KONG Yuanfang, YANG Bin, LI Min, WANG Ning, HU Yulong, LI Xiaofei, JIANG Shiqing, DONG Chunhong

(Henan University of Traditional Chinese Medicine, Zhengzhou 450046, Henan, China)

Abstract: The optimization of the conditions for the microwave extraction of Rehmannia glutinosa polysaccharide(RGP) was investigated using the orthogonal design and response surface methodologies to provide comparison between both optimization methods. Process parameters such as treatment time, temperature, and liquid-to-solid ratio for the microwave extraction method were analyzed. The results of RGP extraction via orthogonal design showed that the best extraction conditions, which afforded a RGP yield of 72.10%, were as follows: extraction time, 150 s; temperature, 80 ℃; liquid-to-solid ratio, 1 g∶10 mL. Meanwhile, the following conditions providing a RGP yield of 74.41% were determined as the best extraction conditions via response surface methodology: extraction time, 137 s; temperature, 89 ℃; liquid-to-solid ratio, 1 g∶8 mL. By comparing these results, it can be concluded that the response surface method offers some advantages over the orthogonal method, i.e., the extraction time is shorter, the ratio of material to liquid is smaller, and the RGP yield is higher. Therefore, the response surface method is more suitable for optimizing the process of RGP microwave extraction.

Keywords: Rehmannia glutinosa polysaccharide (RGP); microwave extraction; orthogonal design; response surface methodology; optimization comparative study

Rehmanniaglutinosa, also known as Dihuang, is an herb of the Scrophulariaceae family. According to different processing methods, it can be divided into fresh Radix Rehmannia, dried Radix Rehmannia and Radix Rehmanniae Praeparata. In records of traditional Chinese medicine,Rehmanniaglutinosaappears as “Shennong’s Herba” and is considered a top-grade herb[1]. The main components ofRehmanniaglutinosaareRehmanniaglutinosapolysaccharides (RGP) and iridoid glycosides[2]. In China,RGPis widely investigated owing to its various pharmacological effects, which include antioxidation, antitumor, and antianxiety activity[3].Fox example, in the past decades, ZHOU studied a purifiedRGPthat exhibits antihyperglycemic and antihyperlipidemic effect via oral administration, and its possible involvement in the streptozotocin (STZ)-induced mechanism in diabetic mice was suggested[4]. TOMODA isolated aRehmanniapolysaccharide SA, which exhibited reticuloendothelial system-potentiating effect in a carbon clearance test[5], from the root ofRehmanniaglutinosa, and revealed that it mainly consisted of arabino-3,6-galactan type structural units[6]. In addition, the effectiveness ofRGPas a mucosal adjuvant for inducing the activation of immune responses in the lung was demonstrated[7].

There are many methods for the extraction of the polysaccharide fraction fromRehmanniaglutinosasuch as microwave extraction, water extract-alcohol precipitation, and ultrasonic extraction. Among the available extraction methods, microwave-assisted extraction offers numerous advantages over other extraction methods, including shorter extraction time, less amount of solvents, higher extraction rate, better product yields, and lower cost[8]The microwave extraction method was first reported by GANZLERetal. in 1986[9]. Since then, microwave extraction has been applied for the extraction of many biologically active compounds. Fox example, using this method, LIetal. extracted tea polyphenols from green tea in 2010[10], Panetal. extracted tanshinones fromSalviamiltiorrhizabunge[11], and Zhangetal. extracted rutin and quercetin from the stalks ofEuonymusalatus(Thunb.)Sieb[12].

Despite these advances, there is no report on the evaluation of theRGPmicrowave extraction efficiency using different analysis method. To improve the yield of extractedRGP, the microwave extraction conditions need to be optimized; to achieve this, orthogonal test method and response surface method are the most commonly used optimization methods. To date, only one of the two optimization methods has been used to optimize theRGPextraction process. For example, QINetal. used the orthogonal design method to optimize theRGPextraction process[13], whereas YUANetal.[14]and CAOetal.[15]reported an optimizedRGPextraction using the response surface methodology.

However, to the best of our knowledge, no comparison between both optimization methods has been made, and it is difficult to conclude which one is better. In this report, the orthogonal design and response surface methodologies were evaluated for the optimization of theRGPmicrowave extraction process. A comparison between the two methods indicates that the efficiency of theRGPmicrowave extraction optimized by response surface methodology is much better than that obtained using the orthogonal design method. In the method optimized by the response surface methodology, the extraction time is shorter, the liquid-to-solid ratio is lower, and theRGPyield is higher. Therefore, the response surface methodology seems to be more suitable for optimizing theRGPmicrowave extraction process.

1 Materials and Methods

1.1 Reagents and apparatus

Rehmanniaglutinosawas collected from its original production area, Jiaozuo. Phenol and concentrated sulfuric acid were purchased from Shanghai Chemicals and Reagents Co. (Shanghai, China). Glucose was purchased from Alfa Chemical Co. (Beijing, China). The electronic balance was a Fuzhou Huazhi Science Instrument Co. Ltd. The microwave chemical reactor, the circular water vacuum pump, and the thermothermal heating magnetic mixer were purchased from Zhengzhou Kaipeng Company. The ultraviolet (UV)-visible spectrophotometer was purchased from Japan Shimadzu Corporation. The electrothermal drum drying box was purchased from Shanghai Yiheng Scientific Instrument Co. Ltd.

1.2 Glucose standard curve

Glucose was analyzed by the phenol-sulfuric acid method. Briefly, the flask was dried before use and 0.1 g of glucose was accurately weighed to prepare the standard stock solution, from which other solutions were prepared by diluting with distilled water.

The standard regression curve for glucose is as follows:

y=10.197 56x+0.038 47(R=0.994 83)

(1)

wherexrepresents the concentration of glucose solution (g/L), andyrepresents the absorbance.

To draw the standard curve of UV light absorbance versus glucose concentration at 450-550 nm, a series of glucose standard solutions with concentrations of 0.04, 0.08, 0.12, 0.16, and 0.2 g/L, containing 5% phenol and 98% concentrated sulfuric acid were prepared. The maximum absorption wavelength of glucose was determined to be 488 nm, and the curve depicted in
Fig.1 was obtained.

Fig.1 Absorbance versus glucose concentration standard curve

1.3 Extraction of Rehmannia glutinosa polysac-charide

Prior to the experiments,Rehmanniaglutinosawas washed, dried at 60 ℃ for 24 h, crushed, and passed through 100 mesh sieves. The following process for extracting polysaccharides fromRehmanniaglutinosawas performed: First, 1.00 g ofRehmanniaglutinosapowder was weighted accurately and put in a dried 50 mL three-necked flask. Then, a certain amount of distilled water was added to perform the microwave extraction. Next, the filtrate was combined and diluted to 100 mL in a volumetric flask to prepare the test solution.

From the as-prepared solution, a sample of 2.0 mL was taken out and diluted to 100 mL, Then, the absorbance of a 1.0 mL sample of the resulting solution was measured, and the concentration of polysaccharides inRehmanniaglutinosawas determined using the glucose standard curve (Fig.1).The yield ofRGPwas obtained according to the formula (2).

RGPyield(%)=(C×V×D/m)×100%

(2)

whereCrepresents the concentration of polysaccharide solution inRehmanniaglutinosa(g/L),Vis the fixed volume of the polysaccharide solution inRehmanniaglutinosa(mL),Drepresents the dilution multiples andmis the amount of theRehmanniaglutinosasamples (g).

2 Results and discussion

2.1 Single factor test

For the optimization of theRGPextraction conditions, we selected temperature, microwave extraction time, the liquid-to-solid ratio and microwave extraction times as the parameters to explore the optimum extraction condition for theRGP.

At first, the microwave extraction times were explored (Fig.2) by investigating the results obtained by performing one, two, three, four and five times. The relationship between the yield and the times of extractions were shown in
Fig.2. From the
Fig.2, we can see that the yield ofRGPonly increased by less than 1% from the two times extraction to three times extraction. Considering the extraction efficiency ofRGPand the experimental cost, we chose two times to extract theRGP.

Fig.2 Rehmannia glutinosa polysaccharide(RGP) extraction yield versus microwave extraction times

Next, we researched the extraction time when the microwave extraction times is defined. During the process, the same amount ofRehmanniaglutinosawas used to extract polysaccharide at different time with the same extraction times and the liquid-to-solid ratio (1 g∶10 mL) at 90 ℃. Five different time (90, 150, 210, 270 and 330 s) was studied. As can be seen from the
Fig.3, as the extraction time increased, theRGPyield first increased, achieving the best result at 150 s, and then remained virtually unchanged. Consequently, 150 s was adopted as the optimum extraction time for the following experiments.

Fig.3 Rehmannia glutinosa polysaccharide (RGP) extraction yield versus microwave extraction time

Then, to explore the optimum extraction temperature, we evaluated the extraction yield at 60, 70, 80, 90 and 100 ℃. The result depicted in
Fig.4 revealed that the best result was obtained at 90 ℃, which was selected at the optimum extraction temperature.

Fig.4 Rehmannia glutinosa polysaccharide (RGP) extraction yield versus microwave extraction temperature

Finally, we investigated the liquid-to-solid ratio. As can be seen from
Fig.5, the best liquid-to-solid ratio was 1∶10. Therefore, the next experiments were performed using this ratio to obtain the best results.

Fig.5 Rehmannia glutinosa polysaccharide (RGP) extraction yield versus liquid-to-solid ratio

2.2 Comparison between orthogonal design and response surface methodologies

With the best extraction conditions in hand, we then studied the orthogonal design by performing an orthogonal experiment of three factors and three levels to analyze the results (Table 1). Time, temperature, and liquid-to-solid ratio were selected as the factors for the orthogonal test. In Level 1, the factors are 90 s, 80 ℃ and a liquid-to-solid ratio of 1∶5. Level 2 represents the factors 150 s, 90 ℃ and a liquid-to-solid ratio of 1∶10. The factors 210 s, 100 ℃, and a liquid-to-solid ratio of 1∶15 corraspond to Level 3.

Table 1 Factors and levels of orthogonal tests L9(33)

Nine groups of experiments according to the orthogonal design were performed (Table 2). In entries 1-3, the factor time was set as 90 s (Level 1), and the temperature and liquid-to-solid ratio were varied, affordingRGPyields of 68.52%, 68.96%, and 70.88% for Level 1, Level 2, and Level 3, respectively. In entries 4-6, 150 s was selected at the factor time (Level 2), andRGPyields of 72.25%, 70.39%, and 66.67% were obtained for Levels 1, 2, and 3, respectively. In entries 7-9, the time was set as 210 s (Level 3), and Levels 1, 2, and 3 producedRGPyields of 69.26%, 68.24% and 68.04%, respectively. Comparing the range of R values of the three factors, the liquid-to-solid ratio had the greatest influence on the yield ofRGP, followed by the microwave extraction temperature, and the smallest was the microwave extraction time.

Table 2 Results of factors and levels of orthogonal tests L9 (33)

Next, the experiments according to the response surface methodology were conducted, establishing the levels -1, 0, 1. In Level -1, the time was 90 s, the temperature was 80 ℃, and the liquid-to-solid ratio was 1∶5. In Level 0, the time was 150 s, the temperature was 90 ℃, and the liquid-to-solid ratio was 1∶10. In Level 1, the time was 210 s, the temperature was 100 ℃, and the liquid-to-solid ratio was 1∶15 (Table 3).

Table 3 Factors and levels of the response surface test

We performed the 17 groups of experiments summarized in Table 4 using Box-Behnken Design software. Among those, 15 groups are experiments and two groups are validation groups. TheRGPyield was calculated using the following formula (3):

Y=73.13-0.60A-1.04B+0.90C+0.50AB+

0.21AC-0.65BC-1.47A2-4.73B2-3.39C2

(3)

whereArepresents the extraction time (s),Bis the temperature (℃),Cindicates the liquid-to-solid ratio (g/mL), andYrepresents theRGPyield (%) according to the Design-Expert 8.06 software.

Table 4 Factors and levels of the response surface test

TheA,B, andCcoefficients reflect the influence on theRGPyield. Using quadratic multiple regression equations as test object to perform analysis of regression variance and significant test, the results displayed in Table 5 were obtained. As can be seen, theFvalue of the regression model was determined to be 58.33 andP<0.000 1, that is, the level of significance is extremely high, indicating that the quadratic multiple regression equation has a very significant regression effect, and the relationship between the various factors and the response values is reflected correctly. TheR2is 0.996 8; the values ofFandPindicate thatB,C,A2,B2, andC2have an effect on the response valuesY.AandBCexhibit a notable effect, whereasABandAChave little effect, i.e., the temperature, the liquid-to-solid ratio and time have significant linear and square effect on theRGPyield. Since the facts have little effect except forBC, the quadratic polynomial regression equation can fit the real response with negligible error; therefore, it is suitable for theRGPmicrowave extraction method.

Table 5 Analysis of regression variance and significance test for the optimization of the Rehmannia glutinosa polysaccharide (RGP) extraction process

The interaction between extraction time, temperature, and liquid-to-solid ratio was also analyzed (Fig.6). The surface presented in the response diagram is concave, suggesting that theRGPyield has a maximum response value. The center of the contour line on the response surface is between -1 and 1, thus expressing that the best process condition forRGPextraction is in the design factor level range.

The response surface curve is relatively flat in
Fig.6a and 6b, which indicates that the three factors have little influence on theRGPyield. In contrast, as the extraction temperature and liquid-to-solid ratio increase, the response surface curves are all relatively sharp, indicating that there is an obvious change on theRGPyield (Fig.6c). It has been previously reported that the interaction between the factors is also reflected in the contours[16]; if the contour approaches the circle, the interaction between the factors is less significant.

Fig.6 Interaction diagram of the response surface test

The best condition for the microwave extraction obtained using the orthogonal design method was the following: time, 150 s; temperature, 80 ℃; liquid-to-solid ratio, 1 g∶10 mL. Using these conditions, the process was repeated three times, affording an averageRGPyield of 72.10%. The results are shown in Table 6.

Meanwhile, the best condition obtained for the microwave extraction using the response surface method was: time, 136.94 s; temperature, 88.69 ℃; liquid-to-solid ratio, 1 g∶8.3 mL. Considering the operational feasibility, this result can be expressed as follows: time, 137 s; temperature, 89 ℃; liquid-to-solid ratio, 1 g∶8 mL. The process was repeated three times using the revised conditions, and an averageRGPyield of 74.41% was obtained (Table 6).

Table 6 Optimal process validation test

3 Conclusion

This study constitutes the first report on a comparative evaluation of the optimization of anRGPmicrowave-assisted extraction process using orthogonal design method and response surface method. TheRGPyield was 74.41% using the extraction process optimized by the response surface method, whereas a slightly lower yield of 72.10% was obtained when using the orthogonal design method. By comparing the two methods, it can be concluded that theRGPextraction method optimized by the response surface methodology is superior than that obtained by the orthogonal design method.