Ultrasonic-extraction of phenolic compounds from Phyllanthus urinaria : optimization model and antioxidant activity

Antioxidants have the ability to suppress, delay, or prevent the degenerative or pathological process of human beings, such as aging, viral infections, coronary heart disease, cancer, allergies, inflammation, and Alzheimer’s disease (Christodouleas et al., 2015; Lin et al., 2008; Ngo et al., 2016), which resulted from the active oxygen free radical-mediated reactions. Antioxidants can also affect and determine the shelf-life, nutritional value, and quality of food products (Christodouleas et al., 2015). Recently, natural antioxidants have become a topic of increasing attention because they can have different uses such as dietary supplements, ingredients in functional foods or replacement of synthetic antioxidants (Arruda et al., 2017). Phenolic compounds from natural sources are an important class of antioxidants. The phenolic compounds have a variety of physiological activity such as antioxidant, antimutagenic, antiallergenic, antiinflammatory, and antimicrobial effects (Martins et al., 2011; Sequeda-Castañeda et al., 2016), and thus are now widely used in the pharmaceutical, nutraceutical and food fields.


Introduction
Antioxidants have the ability to suppress, delay, or prevent the degenerative or pathological process of human beings, such as aging, viral infections, coronary heart disease, cancer, allergies, inflammation, and Alzheimer's disease (Christodouleas et al., 2015;Lin et al., 2008;Ngo et al., 2016), which resulted from the active oxygen free radical-mediated reactions.Antioxidants can also affect and determine the shelf-life, nutritional value, and quality of food products (Christodouleas et al., 2015).Recently, natural antioxidants have become a topic of increasing attention because they can have different uses such as dietary supplements, ingredients in functional foods or replacement of synthetic antioxidants (Arruda et al., 2017).Phenolic compounds from natural sources are an important class of antioxidants.The phenolic compounds have a variety of physiological activity such as antioxidant, antimutagenic, antiallergenic, antiinflammatory, and antimicrobial effects (Martins et al., 2011;Sequeda-Castañeda et al., 2016), and thus are now widely used in the pharmaceutical, nutraceutical and food fields.
In recent years, researchers have focused on finding new effective natural antioxidants, especially from herbal sources.Phyllanthus urinaria belongs to the genus Phyllanthus (Euphorbiaceae) and is widely distributed in Southern America and Asian countries like China and India (Huang et al., 2006;Liu et al., 2015a;Lu et al., 2013).Phyllanthus urinaria has many biological and pharmacological activities including antioxidant, hepatoprotective antiinflammatory, antiallodynic and antioedematogenic, antiviral, and antibacterial (Liu et al., 2015a) and the phenolic compounds from Phyllanthus urinaria are mainly responsible for these activities.
Extraction is considering the first basic and important stage for separating bioactive compounds from raw materials (Salomon et al., 2014).Ultrasonic-assisted extraction (UAE) is one of the rapid and efficient techniques due to its strong cavitation effect and mechanical function (Liu et al., 2015b;Ma et al., 2016;Rocha et al., 2017).Response surface methodology (RSM) is an effective mathematical and statistical tool (Balyan & Sarkar, 2017;He et al., 2016) for modeling and optimization of the extraction parameters and for investigation of the interaction between the extraction parameters (Balyan & Sarkar, 2017;He et al., 2016).Thus, they are usually combined to extract bioactive substances from various natural materials (Al-Juhaimi et al., 2016;Chemat et al., 2017;Maran et al., 2017;Zhu et al., 2015).
In this work, UAE technique and RSM tool were used to optimize the extraction parameters (ethanol concentration, solid-liquid ratio and ultrasonic time) for the phenolic compounds from Phyllanthus urinaria.In addition, the assays of reducing power, DPPH free radicals, ABTS free radicals, and hydroxyl free radicals were performed for investigating the antioxidant capacity of the phenolic compounds.

Ultrasonic assisted extraction of phenolic compounds
The fresh Phyllanthus urinaria was dried in a vacuum oven (DZF-6020, Yiheng, China) at 80 °C for three days and milled into powders by a pulverizer (XS-10B, Longxin, China), and then passed through an 80-mesh sieve to obtain the plant powders (2.13% moisture).The powders (0.2000g) were placed in a 10-mL sealed glass bottle and mixed with ethanol solution.Then the bottle was placed in an ultrasonic bath (500 W, 53 kHz, SK8200H, Kedao, China) to extract the phenolic compounds for different ultrasonic time (10-50 min) at various ethanol concentrations (30-70%) in different solid-liquid ratios (1I:30-1I:70 g/mL).The bath was fixed at 500 W and 53 kHz.

Determination of total phenolic extraction yield
After extraction, the suspension from the extracted bottle was centrifuged by a high-speed centrifuge (H1650-W, Xiangyi, China) at 12000 rpm for 15 min.The phenolic weight in the supernatant was measured by Folin-Ciocalteu method (Sharmila et al., 2016).2.5 mL of tenfold diluted Folin-Ciocalteu reagent, 2 mL of aqueous sodium carbonate (75 g/L) and 0.5 mL of the diluted supernatant were mixed and reacted in a water bath at 50 °C for 5 min, and then the mixture was cooled to room temperature to determine the absorbance at 760 nm by a UV-vis spectrophotometer (UVmini-1240, Shimadzu, Japan).The phenolic concentration was calculated based on the standard curve of gallic acid standard solutions (0-0.5 mg GAE/L) and the phenolic weight was obtained by multiplying the phenolic concentration by volumetric flask volume (50mL).The phenolic extraction yield is calculated as followsI:

Optimization design
Based on the single-factor-test results and the design principle of Box-Behnken (BBD), the extraction parameters were optimized by response surface methodology with ultrasonic time (X 1 ), ethanol concentration (X 2 ) and solid-liquid ratio (X 3 ) as the independent variables and extraction yield (Y) calculated by Equation 1 as the response.The independent variable levels were given in Table 1 and the design experiment results were provided in Table 2.

DPPH radical scavenging ability
The DPPH radical scavenging activity of the phenolic compounds was measured according to Liu (Liu et al., 2015b). 1 mL of 40 μg/mL DPPH solution was added to 3 mL of sample solution and mixed.The mixture was shaken and left in the dark for 30 minutes at room temperature, and the mixture absorbance was determined at 517 nm.The following equation is used to calculate the DPPH radical scavenging effectI: where A 0 is the control solution absorbance without sample (deionized water instead of sample), A 1 is the mixture solution absorbance with sample, and A 2 is the mixture solution absorbance without DPPH (deionized water instead of DPPH).

Hydroxyl radical scavenging ability
The hydroxyl radical scavenging ability of the phenolic compounds was determined according to Liu (Liu et al., 2015b).1.0 mL of 1.5 mmol/L FeSD 4 solution, 0.3 mL of 20 mmol/L sodium salicylate solution, 1.0 mL of sample solution and 0.7 mL of 6 mmol/L H 2 D 2 solution were mixed and reacted for 60 min at 37 °C.Then the mixture absorbance was determined at 520 nm.The following equation is used to calculate the hydroxyl radical scavenging effectI: where A 0 is the control solution absorbance without sample (deionized water instead of sample), A 1 is the mixture solution absorbance with sample, and A 2 is the mixture solution absorbance without sodium salicylate (deionized water instead of sodium salicylate).

ABTS radical scavenging ability
The ABTS radical scavenging ability of the phenolic compounds was determined according to Ye (Ye et al., 2016).7.0 mmol/L ABTS solution and 2.45 mmol/L K 2 S 2 D 8 solution were mixed and reacted in the dark for 12 h at room temperature to prepare ABTS radical solution.The reaction solution was diluted using phosphate buffer solution (pH 7.4, 10 mmol/L) to obtain the absorbance at 734 nm about 0.70 ± 0.02.Then 3.9 mL of the diluted ABTS solution and 0.1 mL of sample solution were mixed and reacted in the dark for 5 min at room temperature to determine the absorbance at 734 nm.The following equation is used to calculate the ABTS radical scavenging effectI: where A 0 is the control solution absorbance without sample (deionized water instead of sample), A 1 is the mixture solution absorbance with sample, and A 2 is the mixture solution absorbance without ABTS (deionized water instead of ABTS).

Reducing power
The reducing power of the phenolic compounds was measured according to Liu (Liu et al., 2015b).2.5 mL of 0.2 mol/L phosphate buffer solution (pH 6.6), 2.5 mL of 1% potassium ferricyanide solution and 1.0 mL of sample solution was mixed and reacted for 20 min at 50 °C.Then 2.5 mL of 10% trichloro acetic acid was added to the mixture and centrifuged at 3000 rpm for 10 min.2.5 mL of the supernatant, 0.5 mL of 0.1% ferric chloride solution and 2.5 mL of distilled water was mixed to react for 5 min, and the mixture absorbance was determined at 700 nm.The following equation is used to calculate the reducing powerI: where A 1 is the mixture solution absorbance with sample, and A 2 is the mixture solution absorbance without FeCl 3 (deionized water instead of FeCl 3 ).

Statistical analysis
All the experiment results were averaged by triplicate.Dne-way ANDVA and Duncan's test were used to determine the significant differences of single-factor levels by SPSS software 20.0.Design Expert software 8.0 was used for the regression analysis and the parameter optimization.Statistical significance was defined at p < 0.05.IC 50 was calculated by Drigin 8.0.The multiple regression analysis for the data (Table 2) was performed by Design-Expert software.The model for the extraction yield (Y) was produced by the quadratic polynomial regression equation of ultrasonic time (X 1 ), ethanol concentration (X 2 ) and solid-liquid ratio (X 3 ).

Effect of ultrasonic time
The effect of ultrasonic time (10-50 min) on the extraction yield is given in Figure 1A with solid-liquid ratio and ethanol concentration fixed at 1I:40 g/mL and 60%, respectively.As the ultrasonic time increases, the extraction yield increases initially and then decreases when a maximum yield at 30 min is achieved.Ultrasound with mechanical agitation effect, cavitation effect, and thermal effect can improve the mass transport and facilitate the release of compounds from the extracted materials (Carrera et al., 2012;Tomšik et al., 2016).However, the long ultrasonic time can degrade the phenolic compounds, leading to the decrease of the extraction yield (Carrera et al., 2012;Ddabas & Koca, 2016).Therefore, the ultrasonic time of 30 min is sensible.

Effect of ethanol concentration
The effect of ethanol concentration (30-70%) on the extraction yield is shown in Figure 1B, when solid-liquid ratio and ultrasonic time are kept at 1I:40 g/mL and 40 min, respectively.The extraction yield increases as ethanol concentration increases but decreases when it reaches a peak at 50%.This is related to the extraction solution polarity.The different ethanol concentrations make the extraction solutions have different polarities (Spigno & Faveri, 2009).The polarity of 50% ethanol solution may be similar to that of the phenolic compounds of Phyllanthus urinaria, resulting in high extraction yield.When the ethanol concentration exceeds 50%, the extraction solution polarity decreases and other substances compete to be extracted from Phyllanthus urinaria, which results in a reduction in extraction yield.Thus, the ethanol concentration is around 50%.

Effect of solid-liquid ratio
The effect of solid-liquid ratio (1I: 30-1I: 70 g/mL) on the extraction yield is shown in Figure 1C with ethanol concentration at 60% and ultrasonic time at 40 min.The extraction yield increases with the solid-liquid ratio increases until a maximum yield reaches at 1I:60 g/mL.The increase of the solid-liquid ratio improves the mass concentration difference between inside and outside the cells, which results in the increase in the mass transport driving force and the internal diffusion rate (Şahin & Şamlı, 2013).The higher solid-liquid ratio means the more phenolic compound release from Phyllanthus urinaria (Prasad et al., 2012), but simultaneously the more cavitation energy from the extraction system is absorbed, resulting in the lower cavitation energy density per unit volume for extraction (Xu et al., 2016).As a result, the solid-liquid ratio of 1I:60 g/mL is suitable.

Model fitting
The equation expressed in the form of coded values is as followsI: 9.53 0.17 -0.12 0.20 -0.022 0.24 0.11 -0.32 -0.14 -0.40 where Y is the dependent variable for extraction yield; X 1 , X 2 , X 3 are the independent variables for ultrasonic time, ethanol concentration and solid-liquid ratio, respectively.
Analysis of variance (ANDVA) is summarized and given in Table 3.As shown in Table 3, the determination coefficient (R 2 ) is 0.9865 which indicates that the model is a reasonable fit for the experimental data (Deng et al., 2016;Han et al., 2016) and the adjusted determination coefficient (R2 adj) is 0.9865 which means the experimental results are highly consistent with the predicted values (Han et al., 2016;Yan et al., 2016).Moreover, the model p-value is 0.0004, which indicates that the regression model is highly significant.The p-value (0.2777) of lack of fit is not significant, which exhibits the fitting model is good for interpreting the data.Simultaneously, the coefficient of variation (C.V. = 0.80%) is relatively low, indicating the better reliability of the experimental values.
It can conclude from the p-value presented in Table 3 that the linear coefficients (X 1 , X 2 and X 3 ), the quadratic term coefficients (X2 1, X2 2 and X2 3), and the interaction term coefficients (X 1 X 3 and X 2 X 3 ) are significant for the p-value < 0.05.These results illustrate that the effect of the three variables on the extraction yield is significant and the effect is non-linear and that the interactions between X 1 X 3 and X 2 X 3 also have a significant influence on the extraction yield.According to the p-value, the order of variables influencing the extraction yield is as followsI: solid-liquid ratio (X 3 ) > ultrasonic time (X 1 ) > ethanol concentration (X 2 ).

Response surface analysis
3D response surface plots are provided as a graphical interpretation for the relationship between independent and dependent variables (Xie et al., 2015).The 3D response surfaces and 2D contours were plotted in Figure 2. From Figures 2A and 2B, as the ultrasonic time (X 1 ) increases of from 20 to 33.96 min, the extraction yield (Y) increases slowly but decreases gradually after 33.96 min.From Figures 2A and 2C, the extraction yield (Y) increases rapidly as the increasing of ethanol concentration (X 2 ) from 40 to 46.48% and reaches a maximum at 46.48%, and then decreases rapidly.Figures 2B and 2D show that the extraction yield (Y) increases quickly at first and then goes down quickly with increasing of solid-liquid ratio (X 3 ), and a peak at 1I:63.23 g/mL is obtained.
Additionally, as shown in Figure 2, all the three response surfaces are convex, indicating the levels of selected parameters are reasonable.By observing the slope variation of the response surfaces and the sparse degree of the contours, the effect significance of the variables on the extraction yield (Y) is in the order of solid-liquid ratio (X 3 ), ultrasonic time (X 1 ) and ethanol concentration (X 2 ).Dbserving the contour plots in Figure 2, it also can conclude that the interactions between ultrasonic time (X 1 ) and solid-liquid ratio (X 3 ), and ethanol concentration (X 2 ) and solid-liquid ratio (X 3 ) is significant.

Predictive model verification
The optimum extraction parameters obtained by Design-Expert software are as followsI: ultrasonic time (X 1 ) of 33.96 min, ethanol concentration (X 2 ) of 46.48%, and solid-liquid ratio (X 3 ) of 1I: 63.23 g/mL.Considering the practical operation convenience, the optimum extraction parameters are adjusted as followsI: ultrasonic time of (X 1 ) of 34 min, ethanol concentration (X 2 ) of 47%, and solid-liquid ratio (X 3 ) of 1I: 63 g/mL.Under these conditions, the extraction yield (9.59%) is close to the predicted value (9.62%), which indicates that the model is reliable and effective for extracting the phenolic compounds from Phyllanthus urinaria.

Antioxidant activity analysis
DPPH radical scavenging ability DPPH radicals are widely used to investigate the radical scavenging activity of various samples for the good stability (Gülçin et al., 2004).The DPPH radical scavenging effect of the phenolic compounds and positive control vitamin C calculated by Equation 2 is shown in Figure 3A.The phenolic compounds and vitamin C exhibits a concentration-dependent manner.The half maximal inhibitory concentration (IC 50 ) of the phenolic compounds is 33.60 μg/mL, which is close to the reported results (28.16 μg/mL) by Mediani (Mediani et al., 2015).IC 50 of vitamin C is 28.43 μg/mL.These results show the phenolic compounds possess a good capacity to scavenge DPPH radicals.

Hydroxyl radical scavenging ability
Hydroxyl radicals are very active and can severely induce tissue damage or cell death (Ge et al., 2014).Therefore, it is of great importance to scavenge the hydroxyl radicals.The phenolic compounds and vitamin C display evident scavenging activity (calculated by Equation 3) against hydroxyl radicals in a concentration-dependent manner (Figure 3B).The IC 50 of the phenolic compounds and vitamin C is 114.34 μg/mL and 92.06 μg/mL, respectively, which exhibits that the phenolic compounds have a fine capacity to scavenge hydroxyl radicals.

ABTS radical scavenging ability
ABTS assay is a decolorization assay applicable to both lipophilic and hydrophilic antioxidants at different pH levels (Raza et al., 2017).The ABTS radical scavenging effect of the phenolic compounds and vitamin C calculated by Equation 4 is displayed in Figure 3C.The scavenging effect of all the samples increases linearly as the concentration increases.The IC 50 of the phenolic compounds and vitamin C is 53.98 μg/mL and 45.42 μg/mL, respectively, which indicates that the phenolic compounds have a significant ability to scavenge ABTS radicals.

Reducing power
The reducing power of a sample can be used as an important index of antioxidant activity (Zheng et al., 2014).The reducing power of the phenolic compounds and vitamin C calculated by Equation 5 is correlated well with the increasing concentration and shows a linear relationship (Figure 3D).In the concentration range of 20-120 μg/mL, the reducing power of the phenolic compounds increases from 0.185 to 0.790, while that of vitamin C is from 0.247 to 1.090.The slope of the phenolic compounds and vitamin C obtained by linear regression is 0.0061 and 0.0084, respectively.These results show the reducing power of the phenolic compounds is equivalent to 76.62% of the reducing power of vitamin C. Therefore, the phenolic compounds have a good reducing power.

Figure 1 .
Figure 1.Effect of different extraction variables on extraction yield.(A) ultrasonic time; (B) ethanol concentration; and (C) solid-liquid ratio.

Figure 2 .
Figure2.Response surface and contour plots showing effect of (A) X 1 and X 2 ; (B) X 1 and X 3 ; (C) X 2 and X 3 .

Figure 3 .
Figure 3. DPPH radical (A), hydroxyl radical scavenging effect (B), ABTS radical (C) and reducing power (D) scavenging effect of phenolic compounds as a function of concentration.

Table 2 .
Box-Behnken design for independent variables and their extraction yield.

Table 1 .
Independent variables and their levels for Box-Behnken design.