518 Chiang Mai J. Sci. 2016; 43(3)

Chiang Mai J. Sci. 2016; 43(3) : 518-533http://epg.science.cmu.ac.th/ejournal/Contributed Paper

Application of Ultrasonic Assisted Extraction ofBioactive Compounds from the Fruits ofAntidesma puncticulatum Miq. and Evaluationof Its Antityrosinase ActivityBancha Yingngam [a,b], Marlene Monschein [a] and Adelheid H. Brantner *[a]

[a] Institute of Pharmaceutical Sciences, Department of Pharmacognosy,

University of Graz, Universitaetsplatz 4/1, A-8010 Graz, Austria.

[b] Faculty of Pharmaceutical Sciences, Department of Pharmaceutical Chemistry and Technology,

Ubon Ratchathani University, Ubon Ratchathani 34190, Thailand.

*Author for correspondence; e-mail: adelheid.brantner@uni-graz.at

Received: 7 March 2014

Accepted: 11 November 2014

ABSTRACT

The ripe fruits of Antidesma puncticulatum are used as commercial sources for phenoliccompounds and anthocyanins for functional foods. The aim of this study was to optimize theprocess parameters for extraction of phenolic and anthocyanin compounds with goodantioxidant activity from the fruits using ultrasonic assisted extraction (UAE). A four-factor,five-level, six-center point central composite design was performed. Effects of solvent-to-solid ratio (X

1: 7.5-37.5 mL/g), acidified ethanol solution (X

2: 0-80%), extraction temperature

(X3: 20-100°C), and extraction time (X

4: 0-20 min) on the recovery of total phenolic and total

anthocyanin contents with good antioxidant activity were investigated. The second orderpolynomial models of the response variables were obtained with an R2 of 0.9455, 0.9308 and0.8998, respectively. The optimal extraction conditions to obtain maximum yields of targetedcompounds were 30 mL/g X

1, 45% (v/v) X

2, 80°C X

3 and 12 min X

2. The experimental

values were in good agreement with the predicted values under the optimized conditions.The major constituents in the optimized extract were organic acids whereas phenoliccompounds were minor components. Delphinidin-3-sambubioside-5-rhamnoside anddelphinidin were identified by HPLC-ESI/MS as major anthocyanins in the extract. The plantextract showed moderate inhibitory activity on mushroom tyrosinase with a non-competitiveinhibitory mechanism. The UAE can be considered as an effective method for extractingthe biologically active compounds from fruits of A. puncticulatum.

Keywords: Antidesma puncticulatum Miq., anthocyanins, antityrosinase activity, phenoliccompounds, response surface methodology, ultrasonic assisted extraction

1. INTRODUCTION

Phenolic compounds-rich plant extractsare considered as good sources for health

promoting effects in functional food industry[1-4]. One of the interesting sources for

Chiang Mai J. Sci. 2016; 43(3) 519

phenolic compounds is tropical fruits. Of 18Thai Antidesma species, Antidesma puncticulatumMiq. (family Phyllanthaceae; synonyms,A. bunius (L.) Spreng. var. thwaitesianum(M ll.Arg.) Trim. and A. thwaitesianumM ll.Arg.) is an indigenous plant in SoutheastAsia, especially in the dipterocarp forest ofthe Phupan Valley in the north-eastern areaof Thailand [5]. The fruits of this plant areclassified to be bigger in size and grow up to8 mm compared with other fruit plantsbelonging to the same genus. Its ripe fruit isof purplish-red or dark-purple whichdepending on the level of ripeness that occursfrom August to October.

Several phenolic compounds havebeen reported in fruits-based wine ofA. puncticulatum . These are caffeic acid,catechin, gallic acid and monogalloylglucoside. The fruits-based wine also containsanthocyanin pigments, such as cyanidin-3-sophoroside, delphinin-3-sambubioside andpelargonidin-3-malonyl glucoside [6]. Notonly the constituents have been identifiedbut also a potential health promotingbenefit related to the above mentionedbioactive components have been reported.The biological activities described in literatureare anti-apoptosis and anti-inflammatoryeffects in human breast epithelial cells [7],antioxidant [6], and α-glucosidase inhibitoryactivity [8].

In order to use these compounds as eitherbioactive components or colorants infunctional foods, an efficient extractionprocedure should be the primary step tomaximize the targeted yields. The use ofultrasonic assisted extraction (UAE) isrecognized as one of the modern techniquesfor the extraction of bioactive compoundsfrom plant material [9]. Compared to othertechniques (solvent extraction, microwaveassisted extraction and supercritical fluidextraction), the UAE offers some advantages

as it is a simple method with low instrumentrequirements and easy to scale up for industries[10]. The principle of UAE to enhance thephenolic compounds recovery is based on theproduction of acoustical cavitation to disruptthe cell wall of the plant material, leadingpenetration of solvent to dissolve desiredconstituents [11]. The use of a lowtemperature for the extraction processprotects degradation of the heat unstablecompounds [9].

To achieve a maximum yield of desiredbioactive compounds, a response surfacemethodology (RSM) was employed togetherwith UAE. This technique has been introducedas an effective statistic method for optimizingcomplex processes. It offers an economicallyefficient procedure compared to using aone-variable-at-a-time methodology which isthe classical method. This is because theRSM can be used to study several processvariables simultaneously interacting with eachother. A lower number of experiments arerequired to predict the optimal conditions ofthe desired system [2].

However, there were no studiesundertaken up to now to maximize the contentof the bioactive compounds in the ripe fruitsof A. puncticulatum. In the present study, someextraction variables of UAE affected theextraction yields of total phenolic content(TPC) and total anthocyanin content (TAC)of A. puncticulatum fruits were conductedusing the RSM. The quantification of organicacids in the extract obtained from theoptimized extraction conditions wasperformed by high-performance liquidchromatography (HPLC). The majoranthocyanin constituents in the extract wereidentified by high-performance liquidchromatography tandem electrosprayionization mass spectrometry (HPLC-ESI/MS). Additionally, in vitro antityrosinase activitywas also tested to support the use of the

520 Chiang Mai J. Sci. 2016; 43(3)

In order to set the range of each variableinto the RSM, effects of relevant independentparameters affected on the yields of thecorresponding response variables werestudied using a one-variable-at-a-timemethod. The TPC and TAC were used asindicators of bioactive compounds. Briefly,the effect of liquid-to-solid ratio was firstlystudied in the range from 5 to 60 mL/g whileother extraction parameters were keptconstant (25% (v/v) ethanol acidified with1% (v/v) acetic acid was fixed in allexperiments, 45°C, 10 min). Second, theimpact of the concentrations of acidifiedethanol in water (0-100% (v/v)) on theextraction yields of the above-mentionedvariables were evaluated. Other extractionparameters were performed at 20 mL/gsolvent-to-solid ratio, 45°C, and 10 min.Third, influence of extraction temperatureswas examined in the range starting from30°C to 90°C. The extraction conditionswere fixed constant for other variables at20 mL/g solvent-to-solid ratio, 25% acidifiedethanol solution, and 10 min extraction time.Finally, the effect of extraction times affectedon extraction yields of the dependentvariables was investigated. The ranges of thisfactor were varied between 5 and 40 minwhereas other process parameters were fixedconstant at 20 mL/g solvent-to-solid ratioand 25% acidified ethanol solution at 45°C.

2.4 Experimental DesignThe RSM was performed using Design

Expert software (version 7.0.0, Stat-Ease, Inc.,Minneapolis, MN, USA) to optimize processparameter for obtaining the maximized yieldof total anthocyanin content and total phenoliccontent. A central composite design (CCD)of RSM was employed in the experimentsby setting the ranges of each independentvariable based on the values obtained from aone-variable-at-a-time method described in

extracts for health promotion benefits.

2. MATERIALS AND METHODS

2.1 Plant MaterialFresh fruits of A. puncticulatum in a fully-

ripe stage were harvested in August 2013from Ubon Ratchathani province, Thailand.The plant was identified by Dr. BanchaYingngam and its voucher specimen wasdeposited at the Herbarium of the Facultyof Pharmaceutical Sciences, UbonRatchathani University (BCY UBU no. 025).The whole fresh fruits were dried in a freezedryer (Martin Christ GmbH, Germany)at -55°C for 72 h and powdered through a60-mesh sieve before extraction.

2.2 ChemicalsThe authentic L-ascorbic acid,

cyanin-3-glucoside, gallic acid, L-3,4-dihydroxyphenylalanine (L-DOPA), 1,1-diphenyl-2-picrylhydrazyl (DPPH),6-hydroxyl-2, 5-7,8-tetramethyl-chroman-2-carboxylic acid (trolox), kojic acid, andmushroom tyrosinase (1881 units/mg) wereobtained from Sigma Chemical Company(St. Louis, MO, USA). All other chemicalsand reagents were of analytical grade.

2.3 UAE ProcedureThe UAE was performed using an

ultrasonic bath (ULTRAsonika™ 57H model,C&A Sales Industrial Supplies, USA). Theultrasonic bath was set a working frequencyof 45 kHz and controlled temperatures bycirculating external from a temperaturecontroller. One gram of A. puncticulatum fruitpowder was subjected to the UAE forpredetermined points of each variable.The obtained samples were then centrifugedat 3500 rpm for 10 min, followed bycollecting the supernatant for furtherquantification of total anthocyanin and totalphenolic contents.

Chiang Mai J. Sci. 2016; 43(3) 521

i=1 i=0

i=0 j =j+1

Table 1. The CCD and experimental data for extraction of the total phenolic content (Y1), the

total anthocyanin content (Y2), and the antioxidant activity (Y

3) from A. puncticulatum fruits.

the section above. Liquid-to-solid ratio (X1),

acidified ethanol concentration (X2), extraction

temperature (X3), and extraction time (X

4)

was chosen as independent variables forthe experiments. The TPC (Y

1) and TAC (Y

2)

and antioxidant activity (Y3) were selected as

the responses for the combination of thoseindependent variables. The CCD wasperformed that consisting of 24 factorialdesigns, 8 axial points, and 6 replicates asindicated in Table 1. The experimentalnumber of factorial designs and axial points

of the study was applied for estimation ofcurvature of the model whereas the replicateexperiments used for estimation of a pureerror sum of squares [12]. The experimentswere randomized to minimize the effects ofunexplained variability affected by extraneousfactors. A second-order polynomial equationcorresponding to the CCD was given asfollows:

Y = β0

+ Σ3 βiX

i + Σ3 β

iiX2 +

Σ2 Σ3 βijX

i X

j(1)

Experimentalnumber

1234567891011121314151617181920212223242526

Independent variablesX

1

(mL/g)

153015301530153015301530153015307.537.522.522.522.522.522.522.522.522.5

X2

(%v/v)

202060602020606020206060202060604040080404040404040

X3

(°C)

40404040808080804040404080808080606060602010060606060

X4

(min)

5555555515151515151515151010101010100201010

Dependent variablesY

1 (mg

GAE/gdry powder)27.04±1.0930.13±1.8930.67±3.3330.84±1.3835.60±2.0743.37±5.0936.51±1.8138.82±5.0228.33±0.8227.41±2.3231.06±1.1631.22±2.2434.65±0.8732.66±0.3738.97±1.3539.12±0.5433.76±0.7641.17±2.8618.88±0.6225.13±2.0232.55±0.7753.60±1.8540.03±0.7239.01±0.6539.85±1.3742.95±1.12

Y2 (mg Cy-3-glu/g drypowder)0.92±0.031.00±0.010.98±0.021.06±0.030.98±0.011.07±0.021.05±0.061.12±0.030.96±0.020.96±0.021.07±0.051.09±0.040.93±0.031.02±0.041.06±0.031.12±0.020.96±0.071.11±0.100.75±0.011.01±0.111.02±0.021.02±0.021.09±0.041.08±0.031.02±0.031.01±0.02

Y3 (mM

TE/gdry powder)

1.24±0.031.61±0.021.44±0.021.80±0.041.46±0.021.71±0.071.72±0.031.89±0.041.51±0.021.93±0.011.68±0.022.54±0.021.92±0.012.80±0.022.01±0.022.38±0.051.61±0.012.66±0.031.04±0.031.59±0.012.17±0.042.66±0.022.29±0.032.49±0.012.55±0.032.49±0.01

522 Chiang Mai J. Sci. 2016; 43(3)

Table 1. Continued.

X1 solvent to solid ratio; X

2 acidified ethanol concentration; X

3 extraction temperature; X

4

extraction time

Experimentalnumber

27282930

Independent variablesX

1

(mL/g)

22.522.522.522.5

X2

(%v/v)

40404040

X3

(°C)

60606060

X4

(min)

10101010

Dependent variablesY

1 (mg

GAE/gdry powder)43.04±1.7239.63±1.1739.88±1.1240.00±1.0

Y2 (mg Cy-3-glu/g drypowder)1.06±0.050.99±0.021.03±0.021.04±0.20

Y3 (mM

TE/gdry powder)

2.76±0.022.43±0.022.59±0.072.51±0.03

where Y is the predicted response ofdependent variable, b

0 is the model constant,

bi, b

ii, and b

ij are the model coefficients

representing the linear, quadratic andinteraction effects of the variables. Finally,the obtained predicted mathematical modelsgiving the maximum recovery yield of Y

1, Y

2

and Y3 were checked to verify the validity

of the obtained mathematical models.For chemical and biological characterizationof the extract, the ethanol of the samplesobtained from the optimum extractionprocess was removed under vacuum bya rotary evaporator (B CHI, Flawil,Switzerland), followed by drying under afreeze-dryer.

2.5 Determination of TPCThe TPC of the plant extracts was

determined using the Folin-Ciocalteu methodas described by Gong et al. [13]. The resultswere expressed as mg gallic acid equivalentper gram of plant powder (mg GAE/g).

2.6 Determination of TACThe TAC in the extract was determined

using a pH differential method according tothe previously reported method [12]. Theamount of total anthocyanins was calculatedin terms of cyanidin-3-O-glucoside equivalentsper gram of dry plant powder (mg Cy-3-glu/g).

2.7 DPPH• Free Radical ScavengingActivity Assay

The antioxidant activity (AA) of A.puncticulatum extract was measured usingDPPH• assay according the method describedby Gong et al. [13]. The antioxidant activityof the plant extract was calculated from thestandard curve of trolox and the results areexpressed as millimolar trolox equivalentper gram of dry plant powder (mM TE/g).

2.8 Quantification of Organic acids in theA. puncticulatum Fruit Extract

The amounts of the five organic acids,citric acid, L-ascorbic acid, L-(+)-tartaric acid,oxalic acid and L-malic acid, in the freezedried extract obtained under the optimumextraction conditions were determinedby HPLC-DAD (Agilent Technologies1100 series, USA). The HPLC instrumentscomprised of a degasser, quaternary pump,automated sample injector, column ovenand a diode array detector. Separation ofthe sample (5 μL) was carried out onLichrosphere C

18 (4.5 mm × 250 mm i.d.,

4.5 μm particle size) using a 50 mM phosphatebuffer solution as mobile phase. The flowrate was set at 0.7 mL/min at 35°C.The detection wavelength of each organicacid was set at 214 nm, except forL-ascorbic acid (254 nm). The results wereexpressed as mg/g freeze dried extract.

Chiang Mai J. Sci. 2016; 43(3) 523

2.9 Identification of Phenolic Compoundsand Anthocyanins

The constituents of phenolic compoundsand anthocyanins in the extract wereidentified using a positive and a negativeionization mode of HPLC-ESI/MS (ThermoFisher Scientific Company, USA). Thisinstrument consisted of a DionexUltiMate 3000 UHPLC coupled with amass spectrometer. The sample (5 μL) wasseparated on a Lichrosphere C

18 column

(250 × 4.6 mm i.d., 4.5 μm particle size) bysetting a flow rate of 0.6 mL/min at25°C. The mobile phase was 0.1% formicacid in water (A) and acetonitrile (B).The gradient elution was 90%A (4 min), from90% to 80%A (6 min), from 80% to 10%A(30 min), constant at 10%A (5 min), and from10% to 90%A (15 min). The positive ionelectrospray was used by fixing nebulizergas (N

2) at a pressure of 6 bars and flow rate

of 1.5 L/min. Heated capillary was set at230°C with a voltage of 1.7 kV. The fullscans mass spectra were measured from m/z50 to 2000. Mass fragments of compoundsin the extract were identified based on theirretention times, UV-VIS spectra and massspectra in comparison to authentic standardsand to literature data.

2.10 Enzymatic Assay for Testing theAntityrosinase Activity2.10.1 Tyrosinase inhibition assay

The tyrosinase inhibition activity wasdetermined according to the method ofChang et al. [14]. The IC

50 of tyrosinase

activity of the samples leading to 50% lossof activity was calculated. Kojic acid was usedas standard reference (0.005-1 mg/mL).

2.10.2 Kinetic analysisThe kinetic reaction of A. puncticulatum

extracts affected on activity of mushroom

tyrosinase was examined [14]. TheLineweaver-Burk plots were constructedand Michaelis-Menten constant (K

m) and

maximal velocity (Vmax

) of the tyrosinasewere calculated.

2.11 Statistical AnalysisAll experiments were performed in

triplicate and the results are expressed asmeans ± standard deviation (SD) (n=3).The RSM data were analyzed using DesignExpert software. Analysis of variance(ANOVA) was used to determine by settingthe p value lower than 0.05 as a minimumstatistically significant difference. The lack offit of the mathematical model was checkedwith a Fisher’s statistical test (F-test). Thefitness of the second order polynomialmodel was determined and expressed interms of the coefficient (R2, R

adj2).

3. RESULTS AND DISCUSSION

3.1 Selection of Relevant Ranges ofTested Independent Variables3.1.1 Effect of solvent-to-solid ratio

As shown in Figure 1A, the extractionefficiency of TPC significantly increasedfrom 29.80±0.87 to 36.50±1.54 mg GAE/gas the ratio of solvent-to-solid increasedwithin the ranges of 5 to 20 mL/g (p<0.05).Like extraction yields of TPC, ratio ofsolvent-to-solid showed similar yields ofTAC (Figure 1B). These results should beattributed to the increase in driving force ofthe mass transfer of targeted compounds [2].Further increasing the ratios of solvent-to-solid did not affect the yields of TPC andTAC. The mass transfer of these chemicalsreached a plateau at the liquid-to-solid ratioof 20 mL/g for TPC and TAC. The solvent-to-solid ratio of 20 mL/g was chosen forthe following experiments.

524 Chiang Mai J. Sci. 2016; 43(3)

Figure 1. Effects of four independentvariables on the yields of total phenoliccontent and total anthocyanin content fromA. puncticulatum fruits: A-B solvent-to-solidratios, C-D acidified ethanol concentrations,E-F extraction temperatures, and G-Hextraction times. Values are means ± standarddeviation of three separate determinations(n = 3).

3.1.2 Effect of acidified ethanolconcentration

Acidified ethanol solution (1% (v/v)acetic acid) was employed as extraction solventbecause anthocyanins usually are extractedunder the mild conditions of acidified solvents[15]. The extraction yields of TPC increasedwith higher percentages of acidified ethanolsolutions from 0% (21.11±1.41 mg GAE/g)

to 25% (36.38±1.89 mg GAE/g) (p<0.001)(Figure 1C). Not any change the yields of TPCwas observed when percentages of acidifiedethanol solutions continued to increase from25% to 50% (p>0.05). On the contrary, a sharpdecrease in the yield of TPC was found withincreasing concentrations of acidified ethanolfrom 50% (36.38±1.89 mg GAE/g) to 75%(23.49±1.16 mg GAE/g) (p<0.001). Yieldsof TAC showed a similar pattern like TPCyields by increasing to a maximum value at25% acidified ethanol solution (Figure 1D).Concentrations of acidified ethanol higherthan 75% sharply dropped the yields of TPC(p<0.001). Thus, 25% acidified ethanol solutionwas fixed constant for the followingexperiments.

3.1.3 Effect of extraction temperatureThe extraction yields of TPC increased

with rising the temperatures up to 75°C,followed by remaining constant up to 90°C(Figure 1E) whereas not any change wasobserved for the TAC yields (Figure 1F).This may be due to the increasing temperaturesenhanced both diffusion coefficient andsolubility of the desired phenolic compounds[2]. In order to avoid the degradation ofsome heat sensitive compounds at hightemperature, 75°C was the preferabletemperature for extraction of the targetedcompounds.

3.1.4 Effect of extraction timeThe extraction yields of dependent

variables increased as the extraction timesincreased from 1 to 10 min (Figure 1G-H).The maximum yield of TPC was found at10 min of extraction time whereas thehighest yield of TAC was observed after5 min. There was no increasing in the yieldsof TPC and TAC when rising the extractiontime up to 40 min. The possible reason forthis phenomenon may be caused by the very

Chiang Mai J. Sci. 2016; 43(3) 525

good solubility of the phenolic compoundsand anthocyanins of A. puncticulatum fruits inhydroalcoholic solutions resulting in a shorterextraction time.

3.2 Statistical Analysis and Fitting to theResponse Surface Models

A total of 30 experimental numbers wasdesigned and the corresponding responsevalues are shown in Table 1. Extracted yieldsof TPC and TAC and measurement of theAA of the plant extract were observed inranges between 18.88-53.60 mg GAE/g,0.75-1.12 mg Cy-3-glu/g, and 1.00-2.80 mMTE/g, respectively. The experimental valuesof both independent and dependent variableswere analyzed to obtain the regression equationthrough various models. The results indicatedthe quadratic model to be the best model forthe extraction process by exhibiting theprobability values (p) lower than 0.0001.The model summary statistics for Y1

, Y2, and

Y3 also showed the maximum values of R2,

Radj

2, and Rpred

2 in comparison to othermodels. The determination coefficient (R2) isa coefficient of the total variation in theresponse predicted model. A high value ofR2 indicates a good adequacy of the model[2]. In this study, the R2 for TPC, TAC andAA were 0.9455, 0.9308 and 0.8998,respectively. These values indicated that theobtained models could predict the responseswithin the range of the independent variables.The tested independent variables and thecorresponding variables were related by asecond-order polynomial model as given interms of the actual factors as follows:

Y1 = 40.90 + 1.07X

1 + 1.27X

2 + 4.38X

3

- 0.48X4 - 0.32X

1X

2 + 0.36X

1X

3 - 1.00X

1X

4 -

0.23X2X

3 + 1.04X

2X

4 - 0.52X

3X

4 - 1.19X2 -

5.05X2 - 0.21X2 - 0.68X2 (2)

Y2 = 1.03 + 0.033X

1 + 0.051X

2 +

0.013X3

+ 0.000416X4

- 0.001875X1X

2 +

0.0008125X1X

3 - 0.009375X

1X

4 -

0.000625X2X

3 + 0.014X

2X

4 - 0.013X

3X

4 +

0.005729X2 - 0.033X2 + 0.001979X2 +0.018X2 (3)

Y3 = 2.56 + 0.24X

1 + 0.099X

2 + 0.13X

3

+ 0.18X4

- 0.01X1X

2 - 0.021X

1X

3 +

0.086X1X

4 - 0.066X

2X

3 - 0.024X

2X

4 +

0.48X3X

4 - 0.14X2 - 0.35X2 - 0.07X2 - 0.076X2

(4)

The ANOVA for those mathematicalmodels was summarized in Table 2.Significance of each variable was evaluatedby ANOVA analysis followed by an F-test.The F-values indicated the influence of eachindependent variable on the correspondingresponses. The model of Y

1, Y

2 and Y

3

showed F-values of 18.59, 14.42, and 9.63with their p-values less than 0.001, indicatingthe model significance [12]. The fitness ofthe models was analyzed through the lack offit and the suitability models to predictaccurately and the variable should have ap-value higher than 0.05 [2]. The lack of fitcoefficients for Y

1, Y

2, Y

3 had F-values of

2.33, 1.37, and 5.10 with non-statisticallysignificant difference (p>0.05) indicating thepredicted models were not relative to the pureerror. The coefficient of variation (C.V.%) wasanalyzed to evaluate the variation of theexperimental data from the predicted models.The results found that these values were 6.31,2.66 and 10.76 for Y

1, Y

2 and Y

3, respectively,

indicating a high precision and a goodreliability of the experiments [12].

The p-values were also used to checkthe effects of four independent variables onthe responses. If the p-values are less than0.05, the term coefficients are significant [2].The results for analysis of the TPC yield (Y

1)

11 2 3 4

1

2 3 4

1 2 3

4

526 Chiang Mai J. Sci. 2016; 43(3)

Table 2. ANOVA and statistical parameters for the response surface models and the lack offit testing of extraction of target compounds from A. puncticulatum fruits.

a Sum of squares, b Degree of freedom, c Mean square; R2 = Coefficient of multiple determination; Radj2 = Adjusted R2;

C.V.% = Coefficient of varianceStatistically significant difference at * p<0.05, # p<0.01, and � p< .001; ns: non-significant

found that the linear coefficients (X1, X

2, X

3)

and the quadratic coefficients (X12, X

22) were

important factors due to their statisticalsignificant differences (p<0.05). In cases ofthe yields of TAC, the linear coefficients(X

1, X

2, X

3), interaction coefficients (X

2X

4),

and the quadratic coefficients (X22, X

42) were

statistically significant (p<0.05). For AA (Y3),

the coefficients that showed statistical

significance were four linear coefficients(X

1, X

2, X

3, X

4) and two quadratic coefficients

(X12, X

22). The remaining coefficients were

not important factors due to the fact thatthere was no significant statistical difference(p>0.05). Therefore, X

1, X

2, X

2 and X

4 were

important factors for the extraction of targetcompounds from A. puncticulatum fruits.

Source

ModelLinear coefficientX

1

X2

X3

X4

Interaction coefficientsX

1X

2

X1X

3

X1X4

X2X3

X2X4

X3X4

Quadratic coefficientsX1

2

X22

X32

X42

ResidualLack of fitPure errorCor totalR2

MeanC.V. %Adeq. Precision

Total phenolic content (Y1)

SS a

1308.90

27.2438.84

460.345.62

1.662.0715.900.8717.244.25

38.79700.68

1.2512.5275.4462.1313.31

1384.340.945535.536.3120.42

DF b

14

1111

111111

11111510529

MS c

93.49

27.2438.84

460.345.62

1.662.0715.900.8717.244.25

38.79700.68

1.2512.525.036.212.66

F-value18.59�

5.42*

7.72#

91.53�

1.12ns

0.33ns

0.41ns

3.16ns

0.17ns

3.43ns

0.85ns

7.71#

139.32�

0.25ns

2.49ns

2.33ns

Total anthocyanin content(Y2)

SS a

0.1500

0.02600.06300.0004

<0.0001

<0.00010.00110.0009

<0.00010.00280.003

0.00090.03000.00010.00910.01100.00810.0029

0.160.9308

1.022.6617.64

DF b

14

1111

111111

11111510529

MS c

0.0110

0.02600.06300.0004<0.0001

<0.00010.00110.0009<0.00010.00280.003

0.00090.03000.00010.00910.00070.00080.0006

F-value14.42�

35.43�

85.90�

5.46*

<0.01ns

0.08ns

1.44ns

1.92ns

0.01ns

4.51*3.76ns

1.23ns

40.75�

0.15ns

12.42#

1.37ns

Antioxidant activity (Y3)

SS a

6.56

1.390.240.410.77

0.00160.0072

0.120.070.0090.036

0.543.270.140.160.730.660.0657.29

0.89982.0510.7610.46

DF b

14

1111

111111

11111510529

MS c

0.47

1.390.240.410.77

0.00160.0072

0.120.070.0090.036

0.543.270.140.160.0490.0660.013

F-value9.63�

28.61�

4.85*

8.34#

15.84�

0.03ns

0.15ns

2.45ns

1.44ns

0.19ns

0.74ns

11.08#

67.19�

2.78ns

3.30ns

5.10ns

3.3 Response Surface AnalysisIn order to understand more details

about the relationships between theindependent variables and their interactionsaffected on the dependent variables, threedimensional surface and contour backgroundplots were constructed from mathematicalmodels given in Eq. (5)-(7). Each plotcomposed of two independent variables

affected on the corresponding responsevariables by fixing the remaining variables atlevel zero.

3.3.1 Effect of extraction parameters onTPC

Figure 2 shows effects of fourindependent variables on the extraction yieldsof TPC (Y

1). X

2 was the most influenced

Chiang Mai J. Sci. 2016; 43(3) 527

parameter affected on Y1 yield (p<0.001),

followed by X2 (p<0.01) and X

1 (p<0.05),

respectively, whereas extraction time did notsignificantly exhibit on such response (p>0.05).No interaction between each independentvariable was observed for Y

1 yield (p>0.05).

Lower values of X1 in ranges of 15-22.5

mL/g resulted in lower yields of Y1.

However, the yield of Y1 increased when

raising the values of X1 from 22.5 to 30 mL/

g, followed remaining constant up to37.5 mL/g (Figure 2A, B, C). This is probablydue to the fact that more solvent can permeatethrough the cell walls of the plant material

Figure 2. Response surface plot showing theeffects of independent variables on the yieldof total phenolic contents: A solvent-to-solidratio and acidified ethanol concentration, Bsolvent-to-solid ratio and extractiontemperature, C solvent-to-solid ratio andextraction time, D acidified ethanolconcentration and extraction temperature, Eacidified ethanol concentration and extractiontime, F extraction temperature and extractiontime.

to dissolve the phenolic compounds [16].These results are in agreement with previousreports describing that the addition ofethanol into water plays a significant rolefor the yield of TPC in the extract [2].

The ability to extract phenoliccompounds with a single solvent was muchlower compared to a combined solventsystem between alcohol and water [16].As expected, increasing acidified ethanolconcentration from 20 to 45% resulted inincreased yields of Y

1 (Figure 2A, D, E).

However, Y1 yields decreased when extracting

with acidified ethanol concentrations higherthan 45%. The reason therefore might bethat the addition of a certain amount ofethanol to the extraction solvent gives anappropriate polarity to phenolic compoundswhereas higher amounts of ethanol in theextraction solvent give a lower polarity [2].These findings are in good agreement withprevious reports [16].

The extraction temperature showed apositive effect on Y

1 yields (Figure 2B, D, F).

Yields of Y1 rapidly increased within an

extraction temperature from 40°C to 80°C.This result should be attributed to a highertemperature which show the ability torupture the phenolic matrix bonds, decreasethe dielectric constant of water and solventviscosity, and improve the solubility anddiffusion rate of phenolic compounds,resulting in improved extraction efficiency[12,17].

3.3.2 Effect of extraction parameters onTAC

Anthocyanins are water-solublecompounds that belong to a parent class ofphenolic compounds. As presented inFigure 3, X

2 showed the strongest effects on

Y2 yield (p<0.001), followed by X

1 (p<0.001)

and X3 (p<0.05), but not X

4 (p>0.05). It was

noticed that X2 and X

4 exhibited positive

528 Chiang Mai J. Sci. 2016; 43(3)

Figure 3. Response surface plot showingeffects of independent variables on the yieldof total anthocyanin contents (refer to Figure3 for (A), (B), (C), (D), (E) and (F) plots).

interactive effects by showing a positiveeffect on the yield of Y

1 (p<0.05) (Figure 3E).

Increasing values of X1 resulted in increased

yields of TAC starting from 15 mL/g up to30 mL/g. Like this, the pattern of Y

2

yields increased with raising X2 values in the

range of 20-45% (Figure 3A, D, E). This isprobably due to the fact that more solventwith the appropriate polarity can enter intothe plant cells to dissolve higher amounts ofanthocyanins [18].

Like above tested variables, the yields ofY

2 also increased with increased X

3 in a

range between 40°C-80°C (Figure 3B, D, F).The reason to describe why raising theextraction temperature increased the TACyield should attribute from several propertiestogether: softening of the plant tissue,disrupting the bonding between anthocyaninsand plant matrix, increasing solubility anddiffusion rate of anthocyanins [12].

3.3.3 Effect of extraction parameters onAA

The results showed that A. puncticulatumfruit extract demonstrated good antioxidantactivity. This may be due to the hydroxylgroups in the antioxidant compounds (organicacid, anthocyanins and other phenoliccompounds) which act as electron donors.These molecules can terminate the radicalchain reaction by converting the DPPH•

radicals into a more stable form [9,19].With respect to effects of extraction

parameters affected on antioxidant activity(AA) of the extracts, X

1 was the most

influenced factor (p<0.001), followed by X4

(p<0.001), X3 (p<0.01), and X

2 (p<0.05).

However, there was no interaction of thosetested independent variables on AA (p>0.05).As shown in Figure 4A -C, higher antioxidantactivity of the extract was observedwhen raising X

1 values from 15 mL/g up

to 30 mL/g. This is probably because ahigher volume of solvent can dissolve moreantioxidant compounds when comparedto a lower volume [16,18].

Next, the effect of acidified ethanolconcentration (X

2) on the AA was evaluated

and the results are shown in Figure 4A, D,and E. This variable influenced the linearand quadratic coefficient of the antioxidantactivity. The AA of the extract increasedwith the increased X

2 value from 20% to 45%,

and decreased when the proportion of suchvariable further increased up to 60%.These results are in agreement with ourprevious study due to the addition ofethanol to water has a great influence onthe antioxidant activity of extracts [2].A correlation between the AA of extractsand the extraction temperature could beobserved. The AA of the extract increasedwith increasing temperature (Figure 4B, D andF). Although several antioxidant compoundsmay degrade when exposed to high

Chiang Mai J. Sci. 2016; 43(3) 529

Figure 4. Response surface plot showingeffects of independent variables on theantioxidant activity (refer to Figure 3 for (A),(B), (C), (D), (E) and (F) plots).

temperature [20], the antioxidant activity ofthe extract in this study did not decrease inthe range of temperature used (40-80°C).This may be due to the rate of extraction ofthermally stable antioxidants which was higherthan the rate of the thermal decompositionof the unstable ones [21].

It could be also noticed that the extractiontemperature exhibited a more significanteffect than the acidified ethanol concentrationon the AA of the extract. This may result fromthe fact that A. puncticulatum contains organicacids as main components as described inthe following section. These compoundsdissolve easier in the hydroalcoholic solventand the extractability of these compoundsimproved with higher temperature. Finally,X

4 showed a significant effect on Y

3 of the

extracts (p<0.001). The Y3 values of the

extract increased with the increasing ofextraction time from 5 min to 12.20 min,

but beyond this level the antioxidant activityreached the plateau region (Figure 4C, E, F).

3.4 Validation of the OptimizedConditions

The optimum conditions for extractingthe target compounds from A. puncticulatumfruits by UAE was obtained at X

1, X

2, X

3

and X4 of 30 mL/g, 44.83%, 79.99°C, and

12.20 min, respectively. These conditionswere re-checked to verify the suitability ofthe obtained mathematical models and toensure that no bias toward the experimentalvalues using the slightly modified conditions:X

1 of 30 mL/g, X

2 of 45%, X

3 of 80°C and

X4

of 12 min. The experimental values ofY

1, Y

2 and Y

3 under such conditions were

44.89±1.03 mg GAE/g, 1.07±0.02 mgCy-3-glu/g, and 2.73±0.19 mM TE/g drypowder, respectively. These experimentalvalues matched well with the predictedvalues (p>0.05), indicating the proposedmathematical models were adequate forextracting the target compounds fromA. puncticulatum fruits.

With respect to the advantages of UAE,similar results were also reported forother plants concerning the improvement ofbioactive compounds extractability [22,23].Applying the UAE caused structuredchanges by destructing the cell walls of theplant material [23]. Thus, the UAE has a higherefficiency than the conventional extractionmethods, i.e., heat reflux extraction, concerningthe reduction of extraction time and energyconsumption.

3.5 Chemical Characterization of FreezeDried A. puncticulatum Extract3.5.1 Quantification of organic acids

The freeze dried samples of A.puncticulatum fruit extracts were purple incolor due to anthocyanin constituents. Organicacids were the major components in such

530 Chiang Mai J. Sci. 2016; 43(3)

extracts and their amounts were analyzed bya HPLC method. The plant extract containedamounts of L-(+)-tartaric acid, citric acid,L-malic acid, ascorbic acid, and oxalic acidof 71.95±0.49, 62.61±3.24, 15.24±1.92,1.05±0.24, and 0.60±0.02 mg/g freezedried extract, respectively.

3.5.2 Identification of phenoliccompounds

The phenolic compounds-containingplant extract was identified by both negativeand positive ionization modes usingHPLC-ESI/MS. Compound 1 (retentiontime, t

R=4.03 min) showing a [M-H]- parent

ion at m/z 191 was identified to be quinicacid. Compound 2 (t

R=13.20 min) exhibited

[M-H]- parent ion at m/z 593 with producinga fragment ion at m/z 285 was identified askaempferol-3-O-rutinoside. Compound 3(t

R=14.10 min) with a [M-H]- parent ion

at m/z 353 was identified to be 5-O-caffeoylquinic acid. Compound 4 (t

R =16.88

min) was identified as guaiacyl (8-5)ferulicacid hexoside. This is the first time thatglycosylated lignans are reported inA. puncticulatum fruit extract. Compound 4(t

R = 18.28 min) showed [M-H]- parent ion at

m/z 301. This ion further produced to formcharacteristic fragment ion at m/z 257 and229 whereas quercetin exhibited fragment ionat m/z 179 and 151. Thus, it was identified asellagic acid, not quercetin. These phenoliccompounds may play an important role incontributing to the antioxidant activity of theextracts.

3.5.3 Identification of anthocyaninsThe anthocyanins in the freeze dried

extract were identified using HPLC-ESI/MSwith a positive ionization mode. Twoanthocyanins were detected in the massspectrograms regarding the data offragmentation patterns, UV-VIS spectra,

sugar moieties and retention times (tR).

The first compound (tR =14.29 min) showed

a [M-H]+ parent ion at m/z 743 (Figure 5A).The MS2 secondary peak at m/z 597 indicatedthe loss of a rhamnose residue (M+-146)from a molecular ion. The MS1 parent ionproduced a MS2 base peak at m/z 303(delphinidin) due to fragmentation of adisaccharide residue (hexose + pentose moiety,162+132). The MS1 secondary peak at m/z465 also supported the sequential loss ofpentose residue (M+-132). Presence ofsambubioside in the chemical structure ofthis compound also produced a greaterpolarity. Thus, this compound was elutedbefore the corresponding aglycone whenseparating by HPLC. This result suggestedthat it was delphinidin-3-sambubioside-5-rhamnoside.

Figure 5. Tentative identification ofanthocyanins in the A. puncticulatum extract: Adelphinidin-3-sambubioside-5-rhamnosideand B delphinidin.

Chiang Mai J. Sci. 2016; 43(3) 531

Figure 6. Dose-response curves for thetyrosinase inhibition effects: A A. puncticulatumextract and B kojic acid. Lineweaver-Burksplots for the inhibition of tyrosinase bydifferent concentrations of A. puncticulatumextract (C).

The latter compound (tR =15.02 min)

showed [M-H]+ parent ion at m/z 303,followed by yielding MS2 and MS3 fragmentsat m/z 257 and 229, respectively (Figure 5B).Thus, this compound was identified asdelphinidin. This result was confirmed bycomparison with its UV-VIS spectra and masscharacteristics and data from literatures [24].However, other anthocyanins that havebeen reported elsewhere as components inwine-based products of this plant (cyanidin-3-sophoroside, delphinin-3-sambubioside,pelargonidin-3-malonyl glucoside) were notbe detected in this study [6]. This is probablydue to the different extraction process, typeof sample and plant variety.

3.6 Inhibitory Activity ofA. puncticulatum Extract on MushroomTyrosinase3.6.1 Effect of A. puncticulatum extracton mushroom tyrosinase

The dose-response curves for thetyrosinase inhibitory effects of theA. puncticulatum extract and kojic acid aredepicted in Figures 6A and B, respectively.The inhibitory diphenolase activities oftyrosinase increased with increase of extractconcentrations. Increasing concentrations ofplant extract from 0.1 to 2 mg/mL resultedin increased percentages of tyrosinaseinhibition from 7.00±3.24% to 88.36±1.42%,respectively. The IC

50 value of plant extract

was calculated to be 0.65±0.03 mg/mL.On the other hand, IC

50 value of kojic acid

was 0.029±0.00 mg/mL which was foundto be 22.41 times more effective thanA. puncticulatum extract. This result revealedthat A. puncticulatum extract had moderateinhibitory effects on mushroom tyrosinase.

3.6.2 Inhibition type of A. puncticulatumextract on mushroom tyrosinase

The type of inhibition of A. puncticulatumextract on mushroom tyrosinase during theoxidation of L-DOPA was studied usingLineweaver-Burk plot analysis. As shown in

532 Chiang Mai J. Sci. 2016; 43(3)

Figure 6C, increasing concentrations ofplant extract resulted in decreased values ofV

max whereas K

m remained constant. This

indicated that the test extract exhibitednon-competitive-type inhibition against thediphenolase activity of mushroom tyrosinase[25]. The apparent K

m and V

max values of

the solution mixture without plant extractwere calculated to be 16.73 mM and 5.41 Uwhereas V

max kinetic parameter decreased to

4.83 U in the presence of plant extract. Thus,the mechanism of action of this plant extractthat act as non-competitive inhibitor shouldbe through the allosteric effect by binding todifferent sites on the tyrosinase comparedwith L-DOPA.

4. CONCLUSIONS

The UAE were performed formaximizing the yield of total phenolics andtotal anthocyanins with good antioxidantactivity from A. puncticulatum ripe fruits. Alltest independent variables play an importantrole affecting the response values. The optimalextraction conditions to achieved maximumyields of such dependent variables wereliquid-to-solid ratio of 30 mL/g, acidifiedethanol concentration of 45% (v/v),extraction temperature of 80°C andextraction time of 12 min. Under theseconditions, the experimental yields of totalphenolics, total anthocyanins, and antioxidantactivity were close to the predicted values.Organic acids were characterized as themajor phenolic compounds in the extract.The mushroom tyrosinase inhibition kineticsshowed that the extract displayed a non-competitive inhibition. Thus, the UAE couldbe a technique for rapid and maximumextraction yields of bioactive compoundsfrom ripe fruits of A. puncticulatum.

ACKNOWLEDGEMENT

This study was partially supported by

post-doctoral research fellowship throughthe Ernst-Mach-Stipendien, Austria (grant no.ICM-2013-04533).

REFERENCES

[1] Gonz lez-Centeno M.R., Knoerzer K.,Sabarez H., Simal S., Rossell C. andFemenia A., Ultrason. Sonochem., 2014; 21:2176-2184.

[2] Yingngam B., Supaka N. andRungseevijitprapa W., J. Food Sci. Technol.,2015; 52: 129-140.

[3] Pukumpuang W., Chansakaow S. andTragoolpua Y., Chiang Mai J. Sci. 2014;41: 568-576.

[4] Yim H.S., Chye F.Y., Liow M.L. and HoC.W., Chiang Mai J. Sci., 2013; 40: 34-48.

[5] Dechayont B., Hansakul P. and Itharat A.,J. Med. Assoc. Thai., 2012; 95: S147-S153.

[6] Nuengchamnong N. and Ingkaninan K.,Food Chem., 2010; 118: 147-152.

[7] Puangpronpitag D., Yongvanit P.,Boonsiri P., Suttajit M., AreejitranusornP., Na H.K. and Surh Y.J., Food Chem.,2011; 127: 1450-1458.

[8] Lawag I.L., Aquinnaldo A.M.,Naheed S. and Mosihuzzaman M.,J. Ethnopharmacol., 2012; 144: 217-219.

[9] ahin S. and amly R., Ultrason. Sonochem.,2013; 20: 595-602.

[10] Barrera V zquez M.F., Comini L.R.,Martini R.E., N ez Montoya S.C.,Bottini S. and Cabrera J.L., Ultrason.Sonochem., 2014; 21: 478-484.

[11] Orphanides A., Goulas V. and Gekas V.,J. Food Eng., 2014; 120: 191-196.

[12] Prakash Maran J., Manikandan S., NivethaC.V. and Dinesh R., Arab. J. Chem., 2013.DOI 10.1016/j.arabjc.2013.02.007.

[13] Gong Y., Hou Z., Gao Y., Xue Y., Liu X.and Liu G., Food Bioprod. Process, 2012; 90:

Chiang Mai J. Sci. 2016; 43(3) 533

9-16.

[14] Chang C.T., Chang W.L., Hsu J.C.,Shih Y. and Chou S.T., Bot. Stud. 2013;54: 1-10.

[15] Chiou A., Panagopoulou E.A., GatzaliF., De Marchi S., Karathanos V.T.,Food Chem., 2014; 146: 157-165.

[16] Wang X., Wu Y., Chen G., Yue W., LiangQ. and Wu Q., Ultrason. Sonochem., 2013;20: 846-854.

[17] Prasad K.N., Hassan F.A., Yang B., KongK.W., Ramanan R.N., Azlan A. and IsmailA., Food Chem., 2011; 128: 1121-1127.

[18] Liu X., Mu T., Sun H., Zhang M. andChen J., Food Chem. 2013; 141: 3034-3041.

[19] Kaewamatawong R. and JoungmunkongZ., J. Sci. Technol. Univ. Ubon Ratchathani,2006; 8: 76-88.

[20] Hossain M.B., Barry-Ryan C., Martin-Diana A.B. and Brunton N.P., Food Chem.,2011; 126: 339-346.

[21] Liyana-Parthirana C. and Shahidi F.,Food Chem., 2005; 93: 47-56.

[22] Fan J.P., Cao J., Zhang X.H., Huang J.Z.,Kong T., Tong S., Tian Z.Y., Xie Y.L.,Xu R. and Zhu J.H., Food Chem., 2012;135: 2299-2306.

[23] Yingngam B., Monschein M. and BrantnerA., Asian Pac. J. Trop. Med., 2014; 7:S497-S505.

[24] Zheng J., Ding C., Wang L., Li G., Shi J.,Li H., Wang H. and Sou Y., Food Chem.,2011; 126: 859-865.

[25] Hu Y.H., Liu X., Jia Y.L., Guo Y.J., WangQ. and Chen Q.X., J. Biosci. Bioeng., 2014;117: 142-146.