Table of Contents Author Guidelines Submit a Manuscript
Journal of Chemistry
Volume 2019, Article ID 2793058, 12 pages
Research Article

Extraction, Enrichment, and Quantification of Main Antioxidant Aglycones of Flavonoids and Tannins from Melastoma Dodecandrum Lour.: Guided by UPLC-ESI-MS/MS

1Institute of Natural Medicine and Health Products, School of Advanced Study, Taizhou University, Taizhou 318000, China
2College of Pharmaceutical Sciences, Zhejiang University of Technology, Hangzhou 310014, China
3College of Ecology, Lishui University, Lishui 323000, China
4Department of Information Engineering, Shijiazhuang Institute of Railway Technology, Shijiazhuang 050041, China
5Chemical Biology Center, Lishui Institute of Agricultural Sciences, Lishui 323000, China

Correspondence should be addressed to Yingpeng Tong; moc.621@661hsif and Kejun Cheng; moc.liamg@nujekgnehc

Received 14 June 2019; Revised 25 September 2019; Accepted 14 October 2019; Published 3 November 2019

Academic Editor: V. Prakash Reddy

Copyright © 2019 Yingpeng Tong et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.


The objective of this work was to determine the main antioxidant aglycones of flavonoids and tannins in antioxidant extracts from Melastoma dodecandrum Lour. (MD), based on its extraction and purification process optimization and component identification by UPLC-ESI-MS/MS. Firstly, the ultrasound-assisted extraction (UAE) process for antioxidants was established by using single factor tests and response surface optimization. Evaluating by DPPH radical scavenging assay, the antioxidant capacity of MD could reach 2742.27 ± 93.86 (μmol Trolox/g) under the optimized conditions including methanol concentration (61%, v/v), extraction time (45 min), liquid-solid ratio (0.46 ml/mg), and extraction temperature (25°C). Then, the antioxidants in the solution were enriched by using macroporous resins. Because of the highest adsorption capacity and desorption ratio, HPD 500 was selected out of 14 macroporous resins for further study based on the static adsorption and desorption tests. The adsorption mechanism of the HPD 500 resin presented that pseudo-first-order kinetics model and Freundlich isotherm model could adequately explain the adsorption process. After the antioxidants were dynamically saturated by HPD 500, the column was eluted with different concentrations of ethanol (0, 25, 50, 75, and 100%, v/v). Among the above five ethanol fractions, 50% ethanol fraction showed the strongest antioxidant activity and more than 95% antioxidants adsorbed in HPD500 could be eluted by 75% ethanol. Guided by analysis results of UPLC-ESI-MS/MS, main antioxidant aglycones of flavonoids and tannins in each ethanol elution fraction were quantitatively detected. Finally, according to the correlation analysis between the antioxidant capacity of each ethanol elution fraction and its content of 10 main flavonoids and phenolic acids, the antioxidant activities of MD could mainly contribute to tannins containing phenolic units such as ellagic acid and/or gallic acid. The results of this work would provide useful information for the production of antioxidants from MD. Thus, the analysis methods could also be a quality control tool for material or products related with MD.

1. Introduction

Melastoma dodecandrum Lour. (MD), an important source material for food and pharmacological industrial in southern China, exhibits a variety of biological activities, such as anti-inflammatory [1], hypoglycemic [2], and hypolipidemic effects [3]. It can also be used to prevent and treat gastrointestinal diseases [4]. MD is rich in flavonoids and tannins with high-antioxidant activity. In our previous study, an ellagitannin named casuarinin was isolated from MD and was found to have significant anti-inflammatory activity [5]. Under physiological conditions in vivo, ellagitannins can be hydrolyzed to ellagic acid by intestinal microbes and then further be metabolized into urolithin A [6], a metabolite that can significantly enhance gut barrier function through the Nrf2 pathway according to a research in 2019 [7]. This innovative research has attracted great attention [810]. In addition, many reports have revealed the antioxidant [11], anti-inflammatory [12], and hypoglycemic activities [13] of flavonoids which are also found to be abundant in MD. Thus, it can be seen the biological activities of MD may be related with its antioxidants including flavonoids and tannins.

However, there is little research on extraction and purification process of antioxidants from MD, which is not conducive to its development and utilization. Furthermore, the reports on detail chemical composition of antioxidants from MD are also very scare. With the aim to solve above problem, the influence of methanol concentration, extraction time, liquid-solid ratio, and extraction temperature on extraction yield of antioxidants from MD was firstly investigated, and an extraction process of antioxidants from MD was established. In addition, following the enrichment of antioxidants from MD by macroporous resins and guided by the analysis result of UPLC-ESI-MS/MS, main aglycones of major flavonoids and tannins from MD were quantified by HPLC-UV, and their contributions to antioxidant capacity of MD were calculated.

2. Materials and Methods

2.1. Materials

The fresh aerial parts of Melastoma dodecandrum Lour. were obtained from Lishui city, Zhejiang province of China, in September 2017 and identified by Professor Ping Wang (cxb20170906033). The materials were dried in an air-circulating oven at 40°C and then ground to sift through a 50-mesh screen stencil, and about 2 kg of powder were collected and kept in a 4°C refrigerator for the following experiments.

2.2. Chemicals

The reference substances including p-coumaric acid, gallic acid, ellagic acid, apigenin, naringenin, luteolin, kaempferol, quercetin, vitexin, and isovitexin were obtained from Chengdu Alfa Biotechnology Co., Ltd. (Chengdu, China). Their purity was all greater than 98% (w/w). Trolox and 1,1-diphenyl-2-picrylhydrazyl radical 2,2-diphenyl-1-(2,4,6-trinitrophenyl) hydrazyl (DPPH), 2.2′-azino-bis(3-ethylbenzenothiazoline6-sulfonic acid) (ABST), and 2,4,6-tripyridyl-s-triazine (TPTZ) were purchased from Yuanye Biological Technology Co., Ltd. (Shanghai, China). Acetonitrile and methanol (HPLC grade) were obtained from Tedia Company (USA). Deionized water was prepared by a Milli-Q plus purification system (Millipore, Billerica, MA, USA). All the other chemicals were reagent grade.

AB-8, ADS-17, D101, HPD100, HPD500, HPD600, HPD950, LSA-10, LSA-21, LX-11, LX-8, NKA-2, X-5, and XDA-8 macroporous resins were supplied by Shaanxi Lebo Biochemical Technology Co., Ltd. (Shaanxi, China), and the detail physical characteristics of these fourteen resins are shown in Table 1. Their moisture contents in Table 1 were determined by drying at 105°C until constant weight. Before using, they were firstly shaken for 24 hours by soaking in 95% ethanol and then washed with deionized water thoroughly.

Table 1: Physicochemical properties and the static adsorption and desorption parameters of antioxidants on fourteen resins.
2.3. Tests for Antioxidant Activity of MD

The antioxidant activity of MD extracts was evaluated by ferric-reducing antioxidant power (FRAP) assay, ABTS•+, and DPPH radical scavenging assay. By using Trolox (TE) as a reference standard, the antioxidant activity of MD was calculated and represented as μmol Trolox equivalents for dry sample (g).

2.3.1. FRAP Assay

It was carried out by Santosh with minor modifications [14]. Firstly, 0.3 M acetate buffer (pH 3.6), 10 mM 2,4,6-tripyridyl-s-triazine (TPTZ) in HCL (40 mM), and FeCl3,6H2O (20 mM) with the ratio of 10 : 1 : 1 were mixed to form FRAP reagent. Then, 0.4 mL of MD extracts were reacted with 3.5 mL of the FRAP reagent for 15 min under room temperature. The absorbance was recorded at 595 nm using a UV1800 spectrophotometer (Shimadzu, Kyoto, Japan).

2.3.2. DPPH Radical Scavenging Assay

The antioxidant capacities of MD extracts were evaluated by DPPH radical scavenging assay described by Maira with some modifications [15]. Firstly, 3.5 mL of 0.07 mM DPPH solution was mixed with 0.4 mL of extracts and then incubated in the dark at room temperature for 30 min. The absorbance at 517 nm was recorded by a UV1800 spectrophotometer (Shimadzu, Kyoto, Japan).

2.3.3. ABTS•+ Radical Scavenging Assay

The ABTS assay was performed following Jorge [16]. 7 mM ABTS solution was mixed with 150 mM potassium persulfate and kept in the dark at room temperature for 12–16 to get ABTS radical cation (ABTS•+). Then, the mixture was diluted with ethanol to obtain an absorbance of about 0.70 at 734 nm. After that, 400 μL of MD extracts was reacted with 4 mL of ABTS•+ solution and incubated at 30°C for 30 min. The absorbance at 734 nm was recorded by a spectrophotometer (Shimadzu, Kyoto, Japan).

2.4. Ultrasound-Assisted Extraction of Antioxidants in MD
2.4.1. Extraction Process

An ultrasonic extractor (DL-720J, Shanghai Zhixin Instrument Co., Ltd.) was used to extract antioxidants from MD. Firstly, about 0.1 g MD powder was weighted accurately and mixed with various volumes (10, 20, 30, 40, and 50 mL) of different concentrations of methanol aqueous solution (0, 25, 50, 75, and 100%, Vmethanol/Vwater), and then the mixtures were treated with different times (15, 30, 45, 60, and 75 min) at certain temperatures (25, 35, 45, and 55°C). After extraction, the samples were cooled to room temperature and centrifuged at 3500 rpm for 15 min. The suspensions were collected to evaluate their antioxidant capacities by DPPH radical scavenging assay (μmol TE/g).

2.4.2. Box–Behnken Design (BBD) Optimization

Based on the single-factor experiments, response surface methodology (RSM) was further applied to optimize the ultrasound-assisted extraction conditions in this work by using Design-expert 11 software. Factors of methanol concentration (X1), extraction time (X2), and liquid-solid ratio (X3) were carried out at three levels to check their influence on the antioxidant capacities of MD extraction solutions. A full quadratic equation formula was utilized to analyze the data from BBD as follows:where Y and represent the response function and regression coefficients and , , and are the linearity, square, and intercept and interaction terms, respectively.

2.5. Antioxidants Enrichment by Using Macroporous Resins
2.5.1. Preparation of Sample Solutions

Based on the optimized extraction conditions by BBD, the extracted solution of MD powder was dried in vacuum at 40°C by a rotary evaporator (Heidolph Instruments GmbH & Co. KG, Germany), then the dried extracts of MD were dissolved by distilled water, and the ratio of distilled water and MD powder was about 100 mL/g.

2.5.2. Screening of Macroporous Resins by Static Adsorption and Desorption Experiments

The fourteen fresh resins (equal to 0.2 g dry resin) were weighted and mixed with 30 mL of sample solutions, and then the flasks were shaken at 160 rpm at 25°C for 180 min in a thermostatic oscillator. After full adsorption, the antioxidant capacities of the filtrate solutions were determined by DPPH radical scavenging assay and represented as μmol Trolox equivalents for 1 mL of sample solution. In desorption experiments, the above fourteen fresh resins were washed with distilled water in triplicate and 10 mL of ethanol was added for desorption at 160 rpm at 25°C for 180 min. Next, the antioxidant capacities of ethanol solutions were analyzed. The adsorption capacity, desorption capacity, and desorption ratio were calculated using the following formulas:where is the adsorption capacity at equilibrium (μmol TE/g dry resin), is the desorption capacity (μmol TE/g dry resin), D is the desorption ratio (%), and are the initial and absorption equilibrium concentrations of antioxidant capacities of sample solutions (μmol TE/mL), and are the volume of the initial sample and desorption solution (mL), and W is the dry resin weight (g).

2.5.3. Adsorption Kinetics on Selected Resin

The fresh selected resin (equivalent to 2.0 g dry resin) mixed with 300 mL of sample solution was added to a flask, shaking on a thermostatic oscillator (160 rpm) at 25°C for 160 min. At each time points of 0, 2, 4, 6, 8, 12, 16, 20, 25, 30, 40, 50, 60, 70, 80, 100, 120, 140, and 160 min, 1 mL of solution was obtained to determine its antioxidant capacity (μmol TE/mL) and the kinetic adsorption curve was plotted.

2.5.4. Adsorption Thermodynamics on Selected Resin

Thirty milliliters of sample solutions with various contents of antioxidants (11.62, 14.52, 17.42, 20.33, and 23.23 μmol TE/mL, respectively) was poured into the flasks which were loaded with hydrated resins (equal to 0.20 g dry resins). Then, the flasks were shaken at 160 rpm at 25, 30, and 35°C for 160 min. Subsequently, the antioxidant capacities in the filtrate solutions were determined.

2.5.5. Enrichment of Antioxidants

The selected resin was packed into a column with a bed volume (BV) of 3 mL, and then 70-fold BV of sample solution was loaded at a flow rate of 0.5 mL/min. Subsequently, the column was eluted with 5-fold BV of distilled water, 25% (v/v), 50% (v/v), 75% (v/v), and 100% (v/v) ethanol sequentially at 0.5 mL/min. After evaluating the antioxidant contents in each fraction, they were, respectively, concentrated in vacuum at 40°C and stored at −20°C with the names of MD-0%, MD-25%, MD-50%, MD-75%, and MD-100%.

2.6. UPLC-ESI-MS/MS Analysis

UPLC-ESI-MS/MS analysis was carried out on an ultra-high-performance liquid chromatography (Waters ACQUITY I Class) coupled with the Waters Xevo G2-XS-Q-TOF mass spectrometer. The phytochemicals were separated by Waters BEH C18 () at a flow rate of 0.43 mL/min at 40°C. The mobile phase was made up by acetonitrile (A) and 0.1% formic acid aqueous solution (B), using a linear gradient program: 0–4 min, from 2 to 5% (A); 4 to 12 min, 5 to 11% (A); 12 to 18 min, 11 to 17% (A); 18 to 28 min, 17 to 98% (A) and then followed by 98% (A) for 7 min. The ESI source was operated in the positive and negative modes with full scan mode from 50 to 1500 m/z, and the conditions were as follows: ion source temperature, 120°C; cone gas flow, 50 L/h; desolvation gas flow, 595 L/h; and capillary voltage, 2 kV.

2.7. Simultaneous Quantification of Main Flavonoids and Phenolic Acids in MD by HPLC-DAD

The extracts named MD-0%, MD-25%, MD-50%, MD-75%, and MD-100% were hydrolyzed with various volumes of 10% hydrochloric acid (HCl) aqueous solution at 80°C for 3 h, and DMSO solutions were added if the extracts could not be dissolved by 10% HCl. After the solutions were cooled to room temperature, they were filtered through 0.25 μm membrane filters and directly injected into an Agilent 1260 HPLC equipped with a diode array detector (DAD) (Agilent Technologies, USA). The column used in this work was an Agilent Zorbax Extend C18 reserved phase column (4 μm, 150 mm × 4.6 mm), the sample injection volume was 5 μl, the flow rate was 1.0 mL/min, the column temperature was 30°C, and the detection wavelength was set from 200–400 nm. The mobile phase was mixed with solvent A (methanol), solvent B (acetonitrile), and solvent D (0.1% formic acid aqueous solution). The gradient program is shown in Table 2.

Table 2: The gradient program for simultaneous quantification of main flavonoids and phenolic acids in MD.

3. Results and Discussion

3.1. Effect of the Extraction Conditions on the Yields of Antioxidants from MD

Experiments were carried out to investigate the influence of methanol concentration on extraction yield. As shown in Figure 1(a), the yield of antioxidants gradually increased from 1997.68 ± 8.97 to 2492.23 ± 122.67 μmol TE/g as the methanol concentration was changed from 0 to 75% (v/v), while the other extraction parameters were set as follows: extraction temperature 25°C, extraction time 30 min, and liquid-solid ratio 0.3 mL/mg; but the yield of antioxidants significantly decreased to 1209.91 ± 77.45 μmol TE/g when the methanol concentration was 100%. This might be due to the low solubility of antioxidants including flavonoids and tannins in high-concentration methanol. In addition, high-concentration of methanol might also cause protein denaturation and further prevent the dissolution of antioxidants [17].

Figure 1: Influence of the methanol concentration (a), extraction temperature (b), extraction time (c), and liquid-solid ratio (d) on the extraction yield of antioxidant contents.

The effect of extraction temperature variations (25, 35, 45, and 55°C) on the yields was inspected and is shown in Figure 1(b) as the other extraction conditions were as follows: methanol concentration 75%, extraction time 30 min, and liquid-solid ratio 0.3 mL/mg; as the temperature increased, the extraction yields gradually decreased and the maximal yield was 2688.55 ± 95.71 μmol TE/g at 25°C, which was close to room temperature. The result demonstrated that polyphenols such as tannins were not stable under high-temperature conditions, which was consistent with Makkar’s report [18].

The influence of extraction time (15, 30, 45, 60, and 75 min) on the yields of antioxidants is displayed in Figure 1(c). These experiments were performed at 25°C with a 0.3 mL/mg liquid-solid ratio by using 75% aqueous methanol solution as extraction solvent. As seen from Figure 1(c), the extraction yield reached a maximal value of 2471.85 ± 74.75 μmol TE/g at 30 min.

By selecting 75% aqueous methanol solution as extraction solvent, experiments were conducted to determine the optimal liquid-solid ratio at 25°C for 30 min. As shown in Figure 1(d), the extraction yields gradually increased from 0.1 to 0.4 mL/mg and reached a maximal value of 2515.34 ± 86.2 μmol TE/g at 0.4 mL/mg; then, the yields were nearly constant from 0.4 to 0.5 mL/mg.

3.2. Optimization of the Extraction Parameters by Response Surface Methodology (RSM)

Three individual factors, including extraction time, methanol concentration, and liquid-solid ratio, were further optimized by the response surface methodology by using Design-expert 11 software. 17 experiments were carried out by Box–Behnken design, and the results are listed in Table 3 [19, 20]. As displayed in Table 3, the extraction yields ranged from 1214.48 ± 211.17 to 2779.49 ± 291.99 μmol TE/g. Moreover, the analysis of variance (ANOVA) of the response surface model is displayed in Table 4. The value was used to test the significance of a coefficient. Basically, a value, smaller than 0.05, indicated that the model terms were significant. As shown in Table 4, the individual factor (X1) and interaction coefficient (X2X3) could significantly influence the extraction yield because the values were smaller than 0.05. The value of the regression model was <0.0001, suggesting that the model was significant. The “Lack of fit value” of 0.9453 > 0.05 indicated that the model, which was chosen and shown as follows, was adequately accurate in predicting the relevant response:

Table 3: Box–Behnken design and result of response surface method analysis.
Table 4: Variance analysis of regression equations.

In addition, the adjusted coefficient of determination (Radj2 = 0.9441) indicated that the model could explain 94.41% of the response value changes. This affirmed that the predicted model was highly significant and could sufficiently match the observed values.

Three-dimensional response surface plots are displayed in Figure 2. From Figures 2(a) and 2(b), we could clearly observe that the extraction time and liquid-solid ratio had a significantly weaker effect on extraction yield than methanol concentration. However, higher yields could be gained by increasing the extraction time at high liquid-solid ratios (Figure 2(c)). The optimal conditions for antioxidants predicted by the response surface software were as follows: 60.75% methanol concentration, 45 min, 0.459 ml/mg, and 25°C. The final conditions were slightly modified: 61% methanol concentration, 45 min, 0.46 ml/mg, and 25°C. Under these conditions, the extraction yield of antioxidants was 2742.27 ± 93.86 μmol TE/g, which was not significantly different from the predicted value of 2789.81 μmol TE/g. The antioxidant capacities of sample solutions extracted under the optimized condition were further evaluated by ABTS•+ radical scavenging assay and FRAP assay. The values for these two assay methods were 2689.33 ± 75.46 and 1892.29 ± 100.14 μmol TE/g, respectively.

Figure 2: Response surface (3D) showing the influence of different extraction parameters on extraction yield.
3.3. Static Adsorption and Desorption Parameters of Antioxidants from MD on Fourteen Resins

Static adsorption and desorption tests were carried out to screen the optimum resin. The antioxidant constituents in MD including flavonoids and tannins process abundant hydrogen groups, exhibiting higher polarity. As shown in Table 1, because of their similar polarity with the target compounds, most of resins with strong-polarity or middle-polarity, including HPD500, HPD600, ADS-17, and LSA-10, had high adsorption capacities while nonpolar resins such as LSA-21 which displayed low adsorption capacities [21]. Considering adsorption and desorption abilities, HPD500 was selected for separation of antioxidants in MD.

3.4. Static Adsorption Kinetics of the Antioxidants from MD on HPD-500 Resin

The curves of the adsorption kinetics for HPD500 are displayed in Figure 3. As we see, the adsorption of antioxidants on HPD500 increased rapidly for first 30 minutes, especially for first 16 minutes. Then, the adsorption elevated slowly in the next 50 minutes and reached equilibrium.

Figure 3: Adsorption kinetics of antioxidants on HPD500.

To explain the adsorption process of antioxidants on HPD500, the pseudo-first-order (equation (4)), pseudo-second-order (equation (5)), and particle diffusion kinetics models (equation (6)) were adopted.

According to the correlation coefficients, both pseudo-first-order (R2 = 0.992) and pseudo-second-order kinetics models (R2 = 0.9804) could describe the adsorption antioxidants on HPD500:where Qt is the adsorption capacity (μmol TE/g) at time t, k1, k2, and kt are the rate constants of pseudo-first-order, pseudo-second-order, and particle diffusion kinetics models in adsorption process, respectively, and C refers to the constant in the particle diffusion kinetics model.

3.5. Static Thermodynamics of Antioxidants on HPD-500 Resin

Langmuir and Freundlich models, two most common theoretical models, were used to fit equilibrated relationship between the concentrations of antioxidants in the fluid phase and HPD500 resin at a given temperature. These two equations are as follows:where Q0 is the maximum adsorption capacity of HPD500 (μmol TE/g), KL reflects the adsorption energy (L/μmol), KF is the parameter related to the adsorption capacity of HPD500 ((μmol TE/g) (L/μmol TE)1/n), and 1/n represents the adsorption intensity of HPD500.

As shown in Table 5, the correlation coefficients in Langmuir and Freundlich models were both higher than 0.9, indicating that both two models could characterize the isothermal adsorption processes of antioxidants on HPD500 resin and the adsorption processes may belong to monolayer coverage adsorption.

Table 5: Langmuir model, Freundlich model, and thermodynamic parameters of the antioxidants in MD on HPD500 resin.

Thermodynamic parameters for the adsorption of antioxidants on HPD500 resin were further estimated to account for the adsorption process. The thermodynamic parameters, including enthalpy (∆H), entropy (∆S), and free energy (∆G), were calculated according to the following equation:where R is the gas constant (J/mol K) and T is the absolute temperature (°K).

As displayed in Table 5, the values of the free energy (∆G) of adsorption were −3.461, −3.639, and −2.452 kJ/mol at 25, 30, and 35°C, correspondingly implying that the adsorption process was spontaneous and feasible. The enthalpy changes (∆H) were also negative, suggesting that the adsorption process was exothermic, and low temperature was favorable to the adsorption process. Moreover, the absolute value of ∆H was less than 43 kJ/mol, showing that the adsorption of antioxidants in MD on HPD500 was controlled by physical mechanisms rather than chemical mechanisms.

3.6. Enrichment of Antioxidants from MD by HPD-500 Resin Column Chromatography

The antioxidant adsorbed on the resin was eluted with different concentrations of ethanol solution. As shown in Table 6, the contents of antioxidants in five fractions were in the order of MD-50% (1406.85 ± 45.37 μmol TE) > MD-25% (570.87 ± 23.94 μmol TE) > MD-75% (540.18 ± 17.66 μmol TE) > MD-100% (82.42 ± 4.09 μmol TE) > MD-0% (68.09 ± 2.74 μmol TE). About 52.72% antioxidants were eluted in the 50% ethanol fraction while only 3.09 and 2.55% antioxidants were in 100% and 0% ethanol fraction, respectively. The contents of antioxidants in the MD-25% and 75% fraction were similar, 21.39% and 20.24%, respectively.

Table 6: The contents of antioxidants and ten main flavonoids and phenolic acids in five fractions from MD.
3.7. Identification of Chemical Constituents of Antioxidants from MD

MD solutions used in UPLC-ESI-MS/MS experiments were prepared under the optimized conditions as shown in Section 3.2. According to a recent paper about the chemical characterization of MD using UPLC-ESI-Q-Exactive Focus-MS/MS [22], various chemical classes including tannins, flavonoids, fatty acids, triterpenoids, sterols, and others were reported. Many researches had presented that polyphenols including tannins and flavonoids had higher antioxidant activities by comparing with fatty acids, triterpenoids, and sterols [2326], so this work mainly focused on tannins and flavonoids from MD. In general, 39 tannins and 16 flavonoids were tentatively identified and shown in Table S1 including their retention times (Rt), molecular formulas, detected accurate mass in negative mode, error in ppm, and the MS/MS fragment ions. The codes of above 54 compounds are marked in Figure 4, the base peak chromatogram in negative ionisation modes. The compounds were identified by comparing the observed MS data and MS/MS spectra in this work with those found in the literatures, especially the constituents from MD and its same botanical family. In addition, the following public databases were also reviewed: ChemSpider (, Spectral Database for Organic Compounds SDBS (, and m/z cloud (

Figure 4: LC-MS chromatograms of extracts from MD in negative-ion mode.
3.7.1. Tannins

Tannins are the most abundant compounds in MD, and they could be eluted in shorter time under reversed-phase chromatographic conditions because they exhibited very strong polarity with many hydroxyl groups. In this work, 39 tannins had been identified from MD. In MS2 spectra of tannins, ions at m/z 169.0135 and/or m/z 301.0004 could be commonly detected, implying the presence of galloyl and/or hexahydroxydiphenolic acid (HHDP) in tannins. This result was further confirmed by the mass loss of 302 or 152 Da in MS2 spectra, due to the loss of the HHDP and/or galloyl units [2730]. Figure 5 shows the characteristic fragmentation pathway of a tannins (A23 in Figure 4.), an isomer mono-O-ellagic-tri-O-galloyl glucoside.

Figure 5: The MS2 spectra and fragmentation pathway of mono-O-ellagic-tri-O-galloyl glucoside.
3.7.2. Flavonoids

Another kind of main antioxidant compounds from MD were flavonoids, and their characterization of fragmentation pathways have been revealed and recognized as the typical retro-Diels–Alder fissions [31]. In this work, 15 flavonoid O-glycosides were determined from MD and their aglycons were considered as apigenin, naringenin, luteolin, kaempferol, or quercetin, due to the observation of ions at m/z 151 and their [M-H] ions [32, 33]. According to our previous research [5], flavonoid C-glycosides, such as vitexin and isovitexin, were the most abundant flavonoids in MD, and their structures were identified based on comparison of MS2 spectra obtained in this work and fragmentation pattern characteristic reported in the literatures [22, 3437].

3.8. Simultaneous Determination of Major Flavonoids and Phenolic Acids in Antioxidants in MD

Based on the result of UPLC-ESI-MS/MS, ten phenolic units including p-coumaric acid, gallic acid, ellagic acid, apigenin, naringenin, luteolin, kaempferol, quercetin, vitexin, and isovitexin were identified. To further confirm the relationship between these 10 phenolic units and antioxidant capacity of MD, the contents of above 10 phenolic units were quantified. In addition to ellagic acid which was dissolved in DMSO, all standards were dissolved in methanol and kept as stock solutions. The above ten stock solutions were mixed in proportion and then diluted to five appropriate concentrations for the construction of calibration curves. Each concentration of solution was injected in triplicate. As demonstrated in Table 6, all calibration curves between the peak areas and the concentration of each standard showed good linear regression (R2 > 0.999) within test ranges. The contents of these 10 compounds in five fractions named MD-0%, MD-25%, MD-50%, MD-75%, and MD-100% were calculated based on the calibration curves.

As shown in Table 6 and Figure 6, gallic acid, vitexin, isovitexin, and ellagic acid could be detected in each fraction, while coumaric acid and naringenin could not be determined due to the low detection limit of HPLC-UV. Flavonoids were mainly concentrated in MD-75% and MD-100%, while the contents of gallic acid and ellagic acid in MD-50% could reach 377.73 ± 5.49 and 1303.79 ± 9.42 μg, respectively, which were much higher than other fractions.

Figure 6: Representative HPLC chromatograms of standard solution and eluted fractions (1, gallic acid; 2, P-coumaric acid; 3, vitexin; 4, isovitexin; 5, ellagic acid; 6, quercetin; 7, naringin; 8, luteolin; 9, kaempferol; 10, apigenin).

Correlation analysis of antioxidant activity versus the contents of ten phenolic molecules illustrated that antioxidant activities of MD fractions had a significant correlation with gallic acid (r = 0.968, ) and ellagic acid (r = 0.929, ). The following two factors may be relevant with the above result: (1) The contents of gallic acid and ellagic acid in the antioxidants absorbed onto HPD500 resin were significantly higher than other flavonoids, which could totally reach 722.16 μg and 2245.22 μg, respectively. (2) More importantly, there was ortho-hydroxyl group in gallic acid and ellagic acid which were considered to be the most important structural feature for anti-DPPH free radicals [38, 39]. That is why the contents of vitexin and isovitexin were similar to gallic acid, but the contributions to antioxidant activities of MD fractions were less. Although flavonoids such as luteolin and quercetin also had an ortho-dihydroxy group, their activities markedly reduced due to glycosylation as displayed in Table S1 [4042]. Based on the above results, the method of correlation analysis was a useful tool to discover the key antioxidant units in complex mixtures without isolation.

4. Conclusions

In summary, an efficient extraction and purification method was established to enrich antioxidants from MD by combining ultrasonic microwave-assisted extraction and macroporous resin chromatography. The optimal extraction conditions for microwave-assisted extraction were as follows: methanol concentration 61%, extraction time 45 min, liquid-solid ratio 0.46 ml/mg, and extraction temperature 25°C. Under these conditions, the extraction yield of antioxidants could reach 2742.27 ± 93.86 (DPPH assay), 2689.33 ± 75.46 (ABST assay), and 1892.29 ± 100.14 μmol TE/g (FRAP assay), respectively. HPD 500 macroporous resin was adopted to purify the antioxidants in MD solution due to its effective adsorption and desorption. More than 95% antioxidants adsorbed in HPD500 could be eluted by 75% ethanol, and the 50% ethanol fraction showed the strongest antioxidant activity. Guided by UPLC-ESI-MS/MS, the contents of ten aglycones of flavonoids and tannins were detected by HPLC-DAD and the key antioxidant phenolic units in MD were found in gallic acid and ellagic acid by correlation analysis. The study provides helpful information for antioxidants production from MD and reveals a new way for the quality control of material or products related with MD. In addition, further research can be carried out to develop functional food or cosmetics using the antioxidants from MD.

Data Availability

The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare that they have no conflicts of interest.


The work was financially supported by the National Key Research and Development Program of China (2017YFC1703903), the National Natural Science Foundations of China (No. 81703688), the Major Science &Technology Project of Zhejiang Province (No. 2012C12014-1), and Lishui Science & Technology Bureau Research Fund (No. 20140212037).

Supplementary Materials

Table S1: the tannins and flavonoids of Melastoma dodecandrum Lour. identified by UPLC-ESI-MS/MS in negative modes. (Supplementary Materials)


  1. R. Ishii, K. Saito, M. Horie, T. Shibano, S. Kitanaka, and F. Amano, “Inhibitory effects of hydrolyzable tannins from Melastoma dodecandrum Lour. On nitric oxide production by a murine macrophage-like cell line, RAW264.7, activated with lipopolysaccharide and interferon-γ,” Biological & Pharmaceutical Bulletin, vol. 22, no. 6, pp. 647–653, 1999. View at Publisher · View at Google Scholar · View at Scopus
  2. L. Li, Z.-P. Luo, H.-D. Zhou, and X.-F. Yang, “Study on hypoglycemic activity of different extracted parts of yao medicine,” Chinese Patent Medicine, vol. 36, no. 5, pp. 1065–1068, 2014. View at Google Scholar
  3. L. Li, Z.-P. Luo, H.-D. Zhou, and X.-F. Yang, “Effects of ethyl acetate extraction of Melastoma dodecandrum Lour. On blood glucose, blood lipid and antioxidation in diabetic mice,” Chinese Journal of Gerontology, vol. 35, no. 6, pp. 3250–3252, 2015. View at Google Scholar
  4. J.-F. Wang, P.-H. Zhang, Y.-T. Wang et al., “Research progress on Melastoma dodecandrum,” Chinese Traditional and Herbal Drugs, vol. 49, no. 5, pp. 1211–1219, 2018. View at Google Scholar
  5. G.-X. Yang, R.-Z. Zhang, B. Lou, K.-J. Cheng, J. Xiong, and J.-F. Hu, “Chemical constituents from Melastoma dodecandrum and their inhibitory activity on interleukin-8 production in HT-29 cells,” Natural Product Research, vol. 28, no. 17, pp. 1383–1387, 2014. View at Publisher · View at Google Scholar · View at Scopus
  6. J. M. Landete, “Ellagitannins, ellagic acid and their derived metabolites: a review about source, metabolism, functions and health,” Food Research International, vol. 44, no. 5, pp. 1150–1160, 2011. View at Publisher · View at Google Scholar · View at Scopus
  7. R. Singh, S. Chandrashekharappa, S. R. Bodduluri et al., “Enhancement of the gut barrier integrity by a microbial metabolite through the Nrf2 pathway,” Nature Communications, vol. 10, no. 1, p. 89, 2019. View at Publisher · View at Google Scholar · View at Scopus
  8. D. Mafra, N. Borges, L. Alvarenga et al., “Dietary components that may influence the disturbed gut microbiota in chronic kidney disease,” Nutrients, vol. 11, no. 3, p. 496, 2019. View at Publisher · View at Google Scholar · View at Scopus
  9. T. P. Wypych, L. C. Wickramasinghe, and B. J. Marsland, “The influence of the microbiome on respiratory health,” Nature Immunology, vol. 20, no. 10, pp. 1–12, 2019. View at Publisher · View at Google Scholar
  10. H. M. Roager and L. O. Dragsted, “Diet-derived microbial metabolites in health and disease,” Nutrition Bulletin, vol. 44, no. 3, pp. 216–227, 2019. View at Publisher · View at Google Scholar
  11. J. S. Park, H. S. Rho, D. H. Kim, and I. S. Chang, “Enzymatic preparation of kaempferol from green tea seed and its antioxidant activity,” Journal of Agricultural and Food Chemistry, vol. 54, no. 8, pp. 2951–2956, 2006. View at Publisher · View at Google Scholar · View at Scopus
  12. V. Khajuria, S. Gupta, N. Sharma et al., “Kaempferol-3-o-β-d-glucuronate exhibit potential anti-inflammatory effect in LPS stimulated RAW 264.7 cells and mice model,” International Immunopharmacology, vol. 57, pp. 62–71, 2018. View at Publisher · View at Google Scholar · View at Scopus
  13. E. de Sousa, L. Zanatta, I. Seifriz et al., “Hypoglycemic effect and antioxidant potential of kaempferol-3,7-O-(α)-dirhamnoside from Bauhinia forficata Leaves,” Journal of Natural Products, vol. 67, no. 5, pp. 829–832, 2004. View at Publisher · View at Google Scholar · View at Scopus
  14. S. B. Parit, V. V. Dawkar, R. S. Tanpure, S. R. Pai, and A. D. Chougale, “Nutritional quality and antioxidant activity of wheatgrass (Triticum aestivum) unwrap by proteome profiling and DPPH and FRAP assays,” Journal of Food Science, vol. 83, no. 8, pp. 2127–2139, 2018. View at Publisher · View at Google Scholar · View at Scopus
  15. M. Casagrande, J. Zanela, A. Wagner et al., “Influence of time, temperature and solvent on the extraction of bioactive compounds of Baccharis dracunculifolia: in vitro antioxidant activity, antimicrobial potential, and phenolic compound quantification,” Industrial Crops and Products, vol. 125, pp. 207–219, 2018. View at Publisher · View at Google Scholar · View at Scopus
  16. J. Hoyos-Arbeláez, L. Blandón-Naranjo, M. Vázquez, and J. Contreras-Calderón, “Antioxidant capacity of mango fruit (Mangifera indica). An electrochemical study as an approach to the spectrophotometric methods,” Food Chemistry, vol. 266, pp. 435–440, 2018. View at Publisher · View at Google Scholar · View at Scopus
  17. F. Dahmoune, B. Nayak, K. Moussi, H. Remini, and K. Madani, “Optimization of microwave-assisted extraction of polyphenols from Myrtus communis L. leaves,” Food Chemistry, vol. 166, pp. 585–595, 2015. View at Publisher · View at Google Scholar · View at Scopus
  18. H. P. S. Makkar and K. Becker, “Effect of pH, temperature, and time on inactivation of tannins and possible implications in detannification studies,” Journal of Agricultural and Food Chemistry, vol. 44, no. 5, pp. 1291–1295, 1996. View at Publisher · View at Google Scholar · View at Scopus
  19. G. Cui, Z. Liu, M. Wei, and L. Yang, “Turpentine as an alternative solvent for the extraction of gutta-percha from Eucommia ulmoides barks,” Industrial Crops and Products, vol. 121, pp. 142–150, 2018. View at Publisher · View at Google Scholar · View at Scopus
  20. Y. Qu, C. Li, C. Zhang, R. Zeng, and C. Fu, “Optimization of infrared-assisted extraction of Bletilla striata polysaccharides based on response surface methodology and their antioxidant activities,” Carbohydrate Polymers, vol. 148, pp. 345–353, 2016. View at Publisher · View at Google Scholar · View at Scopus
  21. Z. Jiang, R. Shi, H. Chen, and Y. Wang, “Ultrasonic microwave-assisted extraction coupled with macroporous resin chromatography for the purification of antioxidant phenolics from waste jackfruit (Artocarpus heterophyllus Lam.) peels,” Journal of Food Science and Technology, vol. 56, no. 8, pp. 3877–3886, 2019. View at Publisher · View at Google Scholar
  22. J. F. Wang, Z. Y. Jia, Z. H. Zhang et al., “Analysis of chemical constituents of Melastoma dodecandrum Lour. By UPLC-ESI-Q-exactive focus-MS/MS,” Molecules, vol. 22, no. 3, p. 476, 2017. View at Publisher · View at Google Scholar · View at Scopus
  23. V. Muccilli, N. Cardullo, C. Spatafora, V. Cunsolo, and C. Tringali, “α-glucosidase inhibition and antioxidant activity of an oenological commercial tannin. Extraction, fractionation and analysis by HPLC/ESI-MS/MS and 1H NMR,” Food Chemistry, vol. 215, pp. 50–60, 2017. View at Publisher · View at Google Scholar · View at Scopus
  24. S. Geng, Y. Chen, A. M. Abbasi, H. Ma, H. Mo, and B. Liu, “Tannin fraction from Ampelopsis grossedentata leaves tea (Tengcha) as an antioxidant and α-glucosidase inhibitory nutraceutical,” International Journal of Food Science & Technology, vol. 51, no. 12, pp. 2692–2700, 2016. View at Publisher · View at Google Scholar · View at Scopus
  25. H. Zhang and R. Tsao, “Dietary polyphenols, oxidative stress and antioxidant and anti-inflammatory effects,” Current Opinion in Food Science, vol. 8, pp. 33–42, 2016. View at Publisher · View at Google Scholar · View at Scopus
  26. Y. Deng, Y. Zhao, O. Padilla-Zakour, and G. Yang, “Polyphenols, antioxidant and antimicrobial activities of leaf and bark extracts of Solidago canadensis L,” Industrial Crops and Products, vol. 74, pp. 803–809, 2015. View at Publisher · View at Google Scholar · View at Scopus
  27. K. J. Meyers, T. J. Swiecki, and A. E. Mitchell, “Understanding the native Californian diet: identification of condensed and hydrolyzable tannins in tanoak acorns (Lithocarpus densiflorus),” Journal of Agricultural and Food Chemistry, vol. 54, no. 20, pp. 7686–7691, 2006. View at Publisher · View at Google Scholar · View at Scopus
  28. J.-P. Salminen, V. Ossipov, E. Haukioja, and K. Pihlaja, “Seasonal variation in the content of hydrolysable tannins in leaves of Betula pubescens,” Phytochemistry, vol. 57, no. 1, pp. 15–22, 2001. View at Publisher · View at Google Scholar · View at Scopus
  29. Y. Y. Soong and P. J. Barlow, “Isolation and structure elucidation of phenolic compounds from longan (Dimocarpus longan Lour.) seed by high-performance liquid chromatography–electrospray ionization mass spectrometry,” Journal of Chromatography A, vol. 1085, no. 2, pp. 270–277, 2005. View at Publisher · View at Google Scholar · View at Scopus
  30. T. H. Hooi, L. S. Kiong, and C. C. Hock, “Characterisation of galloylated cyanogenic glucosides and hydrolysable tannins from leaves of phyllagathis rotundifolia by LC-ESI-MS/MS,” Phytochemical Analysis, vol. 22, no. 6, pp. 516–525, 2011. View at Publisher · View at Google Scholar · View at Scopus
  31. N. Fabre, I. Rustan, E. Hoffmann, and J. Quetin-Leclercq, “Determination of flavone, flavonol, and flavanone aglycones by negative ion liquid chromatography electrospray ion trap mass spectrometry,” Journal of the American Society for Mass Spectrometry, vol. 12, no. 6, pp. 707–715, 2001. View at Publisher · View at Google Scholar · View at Scopus
  32. S. A. O. Santos, C. S. R. Freire, M. R. M. Domingues, A. J. D. Silvestre, and C. P. Neto, “Characterization of phenolic components in polar extracts of Eucalyptus globulus Labill. bark by high-performance liquid chromatography–mass spectrometry,” Journal of Agricultural and Food Chemistry, vol. 59, no. 17, pp. 9386–9393, 2011. View at Publisher · View at Google Scholar · View at Scopus
  33. I. M. Abu-Reidah, M. S. Ali-Shtayeh, R. M. Jamous, D. Arráez-Román, and A. Segura-Carretero, “HPLC–DAD–ESI-MS/MS screening of bioactive components from Rhus coriaria L. (Sumac) fruits,” Food Chemistry, vol. 166, pp. 179–191, 2015. View at Publisher · View at Google Scholar · View at Scopus
  34. B. D. Davis and J. S. Brodbelt, “Determination of the glycosylation site of flavonoid monoglucosides by metal complexation and tandem mass spectrometry,” Journal of the American Society for Mass Spectrometry, vol. 15, no. 9, pp. 1287–1299, 2004. View at Publisher · View at Google Scholar · View at Scopus
  35. M. A. Farag and Z. T. A. Shakour, “Metabolomics driven analysis of 11 Portulaca leaf taxa as analysed via UPLC-ESI-MS/MS and chemometrics,” Phytochemistry, vol. 161, pp. 117–129, 2019. View at Publisher · View at Google Scholar · View at Scopus
  36. A. Wojakowska, J. Perkowski, T. Góral, and M. Stobiecki, “Structural characterization of flavonoid glycosides from leaves of wheat (Triticum aestivum L.) using LC/MS/MS profiling of the target compounds,” Journal of Mass Spectrometry, vol. 48, no. 3, pp. 329–339, 2013. View at Publisher · View at Google Scholar · View at Scopus
  37. P. Quifer-Rada, A. Vallverdú-Queralt, M. Martínez-Huélamo et al., “A comprehensive characterisation of beer polyphenols by high resolution mass spectrometry (LC–ESI-LTQ-Orbitrap-MS),” Food Chemistry, vol. 169, pp. 336–343, 2015. View at Publisher · View at Google Scholar · View at Scopus
  38. T. Yokozawa, C. P. Chen, E. Dong, T. Tanaka, G.-I. Nonaka, and I. Nishioka, “Study on the inhibitory effect of tannins and flavonoids against the 1,1-Diphenyl-2-picrylhydrazyl radical,” Biochemical Pharmacology, vol. 56, no. 2, pp. 213–222, 1998. View at Publisher · View at Google Scholar · View at Scopus
  39. Y.-Z. Cai, S. M. Sun, X. J. Xing, Q. Luo, and H. Corke, “Structure–radical scavenging activity relationships of phenolic compounds from traditional Chinese medicinal plants,” Life Sciences, vol. 78, no. 25, pp. 2872–2888, 2006. View at Publisher · View at Google Scholar · View at Scopus
  40. D. D. Majo, M. Giammanco, M. L. Guardia, E. Tripoli, S. Giammanco, and E. Finotti, “Flavanones in citrus fruit: structure–antioxidant activity relationships,” Food Research International, vol. 38, no. 10, pp. 1161–1166, 2005. View at Publisher · View at Google Scholar · View at Scopus
  41. S. Burda and W. Oleszek, “Antioxidant and antiradical activities of flavonoids,” Journal of Agricultural and Food Chemistry, vol. 49, no. 6, pp. 2774–2779, 2001. View at Publisher · View at Google Scholar · View at Scopus
  42. C. A. Rice-Evans, N. J. Miller, and G. Paganga, “Structure-antioxidant activity relationships of flavonoids and phenolic acids,” Free Radical Biology and Medicine, vol. 20, no. 7, pp. 933–956, 1996. View at Publisher · View at Google Scholar · View at Scopus