Optimization of the Extraction of Anthocyanins from the Fruit Skin of Rhodomyrtus tomentosa (Ait.) Hassk and Identification of Anthocyanins in the Extract Using High-Performance Liquid Chromatography-Electrospray Ionization-Mass Spectrometry (HPLC-ESI-MS)

Anthocyanins are naturally occurring polyphenols that impart bright color to fruits, vegetables and plants. In this study, the extraction of anthocyanins from freeze-dried fruit skin of downy rose-myrtle (Rhodomyrtus tomentosa (Ait.) Hassk var. Gangren) was optimized using response surface methodology (RSM). Using 60% ethanol containing 0.1% (v/v) hydrochloric acid as extraction solvent, the optimal conditions for maximum yields of anthocyanin (4.358 ± 0.045 mg/g) were 15.7:1 (v/w) liquid to solid ratio, 64.38 °C with a 116.88 min extraction time. The results showed good fits with the proposed model for the anthocyanin extraction (R2 = 0.9944). Furthermore, the results of high-performance liquid chromatography-electrospray ionization-mass spectrometry (HPLC-ESI-MS) analysis of the anthocyanins extracted from the fruit skin of downy rose-myrtle revealed the presence of five anthocyanin components, which were tentatively identified as delphinidin-3-glucoside, cyanidin-3-glucoside, peonidin-3-glucoside, petunidin-3-glucoside and malvidin-3-glucoside.


Introduction
Anthocyanins are the principal water-soluble pigments responsible for the red, blue, and purple colors of terrestrial plants. Anthocyanins are omnipresent in our plant-based diet, have little or no known toxicity [1], making them particularly attractive as natural substitutes for synthetic pigments and antioxidants [2]. In addition, an increasing number of studies have demonstrated that anthocyanins have the ability to prevent chronic and degenerative diseases including type 2 diabetes, cardiovascular disease and cancer [3][4][5].
Rhodomyrtus tomentosa (Ait.) Hassk, commonly known as downy rose-myrtle, is a small shrubby tree with pink flowers, and mainly distributed in south-east Asian countries, especially southern parts of China, Japan and Thailand [6]. The juicy berry of downy rose-myrtle turns dark purple when ripe, and is historically regarded as a healthy food and used for making traditional wine, jams and beverages. In downy rose-myrtle fruit, pigments are mainly located in the skin section, which is characterized as dark purple to black in color and probably represents a mixture of anthocyanins [7,8]. However, little information is available concerning the chemical constituents and pharmacologic activity of the downy rose-myrtle. In particular, the anthocyanin compounds in the fruit skin have not been fully elucidated.
Response surface methodology (RSM) is an effective statistical method based on a multivariate non-linear model, and has been widely used for optimizing complex process variables [9]. Several studies on optimizing conditions for the extraction of anthocyanins from different plants using RSM have been published [10,11]. These studies showed that RSM is useful for developing, improving and optimizing processes. In this study, the extraction of anthocyanins from the skin of downy rose-myrtle variety "Gangren" was optimized by applying RSM of three-variable, three-level Box-Behnken design (BBD), aiming to evaluate the influence of factors such as temperature, liquid-solid ratio, and time on the process of extraction. In addition, anthocyanins in the fruit skin of downy rose-myrtle were identified by high-performance liquid chromatography-electrospray ionization-mass spectrometry (HPLC-ESI-MS). The crude extract obtained can be used as food colorants, parapharmaceutical products or for further isolation and purification of a specific anthocyanin. The results of this study will provide valuable information for the exploitation of downy rose-myrtle resources.

Optimization of the Anthocyanin Extraction
(1) Table 1. Box-Behnken design arrangement and response. Table 2 shows the analysis of variance (ANOVA) for the regression equation. The linear term and quadratic term were highly significant (p < 0.01). The model F-value of 138.63 implied that the adequate precision measures the signal to noise ratio. There was only a 0.01% chance that a "model F-Value" this large could occur due to noise [12]. According to the model P-value, X 1 , X 2 , X 3 , X 1 X 2 , X 1 X 3 , X 1 2 , X 2 2 and X 3 2 were significant model terms. A non-significant lack of fit F-value of 2.25 was also observed, which indicated that the quadratic model was valid for the present study. The coefficient of correlation (R 2 ) was 0.9944, showing good fitness of the model. Le Man et al. [13] and Chauhan and Gupta [14] have emphasized the acceptance of any model with R 2 > 0.75. And the "Pred R-Squared" of 0.9407 is in reasonable agreement with the "Adj R-Squared" of 0.9872. In addition, the coefficient of variation (CV) was 0.51%, which implied that the model was reproducible.  Surface and contour plots demonstrating the effects of different variables; two variables varied at time while the third is kept constant, on the response function (anthocyanin yield) are shown in Figure 1. The effect of temperature (X 1 ) and liquid-solid ratio (X 2 ) in the efficiency of extraction are reflected in Figure 1A. When X 1 was fixed, e.g., at 60 °C , as X 2 increased, anthocyanin yield (Y) mounted up, and reached the highest value, and then decreased. When X 2 was fixed, e.g., at 12.5, as X 1 increased, anthocyanin yield increased gradually, and reached the highest value, and then decreased. The similar relationship between Y and X 2 as well as time (X 3 ) was illustrated in Figure 1B. It indicated that the changes of X 2 had more significant effects on Y than the X 3 . The combined effect of temperature (X 1 ) and time (X 3 ) is shown in Figure 1C. It can be seen that, when X 1 was more than 65 °C , the increase of X 3 could cause a decrease of Y. Therefore, it could be concluded that the extraction temperature and liquid-solid ratio played more prominent roles on getting high extraction efficiency of anthocyanins during the extraction process.  The optimal values of the selected variables were obtained by solving the regression equation. After calculation by the Design Expert software, the optimal conditions for extracting anthocyanins from the fruit skin of downy rose-myrtle were 64.38 °C , 116.88 min, 15.7:1 liquid-solid ratio, with the corresponding Y = 4.345 mg/g. To confirm the validity of the statistical experimental strategies, a confirmation experiment with a triplicate set was performed at the selected optimum conditions. Verification tests performed gave 4.358 ± 0.045 mg/g, which clearly showed that the model fitted the experimental data and optimized the extraction process.

Run Temperature (°C ) Liquid-Solid Ratio (v/w) Time (min) Anthocyanin Yield a (mg/g) Observed Predicted
Our results show that the freeze-dried fruit skins of downy rose-myrtle contain anthocyanin at >4.3 mg/g, indicating that the content in the fresh whole fruit is >153 mg/100 g. Consistent with previous reports [4,15], the present results support the fact that dark berries are a superior source of anthocyanins. The anthocyanin contents in typical berries in Finland were systematically higher than those in this study [15]. However, these results are not comparable as such due to the different quantification methods. The anthocyanin contents in the present study are expressed as the weight of a mixture of relevant anthocyanidin 3-glucosides, whereas the results reported by Koponen and co-workers were expressed for the weight of the aglycones (anthocyanidins). In addition, anthocyanin in the pulp was not determined in the present study. Thus, these anthocyanin content values are only estimates and may be misleading; we are interested in systematically exploring anthocyanin contents and components in different berries in South China in our further studies.

Identification of Anthocyanins in the Extract
Gradient reversed-phase HPLC with absorbance detection and MS analysis was used to rapidly identify the main anthocyanins in the fruit skin extracts of downy rose-myrtle. The identification was carried out by comparison retention time and mass spectral data with those of standards. Moreover, previous studies of other researchers were taken as reference for us [11,12,16]. There were no significant differences detected in HPLC-MS results among the extracts at different extraction conditions. Figure 2 shows the representative HPLC chromatogram (A) and MS product ions (B) of the anthocyanins extract at λ = 520 nm. The major anthocyanins were tentatively identified as delphinidin-3-glucoside (peak 1), cyanidin-3-glucoside (peak 2), petunidin-3-glucoside (peak 3), peonidin-3-glucoside (peak 4), and malvidin-3-glucoside (peak 5).

Extraction of Anthocyanins
Freeze-dried powder samples (1.0 g) were added to a 50-mL Erlenmeyer flask and then mixed with different volumes of 60% ethanol containing 0.1% (v/v) hydrochloric acid (pH~1.0) [17] to give liquid-solid ratios ranging from 10 to 20 (v/w), and put in thermostatic water bath at selected temperatures (55-75 °C ) for various durations of time (90-150 min), then centrifuged at 5400 g for 10 min at 25 °C . The supernatant was collected and diluted to 50 mL with the same solvent. All samples were filtered through a 0.45 μm syringe filter (Pall Life Sciences, Ann Arbor, MI, USA). Then, aliquots of those samples were taken to evaluate total anthocyanin content and to identify specific anthocyanin components in the extract.

Total Anthocyanin Content Measurement
The total anthocyanin content was determined according to the spectrophotometric pH-differential method [18]. Briefly, an aliquot (1 mL) of the extract was mixed with 0.025 M potassium chloride buffer (pH 1.0, 4 mL) and 0.4 M sodium acetate buffer (pH 4.5, 4 mL), respectively. The absorbance of the mixture was measured at 510 and 700 nm using a UV-Vis spectrophotometer model UI-trospec 2000 (Amersham Pharmacia Biotech, Dubendorf, Switzerland). Absorbance was calculated as A = [(A 510 − A 700 ) at pH 1.0] − [(A 510 − A 700 ) at pH 4.5] with a molar extinction coefficient of 26,900 for anthocyanin. The total anthocyanin content was calculated as cyanidin-3-glucoside equivalents as Equation (2): 10 (2) where A is absorbance, MW is the molecular weight of cyanidin-3-glucoside (449.2 Da), DF is the dilution factor, V is the final volume (mL), 10 3 is the factor for conversion from g to mg, ε is the cyanindin-3-glucoside molar absorbance (26,900), L is the cell path length (1 cm), and m is the freeze-dried fruit skin weight (g).