Extraction and purification of total flavonoids from Eupatorium lindleyanum DC. and evaluation of their antioxidant and enzyme inhibitory activities

Abstract The health benefits and promising medical treatment potential of total flavonoids from Eupatorium lindleyanum DC. (TFELDC) have been recognized. The process parameters of extracting total flavonoids from Eupatorium lindleyanum DC. by ultrasonic–microwave synergistic extraction (UMSE) were optimized, and they were purified by AB‐8 macroporous resin in the current study. In addition, the antioxidant and enzyme inhibitory activities of the purified TFELDC (PTFELDC) were evaluated. The results showed that the optimal parameters of UMSE were as follows: ethanol volume fraction 71.5%, L/S ratio 12.2 ml/g, microwave power 318 W, and extraction time 143 s. After TFELDC were purified by AB‐8 macroporous resin, the total flavonoid contents of PTFELDC increased from 208.18 ± 1.60 to 511.19 ± 3.21 mg RE/g FDS. Compared with TFELDC, the content of total flavonoids in PTFELDC was increased by 2.46 times. The antioxidant activities of PTFELDC were assessed using DPPH radical, superoxide anion radical, reducing power, and ferric reducing antioxidant power assays, and the IC50 values were found to be 37.13, 19.62, 81.22, and 24.72 μg/ml, respectively. The enzyme inhibitory activities of PTFELDC were measured using lipase, α‐amylase, α‐glucosidase, and acetylcholinesterase assays with the IC50 values 1.38, 2.08, 1.63, and 0.58 mg/ml, respectively. By comparing with their positive controls, it was found that PTFELDC had good antioxidant activities, and lipase, α‐amylase, and α‐glucosidase inhibitory activities, However, the acetylcholinesterase inhibitory activity was relatively weaker. These results suggested that PTFELDC have a promising potential as natural antioxidant, antilipidemic, and hypoglycemic drugs used in functional foods or pharmaceuticals.

And due to the advantages of low toxic and side effects compared with the chemically synthesized drugs, both the crude and purified total flavonoids in plants have attracted attention in nutraceutical and pharmaceutical resource researches (Zhang, Lie et al., 2017).
However, compared with the other plants, the extraction and purification of total flavonoids from Eupatorium lindleyanum DC. (TFELDC) and the evaluation of their antioxidant and enzyme inhibitory activities have been rarely study, which may lead to the negative impact on their development and utilization.
Ultrasonic and microwave radiations have been well known by their effect in accelerating the extracting process and improving the bioactivity of the extraction (Ameer et al., 2017;Ben Ticha et al., 2017;Hsieh et al., 2016;Sökmen et al., 2017). The interest on applying sonochemistry to natural product extraction has increased because of its advantages (e.g., reduction in extraction time, saving in energy, and increased yield) (Rodrigues et al., 2008).
Microwave can heat the extracts quickly and accelerate the extraction process for adsorption and desorption of the targeted compounds from matrix, while its disadvantage is inhomogeneous heating effect (You et al., 2014). Hence, combining microwave with ultrasonic extraction is a complementary technique and may present some more advantages. At present, UMSE has been applied to the extraction of polysaccharides, volatile oil, alkali, and flavonoids and compared with other extraction methods, such as soak extraction, heat extraction, ultrasonic-assisted extraction, and microwave-assisted extraction, which has the following merits: high extraction rate, fast speed, simple technics, pollutionfree, etc (Cheng et al., 2011;Jiang et al., 2019;Sun et al., 2019;Wang et al., 2018).
In this study, UMSE was applied as a rapid method of extraction of TFELDC, and AB-8 macroporous resin was employed in the purification of TFELDC to achieve the purified TFELDC (PTFELDC). In addition, the antioxidant and enzyme inhibitory activities of PTFELDC were evaluated using DPPH radical, superoxide anion radical, reducing power, and ferric reducing antioxidant power assays, and lipase, α-amylase, α-glucosidase, and acetylcholinesterase assays. Findings in this study are helpful to obtain an effective and nontoxic natural antilipidemic, hypoglycemic, and antitumor source for functional foods and pharmaceuticals.

| Plant material
Eupatorium lindleyanum DC. was purchased from Bozhou Haiyitang medicines procurement Co., Ltd. and identified by Prof. Weidong Wang, Department of Food Sciences, Xuzhou University of Technology on 20 March 2018, which was taxonomically authenticated and deposited with voucher number no. 20180322-2 at the herbarium of Food Department, Xuzhou University of Technology, Xuzhou. It was dried in an oven Boxun,Shanghai,China) with air circulation at 60°C, and the dried material was ground into the fine and homogeneous powder using a grinder Jingcheng,Qingzhou,Shandong,China). Then, the sample powder was sieved through a 60 mesh sieve and kept in a sealed plastic bag at room temperature prior to use.

| Ultrasonic-microwave synergistic extraction (UMSE)
The ultrasonic-microwave synergistic extraction apparatus (CW-2000, Xintuo) equipped with an ultrasonic control device (power from 10 to 800 W) and a microwave control device (power 50 W, frequency 40 kHz). 4.0 g of the sample powder was weighed exactly and placed in a 100 ml quartz extraction cell equipped with reflux system. After the apparatus was turned on, extraction time was counted and the extraction was carried out continuously at the preset parameters. When the extraction was completed, and the extracts were collected, filtrated, and fixed for the further determination of total flavonoid content. After the analysis, the remaining extract was concentrated, freeze-dried, and stored at −20°C.

| Experimental design
The effects of ethanol volume fraction, liquid-to-solid ratio (L/S ratio), microwave power, and extraction time on the extraction yield of TFELDC were firstly investigated. And then, on the basis of results above, the Box-Behnken design (BBD) (Samaram et al., 2015) with four independent variables at three levels was used to optimize the extraction conditions of UMSE of TFELDC.
The levels of the three factors were as follows: ethanol volume fraction (x 1 ) of 70%, 80%, and 90%; L/S ratio (x 2 ) of 10, 12, and 14 ml/g; microwave power (x 3 ) of 200, 300, and 400 W; and extraction time (x 3 ) of 120, 140, and 160 s. The extraction yield was taken as the dependent variable.
The generalized second-order polynomial model used in the response surface analysis was as follows: where β 0 is the constant coefficient; β i , β ij , and β ii are the regression coefficients for the linear, interaction, and quadratic terms; x i and x j are the independent variable actual value.

| Determination of extraction yield
The extraction yield of TFELDC was calculated according to the following equation by the method of Alara et al. (2018) with some modifications.
where W e is the mass of rutin equivalents extracted in the solution, and W t is the mass of E. lindleyanum DC.

| Purification of TFELDC
The purification of TFELDC was carried out according to the method of  with some modifications. Dynamic adsorption and desorption tests were carried out on a column (2.6 × 60 cm) wet-packed with AB-8 macroporous resin, and the bed volume (BV) was 290 ml. The dynamic adsorption tests were performed as follows: The pH value of sample solution (2 g of crude extract dissolved in 1,000 ml of deionized water) was adjusted to 4.6 with 1 M hydrochloric acid and loaded onto the column at a flow rate of 1 BV/h. The dynamic desorption tests were performed as follows: After sample being loaded, distilled water was used for elution until the eluent was colorless, and then the column was eluted by 3.5 BV of 70% ethanol at a flow rate of 0.5 BV/h. At last, the elution solution was collected, concentrated, and lyophilized for studying its phytochemical properties and biological activities in vitro, respectively.

| Determination of total flavonoid content
The total flavonoid content was determined according to the colorimetric method developed by Benabderrahim et al. (2019) with minor modifications. Briefly, 5.4 ml of different concentrations of rutin and 0.3 ml of 5% (w/v) sodium nitrite were mixed in a 10-ml colorimetric tubes. After the mixture was incubated for 6 min in the dark at 37°C.
Following the addition of 0.3 ml of 10% (w/v) aluminum nitrate, the mixture was incubated for 6 min in the dark at 37°C. Subsequently, 4 ml of 1 M sodium hydroxide was added into them, and the mixture was incubated for 10 min in the dark at 37°C. The absorbance of the mixture at 510 nm was measured using a UV-Vis spectrophotometer (Shanghai Spectrum Instrument Co. Ltd.). The total flavonoid contents in TFELDC or PTFELDC were expressed as mg rutin equivalents per gram of freeze-dried sample (mg RE/g FDS) through the calibration curve of rutin.

| DPPH radical scavenging assay
The DPPH radical scavenging activity of PTFELDC was assessed using the method in the research by Chen and Huang (2019) with minor modifications. Briefly, 50 µl of different concentrations of PTFELDC was mixed with 150 µl of 0.15 mM DPPH (in ethanol) in 96-well plates. After the mixture was incubated for 60 min in the dark at 37°C, the absorbance of the mixture at 517 nm was measured using a microplate reader (Synergy H1, Bio-Tek). Ascorbic acid was used as a positive control. The scavenging rate was calculated as follows: where A i is the absorbance in the presence of the sample, A j is the absorbance without DPPH, and A c is the absorbance of the control (without sample).

| Superoxide anion radical scavenging assay
The superoxide anion radical scavenging activity of the samples was assessed using the method in the research by Sfahlan (3) Scavenging rate (\% ) = (1 - Zhang et al. (2018) and Zhang et al. (2015)  where A rp is the reducing power, A i is the absorbance in the presence of the sample, and A c is the absorbance of the control (without sample).

| Ferric reducing antioxidant power (FRAP) assay
The antioxidant potential of PTFELDC was assessed by using the for 10 min in the dark at 37°C, the absorbance at 593 nm was measured using a microplate reader. Ascorbic acid was used as a positive control. A frap was the ferric reducing antioxidant power, which was calculated as Equation (4).

| Lipase inhibitory assay
The lipase inhibitory activity of PTFELDC was measured using the method in the research by Spínola et al. (2018) with minor modifications. Briefly, 50 μl of different concentrations of PTFELDC was mixed with 50 μl of 1.2 U/ml pancreatic lipase solution in PBS (0.1 M, pH 8.0). After the mixture was incubated for 10 min in the dark at 37°C, 100 μl of 0.2 mM p-NPP (in PBS), which was preincubated at 37°C for 10 min, was added. Subsequently, the mixture was incubated for 20 min in the dark at 37°C. The absorbance at 405 nm was measured using a microplate reader. Orlistat was used as a positive control. The inhibitory rate was calculated as follows: where A i is the absorbance in the presence of sample, enzyme, and substrate, A j is the absorbance in the presence of sample, PBS (instead of enzyme), and substrate; A c is the absorbance in the presence of PBS (instead of sample), enzyme, and substrate; A d is the absorbance in the presence of PBS (instead of sample and enzyme) and substrate.

| α-amylase inhibitory assay
The α-amylase inhibitory activity of PTFELDC was measured using the method in the research by Guo et al. (2019) with minor modifications. Briefly, 700 μL of varying concentrations of sample solution was mixed with 600 μL of 1% (m/v) soluble starch in test tubes. After the mixture was incubated at 37°C in the dark for 5 min, 200 μL of 20 U/ml α-amylase solution (in deionized water), which was preincubated at 37°C in the dark for 5 min, was added and incubated at 37°C in the dark for 5 min. Subsequently, 500 μL of 1% (m/v) dinitrosalicylic acid reagent was added and the mixture was incubated at 100°C in the dark for 5 min. The mixture was cooled down to room temperature and diluted to 10 ml with deionized water. Finally, 200 μL from each sample was seeded in a 96-well plate, and the absorbance at 540 nm was measured using a microplate reader. Acarbose was used as a positive control. The inhibitory rate was calculated as Equation (5).

| α-glucosidase inhibitory assay
The α-glucosidase inhibitory activity of PTFELDC was meas- incubated for 10 min in the dark at 37°C. Then, 20 μl of 40 U/ml a-glucosidase (in PBS), which was preincubated at 37°C for 5 min, was added to the above mixture. Subsequently, the mixture was incubated for 20 min in the dark at 37°C. Finally, the reaction was terminated by the addition of 100 μl of 0.3 mM sodium carbonate, and the absorbance at 405 nm was measured using a microplate reader. Acarbose was used as a positive control. The inhibitory rate was calculated as Equation (5).

| Acetylcholinesterase inhibitory assay
The acetylcholinesterase inhibitory activity of PTFELDC was 40 μl of 1.875 mM acetylthiocholine iodide (in PBS), which were preincubated at 37°C for 5 min, were added. Subsequently, the mixture was incubated for 20 min in the dark at 37°C. The absorbance at 405 nm was measured using a microplate reader.
Galantamine was used as a positive control. The inhibitory rate was calculated as Equation (5).

| Statistical analysis
All experiments were performed in triplicate and expressed as

| Effects of ethanol volume fraction
Different concentrations of solvent have an effect on the polarity of the solvent, so it is very important to find a optimal concentration in order to get a higher extraction rate (Yuan et al., 2014). In general, the solubility of total flavonoids in a certain concentration of ethanol is excellent. Therefore, ethanol was applied to extract TFELDC and the results were depicted in Figure 1a. The extraction rates of TFELDC firstly increased from 0.667 ± 0.015% to 0.759 ± 0.012% and then declined to 0.690 ± 0.012% with the increase in ethanol volume fraction from 60% to 100%. This was because that the polarity of the mixer changes when water and ethanol were mixed together. At a level of 80% (v/v) ethanol, the polarities of the mixer and total flavonoids were in close proximity. Based on the theory of "similarity and intermiscibility," the extraction rate reached its peak value under the conditions. Based on the current results, 80% was considered to be the optimum ethanol volume fraction in the present experiment.
3.1.2 | Effects of L/S ratio L/S ratio is an important extraction parameter affecting the extraction rate of flavonoids (Wang et al., 2010), and the results of the effects of L/S ratio on the extraction rates of TFELDC were depicted in Figure 1b.
The extraction rates of TFELDC increased from 0.642 ± 0.013% to 0.825 ± 0.016%, as the L/S ratios increased from 6:1 to 12:1 ml/g, but a further increase in the L/S ratio decreased the extraction rate to 0.794 ± 0.016%. This was because that a larger L/S ratio implied greater concentration difference between the plant material and the exterior extraction solvent, and so the diffusion of target compounds were more rapidly . Additionally, the higher L/S ratio could reduce the viscosity of the extraction solvent, which can make the diffusion of target components easier and facilitate acoustic cavitation (Tsiaka et al., 2015). However, when L/S ratio was too high, both of the microwave and ultrasonic intensity imposed on the average vegetal tissue were weakened, which was not conducive for the fragmentation of raw material (Xu et al., 2014). At the same time, from an economic perspective, using a large amount of extraction solvent was not considered cost-effective due to the high operating cost of extraction solvents and energy consumption (Yang et al., 2013). So, L/S ratio of 12:1 ml/g was selected for further study.

| Effects of microwave power
Microwave power significantly influenced the target components yield in the extraction process (Dahmoune et al., 2014). As shown in Figure 1c, with the increase in microwave power, the extraction yields of TFELDC increased. This was because microwave can crack the cell wall, disrupt the tissue structure of material, and accelerate the dissolution of target components (Ji et al., 2007), resulting in high extraction yield of TFELDC in earlier stage of UMSE in the study. However, when microwave power was above 300 W, the extraction yield declined. This was because overexposure under microwave irradiation resulted in local high temperature, which might decompose target components (Chen et al., 2017). From the above results, 300 W was chosen as the suitable microwave power in the further discussion.

| Effects of extraction time
The extraction time has also a significant effect on the extraction yield (Setyaningsih et al., 2015). In order to investigate the extraction efficiency and optimize the extraction process of UVSE, the extraction time was investigated and the results were depicted in Figure 1d. The extraction rates of TFELDC increased rapidly from 0.812 ± 0.019% to 1.032 ± 0.009%, when extraction time varied from 100 to 140 s, and reached a maximum at 140 s. However, the extraction rate was slowly decreased to 0.801 ± 0.019% with the prolonging of extraction time.
This was because that as extraction time increases, both of microwave and ultrasonic wave could effectively disrupt the cell walls, leading to the release of target compounds into the exterior extraction solvent and the increase in extraction rates at the early period of extraction (Maran et al., 2015). But the exposure of microwave or ultrasonic treatment for long time caused the decomposition of certain sensitive bioactive compounds, which reduced the extraction rates (Chen et al., 2017;Xu et al., 2016). Consequently, it can be obtained that 140 s was chosen as the suitable extraction time.

| Establishment of a model and analysis of response surface
3.2.1 | Fitting the model A regression analysis was carried out to fit mathematical models to the experimental data (Table 1) aiming at an optimal region for the responses studied. The predicted model could be described by the following equation in terms of coded values: The significance of each coefficient was determined using the Fvalue and p-value in Table 2. The corresponding variables would be more

| Analysis of response surface
The regression model was able to predict the effects of the four factors on the extraction yield of TFELDC. The relationship between independent and dependent variables was illustrated in 2-D contour plots and 3-D response surface plots, which were generated by the model (Figure 2). Two variables were depicted in 3-D response surface plots while the other two variables kept constant. The shape of 2-D contour plots reflected the degree of the mutual effect. The oval suggested that the effect is significant while roundness insignificant. As shown in the Figure 2, the mutual effects between any two variables were insignificant at 0.05 level.

| Verification experiments
The optimal extraction conditions for achieving the maximal extraction yield of TFELDC obtained by the regression Equation The predicted extraction yield was 1.044%. Under these conditions, the experimental extraction yield was 1.050 ± 0.006% (n = 3), which was well-matched with the predicted value. The

| Purification of TFELDC
AB-8 macroporous resin, a type of weak polar resin, has been extensively applied in the purification of plant flavonoids, because of its physicochemical stability, adsorption/desorption selectivity, and recyclability (Gu et al., 2016;. After TFELDC were purified by AB-8 macroporous resin, the total flavonoid contents of PTFELDC increased from 208.18 ± 1.60 to 511.19 ± 3.21 mg RE/g FDS. The content of total flavonoids was increased by 2.46 times. Therefore, it could be concluded that the purification of TFELDC by AB-8 macroporous resin was an effective method. The results were in agreement with earlier studies (Hamed et al., 2019;Zhang et al., 2011).

| Antioxidant activities
3.4.1 | DPPH radical scavenging activity DPPH radical scavenging assay, based on the reduction of DPPH solution in the presence of a proton-donating substance, has been extensively used to evaluate the free radical scavenging ability of varied samples (Chen et al., 2012). The scavenging activity of PTFELDC on DPPH radicals was shown in Figure 3a with ascorbic acid as a control standard. It was obvious that the scavenging rates of PTFELDC on DPPH radicals were positively correlated with increasing concentrations. At the concentration of 10 μg/ml, PTFELDC showed the scavenging rates of 30.12 ± 0.94% on DPPH radical, and at 100 μg/ml, the scavenging rate increased to 76.08 ± 0.70%. The parameter used to compare the DPPH radical scavenging activity of PTFELDC and ascorbic acid is IC 50 value, defined as the concentrations of the sample at which the scavenging rate reaches 50% (Guo et al., 2011) and calculated by the probit regression (de Lima et al., 2015). Practically, a lower IC 50 value corresponds to stronger antioxidant activity of tested sample (Kozarski et al., 2011). The IC 50 value of PTFELDC was 37.13 μg/ml, which was 6.84 times that of ascorbic acid (5.43 μg/ml).
Although the IC 50 value for the standard ascorbic acid in the assay is significantly different from those (90 or 140 μg/ml) in the literature reported by  and , it is basically equivalent to those (5.28 or 4.11 μg/ml) in the literature reported by Dong et al. (2019) and . This is because that the IC 50 for the standard ascorbic acid is related to the determination conditions (such as temperature, time, and optical path), calculation method (such as whether to remove the background and the choice of experimental point).

| RP activity
The reducing power assay measures the electron-donating ability of antioxidants by the potassium ferricyanide reduction method. The presence of antioxidants would result in the reduction capacity of potassium ferricyanide (Fe 3+ ) to potassium ferrocyanide (Fe 2+ ), which further reduces the production of ferric ferrocyanide. The reducing capacity of samples may serve as an index to evaluate the potential antioxidant properties (Yang et al., 2011). As shown in Figure 3c, the A rp values of PTFELDC showed a concentration-dependent manner with increasing concentrations. At the concentration of 10 μg/ml, PTFELDC showed the A rp was 0.070 ± 0.004, and at 100 μg/ml, the A rp increased to 0.617 ± 0.012. IC 50 value is defined as the effective concentration at which the absorbance is 0.5 for reducing power and was calculated by interpolation from linear regression analysis (Guo et al., 2011). The IC 50 values of PTFELDC and ascorbic acid were 81.22 and 6.78 μg/ml, which indicated that PTFELDC has better reducing power.

| FRAP activity
The FRAP activity measures the reduction capacity of the TPTZ-Fe(III) complex to its ferrous TPTZ-Fe(II) form using antioxidants (Sokamte et al., 2019). As shown in Figure 3d, the A frap values of PTFELDC showed a concentration-dependent manner with increasing concentrations. At the concentration of 3 μg/ml, PTFELDC showed the A frap was 0.072 ± 0.003, and at 30 μg/ml, the A frap increased to 0.587 ± 0.008. The IC 50 values of PTFELDC and ascorbic acid were 24.72 and 2.63 μg/ml, which indicated that PTFELDC has better ferric reducing antioxidant power.

| Lipase inhibitory activity
Lipase is responsible for the metabolism of triglycerides, which is the key enzyme in the process of fat hydrolyzate in organism (Spínola et al., 2020). When the activity of lipase is too high, the organism will produce excessive monoglycerides and fatty acids, which will lead to the hyperlipidemia. By inhibiting the activity of lipase, the lipase inhibitor can reduce the content of lipid hydrolyzate and accelerate lipid exclusion from the body so as to achieve the goal of losing weight. At present, the commonly used lipase inhibitors are orlistat. However, the drug orlistat, a potent and specific longterm gastrointestinal lipase inhibitor, has serious side effects, such as oily stools, diarrhea, abdominal pain, and fecal spots (Filippatos et al., 2008). Therefore, it is hoped to discover new alternatives by scholars at home and abroad, which could increase the therapeutic effect and reduce the side effects. To this day, many natural plant extracts have been reported to possess lipase inhibitory activity (Herrera et al., 2019;Inthongkaew et al., 2017;Sosnowska et al., 2018;Świerczewska et al., 2019). The lipase inhibitory activity of PTFELDC was plotted in Figure 4a. The PTFELDC were able to inhibit the lipase activity dose-dependently. When the concentrations of PTFELDC were in the range of 0.40-3.20 mg/ml, the effects of them, expressed as inhibition rates, increased from 14.26 ± 3.27% to 89.16 ± 2.15%. Orlistat used as a positive control exhibited the lipase inhibition rates from 11.89 ± 0.79% to 93.91 ± 4.54%, when its concentrations changed from 0.04 to 0.32 mg/ml. The results showed that the IC 50 values of PTFELDC and orlistat were 1.38 and 0.16 mg/ml, respectively. However, this does not prevent PTFELDC from becoming a good lipase inhibitor.

| α-amylase inhibitory activity
Inhibition of α-amylase, an enzyme that plays a key role in digestion of starch, is considered as a strategy in the treatment of type II diabetes (Berrout et al., 2018). The α-amylase inhibitory activity of PTFELDC was plotted in Figure 4b. The PTFELDC (from 0.90 to 7.20 mg/ml) were able to inhibit the α-amylase activity dosedependently, and the inhibition rates of them increased from 31.93 ± 0.54% to 95.97 ± 1.61%. Acarbose used as a positive control exhibited the α-amylase inhibition rates from 10. is basically equivalent to that (acarbose, 1.89 mg/ml) in the literature reported by Oyedeji-Amusa & Ashafa (2019). This is because that the IC 50 value for the standard is related to types of standard and the detection system.

| α-glucosidase inhibitory activity
The α-glucosidase located on the brush border surface membrane of intestinal cells is responsible for degrading oligosaccharides into monosaccharides, thus leading to an increase in blood glucose level (Baessa et al., 2019;Kim et al., 2010). Therefore, α-glucosidase has been recognized as a potential targets for the management of diabetes (Beidokhti et al., 2020). The α-glucosidase activity of PTFELDC was plotted in Figure 4c. The PTFELDC (from 0.60 to 4.80 mg/ml) were able to inhibit the α-glucosidase activity dose-dependently, and the inhibition rates of them increased from 29.46 ± 3.23% to 78.99 ± 1.71%. Acarbose used as a positive control exhibited the α-glucosidase inhibition rates from 16.67 ± 1.18% to 61.96 ± 1.67%, when its concentrations changed from 0.15 to 1.20 mg/ml. The results showed that the IC 50 values of α-glucosidase activity inhibition of PTFELDC and acarbose were 1.63 and 0.94 mg/ml, respectively. The disparity between the effects of PTFELDC and acarbose on α-glucosidase inhibitory activity was less than 2 times, so PTFELDC was a good αglucosidase inhibitor. The IC 50 value for the standard acarbose in the assay is significantly different from that (voglibose, 10.47 mg/ ml) in the literature reported by Ng and Rosman (2019), and it is basically equivalent to that (acarbose, 1.02 mg/ml) in the literature reported by Oyedeji-Amusa and Ashafa (2019). The reason for this phenomenon is similar to that in α-amylase inhibition assay.

| Acetylcholinesterase inhibitory activity
Acetylcholinesterase is a kind of transmitter hydrolase, which mainly exists in the postsynaptic membrane of human neuromuscular tissue (Oliveira et al., 2011). Its major biological function is to hydrolyze acetylcholine at the cholinergic synapse rapidly, prevent acetylcholine accumulation, and maintain the normal physiological function of the nervous system (Masuoka et al., 2019). Thus, inhibition of acetylcholinesterase has been considered as a promising approach for the treatment of Alzheimer's disease and possible therapeutic applications for the treatment of Parkinson's disease, aging, and myasthenia gravis (Bianco et al., 2015;Masondo et al., 2019). The acetylcholinesterase activity of PTFELDC was plotted in Figure 4d.
The PTFELDC (from 150 to 1,200 μg/ml) were able to inhibit the acetylcholinesterase activity dose-dependently, and the inhibition rates of them increased from 24.04 ± 2.79% to 71.48 ± 4.63%.
Galantamine used as a positive control exhibited the acetylcholinesterase inhibition rates from 38.56 ± 1.83% to 77.00 ± 1.33%, when its concentrations changed from 0.008 to 0.064 μg/ml. The results showed that the IC 50 values of PTFELDC and galantamine were 0.58 and 1.56 × 10 -5 mg/ml, respectively. Compared with galantamine, PTFELDC displayed a weak inhibitory activity on acetylcholinesterase, which was about 3.72 × 10 4 times lower.

| CON CLUS IONS
The efficient UMSE technique was used to extract TFELDC, and a Box-Behnken design was employed to optimize the extraction process. The optimal parameters were as follows: ethanol volume fraction 71.5%, L/S ratio 12.2 ml/g, microwave power 318 W, and Extraction time 143 s. Under the optimized conditions, the experimental extraction yield achieved to 1.050%, which was well-matched with the predicted value (1.044%), indicating that the model was adequate for the extraction process. After TFELDC were purified by AB-8 macroporous resin, the total flavonoid contents of PTFELDC increased from 208.18 ± 1.60 to 511.19 ± 3.21 mg RE/g FDS, which was increased by 2.46 times. Thus, it can be concluded that the purification of TFELDC by AB-8 macroporous resin is an effective method. The evaluation of the antioxidant and enzyme inhibitory activities of PTFELDC suggested that they could be developed as naturally potential antioxidants, antilipidemic, or hypoglycemic drugs for functional foods and pharmaceuticals, while the mechanism of antioxidant, antilipidemic, or hypoglycemic activities of PTFELDC in vivo need to be further elucidated. In this study, the result did not show that they were suitable to be used as an acylchoinesterase inhibitory agent, which need further discussion.

ACK N OWLED G M ENTS
We gratefully acknowledge the financial support received from the

CO N FLI C T O F I NTE R E S T
There is no conflict of interest recorded for this work.

AUTH O R CO NTR I B UTI O N
Chao Li designed the experiments, wrote and edited the original draft, and analyzed the data; Yingbin Shen: designed the experiments, wrote and edited the original draft, and involved in formal analysis; Shanglong Chen extracted the data; Jin Sha contributed to antioxidant assay; Jue Cui analyzed the data; Juping He contributed to enzyme inhibitory activities; Junning Fu wrote-review and editing.