Optimizing the extraction of anti-tumor alkaloids from the stem of Berberis amurensis by response surface methodology

Industrial Crops and Products 69 (2015) 68–75

Contents lists available at ScienceDirect

Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop

Optimizing the extraction of anti-tumor alkaloids from the stem of Berberis amurensis by response surface methodology Junkai Wu a , Dan Yu a , Huifeng Sun a , Yu Zhang a , Wenwei Zhang b , Fanjia Meng b , Xiaowei Du a,∗ a Key laboratory of Chinese Materia Medica, Ministry of Education, Pharmaceutical College, Heilongjiang University of Chinese Medicine, 24 Heping Road, Harbin 150040, China b Teaching and Research Center, Heilongjiang University of Chinese Medicine, 24 Heping Road, Harbin 150040, China

a r t i c l e

i n f o

Article history: Received 9 October 2014 Received in revised form 4 January 2015 Accepted 28 January 2015 Keywords: Alkaloids of Berberis amurensis Rupr. (BAAs) Response surface methodology Extraction optimization Anti-tumor activity

a b s t r a c t Response surface methodology (RSM) using a Box–Behnken design (BBD) was employed to optimize the conditions for extraction of anti-tumor alkaloids from the stem of Berberis amurensis Rupr. (BAAs). Four independent variables (ethanol concentration, pH value, ratio of liquid to material and extraction time) were investigated and the optimal conditions for BAAs were evaluated by means of response surface methodology (RSM). Moreover, the in vitro anti-tumor activity of BAAs was investigated. The results showed that the experimental data could be fitted to a quadratic polynomial model using correlation analysis of the mathematical regression model. Response surface plots showed that all independent variables significantly influenced the extraction yield of BAAs. The optimum extraction conditions were as follows: ethanol concentration of 67.28%, pH value of 1.81, the ratio of liquid to material of 11.24:1 (mL/g), and extraction time of 1.58 h. The average experimental BAAs yield under the optimum conditions was found to be 24.63 ± 0.28 mg/g, which agreed with the predicted value of 24.33 mg/g. UPLC-PDA analysis showed that berberine was the principal alkaloid compound in B. amurensis Rupr stem. Additionally, BAAs could inhibit MCF-7 and HEPG2 cell proliferation in vitro, and the 50% inhibitory concentration (IC50) at 48 h was around 402.25 and 477.17 ␮g/mL, respectively. The anti-tumor activity of BAAs were dose-dependent. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Protoberberine alkaloids belong to benzyltetrahydroisoquinolines alkaloids, which widely distribute in the vegetable kingdom, e.g., in Ranunculaceae plant Coptis chinensis Franch., Berberidaceae plant Berberis amurensis Rupr., Rutaceae plant Phellodendron amurense Rupr., etc. (Chen et al., 2008; Li et al., 2006; Xu et al., 2006). In recent years, an increasing number of protoberberine alkaloids have been used in medicine and health-care food due to their various biological activities, such as antidiabetic (Hsu et al., 2013), antidepressive (Sun et al., 2013a), memory-enhancing (Dhingra and Kumar, 2012), and neuroprotective effects (Luo et al., 2012). At the same time, a growing amount of research has

Abbreviations: BBD, Box–Behnken design; RSM, response surface methodology; BBAs, alkaloids of Bereris amurensis Rupr. ∗ Corresponding author at: Pharmaceutical College, Heilongjiang University of Chinese Medicine, 24 Heping Road, Xiangfang Dist., Harbin 150040, PR China. Tel.: +86 451 87267031. E-mail address: [email protected] (X. Du). http://dx.doi.org/10.1016/j.indcrop.2015.01.053 0926-6690/© 2015 Elsevier B.V. All rights reserved.

shown that protoberberine alkaloids possess anti-tumor effect by inducing cell cycle arrest, p53-dependent apoptosis, inhibiting AP-1 activity or other means (Cao et al., 2013). B. amurensis Rupr. is an important medicinal plant that grows mainly in Northeast and North China, the middle and northern part of Korean peninsula and the far East of Russia. The roots, stems, fruits and leaves of B. amurensis have long been used as traditional medicinal herb for the treatment of hypertonia, dysentery, eczema and diseases of liver, intestine and also as a hemostatic agent (Lee and Kim, 1997; Yusupov et al., 1993). To date, more than ten protoberberine alkaloids, including berberine, jatrorrhizine, palmatine, have been isolated from B. amurensis (Lee and Kim, 1997; Karimov, 1993; Yusupov et al., 1993). In addition, eight phenolic constituents including six lignan derivatives and two phenylpropanoids have been isolated from this plant (Park et al., 2009). Generally, the prepared sample of crude drug could be affected by some parameters although the quality of raw material is constant. For example, pH value is a critical factor for extraction of total alkaloids owing to the special structure of the compounds. Moreover, a higher content could be obtained by increasing time,

J. Wu et al. / Industrial Crops and Products 69 (2015) 68–75

but the value may be invariant or decreased when the time reaches to a stable level. In the course of extraction, the content of active constituents would be decreased and the total volume of prepared samples would be increased along with the ratio of liquid to raw material and extraction number increasing. Finally, ethanol concentration is also a significant factor for extraction of alkaloids and appropriate concentration of ethanol should be chosen. Therefore, the five parameters exhibit important effects on the extraction of alkaloids. In order to confirm the optimal operating conditions while the others are kept at a constant level. One-factor-at-a-time technique has been widely used to achieve higher extraction yields. The major disadvantage of this method is that it could not shows interactive effects among the variables and depict the optimal parameters on the process (Bas¸ and Boyacı, 2007). Compared with a one-factor-ata-time design, mathematical and statistical techniques are mainly adopted for designing experiments, building models, evaluating the effects of parameters and confirming optimum condition of factors for desirable responses in response surface methodology. The most common design, Box–Behnken design (BBD), has been widely used to optimize technical parameters. In terms of these advantages, BBD is commonly employed in optimizing the extraction of bioactive constituents including alkaloids (Teng and Choi, 2014), flavonoids (Sheng et al., 2013), ginsenosides (Sun et al., 2013b), and polysaccharides (Zhang et al., 2014). Up to now, there is no information on optimization of protoberberine alkaloids extraction from the stem of B. amurensis. In addition, there are no detailed investigations to explore the antitumor activity of alkaloids in the stem of B. amurensis (BAAs). In the present study, the alkaloids content was considered as response value while pH value, ethanol concentration, the ratio of liquid to material, extraction time and number were selected as optimization parameters. BBD, followed by canonical and ridge analysis, was employed to optimize the process parameters of BAAs extraction. Furthermore, the anti-tumor effect of BAAs on MCF-7 and HEPG2 cell was evaluated in the search for high quality bioactive constituents for use in the pharmaceutical industry. 2. Materials and methods

69

material, extraction time and number. The extraction solution was centrifuged at 4000 rpm for 10 min to collect the supernatant. After repeated extraction, all the supernatant was filtrated and diluted to 100 mL for determination of BAAs. 2.4. Determination of BAAs Content of three protoberberine alkaloids was measured applying a previously described method (Kong et al., 2009) with some modification. Briefly diluted sample (1 mL) was evaporated to dryness in an evaporating dish in water bath (30 ◦ C). Then the residue was dissolved with MeOH–HCl (100:1, v/v) in a 100 mL volumetric flask and filtered through a 0.22 ␮m nylon filter membrane prior to injection into the UPLC system. 2.5. UPLC analysis of BAAs In the present work, BAAs were quantified simultaneously by the Waters ACQUITY UPLC® system (Waters Corporation, Milford, USA). Chromatographic separation was performed on an ACQUITY UPLC BEH C18 column (2.1 mm × 100 mm, 1.7 ␮m, Waters Corporation, Milford, USA) with the oven temperature of 35 ◦ C. The chromatographic peaks were identified with the Waters PDA e Acquity UPLC Detector. The detection wavelength was set at 345 nm. All chromatographic experiments were performed in an isocratic mode. The mobile phase consisted of a mixture of solvents of chromatographic purity, acetonitrile/water (2 mM ammonium acetate and 0.05% formic acid, 25:75, v/v, pH 3.20) (Chen et al., 2013; Qiu et al., 2012). The flow rate was set to 0.4 mL/min. Compound samples were weighed on analytical scales with an accuracy of 0.01 mg. All the samples were prepared in 5, 10, 25 and 100 mL volumetric flasks. Solution volume injections of 2.0 ␮L were performed with the use of an autosampler. Each sample was analyzed three times and the run time was 5 min. All solutions were filtered through a 0.22 ␮m membrane filter. The final result was presented as an arithmetical mean. The control of the UPLC system and detection was achieved by the Empower II Waters® software. The method was validated in terms of precision, accuracy, specificity and stability for BAAs.

2.1. Plant materials The B. amurensis Rupr. was collected from the Shangzhi city (N45◦ 14 33.62 , E127◦ 33 54.23 ) of Heilongjiang Province, China in October, 2013 and authenticated by professor Chen Wang, biological department, Harbin Normal University, People’s Republic of China. The material was dried in air and cut into slices, then ground to powder with an approximate size of 0.2–0.5 mm. The powder was kept in sealed polyethylene bags at 4 ◦ C until required. 2.2. Chemicals and reagents Acetonitrile of chromatographic grade was purchased from Merck Serono Pharmaceutical R&D Co., Ltd. (Beijing, China). Ethanol and ammonium acetate were of analytical grade and purchased from Beijing Chemical Reagents Co. (Beijing, China). Reverse osmosis Milli-Q water (Millipore, Bedford, MA, USA) was used for all solutions and dilutions. Standards of berberine hydrochloride, jatrorrhizine hydrochloride and palmatine hydrochloride, all in 98% purity, were obtained from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). 2.3. Extraction of BAAs Samples of three grams were extracted by ethanol solvent in a designed extraction concentration, pH value, the ratio of liquid to

2.6. Experimental design and statistical analyzes The main factors affecting extraction efficiency, including pH value, ethanol concentration, extraction time, the ratio of liquid to material and extraction number, had significant effects on BAAs production, so the five parameters were screened by single-factor experiment. Based on the preliminary results, the proper range of the extraction variables including X1 (ethanol concentration), X2 (pH value), X3 (ratio of liquid to material), and X4 (extraction time) was determined. Then, a three-level-four-factor BBD (Design Expert software, Trial Version 7.0.0, Stat–Ease Inc., Minneapolis, MN, USA) was applied to determine the best combination of extraction variables for the yields of BAAs. The range of independent variables and

Table 1 Variables and experimental design levels for response surface. Independent variables

Coded symbols

Ethanol concentration (%) pH value Ratio of liquid to material (mL/g) Extraction time (h)

X1 X2 X3 X4

Levels −1

0

1

60 1 8 1

65 2 10 1.5

70 3 12 2

70

J. Wu et al. / Industrial Crops and Products 69 (2015) 68–75

their levels were shown in Table 1. The three variables were coded according to the following equation: Xi =

xi − x0 ,i = 1 − 4 x

(1)

where Xi was a coded value of the variable; xi was the actual value of the variable; x0 was the actual value of the independent variable at the center point; x was the step change of the variable. The values for BAAs content in each trial were average of triplicates. The data were expressed as mean ± standard deviation (SD). On the basis of the experimental data, a second-order polynomial model corresponding to the BBD was applied to correlate the relationship between the independent variables and the response (yields of BAAs) to predict the optimized conditions. The non-linear computer-generated quadratic model was given as Y = ˇ0 +

4  i=0

ˇi Xi +

4  j=0

ˇii Xi2 +

4 4  

ˇij Xi Xj

(2)

i=0 j=0

where Y was the response function; ˇ0 was a constant; ˇi , ˇii and ˇij were the linear, quadratic and interactive coefficients, respectively; Xi was the coded levels of independent variables. The terms Xi Xj and Xi 2 represented the interaction and quadratic terms, respectively. Statistical analysis of the single-factor experimental data was evaluated by SPSS 18.0 software (SPSS Inc., Chicago, IL, USA). Design Expert software was used for the experimental design and optimized conditions. F-test and p-value were used to check the significance of the regression coefficient. The adequacy of the model was determined by evaluating the lack of fit, the coefficient of determination (R2 ) and the F-test value, which were obtained from the analysis of variance (ANOVA). Data were expressed as the means (SEM) of three replicated determinations.

incubation, the rate of cell proliferation was measured with the same method as Section 2.7.1. 3. Results and discussion 3.1. The effect of ethanol concentration on the BAAs In order to study the effect of extraction solvent, different concentrations of ethanol (55, 60, 65, 70, 75%) were prepared when other experimental parameters were set as follows: the ratio of liquid to material 10:1 (mL/g), pH value 2.0, extraction time 1.5 h and extraction number 1. It could be seen from Fig. 1a that different ethanol concentrations had important effects on yield of BAAs. The extraction yield reached the maximum (19.36 ± 0.36 mg/g) when 65% ethanol was used. It was enhanced along with the increased concentration of ethanol before it reached 65%, but declined with that from 65 to 75%. So 65% ethanol was selected as the center point for further RSM experiment. 3.2. The effect of pH value on the BAAs pH value was an critical parameter of the total alkaloids extraction (Johansen et al., 2012; Ricard et al., 2012). It was set at 1.0, 2.0, 3.0, 4.0 and 5.0, respectively, to examine the influence of that on the yield of the BAAs when other reaction conditions were as follows: ethanol concentration 65%, extraction time 1.5 h, ratio of liquid to raw material 10:1 (mL/g) and extraction number 1. Fig. 1b indicates that the yield of the BAAs reached a maximum value of 18.47 mg/g when pH value was 2.0. Consequently, pH value of 1.0–3.0 was considered to be optimal acidity in the experiment.

2.7. Evaluation of BAAs anti-tumor activities in vitro

3.3. The effect of ratio of liquid to material on the BAAs

2.7.1. The effects of BAAs on MCF-7 cell proliferation Human breast cancer MCF-7 cell line was obtained from Center for Experimental Animals of Sun Yat-sen University, Guangzhou, China. They were incubated in DMEM, supplemented with 10% fetal bovine, 100 units/mL penicillin, and 100 mg/mL streptomycin, at 37 ◦ C in a 5% CO2 environment. MCF-7 cells were seeded in 96-well plates at 5.0 × 103 cells per well and cultured for 24 h. The medium was then replaced with the same composition containing BAAs at concentrations of 10, 25, 50,100, 200, 400, 800 and 1000 ␮g/mL for 48 h. Control cells were treated with an equal amount of DMSO. Following incubation, the medium was removed and 100 ␮L of fresh medium containing 50 ␮g MTT were then added into each well for 4 h incubation. Medium was then removed, and 150 ␮L DMSO were added for 15 min. A microplate reader (Thermo Multiskan MK3, USA) was applied to test absorbance at 570 nm (Mosmann, 1983). The rate of cell proliferation was calculated according to the absorbance measured relative to the absorbance of control cells exposed to the vehicle alone.

The screening of the ratio of liquid to raw material was one more important step (Luo and Chen, 2010; Zheng et al., 2009; Yin and Dang, 2008). Total alkaloids in raw material could not be extracted completely because of a low ratio of liquid to raw material, while excessive amount of solvent would cause higher cost, adversely. Therefore, suitable ratio of liquid to raw material should be selected for extraction of targeted alkaloids. In the study, effect of different ratio of liquid to material (6:1, 8:1, 10:1, 12:1 and 14:1) on the extraction yield was investigated when the other reaction conditions were set as follows: ethanol concentration 65%, pH value 2.0, extraction time 1.5 h, and extraction number 1. As seen from Fig. 1c, the yield of the BAAs reached the critical value 18.21 mg/g at the ratio of 10:1 (mL/g), and then started to maintain a dynamic equilibrium. Therefore, the range of ratios of 8–12:1 (mL/g) was used in the present work.

2.7.2. The effects of BAAs on HEPG2 cell proliferation HepG2 cells (Center for Experimental Animals of Sun Yat-sen University, Guangzhou, China) were cultured in DMEM, supplemented with 10% fetal bovine serum, 100 units/mL penicillin, 100 mg/mL streptomycin, and 2 mM l-glutamine in a humidified atmosphere of 5% CO2 at 37 ◦ C. HepG2 cells were seeded into 96-well plates at 5 × 103 cells per well. 24 h after plating, the medium was discarded and then the fresh medium containing BAAs at different concentrations of 10, 25, 50,100, 200, 400, 800 and 1000 ␮g/mL was added for 48 h. Control cells were treated with an equal amount of DMSO. Following

3.4. The effect of extraction time on the BAAs Under the above optimal conditions of ethanol concentration (65%), pH value (2.0) and the ratio of liquid to material (10:1), effects of extraction time (0.5, 1, 1.5, 2, 2.5 h) on the yield of total alkaloids were tested with extraction number 1. The results were displayed in Fig. 1d. The yield of the BAAs reached a maximum value of 18.69 mg/g, then it maintained a mild slope. Therefore, the center point of extraction time chosen for RSM was 1.5 h. 3.5. The effect of extraction number on the BAAs Extraction number was also a very essential factor for extraction of active components from plant materials. In our study, effect of

J. Wu et al. / Industrial Crops and Products 69 (2015) 68–75

71

Fig. 1. Effect of different extraction parameters (ethanol concentration, %; pH value; the ratio of liquid to material, mL/g; extraction time, h and extraction number) on yield of BAAs.

different extraction number (1, 2, 3, 4 and 5) on the yield was investigated. Five groups of samples were extracted with the optimal parameters obtained above. The result manifested that the yield of the BAAs increased with the number of extraction cycles, but there was no significant increase after 3 cycles (Fig. 1e). Taking into account of extraction efficiency, the extraction number 3 was adopted in this work.

3.6. Optimization of the procedure 3.6.1. The model fitting and statistical analysis The extraction of the BAAs was optimized through RSM approach. The twenty nine designed experiments for optimizing the four individual parameters in the current BBD were shown in Table 2, and five replicates (runs 25–29) at the center of the design were estimated by a pure error sum of squares. By applying multiple regression analysis on the experimental data, the response variable and the test variables were related by the following second-order polynomial equation:

where X1 , X2 , X3 and X4 were the coded values of ethanol concentration, pH value, the ratio of liquid to material and extraction time, respectively. Statistical testing of regression equation was checked by F-test, and ANOVA for the fitted quadratic polynomial model of extraction of BAAs were shown in Table 3. p-Value of the model was smaller than 0.0001 which indicated that the model was suitable in this experiment. The determination coefficient (R2 = 0.9259) was close to 1, manifesting that the model could explain 92.59% of the response value changes. Meanwhile, a low value 0.0363 of coefficient of the variation (C.V.) showed a high degree of precision and a good deal of reliability of the experimental value. The results showed that the model was adequate to represent the relationship between the response and the independent variables. Furthermore, the lack of fit was used to measured the adequacy of the fit. The F-value of 4.16 and p-value of 0.0911 represented that the lack of fit was insignificantly to the pure error. Adequate precision compared the range of the predicted values at the design points to the average prediction error. A ratio greater than 4 indicated

Y = −273.95125 + 7.71957X1 + 3.903X2 + 1.98917X3 + 30.229X4 − 0.061747X12 − 2.16242X22 − 0.26748X32 − 4.93467X42 + C +0.0695X1 X3 − 0.143X1 X4 + 0.13375X2 X3 + 0.745X2 X4 − 0.565X3 X4

(3)

72

J. Wu et al. / Industrial Crops and Products 69 (2015) 68–75

Table 2 The Box–Behnken experimental design with four independent variables. No.

X1 (concentration of ethanol)

X2 (pH value)

X3 (ratio of liquid to material)

X4 (extraction time)

Y (extraction yield, mg/g)

Content of jatrorrhizine (mg/g)

Content of palmatine (mg/g)

Content of berberine (mg/g)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

60 60 60 65 65 70 65 65 65 70 70 65 70 65 65 65 70 65 65 65 60 70 60 60 65 65 65 65 65

2 2 2 3 1 2 2 3 1 2 2 2 3 3 1 2 2 3 1 2 1 1 2 3 2 2 2 2 2

12 12 10 10 10 10 12 12 8 8 10 12 10 8 10 8 12 10 12 8 10 10 8 10 10 10 10 10 10

2 1.5 1 1 1 1 1 1.5 1.5 1.5 2 2 1.5 1.5 2 1 1.5 2 1.5 2 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5 1.5

19.96 20.31 18.51 17.41 20.78 21.81 22.21 20.43 20.25 20.18 21.83 24.13 20.57 17.6 22.16 17.74 22.86 20.28 22.01 21.92 19.81 22.18 20.41 17.83 23.19 23.78 24.11 23.08 23.61

2.52 2.71 2.39 2.27 2.17 2.52 2.78 2.29 2.31 2.47 2.55 2.89 2.5 1.98 2.91 2.35 3.18 2.48 2.77 2.59 2.32 2.83 2.37 2.27 2.67 2.81 3.02 2.68 2.59

2.16 2.22 2.01 1.88 2.13 2.25 2.61 2.11 2.16 2.27 2.23 2.63 2.17 1.79 2.64 2.06 2.87 2.17 2.61 2.44 2.12 2.65 2.14 2.13 2.39 2.28 2.77 2.44 2.36

15.28 15.38 14.11 13.26 16.48 17.04 16.82 16.03 15.78 15.44 17.05 18.61 15.9 13.83 16.61 13.33 16.81 15.63 16.63 16.89 15.37 16.7 15.9 13.43 18.13 18.69 18.32 17.96 18.66

adequate model discrimination. In the present study, the value of 11.3 showed an adequate signal. This model could be used to navigate the design space. The significance of each coefficient was also determined using F-value and p-value. The results were given in Table 3. From the pvalues of each model term, it could be concluded that the variables with the largest effect were the linear terms of ethanol concentration (X1 ), pH value (X2 ), ratio of liquid to material (X3 ), extraction time (X4 ), the quadratic term of ethanol concentration (X1 × X1 ), and pH value (X2 × X2 ) (p < 0.001). Besides, the quadratic term of ratio of liquid to material (X3 × X3 ) and extraction time (X4 × X4 )

were also found significant (p < 0.01). Meanwhile, the ratio of liquid to material (X3 ) was the major factor affecting the yield. However, the other term coefficients were not significant (p > 0.05). 3.6.2. Analysis of the response surface To provide a better visualization of the statistically significant factors derived from the statistical analysis, three-dimensional response surface plots for the effects of independent variables on the extraction of BAAs were given in Fig. 2a–f. These types of plots showed effects of two factors on the response at one time while the other two factors were kept at zero level in all figures.

Table 3 ANOVA for response surface quadratic model analysis of variance table. Source

Sum of squares

Degree of freedom

Mean square

F-Value

p-Value (Prob > F)

Model X1 X2 X3 X4 X1 X2 X1 X3 X1 X4 X2 X3 X2 X4 X3 X4 X1 2 X2 2 X3 2 X4 2 Residual Lack of fit Pure error Cor total R2 Adj. R2 Pred. R2 Adequate precision

102.62 13.23 14.24 15.99 11.64 0.034 1.93 0.51 0.29 0.56 1.28 15.46 30.33 7.43 9.87 8.21 7.49 0.72 110.83 0.9259 0.8519 0.6006 11.3

14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 14 10 4 28

7.33 13.23 14.24 15.99 11.64 0.034 1.93 0.51 0.29 0.56 1.28 15.46 30.33 7.43 9.87 0.59 0.75 0.18

12.5 22.56 24.28 27.26 19.86 0.058 3.3 0.87 0.49 0.95 2.18 26.36 51.73 12.66 16.84

<0.0001 0.0003 0.0002 0.0001 0.0005 0.8126 0.091 0.3663 0.4962 0.3471 0.1622 0.0002 <0.0001 0.0031 0.0011

4.16

0.0911

J. Wu et al. / Industrial Crops and Products 69 (2015) 68–75

73

Fig. 2. Response surface (3D) showing the effect of different extraction parameters (X1 : ethanol concentration, %; X2 : pH value; X3 : the ratio of liquid to material, mL/g and X4 : extraction time, h) added on the response Y.

Fig. 2a represents the effects of ethanol concentration and pH value on the yield of BAAs. Both extraction time and liquid to solid ratio had positive effects on the response. It can be seen that by increasing the ethanol concentration BAAs yield increased as well, reached a maximum value, and the further increase had slightly effect. However, when the ethanol concentration was increased from 60% to 67%, the yield ascended greatly with the enhanced acidity. Fig. 2b depicts the interaction effect of ethanol concentration and the ratio of liquid to material on the yield. As the ratio of liquid to material reached to a high level, the yield increased with an rising of ethanol concentration. Therefore, a slightly higher temperature was required to achieve maximum increase. Fig. 2c shows that extraction time exhibited a weaker effect whereas ethanol concentration represented a significant effect on BAAs yield. The yield could be significantly achieved with the increases of ethanol concentration, especially at high levels of extraction time. According to Fig. 2d, as the ratio of liquid to material increased from 8:1 to 11:1 (mL/g), BAAs yield increased. In addition, the result

was consistent with the preliminary experimental result and could determine the accurate value of the parameter. The effects of pH value and extraction time on the yield of BAAs were shown in Fig. 2e. The yield mainly depended upon pH value that resulted in a curvilinear increase until zero level 2.0, and then decreased in alkaloids yield. The yield increased greatly when extraction time range was from 1.0 to 1.65 h, and the time curve indicated that 1.65 h was required to achieve maximum increasement. The effects of the ratio of liquid to material and extraction time on the yield of BAAs could be seen in Fig. 2f. It was obvious that the lower the ratio of liquid to material resulted in lower yield. The effect of extraction time was nearly equivalent to that of ratio of liquid to material. 3.6.3. Validation of predictive model The optimum conditions for extraction of BAAs were obtained from Design-Expert 7.0.0 software. The optimum conditions for independent variables and the predicted values of the responses were also presented as follows: ethanol concentration of 67.28%,

74

J. Wu et al. / Industrial Crops and Products 69 (2015) 68–75

Table 4 Predicted and experimental values of the responses at optimum conditions. Optimum condition

Extraction yield (mg/g)

Concentration of ethanol (%)

pH value

Ratio of liquid to material (mL/g)

Extraction time (h)

Experimental

Predicted

67.28

1.81

11.24

1.58

24.63 ± 0.28

24.33

Fig. 4. Effects of BAAs on cell proliferation (mean ± SD, n = 10); (a) MCF-7 cell; (b) HEPG2.

3.7. In vitro assays

Fig. 3. The HPLC chromatograms of standard samples (a) and purified alkaloids (b) at 345 nm; 1. jatrorrhizine; 2. palmatine; 3. berberine.

the pH value of 1.81, the ratio of liquid to material of 11.24:1 (mL/g), extraction time of 1.58 h and extraction number of 3, respectively. Under these conditions, the model gave predicted values of extraction yield of 24.33 mg/g. To verify the suitability of the equation model used for predicting the optimum response values, the optimization study was performed. Under the optimal conditions, the experimental yield of BAAs was 24.63 ± 0.28 mg/g (N = 3), which was close to the predicted value. The difference between the real value and the predicted one was not significant (p > 0.05), which indicated that the model was adequate for the extraction process (Table 4).

3.6.4. Alkaloid compounds identification Fig. 3 shows the UPLC chromatogram of standards, including jatrorrhizine (21.76 ng/mL), palmatine (17.68 ng/mL), berberine (18.32 ng/mL), and extract. By comparing relative retention time and UV–vis spectra with those of reference standards, three alkaloid compounds were identified. It can be observed that the berberine is the predominant alkaloid.

3.7.1. Suppression of MCF-7 cell proliferation by BAAs We measured proliferation of MCF-7 cell after treatment with multiple concentrations of BAAs for 48 h and found that it was suppressed to varying degrees (Fig. 4). BAAs at doses of 25 ␮g/mL and below did not exhibit significant inhibition. MCF-7 cell proliferation was significantly inhibited by BAAs at concentrations of 50 ␮g/mL and above (p < 0.05) in a dose-dependent manner. At the concentration of 1000 ␮g/mL, BAAs inhibited proliferation of MCF-7 cell by 80.69% (p < 0.05). The 50% inhibitory concentration (IC50) at 48 h was around 402.25 ␮g/mL. 3.7.2. Suppression of HEPG2 cell proliferation by BAAs The proliferation of HEPG2 cells was suppressed by multiple concentrations of BAAs to varying degrees after 48 h treatment (Fig. 4). It was inhibited slightly at doses below 50 ␮g/mL, however, it was extremely significantly inhibited at concentrations above 100 ␮g/mL. Markedly, the degree of inhibition increased in a dose-dependent manner. At the concentration of 1000 ␮g/mL, BAAs inhibited HEPG2 cell proliferation by 68.75% (p < 0.05). The IC50 at 48 h was around 477.17 ␮g/mL. 4. Conclusions In this study, the extraction conditions for BAAs were optimized by BBD, and a quadratic polynomial model was obtained from RSM. It was adequate for estimating the effect of four main independent variables (ethanol concentration, pH value, ratio of liquid to mate-

J. Wu et al. / Industrial Crops and Products 69 (2015) 68–75

rial and extraction time) by using the contour and surface plots in RSM. The confirmatory experimental optimum conditions for the alkaloids were as follows: ethanol concentration of 67.28%, the pH value of 1.81, the ratio of liquid to material of 11.24:1 (mL/g), extraction time of 1.58 h and three extraction cycles. Under these conditions, the experimental yield of BAAs was 24.63 ± 0.28 mg/g, which was highly close with the predicted yield value. This process could be considered as a sustainable alternative for the industry since it allowed simplified handling and the quantity of targeted extracts to be improved. The anti-tumor activities of the BAAs in vitro including inhibition proliferation of MCF-7 and HEPG2 cells, were evaluated, which suggested that BAAs could inhibit the two cell lines at different doses. MCF-7 cell proliferation was significantly inhibited by BAAs at concentration of 50 ␮g/mL, and inhibition ratio would reach at 80.69% at a concentration of 1000 ␮g/mL. The 50% inhibitory concentration (IC50) at 48 h was around 402.25 ␮g/mL. HEPG2 cell proliferation was highly significantly inhibited by BAAs when the concentrations above 100 ␮g/mL. At the concentration of 1000 ␮g/mL, BAAs inhibited HEPG2 cell proliferation by 68.75% (p < 0.05). The IC50 at 48 h was around 477.17 ␮g/mL. Further research on the chemical structure and anti-tumor mechanism of BAAs will be carried out in the future. Acknowledgments This research was supported by grants from Key laboratory of Chinese Materia Medica of Heilongjiang University of Chinese Medicine (no. 20110101), Outstanding young teacher support plan of Heilongjiang University of Chinese Medicine. References Bas¸, D., Boyacı, I˙ .H., 2007. Modeling and optimization I: usability of response surface methodology. J. Food Eng. 78, 836–845. Cao, H., Song, S., Zhang, H., Zhang, Y., Qu, R., Yang, B., Jing, Y., Hu, T., Yan, F., Wang, B., 2013. Chemopreventive effects of berberine on intestinal tumor development in Apcmin/+mice. BMC Gastroenterol. 13, 161–163. Chen, J.H., Wang, F.M., Liu, J., Lee, F.S., Wang, X.R., Yang, H.H., 2008. Analysis of alkaloids in Coptis chinensis Franch by accelerated solvent extraction combined with ultra performance liquid chromatographic analysis with photodiode array and tandem mass spectrometry detections. Anal. Chim. Acta 613, 184–195. Chen, J.L., Zhang, Y.L., Dong, Y., Cai, G.Z., Wang, H., Gong, J.Y., Cui, H.M., 2013. Simultaneous determination of four kinds of alkaloid from Rizoma Coptidis in rat plasma by LC–MS/MS. Chin. J. Exp. Trad. Med. Formulae 19, 174–178. Dhingra, D., Kumar, V., 2012. Memory-enhancing activity of palmatine in mice using elevated plus maze and morris water maze. Adv. Pharmacol. Sci. 2012, 1–7. Hsu, Y.Y., Tseng, Y.T., Lo, Y.C., 2013. Berberine a natural antidiabetes drug, attenuates glucose neurotoxicity and promotes Nrf2-related neurite outgrowth. Toxicol. Appl. Pharmacol. 272, 787–796.

75

Johansen, K.T., Ebild, S.J., Christensen, S.B., Godejohann, M., Jaroszewski, J.W., 2012. Alkaloid analysis by high-performance liquid chromatography-solid phase extraction-nuclear magnetic resonance: new strategies going beyond the standard. J. Chromatogr. A 1270, 171–177. Karimov, A., 1993. Berberis alkaloids. Chem. Nat. Compd. 29, 415–438. Kong, W.J., Zhao, Y.L., Xiao, X.H., Jin, C., Li, Z.L., 2009. Quantitative and chemical fingerprint analysis for quality control of Rhizoma Coptidis chinensis based on UPLC-PAD combined with chemometrics methods. Phytomedicine 16, 950–959. Lee, H.Y., Kim, C.W., 1997. Studies on the constituents of Berberis amurensis Ruprecht. Kor. J. Pharmacogn. 28, 257–263. Li, C.Y., Lu, H.J., Lin, C.H., Wu, T.S., 2006. A rapid and simple determination of protoberberine alkaloids in cortex phellodendri by 1 H NMR and its application for quality control of commercial traditional Chinese medicine prescriptions. J. Pharm. Biomed. Anal. 40, 173–178. Luo, C., Chen, Y.S., 2010. Optimization of extraction technology of Se-enriched Hericium erinaceum polysaccharides by Box–Behnken statistical design and its inhibition against metal elements loss in skull. Carbohydr. Polym. 82, 854–860. Luo, T., Jiang, W., Kong, Y., Li, S., He, F., Xu, J., Wang, H.Q., 2012. The protective effects of jatrorrhizine on beta-amyloid (25–35)-induced neurotoxicity in rat cortical neurons. CNS Neurol. Disord. Drug Targets 11, 1030–1037. Mosmann, T., 1983. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J. Immunol. Methods 65, 55–63. Park, H.B., Lee, K.H., Kim, K.H., Lee, I.K., Noh, H.J., Choi, S.U., Lee, K.R., 2009. Lignans from the roots of Berberis amurensis. Nat. Prod. Sci. 15, 17–21. Qiu, L.L., Chen, L.H., Yan, D., Zhang, P., Tan, M.R., Du, X.X., Xiao, X.H., 2012. Simultaneous determination of six quaternary ammonium alkaloids in Coptidis Rhizoma by UPLC. Chin. Herb. Med. 4, 164–169. Ricard, F., Abe, E., Duverneuil-Mayer, C., Charlier, P., de la Grandmaison, G., Alvarez, J.C., 2012. Measurement of atropine and scopolamine in hair by LC–MS/MS after Datura stramonium chronic exposure. Forensic Sci. Int. 223, 256–260. Sheng, Z.L., Wan, P.F., Dong, C.L., Li, Y.H., 2013. Optimization of total flavonoids content extracted from Flos Populi using response surface methodology. Ind. Crops Prod. 43, 778–786. Sun, C.P., Gao, W.P., Zhao, B.Z., Cheng, L.Q., 2013a. Optimization of the selective preparation of 20(R)-ginsenoside Rg3 catalyzed by d,l-tartaric acid using response surface methodology. Fitoterapia 84, 213–221. Sun, S., Wang, K., Lei, H., Li, L., Tu, M., Zeng, S., Zhou, H., Jiang, H., 2013b. Inhibition of organic cation transporter 2 and 3 may be involved in the mechanism of the antidepressant-like action of berberine. Prog. Neuropsychopharmacol. Biol. Psychiatry 49, 1–6. Teng, H., Choi, Y.H., 2014. Optimization of ultrasonic-assisted extraction of bioactive alkaloid compounds from rhizoma coptidis (Coptis chinensis Franch) using response surface methodology. Food Chem. 142, 299–305. Xu, R., Dong, Q., Yu, Y., Zhao, X., Gan, X., Wu, D., Lu, Q., Xu, X., Yu, X.F., 2006. Berbamine: a novel inhibitor of bcr/abl fusion gene with potent anti-leukemia activity. Leukemia Res. 30, 17–23. Yin, G., Dang, Y., 2008. Optimization of extraction technology of the Lycium barbarum polysaccharides by Box–Behnken statistical design. Carbohydr. Polym. 74, 603–610. Yusupov, M.M., Karimov, A., Shakirov, R., Gorovoi, P.G., Faskhutdinov, M.F., Levkovich, M.G., Abdullaev, N.D., 1993. Berberis alkaloids. XXVI. An investigation of the alkaloids of Berberis amurensis. Chem. Nat. Compd. 29, 338–340. Zhang, A.Q., Li, X.Q., Xing, C., Yang, J.H., Sun, P.L., 2014. Antioxidant activity of polysaccharide extracted from Pleurotus eryngii using response surface methodology. Int. J. Biol. Macromol. 65, 28–32. Zheng, X., Wang, X., Lan, Y., Shi, J., Xue, S.J., Liu, C., 2009. Application of response surface methodology to optimize microwave-assisted extraction of silymarin from milk thistle seeds. Sep. Purif. Technol. 70, 34–40.