Efficient Approach for the Extraction and Identification of Red Pigment from Zanthoxylum bungeanum Maxim and Its Antioxidant Activity

Red pigment (RP) was extracted from the peels of Zanthoxylum bungeanum Maxim (PZB) by ultrasonic-assisted extraction (UAE) in this work. Box–Behnken design–response surface methodology (BBD-RSM) was employed to research the efficiency of the RP extraction. Based on the optimization of RSM, results showed that the optimal extraction conditions were as follows: liquid–solid ratio of 31 mL/g, extraction time of 41 min, and extraction temperature of 27 °C, and under these conditions, the actual absorbance value was 0.615 ± 0.13%, highly agreeing with the predicted value by the model. Furthermore, ultra-performance liquid chromatography–mass spectrometry (UPLC-MS) was used to separate and analyze RP. The compounds of RP were mainly flavonoids, and there were five compounds detected for the first time in PZB. In addition, RP showed significant antioxidant activities in vitro, which could be developed for anti-aging candidate drugs and functional foods. In conclusion, ultrasound-assisted extraction with BBD-RSM and chromatographic separation technology with UPLC-MS are efficient strategies for the isolation and identification of RP from PZB.


Introduction
Zanthoxylum bungeanum Maxim (Rutaceae), a known traditional spice in China, is commonly applied in cooking for its special flavor and numbing taste. Z. bungeanum has also been used as a common traditional Chinese medicine (TCM) for treating various diseases for decades. Z. bungeanum is widely distributed in China, including Sichuan, Shanxi, Gansu, Shandong, Hebei, etc. [1,2]. Increasing evidence has demonstrated that Z. bungeanum shows diversified pharmacological activities, such as insecticidal, antifungal, antioxidant, anti-hypertension, and anti-inflammatory effects, etc. [3][4][5][6][7]. In addition, previous investigations also reported that Z. bungeanum contains many compounds, including alkaloids, terpenoids, flavonoids, free fatty acids, etc. [8]. Furthermore, amides (the characteristic alkaloids in this plant), essential oil, and flavonoids are reported to be the major active constituents of Z. bungeanum [4,9]. absorption wavelength of RP, the absorbance value of RP at 520 nm manifested the extraction efficiency of RP and was used as the response value for the experimental design.

Effects of Solvents on RP Extraction
The effects of solvents on RP extraction are shown in Table 1. In this study, the extraction efficiency of different extraction solvents was investigated with the other factors set at the central point. It was found that the extraction efficiency of 0.1% (v/v) hydrochloric acid in methanol was the best according to the color and absorbance value of RP. Further, it is advantageous to extract anthocyanin pigments in acidic solvents such as hydrochloric acid, as it can keep the pigment stable and reduce its degradation [28]. Hence, 0.1% (v/v) hydrochloric acid in methanol was chosen as the optimum extraction solvent of RP and used for further experiments.

Effects of Liquid-Solid Ratio on RP Extraction
The effects of liquid-solid ratio (15,20,25,30,35 mL/g) on the extraction efficiency of RP were investigated, and the results are displayed in Figure 2A. Meanwhile, other parameters were fixed as follows: the extraction solvent was 0.1% (v/v) hydrochloric acid in methanol, the extraction time 30 min, and temperature 25 • C. The absorbance value of RP obviously increased before the liquid-solid ratio reached 30 mL/g, but after that, the absorbance value of RP changed slightly. The results showed that high liquid-solid ratios would lead to the waste of extraction solvent instead of improving the extraction efficiency significantly. Based on our results above, the liquid-solid ratios of 25-35 mL/g were selected for further experiments.

Effects of Extraction Time on RP Extraction
The extraction times of 10,20,30,40, and 50 min were studied to analyze the influence of extraction time on the extraction efficiency of RP while the other parameters were kept constant (the extraction solvent was 0.1% (v/v) hydrochloric acid in methanol, liquid-solid ratio 30 mL/g, and extraction temperature 25 • C), and the results are reported in Figure 2B. With the increase of extraction time, the absorbance value of RP increased at first and decreased afterwards, reaching a maximum at 40 min. This finding indicated that the extraction efficiency began to keep a dynamic equilibrium with the increase of extraction time. However, overlong extraction time would cause the degradation of RP [29]. Consequently, an extraction time of 30-50 min was used in the RSM experiments.

Effect of Extraction Temperature on RP Extraction
The effects of extraction temperatures 15, 20, 25, 30, and 35 • C are reported in Figure 2C for the liquid-solid ratio of 30 mL/g and the extraction time of 30 min. The extraction efficiency of RP increased significantly when temperature changed from 15 • C to 25 • C, and reached a maximum level at 25 • C. This phenomenon indicated that a higher temperature initially increased the dissolution of RP, but when the temperature continued to rise, some heat labile components in RP were destroyed. Usually, the extraction temperature of anthocyanins should be controlled under approximately 30 • C [30]. Therefore, 20-30 • C was considered to be the extraction temperature range of the RSM experiments.

Model Fitting Analysis
On the basis of the previous single-factor experiments, liquid-solid ratio (X 1 ), extraction time (X 2 ), and extraction temperature (X 3 ) were chosen as the three factors of the response surface methodology (the extraction solvent was 0.1% (v/v) hydrochloric acid in methanol). As illustrated in Table 2, seventeen experiments were designed with BBD. The actual absorbance value of RP varied from 0.460 to 0.620, reaching a maximum value at a liquid-solid ratio of 30 mL/g, extraction time of 40 min, and extraction temperature of 25 • C. The predicted absorbance value of RP was gained from the second-order polynomial Equation (1): (1) According to the results in Table 2, the analysis of variance (ANOVA) conducted to analyze the suitability and adequacy of the model is shown in Table 3. The regression models were tremendously significant from the perspective of high F-value (33.97) and low p-value (<0.0001). The lack of fit (F-value = 1.70, p-value = 0.3041 > 0.05) was not significant and indicated that the model could be applied to predict the absorbance value of RP. Both determination coefficient (R 2 = 0.9776) and adjusted determination coefficient (Adj R 2 = 0.9488) were significant and high, which showed that the predicted values were correlated with the experimental values. The low coefficient of variation (C.V.) (2.33%) and high Adeq precision (15.210) indicated that there was relatively high precision and reliability among the experimental values. In addition, the significance of each coefficient can be evaluated from the F-value (positively correlated with significance) and p-values (negatively correlated with significance). When the p-value of the model was less than 0.05, the model was significant and suited to optimize the extraction parameters. The independent variables (X 1 , X 2 , and X 3 ) and the quadratic terms (X 1 2 , X 2 2 , and X 3 2 ) were significant. However, the other coefficients (X 1 X 2 , X 1 X 3 , X 2 X 3 ) were not significant.

Response Surface Analysis
As shown in Figure 3, a 3D response surface and 2D contour plots were used to analyze the interaction effects of three variables on the absorbance values of the RP on the basis of the regression equation. When the shape of contour plot is elliptic instead of round, the interaction between variables is significant. Otherwise, the interaction is insignificant. Besides, when the shape of the response surface is convex, the range of variables was set reasonably [31]. The effects of liquid-solid ratio, extraction time, and their interaction on the absorbance value of RP are revealed in Figure 3A,B. The RP absorbance value increased when the liquid-solid ratio increased in the range of 25-31.35 mL/g, and the extraction time increased in the range of 30-41.05 min, and then the absorbance value declined after 31.35 mL/g and 41.05 min. The circular contour shapes showed that the interaction of the liquid-solid ratio and extraction time was insignificant, which agrees with the results in Table 3. The effects of liquid-solid ratio and extraction temperature on the absorbance value of RP are shown in Figure 3C,D. It can be found that when the liquid-solid ratio was 30.99 mL/g and extraction temperature was 27.71 • C, the absorbance value of RP reached a maximum (approximately 0.618). The shape of the contour plot was circular, which indicates that the interactions of the two variables were not significant. As reported in Figure 3E,F, the effects of extraction time and extraction temperature on the absorbance of RP were evaluated. A maximum absorbance value of RP was obtained when the extraction time and extraction temperature was approximately 40.68 min and 27.85 • C. Additionally, the circular contour plots showed that the interaction between extraction time and extraction temperature was insignificant.

Verification of Extraction Conditions
The feasibility of the model for calculating the optimum response values were tested with the optimum conditions as follows: the liquid-solid ratio was 31.14 mL/g, extraction time was 40.93 min, and the extraction temperature was 27.44 • C. Validation tests (n = 3) were executed under the aforesaid conditions with slight modifications: the liquid-solid ratio was 31 mL/g, the extraction time was 41 min, and the extraction temperature was 27 • C. Under these conditions, the actual absorbance of RP obtained was 0.615 ± 0.13%, highly matching with the predicted value and consequently indicating that the RSM model is accurate and adequate.

UPLC-MS Analysis
Chemical constituents of the RP from PZB were separated and identified using UPLC-MS at negative and positive modes, respectively, and the total ion chromatograms are presented in Figure 4. From our results, twenty-one constituents were detected, including phenylpropanoids, anthocyanins, and flavonoids. The related information is reported in Table 4, and the chemical structure is shown in Figure 4. Among them, five constituents (2, 5, 7, 8, 18) were identified for the first time in PZB. The compounds analysis was established on the MS, MS 2 spectral data and the mass spectrometric data interpretation reported in the literature.
As shown in Table 4, [34]. Epicatechin belongs to catechins mainly sourced from tea polyphenols which has various bioactivities, such as antioxidant, anti-inflammatory, and antidiabetic effects, and so on [35].
There were two anthocyanins detected in the experiment. Seven quercetin glycosides were identified as displayed in Table 4. Compound 9 and compound 10 manifested molecular ions at m/z 773.33 and m/z 627.25, both of which had obvious product ions at m/z 303 and matched with quercetin, revealing that compounds 9 and 10 were quercetin glycosides. However, there was no more fragmentation information to confirm sugars existing in the above quercetin glycosides. According to the displayed fragment information, compound 9 might be quercitrin triglycoside, and compound 10 was likely to be quercitrin bioside. The former lost 470 µ, possibly corresponding to two hexoses and one deoxyhexose. The latter lost 324 µ, probably conforming to two hexoses. The precursor ions at m/z 609.34, 463.28, 463.33, and 433.25 all generated product ions at m/z 300, which corresponds to a quercetin moiety. Combined with the reported components of PZB in the literature, compound 13 losing rutinose was identified as rutin.    1, 6, 11, 12, 16, 19, and 20) have not yet been identified. There were no mass spectral data in the databases or reported literature to characterize the above compounds. Further mass spectrometry analysis and NMR identification might be employed to determine these compounds.

Antioxidant Activity of RP
2.5.1. 1,1-Diphenyl-2-picrylhydrazyl (DPPH)· Radical Scavenging Capacity of RP The DPPH· radical scavenging capacity assay is widely applied to evaluating the antioxidant activity of candidate agents, and is relatively convenient and rapid compared with other methods [38]. As Figure 5A indicates, in the concentration range of 0.02-0.5 mg/mL, the DPPH· scavenging rate of RP ranged from 30.45% to 90.50%, and the EC 50 (efficient concentration to scavenge 50% radicals) of RP was 60 µg/mL. Therefore, our results revealed that RP has a potential scavenging capacity for DPPH· radicals.

2,2'-Azinobis-(3-ethylbenzthiazoline-6-sulphonate)(ABTS)· + Radical Scavenging Capacity of RP
The ABTS· + radical scavenging assay is another common method used to evaluate the antioxidant activity of chemical components [39]. As shown in Figure 5B, the ABTS· + radical scavenging capacity of RP varied from 29.04% to 93.20% at the concentration of 0.02-0.5 mg/mL with an EC 50 value of 42 µg/mL. The results showed that RP had a prominent ABTS· + radical scavenging potential.

Hydroxyl Radical Scavenging Capacity of RP
The hydroxyl radical scavenging assay is also adopted frequently for antioxidant activity evaluation [40]. As shown in Figure 5C, the hydroxyl radical scavenging capacity of RP ranged from 10.16% to 56.11% at the concentration of 0.02-0.5 mg/mL with an EC 50 value of 268 µg/mL. The results indicated that RP had hydroxyl radical scavenging potential. However, the hydroxyl radical scavenging capacity of RP was poorer than the radical scavenging capacity of DPPH· and ABTS· + .
Aging is a complicated physiological process, and usually causes cognitive dysfunctions, memory loss, schizophrenia, Alzheimer's and Parkinson's diseases, etc. The biological mechanisms for explaining the aging process are not clear yet, and the free radical theory is one of the most approved theories [41]. Hence, antioxidant substances may be advantageous for delaying aging and extending the life span. Collectively, the results discussed above suggest that RP could be beneficial for scavenging body radicals, which could be further developed for anti-aging candidate drugs and foods.

Ultrasound-Assisted Extraction (UAE)
In this study, PZB was pretreated with ethyl acetate to remove impurities and then air-dried. PZBs (2.0 g) were put into an Erlenmeyer flask with some extraction solution. Then, the Erlenmeyer flask with PZB and extraction solution was put in an ultrasonic extraction machine (Tianjin Autoscience Instrument Co., Ltd., Tianjin, China) with a power of 300 W and a frequency of 40 kHz and extracted at an appropriate extraction time and temperature. After extraction, the extract mixture was separated by filtering with a funnel, and the solution concentrated to 50 mL by rotary evaporator (Yarong Biochemical Instrument, Shanghai, China) under reduced pressure at 45 • C. Thus, the original pigment extraction for absorbance detection was obtained. In addition, the original pigment extraction was concentrated further to remove solvents by rotary evaporator, and an appropriate volume of ethyl acetate was put into the concentrates to get rid of amides, essential oil, and other compositions further. Then, primary purified RP was collected for the subsequent determination of antioxidant activity and identification.

Spectral Characteristics Analysis
To obtain the maximum absorption wavelength of RP, the spectral characteristics of the above original pigment extraction was detected by UV-Vis spectrophotometer (Purkinje General Instrument Co., Ltd., Beijing, China) in the range of 400 to 800 nm. Then, the absorbance of the original pigment extraction was determined at the maximum absorption wavelength to evaluate the extraction efficiency of RP according to the method previously reported by Liu et al. with some modifications [42].

Optimization of Extraction Technology by BBD-RSM
Based on the preliminary single-factor experiments, the liquid-solid ratio, extraction time, and extraction temperature were selected to be the three factors of the RSM design with the optimal solvents (0.1% (v/v) hydrochloric acid in methanol). A Box-Behnken design (BBD) of three factors at three levels was set to perform RSM by using Design Expert software (version 8.0.6.1, Stat-Ease Inc., Minneapolis, MN, USA). Three major influence factors covered liquid-solid ratio (X 1 , mL/g), extraction time (X 2 , min), and extraction temperature (X 3 , • C), and three levels coded −1, 0, and +1 were respectively built based on these factors (Table 5).

Separation and Identification of RP by UPLC-MS
In the research, UPLC-MS was applied to the identification of RP. The analysis was operated on an Acquity UPLC-Quattro Premier XE system (Waters, Massachusetts, USA). The sample separation was performed applying a UPLC HSS C18 analytical column (2.1 mm × 100 mm, 1.7 µm) (Waters, Massachusetts, USA) at 45 • C, the flow rate was 0.25 mL/min, and the injection volume was 2 µL. The mobile phase included methanol (solvent A) and 0.1% (v/v) formic acid in water (solvent B). The elution gradient started with 10-35% A to 15 min, 35-45% A to 40 min (end analysis). Mass spectra parameters were set as follows: negative and positive ion mode for a mass range of m/z 180 to m/z 800, capillary voltage 2.8 kV, cone voltage 20 V, source temperature 100 • C, desolvation temperature 250 • C, cone gas flow 40 L/h, desolvation gas flow 500 L/h.

DPPH· Radical Scavenging Assay
The DPPH· radical scavenging activity was analyzed according to a method reported with some amendments [43]. A series of concentrations of RP (0.02-0.5 mg/mL) were prepared, and DPPH· solution (1 × 10 −4 mol/L, 2 mL, in ethanol) was mixed with the sample solutions (2 mL), respectively. The mixture was kept in the dark at room temperature for 30 min, the absorbance was then measured at 517 nm, and ascorbic acid served as the positive control. The DPPH· free radical scavenging capability was calculated as the following Equation (2): where A 0 is the absorbance of methanol plus DPPH· solution. A 1 is the absorbance of samples plus DPPH·. A 2 is the absorbance of samples plus ethanol.

ABTS· + Radical Scavenging Activity
The ABTS· + radical scavenging capacity was evaluated on the basis of a previous report with some modifications [44]. A series of concentrations of RP (0.02-0.5 mg/mL) were prepared, ABTS· + was generated after the chemical reaction between potassium persulphate solution (2.5 mM) and ABTS solution (7 mM) in the dark for 12-16 h at room temperature. The prepared ABTS· + solution was diluted with purified water to the absorbance of 0.70 ± 0.02 at 734 nm. Then, the diluted ABTS· + solution (3.5 mL) was added into sample solution (0.5 mL) and interacted for 5 min at room temperature. The absorbance was measured at 734 nm and the ascorbic acid served as a positive control. The ABTS· + radical scavenging capacity was evaluated according to Equation (3): where A 0 is the absorbance of methanol plus ABTS· + solution. A 1 is the absorbance of samples plus ABTS· + . A 2 is the absorbance of samples plus purified water.

Hydroxyl Radical Scavenging Activity
The hydroxyl radical scavenging activity was evaluated based on the report described previously with some amendments [45]. Hydroxyl radicals, generated from the chemical reaction between FeSO 4 and H 2 O 2 , have the ability to hydroxylate salicylate and the absorbance can be measured at 510 nm. A series of concentrations of RP (0.02-0.5 mg/mL) were prepared. RP solution (1 mL) was mixed with FeSO 4 solution (6 mM, 1 mL), H 2 O 2 solution (6 mM, 1 mL), and salicylic acid solution (6 mM, 1 mL, in ethanol), and reacted at 37 • C for 30 min. The absorbance was measured at 510 nm and ascorbic acid was used as the positive control. The hydroxyl radical scavenging activity was calculated as follows: Hydroxyl free radical scavenging capability(%) = (1 − where A 0 is the absorbance of methanol plus FeSO 4 , H 2 O 2 , and salicylic acid solution. A 1 is the absorbance of samples plus FeSO 4 , H 2 O 2 , and salicylic acid solution. A 2 is the absorbance of samples plus FeSO 4 , H 2 O 2 , and ethanol.

Statistical Analysis
Three parallel tests were carried out in each experiment, and results were represented as means ± S.D. Analysis of variance (ANOVA) was utilized to analyze the data, and p < 0.05 was deemed statistically significant.

Conclusions
In this research, the RP from PZB was extracted efficiently by utilizing an ultrasound-assisted method in combination with BBD-RSM. The optimal conditions were as follows: liquid-solid ratio of 31 mL/g, extraction time of 41 min, and extraction temperature of 27 • C. The actual absorbance value was 0.615 ± 0.13%, highly agreeing with the value predicted by the model.
In addition, the chromatographic separation and characterization of the RP from PZB were investigated tentatively for the first time in the present study. The RP from PZB was separated and identified by UPLC-MS. The results indicated that compounds of RP from PZB including phenylpropanoids, anthocyanins and flavonoids were detected, and quercetin glycosides were the major compounds. Neochlorogenic acid, p-coumaroylquinic acid, cyanidin-3-glucoside, cyanidin-3-rutinoside, and quercetin-3-rhamnoside were detected from PZB for the first time.
Furthermore, the in vitro antioxidant activity suggested that the RP of PZB has significant free radical scavenging capacity and could be developed for anti-aging drugs and functional foods. In conclusion, this research provides scientific data for the highly-efficient extraction and identification of RP from PZB.