-Pak® (SPE) cartridges by C18 column Sep-Pak chromatographycartridges column and semi-preparative chromatography HPLC and (Prep-HPLC).semi-preparative α-Glucosidase HPLC (Prep-HPLC). inhibitoryα -Glucosidaseactivities and inhibitoryantioxidant activities activities and of antioxidant the purified activities flavonoid of glycosides,the purified fractions, flavonoid and glycosides, extracts of fractions, six bayberry and cultivars extracts were of six also bayberry evaluated. cultivars were also evaluated.

FigureFigure 1. Chemical structures of fourfour mainmain ﬂavonoidflavonoid glycosidesglycosides fromfrom ChineseChinese bayberrybayberry pulppulp extract: extract: cyanidin-3-cyanidin-3-OO-glucoside-glucoside (C3G) (C3G) (1); ( 1myricetin-3-); myricetin-3-O-rhamnosideO-rhamnoside (M3R) (M3R) (2); quercetin-3- (2); quercetin-3-O-galactosideO-galactoside (Q3Gal) ((Q3Gal)3); and quercetin-3- (3); and quercetin-3-O-rhamnosideO-rhamnoside (Q3R) (4). (Q3R) (4).

2. Results and Discussion 2. Results and Discussion 2.1. Purification of Four Flavonoid Glycosides from BQ Pulp 2.1. Purification of Four Flavonoid Glycosides from BQ Pulp A simple and efficient method by the combination of column chromatography and Prep-HPLC was A simple and efficient method by the combination of column chromatography and Prep-HPLC established for rapid isolation and purification of four flavonoid glycosides from BQ pulp. was established for rapid isolation and purification of four flavonoid glycosides from BQ pulp. Based on our preliminary experiment, the SPE column showed good isolation between C3G and three Based on our preliminary experiment, the SPE column showed good isolation between C3G and flavonols (M3R, Q3Gal, and Q3R). The dynamic adsorption and desorption tests of C3G were first carried outthree due flavonols to their (M3R,high quantity Q3Gal, and(Figure Q3R). 2). The dynamic leakage adsorption curves and of desorption C3G on the tests SPE of column C3G were were obtainedfirst carried based out on due the to volume their high of effluent quantity and (Figure the initial2). The concentration dynamic leakage of the curvessample of solution. C3G on The the initial SPE feedcolumn concentration were obtained of C3G based on on SPE the column volume was of effluent 1.18 mg/mL. and the Dynamic initial concentration adsorption and of the desorption sample experimentssolution. The showed initial feedthat the concentration breakthrough of C3Gvolume on SPEof C3G column on SPE was column 1.18 mg/mL. was about Dynamic 4 bed volume adsorption (BV) (Figureand desorption 2A), and experiments the majority showedof C3G thatabsorbed the breakthrough by SPE column volume was eluted of C3G by on 10% SPE methanol column was(with about 0.5% formic4 bedvolume acid) solution (BV) (Figure (Figure2A), 2B). and Therefore, the majority 10% ofme C3Gthanol absorbed aqueous by solution SPE column was used was elutedas the eluent by 10% in isocraticmethanol elution, (with 0.5%where formic complete acid) desorption solution (Figure was done2B). Therefore,with an elution 10% methanolvolume of aqueous about 120 solution mL (10 was BV) (Figureused as 2C). the eluentThe eluents in isocratic were elution,collected where as 10% complete fractions, desorption the SPE wascolumn done was with then an elutionwashed volume with 50% of methanolabout 120 aqueous mL (10 BV)solution, (Figure and2 C).the Theeluents eluents were werecollected collected as a 50% as 10%fraction. fractions, the SPE column was thenThe washed 10% withfraction 50% was methanol rich in aqueousC3G (peak solution, 1) and andthe 50% the eluentsfraction were mainly collected contained as a three 50% fraction.flavonols, namely,The M3R 10% (peak fraction 2), Q3Gal was rich (peak in C3G 3), and (peak Q3R 1) (peak and the4) (Figure 50% fraction 3). Four mainly flavonoid contained glycosides three were flavonols, further separatednamely, M3R and (peakpurified 2), Q3Galby two (peak different 3), and runs Q3R of (peakPrep-HPLC. 4) (Figure The3). flow Four diagram flavonoid of glycosides preparation were and purificationfurther separated of four and flavonoid purified glycosides by two differentfrom BQ runspulp ofwas Prep-HPLC. shown in Figure The flow 3. diagram of preparation and purification of four flavonoid glycosides from BQ pulp was shown in Figure3.

Molecules 2016, 21, 1148 3 of 12 Molecules 2016, 21, 1148 3 of 11

Molecules 2016, 21 , 11481.2 A 36 B 3 of 11 1.0 30 0.8 24 1.2 A 36 B

0.61.0 3018 C3G (mg) (mg/mL) 0.40.8 2412

0.20.6 186 Concentration of C3G of Concentration C3G (mg) 0.0 (mg/mL) 0.4 120

0.2 012345678 6 0 102030405060 Concentration of C3G of Concentration Feed volume (BV) Concentration of methanol (%) 0.0 0

012345678 0 102030405060 Feed volume35 (CBV) Concentration of methanol (%) 28 35 C 21

28 14 C3G (mg) 21

7 14 C3G (mg) 07 0 20406080100120 0 Elution volume (mL) 0 20406080100120 Elution volume (mL) FigureFigure 2. Dynamic 2. Dynamic leakage leakage curve curve (A); (A gradient); gradient elution elution curve curve (B ();B and); and isocratic isocratic de desorptionsorption curve curve (C (C) of) of C3G ® on FigureaC3G Sep-Pak on 2. aDynamic Sep-Pak. leakage®. curve (A); gradient elution curve (B); and isocratic desorption curve (C) of C3G on a Sep-Pak®.

FigureFigure 3. HPLC 3. chromatogramHPLC chromatogram of SPE column of SPEand semi-prepa column andrative semi-preparative HPLC (Prep-HPLC) HPLC purification (Prep-HPLC) procedure. Figure 3. HPLC chromatogram of SPE column and semi-preparative HPLC (Prep-HPLC) purification procedure. puriﬁcation procedure.

Molecules 2016, 21, 1148 4 of 12

Molecules 2016, 21, 1148 4 of 11 The isolated compounds were further identified according to their retention time and UV absorption characteristics by HPLC-DAD analysis (Figure3). Further structure identification of The isolated compounds were further identified according to their retention time and UV absorption the purified compounds was confirmed by LC-MS2 (Figure4). For peak 1, the ion at m/z 449.2 characteristics by HPLC-DAD analysis (Figure 3). Further structure identification of the purified [M + H]+ suggested the molecular weight of C3G, and the ion at m/z 287.1 [M − 162]+ represented the compounds was confirmed by LC-MS2 (Figure 4). For peak 1, the ion at m/z 449.2 [M + H]+ suggested the glycoside cleavage of one hexoside residue from cyanidin aglycone. For peak 2, the ion at m/z 463.0 molecular weight of C3G, and the ion at m/z 287.1 [M − 162]+ represented the glycoside cleavage of one [M − H]− suggested the molecular weight of M3R, and the ion at m/z 316.0 [M − 146 − 2H]−, 317.0 hexoside residue from cyanidin aglycone. For peak 2, the ion at m/z 463.0 [M − H]− suggested the molecular − − − weight[M of146 M3R,H] andrepresented the ion at m/z the glycoside316.0 [M − cleavage 146 − 2H] of−, one 317.0 hexoside [M − 146 residue − H]− fromrepresented myricetin the aglycone. glycoside − cleavageFor peak of 3one and hexoside peak 4, theresidue ion atfromm/ zmyricetin463.0 [M aglycone.− H] suggested For peak the3 and molecular peak 4, the weight ion at of m/z Q3Gal, 463.0 the[M − − H]ion− suggested at m/z 447.1 the molecular [M − H] weightsuggested of Q3Gal, the molecularthe ion at m/z weight 447.1 of [M Q3R, − H] and− suggested the ion atthem molecular/z 300.0 [M weight− 162 of − − Q3R,− 2H] and ,the 301.0 ion [Mat m/z− 162 300.0− [MH] − 162resulted − 2H]− from, 301.0 the [M cleavage − 162 − H] of− aresulted hexoside from or the deoxyhexoside cleavage of a residuehexoside orfrom deoxyhexoside quercetin aglycone. residue from Compared quercetin with aglycone. the previous Comp reportsared with [13 the], the previous purified reports compounds [13], the could purified be 2 compoundsconfirmed could subsequently be confirmed by our subsequently LC-MS data. by our LC-MS2 data.

FigureFigure 4.4. LC-MSLC-MS22 spectrum of of four four purified puriﬁed products. products.

Molecules 2016, 21, 1148 5 of 12

The purities, recoveries, and yields of four flavonoid glycosides in different procedures were summarized in Table1. The purities of C3G, M3R, Q3Gal, and Q3R in the crude extract of BQ pulp were as low as 0.74%, 0.02%, 0.06%, and 0.03%, respectively. According to the HPLC chromatograms, the crude extract of bayberry fruit was separated into two fractions (Figure3). After one-step SPE column purification, the purity of C3G, M3R, Q3Gal, and Q3R increased to 54.12%, 2.42%, 7.75%, and 3.84%, respectively, which were about 100-fold enriched than those of the crude extract. The recoveries were all more than 70% (Table1). Further Prep-HPLC purification resulted in 102.4 mg C3G with 99.17% purity and the recovery rate was 78.18% from 240 mg 10% fraction, and also 4.6 mg M3R with 90.38% purity and the recovery rate was 71.58%, 15.6 mg Q3Gal with 94.57% purity and the recovery rate was 79.32%, 7.6 mg Q3R with 93.61% purity and the recovery rate was 77.20% from 240 mg 50% fraction (Table1). Such purification procedure was proved as time saving and high efficiency with high yield. The purified compounds can be used for further pharmaceutical research.

Table 1. Purities and recoveries of four ﬂavonoid glycosides in the two-step puriﬁcation procedure.

Compounds Puriﬁcation Step Purity (%) Recovery (%) Yield (mg)

1 Crude extract 0.74 / / C3G 2 SPE column 54.12 90.75 620.4 1 3 Prep-HPLC 99.17 78.18 102.4 2

1 Crude extract 0.02 / / M3R 2 SPE column 2.42 72.24 298.5 1 3 Prep-HPLC 90.38 71.58 4.6 3

1 Crude extract 0.06 / / Q3Gal 2 SPE column 7.75 77.11 298.5 1 3 Prep-HPLC 94.57 79.32 15.6 3

1 Crude extract 0.03 / / Q3R 2 SPE column 3.84 76.42 298.5 1 3 Prep-HPLC 93.61 77.20 7.6 3 1 The amount of sample was obtained from 50 g freeze dried raw material by the SPE column; 2 the amount of compound was obtained from 240 mg SPE column sample (10% fraction) by Pre-HPLC; 3 the amount of compound was obtained from 240 mg SPE column sample (50% fraction) by Pre-HPLC.

2.2. Antioxidant Activities Assay The antioxidant activity of crude extract, fractions and purified compounds from BQ pulp were evaluated by using both 2.2-diphenyl-1-picrylhydrazyl (DPPH) and 2.20-azinobis (3-ethylbenzothiazoline -6-sulfonic acid) diammonium salt (ABTS) free radical scavenging assays. Apparent dose-dependent antioxidant activities were observed for all the samples with serial concentrations tested (Figure5). The IC 50 value of DPPH scavenging activity of the crude extract was 148.11 µg/mL, while the IC50 values for the 10% fraction and 50% fraction were 8.17 µg/mL and 3.45 µg/mL, respectively, for the same assay. For the four purified compounds, higher DPPH scavenging activities were found than those of the crude extract or 10% or 50% fractions, since the IC50 values for C3G, M3R, Q3Gal, and Q3R were 3.22, 2.81, 2.43, and 2.55 µg/mL, respectively (Figure5A). The data from the ABTS scavenging assay showed similar results with the DPPH assay (Figure5B). Therefore, these four flavonoid glycosides may all contribute to the antioxidant activity of the BQ pulp crude extract. Antioxidant activity has been proposed by many studies as a key mechanism for the prevention of hyperglycemia associated with diabetes. Recently, our results showed that the extract of BQ fruit could attenuate oxidative stress-induced pancreatic β cell necrosis and protected the insulin secretion ability of the surviving β cell in diabetic mice, which was possibly due to its high antioxidant activity [11,14]. In addition, the glucose consumption activities of the bayberry fruit from different cultivars were significantly correlated to their antioxidant activities [12]. Therefore, the four flavonoid glycosides with high antioxidant activities may play important role to modulate glucose metabolism in vivo. Molecules 2016, 21, 1148 6 of 12 Molecules 2016, 21, 1148 6 of 11

Molecules 2016, 21, 1148 6 of 11 A IC (μg/mL) 120 50 148.11±19.66 8.17±1.25 3.45±0.40 3.22±0.23 2.81±0.02 2.43±0.10 2.55±0.12 A IC (μg/mL) 90 50

120 148.11±19.66 8.17±1.25 3.45±0.40 3.22±0.23 2.81±0.02 2.43±0.10 2.55±0.12 60 90

30 60 0 30 0 5 0 0 5 0 μ .0 20 0 0 0 .0 50 0 .0 .00 .00 .00 .25 .5 .00 .2 .5 .0 .2 .5 .00 g/mL 60 1 24 2.5 5.0 10 2. 5.0 10 2 4 8 1 2 5 1 2 5 1 2 5

DPPH scavenging activity (%) activity scavenging DPPH 0 Crude extract 10% fraction 50% fraction C3G M3R Q3Gal Q3R 0 5 0 0 5 0 μ .0 20 0 0 0 .0 50 0 .0 .00 .00 .00 .25 .5 .00 .2 .5 .0 .2 .5 .00 g/mL B 60 1 24 2.5 5.0 10 2. 5.0 10 2 μ 4 8 1 2 5 1 2 5 1 2 5

DPPH scavenging activity (%) activity scavenging DPPH IC ( g/mL) 50 120 Crude extract 10% fraction 50% fraction C3G M3R Q3Gal Q3R 119.87±6.12 6.22±0.52 3.52±0.44 4.19±0.16 3.24±0.02 2.09±0.01 2.28±0.05 B IC (μg/mL) 12090 50

119.87±6.12 6.22±0.52 3.52±0.44 4.19±0.16 3.24±0.02 2.09±0.01 2.28±0.05 6090

3060

300 μ .0 0 0 3 5 .5 3 5 .5 0 00 .0 56 3 25 .94 88 75 94 88 5 g/mL 75 15 30 3.1 6.2 12 3.1 6.2 12 2.5 5. 10 1. 3.1 6. 0 1. 3. 0. 1. 3.7 ABTS scavenging activity (%) activity scavenging ABTS 0 Crude extract 10% fraction 50% fraction C3G M3R Q3Gal Q3R μ .0 0 0 3 5 .5 3 5 .5 0 00 .0 56 3 25 .94 88 75 94 88 5 g/mL 75 15 30 3.1 6.2 12 3.1 6.2 12 2.5 5. 10 1. 3.1 6. 0 1. 3. 0. 1. 3.7 ABTS scavenging activity (%) activity scavenging ABTS FigureFigure 5. 5.2.2-diphenyl-1-picrylhydrazyl2.2-diphenyl-1-picrylhydrazylCrude extract 10% fraction 50% (DPPH) fraction (DPPH) radical radicalC3G scavenging M3R activityQ3Galactivity (A(Q3RA) and) and 2.2 0-azinobis2.2 ′-azinobis (3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS) radical scavenging activity (B) of crude Figure(3-ethylbenzothiazoline-6-sulfonic 5. 2.2-diphenyl-1-picrylhydrazyl acid) (DPPH) diammonium radical salt scavenging (ABTS) radical activity scavenging (A) and activity 2.2′-azinobis (B) of extract, fractions, and isolated and four purified compounds from BQ pulp. Data are presented as the (3-ethylbenzothiazoline-6-sulfoniccrude extract, fractions, and isolated acid) anddiammonium four puriﬁed salt compounds (ABTS) radical from scavenging BQ pulp. Dataactivity are ( presentedB) of crude mean ± SD (n = 3)± and IC50 was calculated for each sample for both assays. extract,as the fractions, mean SDand (n isolated= 3) and an ICd50 fourwas purified calculated compounds for each sample from forBQ bothpulp. assays. Data are presented as the

mean ± SD (n = 3) and IC50 was calculated for each sample for both assays. 2.3. 2.3.α-Glucosidaseα-Glucosidase Inhibitory Inhibitory Activity Activity of Different of Different Fractions Fractions from from BQ BQPulp Pulp 2.3. Inα-Glucosidase theIn thepresent present Inhibitory study, study, different Activity different offractions Different fractions of Fractions BQ of BQpu lp pulpfrom extract BQ extract Pulp as well as well as asthe the purified puriﬁed compounds compounds were testedwereIn for the tested the present α for-glucosidasethe study,α-glucosidase different inhibitory fractions inhibitory activity of BQ activity (Figure pulp (Figureextract 6). Among 6as). well Among four as the fourpurified purified puriﬁed compounds, compounds compounds, thewere α - glucosidase inhibitory activities were M3R (IC50 = 418.8 μg/mL, 902.6 μmol/L) > Q3R (IC50 = 491.8 μg/mL, testedthe αfor-glucosidase the α-glucosidase inhibitory inhibitory activities activity were M3R(Figure (IC 6).50 =Among 418.8 fourµg/mL, purified 902.6 compounds,µmol/L) > the Q3R α- 1097.8 μmol/L) > Q3Gal (IC50 = 556.4 μg/mL, 1199.1 μmol/L) > C3G (IC50 = 1444.3 μg/mL, 3216.7 μmol/L), glucosidase(IC50 = 491.8 inhibitoryµg/mL, activities 1097.8 wereµmol/L) M3R (IC >50 Q3Gal = 418.8 (IC μg/mL,50 = 556.4902.6 µμg/mL,mol/L) > 1199.1 Q3R (ICµmol/L)50 = 491.8 > μ C3Gg/mL, which were comparable to the positive control acarbose (IC50 = 383.2 μg/mL, 593.6 μmol/L). 1097.8(IC50 μ=mol/L) 1444.3 > Q3Galµg/mL, (IC 3216.750 = 556.4µmol/L), μg/mL, which1199.1 wereμmol/L) comparable > C3G (IC to50 = the 1444.3 positive μg/mL, control 3216.7 acarbose μmol/L), (IC = 383.2 µg/mL, 593.6 µmol/L). 50 which50 were comparable to the positive control acarboseIC ( μ(ICg/mL )= 383.2 μg/mL, 593.6 μmol/L). 120 50 383.2 ± 33.9 2433.4 ± 75.2IC (μg/m207.2L) ± 22.6 15.4 ± 1.1 12090 50 383.2 ± 33.9 2433.4 ± 75.2 207.2 ± 22.6 15.4 ± 1.1 6090

3060

300

.0 5 0 0.0 00 5.0 00 2.5 2 μg/mL 50 000 2 5 000 6 00 0 2 50 10 2 6 1250 2 5 125 250 5 6. 12.5 25. 50.0 -glucosidase (%)

α .0 5 0 Acarbose0.0 00 Crude5.0 extract00 2.5 50%2 fraction μg/mL 50 000 2 5 000 610% fraction00 2 50 10 2 6 1250 2 5 125 250 5 6. 12.5 25. 50.0 -glucosidase (%)

α IC (μg/mL) 120 Acarbose Crude extract50 10% fraction 50% fraction 1444.3 ± 47.4 418.8 ± 6.0IC (μg/mL556.4) ± 29.9 491.8 ± 17.8 12090 50 1444.3 ± 47.4 418.8 ± 6.0 556.4 ± 29.9 491.8 ± 17.8 Inhibition of 6090

Inhibition of 60

300 .0 0 .0 00 00 00 . μg/mL 000 0.0 000 0.0 000 0 250.0 500 1 20 25 500.0 1 20 25 500.0 1 20 250 500 1000 2000 0 0 .0 .0 μ .0 00 0.0 0 0.0 0 g/mL C3G000 M3R000 Q3Gal000 Q3R 2000 250.0 500 1 20 25 500.0 1 20 25 500.0 1 20 250 500 1000 C3G M3R Q3Gal Q3R Figure 6. Dose-dependent changes in α-glucosidase inhibition of crude extract, different fractions, and four purifiedFigureFigure 6.compounds Dose-dependent 6. Dose-dependent from BQchanges changespulp. in Da α in-glucosidasetaα -glucosidaseare presented inhibition inhibition as mean of crude of± crudeSD extract,(n extract,= 3) differentand different acarbose fractions, fractions, is used and and as four the positivepurifiedfour control. puriﬁed compounds compounds from BQ from pulp. BQ Da pulp.ta are Data presented are presented as mean as mean± SD ±(n SD= 3) (n and= 3) acarbose and acarbose is used isused as the positiveas the control. positive control.

Molecules 2016, 21, 1148 7 of 12

Molecules 2016Several, 21, 1148 studies have reported that C3G have beneficial effects for diabetes. The possible7 of 11 action mechanisms include improving insulin sensitivity via modulating the c-Jun N-terminal kinase/forkSeveral studies head have box O1reported signaling that pathway C3G have [15 ],beneficial ameliorating effects hyperglycemia for diabetes. viaThe the possible reduction action mechanismsof retinol include binding improving protein 4 insulin expression sensitivity in KK-A via ymodulatingmice [16], the or c-Jun exerting N-terminal insulin-like kinase/fork activities head boxby O1 PPAR signalingγ activation pathway [ 17[15],]. Q3Galameliorating and Q3R hyperglycemia could enhance via the glucose reduction uptake of retinol through binding stimulating protein 4 expressionthe insulin-independent in KK-Ay mice [16], AMP-activated or exerting insulin-like protein kinaseactivities pathway by PPAR [18γ ,activation19]. In the [17]. present Q3Gal study,and Q3R couldour enhance findings glucose suggest uptake that C3G, through M3R, Q3Gal,stimulating Q3R alsothe hadinsulin-independent significant α-glucosidase AMP-activated inhibitory protein activities. kinase pathwayWhether [18,19]. they In showed the present similar study, activity our findingsin vivo and suggest whether that C3G, they M3R, have synergisticQ3Gal, Q3R effects also had with significant each α-glucosidaseother or with inhibitory other compounds activities. deserveWhethe furtherr they showed investigations. similar activity in vivo and whether they have synergisticInterestingly, effects with the each 50% other fraction or with from other BQ compounds pulp extract deserve had the further highest investigations.α-glucosidase inhibitory activityInterestingly, (IC50 =the 15.4 50%µ fractiong/mL) thanfrom BQ the pulp 10% extract fraction had (IC the50 highest= 207.2 αµ-glucosidaseg/mL), the inhibitory crude extract activity (IC50(IC = 5015.4= 2433.4 μg/mL)µg/mL) than the and 10% the fraction four purified (IC50 = compounds, 207.2 μg/mL), which the crude were alsoextract more (IC potent50 = 2433.4 than μ acarboseg/mL) and the (Figurefour purified6). compounds, which were also more potent than acarbose (Figure 6).

2.4.2.4. α-Glucosidaseα-Glucosidase Inhibitory Inhibitory Activity Activity of Six of SixBayberry Bayberry Cultivars Cultivars ApparentApparent dose-dependent dose-dependent α-glucosidaseα-glucosidase inhibitory inhibitory activities activities were wereobserved observed for six for bayberry six bayberry cultivars withcultivars serial concentrations with serial concentrations tested (Figure tested 7). The (Figure crude7 extracts). The crude of six extractsbayberry of cultivars, six bayberry i.e., BQ, cultivars, Dongkui (DK),i.e., Muye BQ, Dongkui (MY), Wuzi (DK), (WZ), Muye Zaodamei (MY), Wuzi (ZDM), (WZ), Dayexidi Zaodamei (DYXD), (ZDM), demonstrated Dayexidi (DYXD), α-glucosidase demonstrated inhibitory activityα-glucosidase with IC50 values inhibitory ranging activity from with 2083.7–3167.0 IC50 values μg/mL ranging and from the 10% 2083.7–3167.0 fractions withµg/mL IC50 andvalues the ranging 10% fromfractions 245.5–343.8 with ICμ50g/mL.values Furthermore, ranging from 50% 245.5–343.8 fractionsµ g/mL.of six Furthermore,cultivars showed 50% fractionsthe potent of sixα-glucosidase cultivars inhibitoryshowed activity the potent withα IC-glucosidase50 values ranging inhibitory from activity 23.2–69.5 with μg/mL, IC50 values all of rangingwhich were from better 23.2–69.5 than µthatg/mL, of the positiveall of control which wereacarbose. better than that of the positive control acarbose.

IC (μg/mL) 50 100 DK 2545.5±29.3 270.0±13.1 23.2±1.1

0 100 MY 2083.7±38.9 307.5±24.8 30.1±2.3

0 100 WZ 2556.5±53.4 307.8±17.4 28.4±2.6

50 -glucosidase activity (%) -glucosidase α

0 100 ZDM 2891.1±90.2 245.5±23.8 51.9±2.7

50 Inhibition of

0 100 XYXD 3167.0±45.8 343.8±25.8 69.5±2.9

0 5 50 00 00 .5 25 50 00 25 2.5 5.0 .0 μg/mL 62 12 25 50 62 1 2 5 6. 1 2 50 Crude extract 10% fraction 50% fraction FigureFigure 7. Inhibition 7. Inhibition of ofα-glucosidaseα-glucosidase by by crude crude extract, differentdifferentfractions fractions from from ﬁve five Chinese Chinese bayberry bayberry

cultivars.cultivars. IC50 ICvalues50 values are presented are presented as mean as mean ± SD± (nSD = 3). (n = 3).

The relatively higher α-glucosidase inhibitory activities in the 50% fractions and 10% fractions than those of the purified compounds indicated the involvement of other compounds or the synergistic effects of different bioactive compounds for the α-glucosidase inhibitory effect, which merits further analysis in

Molecules 2016, 21, 1148 8 of 12

The relatively higher α-glucosidase inhibitory activities in the 50% fractions and 10% fractions than those of the puriﬁed compounds indicated the involvement of other compounds or the synergistic effects of different bioactive compounds for the α-glucosidase inhibitory effect, which merits further analysis in the future. Additionally, this is the ﬁrst report about the in vitro α-glucosidase inhibitory activities of different bayberry extracts, and the potent α-glucosidase inhibitory effect of the 50% fraction may provide potential alternatives for the hypoglycemic agents.

3. Materials and Methods

3.1. Chemicals and Reagents The authentic C3G, DPPH, dimethyl sulfoxide (DMSO), α-glucosidase (EC 3.2.1.20), methanol, and acetonitrile of chromatographic grade were purchased from Sigma-Aldrich (St. Louis, MO, USA). M3R, Q3Gal, Q3R, acarbose, and 4-nitrophenyl-α-D-glucopyranoside (PNPG) were from J and K Scientiﬁc (Shanghai, China). ABTS was bought From Aladdin Inc. (Shanghai, China). Bovine serum albumin (BSA) was purchased from Sangon Biotech (Shanghai, China) Co., Ltd. (Shanghai, China). C18 Sep-Pak® cartridge (12 cc/2 g) columns were obtained from Waters (#WAT036915, Milford, MA, USA). Double-distilled water (ddH2O) was used in all experiments and samples for HPLC–DAD and LC-ESI-MS were ﬁltered through a 0.22 µm membrane before injection. All of the other reagents were of analytical grade bought from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China).

3.2. Fruit Materials Chinese bayberry cultivars BQ, DK, MY, WZ, ZDM, and XYXD fruit were harvested at commercial maturity in June 2014 from Zhejiang province, China. The fruit were selected for uniformity of shape and color, and absence of disease and mechanical damage. The pulp was frozen in liquid nitrogen. After freeze-drying (FM 25EL-85, VirTis, Gardiner, NY, USA), it was ground into a ﬁne powder and stored at −80 ◦C until extraction and analysis.

3.3. HPLC Analysis of Four Flavonoid Glycosides Flavonoid glycosides were analyzed according to Schieber, Keller, and Carle [20] with some modifications. An HPLC system (2695 pump, 2996 diode array detector, Waters) coupled with an ODS C18 analytical column (4.6 mm × 250 mm) was used for characterizing individual flavonoids compounds. The compounds were detected between 200 and 600 nm. The mobile phase of HPLC consisted of 0.1% (v/v) formic acid in water (eluent A) and of acetonitrile: 0.1% formic acid (1:1, v/v) (eluent B). The gradient program was as follows: 0–40 min, 10%–38% of B; 40–60 min, 38%–48% of B; 60–70 min, 48%–100% of B; 70–75 min, 100%–10% of B; 75–80 min, 10% of B. The column was operated at a temperature of 25 ◦C, the flow rate was 1 mL/min, and the injection volume was 10 µL. Flavonoid glycosides were then quantified based on peak area and comparison with the standard curves. In order to plot the standard curve, independent dilution was made from 2.0 mg/mL with methanol (C3G was with 0.5% formic acid into it) to prepare standard solutions at concentrations of 0.1, 0.2, 0.4, 0.6, 0.8, 1.0, and 2 mg/mL.

3.4. Preparation of the Crude Extract Lyophilized BQ pulp powder (40 g) was extracted with 800 mL of 80% aqueous methanol (with 0.5% formic acid) by sonication for 30 min, while the lyophilized pulp powders (10 g) of the other ﬁve (DK, MY, WZ, ZDM, and DYXD) cultivars were extracted with 200 mL of 80% aqueous methanol (with 0.5% formic acid) by sonication for 30 min. The ultrasonic frequency and power were 60 kHz and 30 W, respectively. The crude extract was centrifuged at 10,000 rpm for 10 min at room temperature and the residue was extracted twice as above. All of the supernatants were combined and evaporated by a rotary evaporator under reduced pressure at 37 ◦C to remove methanol, and then dissolved in ddH2O (with 0.5% formic acid). Molecules 2016, 21, 1148 9 of 12

3.5. Dynamic Adsorption and Desorption Tests In the present study, an SPE column was selected to elute the target compounds. The SPE column was washed by 2 BV of methanol and ddH2O successively, and then the crude extract was loaded on the column. The initial feed concentration was 1.81 mg/mL C3G, 0.04 mg/mL M3R, 0.14 mg/mL Q3Gal, and 0.07 mg/mL Q3R. Dynamic adsorption and desorption experiments were first carried out to determine the loading volume and the elution conditions. The column was washed thoroughly by ddH2O to remove the impurities. Gradient elution from the SPE column was carried out by increased concentrations of methanol (10%, 20%, 30%, 40%, 50%, and 60% with 0.5% formic acid) successively. Each effluent was collected and analyzed by HPLC. Samples with enriched C3G (10% fraction) and enriched flavonols (50% fraction) were evaporated at 37 ◦C under vacuum for further Prep-HPLC purification.

3.6. Prep-HPLC Purification The 10% fraction was used to purify individual C3G and the 50% fraction was used to purify the three flavonols by Prep-HPLC (2535 pump, 2998 detector, 2707 auto-sampler, Waters, Milford, MA, USA) with a SunFireTM prep C18 OBDTM (5 µm, 19 mm × 250 mm) column, respectively. The mobile phase solvents were the same as that of HPLC. The C3G gradient program was as follows: 0–10 min, 10% B; 10–30 min, 10%–30% B; 30–50 min, 30%–45% B; 50–60 min, 45%–60% B. The flavonols gradient program was as follows: 0–35 min, 20%–38% B; 35–60 min, 38%–48% B, 60–80 min, and 48%–100% B. The flow-rate was 5.0 mL/min, the injection volume was 300 µL and the concentration was 100 mg/mL. The effluent from the outlet of the column was monitored at 280 nm and 350 nm. The peak fractions were collected according to the chromatogram.

3.7. LC-MS Analysis Mass spectrometric analyses were performed by an Agilent 6460 triple quadrupole mass spectrometer equipped with an ESI source (Agilent Technologies, Santa Clara, CA, USA) that operated in both positive ionization (C3G) and negative ionization (flavonols) mode. The nebulizer pressure was set to 45 psi and the flow rate of drying gas (N2) was 5 L/min. The flow rate and the temperature of the sheath gas were 11 L/min and 350 ◦C, respectively. Chromatographic separations were done as the previous HPLC method. The eluent was split and approximately 0.3 mL/min was introduced into the mass detector. An Agilent Mass Hunter Workstation was used for data acquisition and processing.

3.8. Antioxidant Activity DPPH radical scavenging activity was measured according to Brand-Williams, Cuvelier, and Berset [21] with modiﬁcations. The reaction for scavenging DPPH radical was carried out by adding 2 µL sample to 198 µL 25 µg/mL DPPH solution at room temperature. Absorbance at 515 nm before (A0) and after (A1) the reaction (1 h) was recorded using a micro-plate reader (BioTech, Synergy H1, Winooski, VT, USA). The radical scavenging activity was calculated as:

Scavenging rate (%) = (A0 − A1)/A0 × 100 (1)

The concentration (µg/mL) that inhibited 50% DPPH radicals was calculated as the IC50 value. ABTS assay was carried out using a spectrophotometer as reported [22]. The ABTS radical cation was generated by reacting 7 mmol/L ABTS with 2.45 mmol/L potassium persulfate, and the mixture was allowed to stand in the dark at room temperature for 16 h before use. Before analysis, the ABTS solution was diluted with ethanol to an absorbance of 0.70 ± 0.05 at a wavelength of 734 nm. The absorbance at 734 nm was recorded for 6 min after mixing of 0.1 mL of the tested samples dissolved in ethanol with 3.9 mL of ABTS solution. The radical scavenging activity was calculated as the Equation (1) above. The concentration (µg/mL) that inhibited 50% ABTS radicals was calculated as the IC50 value. Molecules 2016, 21, 1148 10 of 12

3.9. Inhibition of α-Glucosidase Activity Inhibitory α-glucosidase activity was determined spectrophotometrically in a 96-well microtiter plate based on PNPG as a substrate following the slightly modiﬁed method described in [23,24]. α-Glucosidase (20 µL, 0.2 U/mL in 0.01 mol/L potassium phosphate buffer containing 0.2% of BSA) and 0.1 mol/L potassium phosphate buffer (pH 6.8, 112 µL) were mixed with a test sample (8 µL). After pre-incubating at 37 ◦C for 15 min, 2.5 mmol/L PNPG (50 µL) was added. The reaction was ◦ incubated at 37 C for 15 min and stopped with 0.1 mol/L Na2CO3 (80 µL). 4-Nitrophenol absorption was measured at 405 nm. Solution with DMSO instead of the sample was used as negative control. Solution without substrate was used as a blank. Acarbose was also assayed as a standard reference. The percent inhibition of α-glucosidase was calculated as follows:

OD − OD Inhibition (%) = (1 − test blank ) × 100 (2) controlODtest − controlODblank

3.10. Statistical Analysis Experiments were performed in triplicate and data were expressed as the mean ± standard deviation (SD). Data were analyzed using SPSS statistics Version 22 (SPSS Inc., Chicago, IL, USA). All graphical representations were performed using Origin Pro 8.0 software packages (Origin lab Corporation, Northampton, MA, USA).

4. Conclusions An efficient simultaneous purification method combining SPE chromatography with Prep-HPLC was established to isolate four flavonoid glycosides from bayberry fruit in the present study. Purified flavonoid glycosides were identified by HPLC and LC-MS. In addition, four purified compounds were important contributors to the antioxidant and α-glucosidase inhibitory activity of bayberry. Furthermore, the potential high α-glucosidase inhibitory activity of 50% fraction of fruit extracts suggested that the consumption of bayberry fruit after diet might be useful to control the rapid increase of postprandial blood level. Such results may provide an important foundation for further investigation of the pharmaceutical effects of bayberry fruit, as well as the mechanisms acted by different bioactive compounds.

Acknowledgments: This work was supported by the Special Scientific Research Fund of Agricultural Public Welfare Profession of China (201203089), the Program of International Science and Technology Cooperation (2014DFE30050) and the Fundamental Research Funds for the Central Universities. Author Contributions: Shuxia Yan, Xian Li and Changjie Xu designed the experiments; Shuxia Yan, Qiang Lv, Xianan Zhang and Xin Wen performed the experiments; Shuxia Yan, Xianan Zhang and Xian Li analyzed the data; Shuxia Yan, Xian Li, Changjie Xu and Chongde Sun contributed reagents, materials and analytical tools; Shuxia Yan and Xian Li wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.

References

1. American Diabetes Association. Diagnosis and classiﬁcation of diabetes mellitus. Diabetes Care 2014, 37, 81–90. 2. Bischoff, H. Pharmacology of α-glucosidase inhibition. Eur. J. Clin. Investig. 1994, 24, 3–10. 3. Baron, A.D. Postprandial hyperglycaemia and alpha-glucosidase inhibitors. Diabetes Res. Clin. Pract. 1998, 40, S51–S55. [CrossRef] 4. Fujisawa, T.; Ikegami, H.; Inoue, K.; Kawabata, Y.; Ogihara, T. Effect of two α-glucosidase inhibitors, voglibose and acarbose, on postprandial hyperglycemia correlates with subjective abdominal symptoms. Metabolism 2005, 54, 387–390. [CrossRef][PubMed] Molecules 2016, 21, 1148 11 of 12

5. Laar, F.A.V.D.; Lucassen, P.L.; Akkermans, R.P.; Lisdonk, E.H.V.D.; Rutten, G.E.; Weel, C.V. α-Glucosidase inhibitors for patients with type 2 diabetes results from a Cochrane systematic review and meta-analysis. Diabetes Care 2005, 28, 154–163. [CrossRef][PubMed] 6. Madar, Z. The effect of acarbose and miglitol (BAY-M-1099) on postprandial glucose levels following ingestion of various sources of starch by nondiabetic and streptozotocin-induced diabetic rats. J. Nutr. 1989, 119, 2023–2029. [PubMed] 7. Zhou, S.H.; Fang, Z.X.; Yuan, L.; Chen, J.C.; Liu, D.H.; Ye, X.Q. Phenolics and antioxidant properties of bayberry (Myrica rubra Sieb. et Zucc.) pomace. Food Chem. 2009, 112, 394–399. [CrossRef] 8. Fang, Z.; Zhang, M.; Wang, L. HPLC-DAD-ESIMS analysis of phenolic compounds in bayberries (Myrica rubra Sieb. et Zucc.). Food Chem. 2007, 100, 845–852. [CrossRef] 9. Bao, J.S.; Cai, Y.Z.; Sun, M.; Wang, G.Y.; Corke, H. Anthocyanins flavonols and free radical scavenging activity of Chinese bayberry Myrica rubra extracts and their color properties and stability. J. Agric. Food Chem. 2005, 53, 2327–2332. [CrossRef][PubMed] 10. Sun, C.D.; Huang, H.Z.; Xu, C.J.; Li, X.; Chen, K.S. Biological activities of extracts from Chinese bayberry (Myrica rubra Sieb. et Zucc.): A review. Plant Foods Hum. Nutr. 2013, 68, 97–106. [CrossRef][PubMed] 11. Sun, C.D.; Zhang, B.; Zhang, J.K.; Xu, C.J.; Wu, Y.L.; Li, X.; Chen, K.S. Cyanidin-3-glucoside-rich extract from Chinese bayberry fruit protects pancreatic β cells and ameliorates hyperglycemia in streptozotocin-induced diabetic mice. J. Med. Food 2012, 15, 288–298. [CrossRef][PubMed] 12. Zhang, X.N.; Huang, H.Z.; Zhao, X.Y.; Lv, Q.; Sun, C.D.; Li, X.; Chen, K.S. Effects of flavonoids-rich Chinese bayberry (Myrica rubra Sieb. et Zucc.) pulp extracts on glucose consumption in human HepG2 cells. J. Funct. Foods 2015, 14, 144–153. [CrossRef] 13. Zhang, X.N.; Huang, H.Z.; Zhang, Q.L.; Fan, F.J.; Xu, C.J.; Sun, C.D.; Li, X.; Chen, K.S. Phytochemical characterization of chinese bayberry (Myrica rubra Sieb. et Zucc.) of 17 cultivars and their antioxidant properties. Int. J. Mol. Sci. 2015, 16, 12467–12481. [CrossRef][PubMed] 14. Zhang, B.; Kang, M.X.; Xie, Q.P.; Xu, B.; Sun, C.D.; Chen, K.S.; Wu, Y.L. Anthocyanins from Chinese bayberry extract protect β cells from oxidative stress-mediated injury via HO-1 upregulation. J. Agric. Food Chem. 2011, 59, 537–545. [CrossRef][PubMed] 15. Guo, H.; Min, X.; Zou, T.; Ling, W.; Zhong, R.; Zhang, W. Cyanidin 3-glucoside attenuates obesity-associated insulin resistance and hepatic steatosis in high-fat diet-fed and db/db mice via the transcription factor foxo1. J. Nutr. Biochem. 2012, 23, 349–360. [CrossRef][PubMed] 16. Sasaki, R.; Nishimura, N.; Hoshino, H.; Isa, Y.; Kadowaki, M.; Ichi, T.; Tanaka, A.; Nishiumi, S.; Fukuda, I.; Ashida, H.; et al. Cyanidin 3-glucoside ameliorates hyperglycemia and insulin sensitivity due to downregulation of retinol binding protein 4 expression in diabetic mice. Biochem. Pharmacol. 2007, 74, 1619–1627. [CrossRef][PubMed] 17. Scazzocchio, B.; Varì, R.; Filesi, C.; D’Archivio, M.; Santangelo, C.; Giovannini, C. Cyanidin-3-O-β-glucoside and protocatechuic acid exert insulin-like effects by upregulating PPARγ activity in human omental adipocytes. Diabetes 2011, 60, 2234–2244. [CrossRef][PubMed] 18. Eid, H.M.; Martineau, L.C.; Saleem, A.; Muhammad, A.; Vallerand, D.; Benhaddou-Andaloussi, A.; Nistor, L.; Afshar, A.; Arnason, J.T.; Haddad, P.S. Stimulation of AMPK-activated protein kinase and enhancement of basal glucose uptake in muscle cells by quercetin and quercetin glycosides, active principles of the antidiabetic medicinal plant Vaccinium vitis-idaea. Mol. Nutr. Food Res. 2010, 54, 991–1003. [CrossRef] [PubMed] 19. Zhou, J.; Yoshitomi, H.; Liu, T.; Zhou, B.; Sun, W.; Qin, L.; Guo, X.; Huang, L.; Wu, L.; Gao, M. Isoquercitrin activates the AMPK-activated protein kinase (AMPK) signal pathway in rat H4IIE cells. BMC Complement. Altern. Med. 2014, 14, 1–10. [CrossRef][PubMed] 20. Schieber, A.; Keller, P.; Carle, R. Determination of phenolic acids and flavonoids of apple and pear by high-performance liquid chromatography. J. Chromatogr. A 2001, 910, 265–273. [CrossRef] 21. Brand-Williams, W.; Cuvelier, M.E.; Berset, C. Use of a free radical method to evaluate antioxidant activity. LWT Food Sci. Technol. 1995, 28, 25–30. [CrossRef] 22. Re, R.; Pellegrini, N.; Proteggente, A.; Pannala, A.; Yang, M.; Rice-Evans, C. Antioxidant activity applying an improved ABTS radical cationdecolorization assay. Free Radic. Biol. Med. 1999, 26, 1231–1237. [CrossRef] 23. Li, T.; Zhang, X.D.; Song, Y.W.; Liu, J.W. A microplate-based screening method for alpha-glucosidase inhibitors. J. Clin. Pharmacol. Ther. 2005, 10, 1128–1134. Molecules 2016, 21, 1148 12 of 12

24. Zhang, J.; Zhao, S.; Yin, P.; Yan, L.; Han, J.; Shi, L.; Zhou, X.; Liu, Y.; Ma, C. α-Glucosidase inhibitory activity of polyphenols from the burs of Castanea mollissima Blume. Molecules 2014, 19, 8373–8386. [CrossRef][PubMed]

Sample Availability: Samples of the compounds 1–4 are available from the authors.

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).