Hesperidin, which is the most abundant flavonoid of
Citrus unshiu
(Rutaceae), has been reported to possess diverse activities and widely used as functional foods and cosmetics. For the development of functional products, extraction procedure is indispensable. Extraction conditions affect the composition of extract as well as its biological activity. Therefore, we tried to optimize extraction conditions such as extraction solvent, extraction time and extraction temperature for maximum yield of hesperidin using response surface methodology with threelevelthreefactor BoxBehnken design (BBD). Regression analysis showed a good fit of the experimental data and the optimal condition was obtained as ethanol concentration, 59.0%; temperature 71.5 ℃ and extraction time, 12.4 h. The hesperidin yield under the optimal condition was found to be 287.8 μg per 5 mg extract, which was well matched with the predicted value of 290.5 μg. These results provides optimized extraction condition for hesperidin and might be useful for the development of hesperidin as functional products like health supplements, cosmetics and medicinal products.
Introduction
Citrus unshiu
Marcov, which belongs to the family of Rutaceae, is a seedless and easypeeling Korean citrus fruit that accounts for 30% of the total fruits produced in Korea. Its dried peels have been widely used as a folk medicine in Asian countries to improve bronchial and asthmatic conditions or blood circulation.
1

4
Citrus peel, a byproduct of the citrus juice industry, contains a large amount of pectin and flavonoids. Hesperidin (
Fig. 1
), a flavanone glycoside, is one of the abundant flavonoid of
C. unshiu
.
5
It has been reported to possess various biological activities such as antioxidant, antiinflammatory, antiallergic, anticancer, antiobesity and hypolipidemic effects.
6

10
Therefore, hesperidin has been widely used as functional foods, cosmetics and medicinal products.
Chemical structure of hesperidin.
For development of functional products from
C. unshiu
, extraction procedure is indispensable. Many factors such as extraction solvent, extraction time and extraction temperature affect the composition of extract as well as its biological activity.
11

13
Therefore, optimization of extraction condition is essential for maximum efficacy. Response surface methodology can derive optimal condition by taking into several factors together. Thus, it is effective for optimization of extraction condition, especially in case of several variables.
14
,
15
In the present study, we tried to optimize extraction conditions such as extraction solvent, extraction time and extraction temperature for maximum yield using response surface methodology with threelevelthreefactor BoxBehnken design (BBD).
Experimental
Plant material
− The peels of
C. unshiu
were purchased from a local herbal market in Chungbuk, Korea in November 2013. They were identified by the herbarium of College of Pharmacy at Chungbuk National University, where a voucher specimen was deposited (CBNU201311CU). Hesperidin was purchased from SigmaAldrich Chemical Co.
Experimental design for response surface methodology
−A BoxBehnken design (BBD) with three variables and three levels was used to optimize the extraction conditions for hesperidin from
C. unshiu
peels. Extraction solvent (
X
_{1}
), extraction temperature (
X
_{2}
) and extraction time (
X
_{3}
) were chosen for independent variables. The ranges of these variables were determined as extraction solvent (
X
_{1}
, ethanol ratio as 0, 50 and 100%), extraction temperature (
X
_{2}
, 30, 60, and 90 ℃) and extraction time (
X
_{3}
, 2, 13 and 24 h) on the basis of a preliminary single factor experiment. The variables were coded at three levels (−1, 0, and 1) and the complete design consisted of 15 experimental points including three replication of the center points (all variables were coded as zero), as shown in
Table 1
. The hesperidin yields per
C. unshiu
extract were selected as the dependent responses.
A BoxBehnken design for independent variables and their responses
A BoxBehnken design for independent variables and their responses
Regression analysis was performed according to the experimental data; the mathematical model can be explained by the following equation:
Y
is the response,
β
_{0}
is the constant coefficient,
β_{i}
are the linear coefficients,
β_{ii}
are the quadratic coefficients and
β_{ij}
are the interaction coefficients. The statistical significance of the coefficients in the regression equation was checked by analysis of variance (ANOVA). The fitness of the polynomial model equation to the responses was evaluated with the coefficients of
R
^{2}
and the lack of fit was evaluated using
F
test.
HPLC conditions for quantitation of hesperidin
− Analysis was performed using a Waters HPLC system equipped with Waters 600s pumps, a 996 photodiode array detector, and Waters Empower software using Phenomenex GeminiNX 5μ C18 110A (150 × 10.0 mm) for quantitation. Chromatographic separation was accomplished using methanol with water (40:60) at a flow rate of 2.0 ml/min. The wavelength for detection was set at 280 nm.
Stock standard solution of hesperidin was prepared in DMSO at a concentration of 1.0 mg/ml. Standard working solutions were prepared with serial dilution of 1.0, 0.5, 0.25, 0.125, 0.0625 and 0.03125 mg/ml and used for a calibration curve. A good linearity of calibration curve for hesperidin was achieved with a correlation coefficient of 0.9938.
For the preparation of
C. unshiu
peel extract, 1 g of the dried powdered peels of
C. unshiu
were weighed accurately and extracted with 10 ml extraction solvent as indicated. The hesperidin yield was expressed as μg hesperidin per 5 mg
C. unshiu
peel extract.
Results and Discussion
To develop hesperidin as functional product ingredients, efficient production of hesperidin is required. Hesperidin is known to be a major compound of
C. citrus
peels and HPLC analysis clearly showed the presence of hesperidin in the total extract of
C. citrus
peels (
Fig. 2
). Therefore, the extraction conditions of
C. citrus
peels for the maximum yield of hesperidin were investigated by quantitation of hesperidin using HPLC analysis. The effects of three extraction variables, such as extraction solvent, extraction temperature and extraction time were tested in this study. Extraction solvent was chosen as ethanol because ethanol is safer than other alcohols for function foods. The ranges of these variables were selected through a preliminary single factor experiment as extraction solvent (
X
_{1}
, ethanol ratio as 0  100%), extraction temperature (
X
_{2}
, 30  90℃) and extraction time (
X
_{3}
, 2  24 h). Hesperidin yield was evaluated as hesperidin content per dried
C. unshiu
peel extract using a BoxBehnken design (BBD) with threelevel threefactor, as shown in
Table 1
.
(A) HPLC chromatogram of hesperidin and (B) HPLC chromatogram of C. citrus peels extract.
Multiple regression analysis of the experimental data yielded this secondorder polynomial regression equations as follows:
The significance of each coefficient was determined using
t
tests and
p
values (
Table 2
). ANOVA analysis of the regression equation was also used for the determination of significance and suitability (
Table 3
). Greater
F
values and smaller
p
values were considered significant. The quality of the model was also determined by lack of fit.
Regression coefficients and their significances in the secondorder polynomial regression equation for hesperidin yield
Regression coefficients and their significances in the secondorder polynomial regression equation for hesperidin yield
ANOVA for response surface regression equation
R^{2} = 0.958, adjusted R^{2} = 0.882
As shown in
Table 2
, the linear term of ethanol concentration (
X
_{1}
) and extraction temperature (
X
_{2}
) and the quadratic term of ethanol concentration (
X
_{1}
^{2}
) and extraction temperature (
X
_{2}
^{2}
) showed great importance for the hesperidin yield. However, other factors such as the linear terms of extraction time (
X
_{3}
), the quadratic terms of extraction time (
X
_{3}
^{2}
), and the interaction terms of variables,
X
_{1}
X
_{2}
,
X
_{1}
X
_{3}
and
X
_{2}
X
_{3}
did not show any significant effects. Determination of the suitability of the prediction of hesperidin yield per dried extract was confirmed by an
F
value of 12.64 and a
p
value of 0.006. The values of coefficient determination (R
^{2}
) and the adjusted coefficient determination (adj. R
^{2}
) of the predicted model in this response were 0.958 and 0.882, respectively, which suggested a high degree of correlation between observed and predicted values (
Table 3
).
The relationship between every two variables in the hesperidin yield was shown in threedimensional response surface plots based on regression equations of hesperidin yield per dried extraction (
Fig. 3
). Hesperidin yield per dried extract is mostly affected by the ethanol ratio (
X
_{1}
), which followed by extraction temperature (
X
_{2}
) and extraction time (
X
_{3}
).
Response surface plot analysis of extraction solvent (X_{1}), extraction temperature (X_{2}) and extraction time (X_{3}) on hesperidin yield.
Based on these results, the optimal extraction conditions for maximum hesperidin yield were suggested to be an extraction solvent (ethanol concentration) of 59.0%; an extraction temperature of 71.5 ℃; and an extraction time of 12.4 h for maximum hesperidin, which predicted 290.5 μg hesperidin/5 mg extract. An extract of
C. citrus
peels prepared under these conditions gave 287.8 μg hesperidin/5 mg extract, which was wellmatched with predicted values (
Table 4
).
Predicted and observed values of hesperidin yield under optimized conditions
Predicted and observed values of hesperidin yield under optimized conditions
For development of functional products using hesperidin, extraction procedure from
C. unshiu
is indispensable. In the present study, the extraction condition was optimized by taking into three factors such as extraction solvent, extraction time and extraction temperature using response surface methodology. Our present results clearly showed importance of extraction condition, especially the extraction solvent, for the maximum yield of hesperidin. Therefore, the optimized extraction condition for hesperidin might be useful for the development of hesperidin as functional products like health supplements, cosmetics and medicinal products.
Acknowledgements
This work was supported by the research grant of Chungbuk National University in 2013.
Lee S.
,
Ra J.
,
Song J. Y.
,
Gwak C.
,
Kwon H. J.
,
Yim S. V.
,
Hong S. P.
,
Kim J.
,
Lee K. H.
,
Choi J. J.
,
Park Y. S.
,
Park C. S.
,
Ahn H. J.
2011
J. Ethnopharmacol.
133
973 
979
DOI : 10.1016/j.jep.2010.07.018
Ma Y. Q.
,
Ye X. Q.
,
Fang Z. X.
,
Chen J. C.
,
Xu G. H.
,
Liu D.H.
2008
J. Agric. Food Chem.
56
5682 
5690
DOI : 10.1021/jf072474o
Parhiz H.
,
Roohbakhsh A.
,
Soltani F.
,
Rezaee R.
,
Iranshahi M.
2015
Phytother. Res.
29
323 
331
DOI : 10.1002/ptr.5256
Yumnam S.
,
Park H. S.
,
Kim M. K.
,
Nagappan A.
,
Hong G. E.
,
Lee H. J.
,
Lee W. S.
,
Kim E. H.
,
Cho J. H.
,
Shin S. C.
,
Kim G. S.
2014
PLos One
9
e101321 
DOI : 10.1371/journal.pone.0101321
Kim D. K.
,
Lee K. T.
,
Eun J. S.
,
Zee O. P.
,
Lim J. P.
,
Eum S. S
,
Kim S. H.
,
Shin T. Y.
1999
Arch. Pharm. Res.
22
642 
645
DOI : 10.1007/BF02975340
Lee E. S.
,
Lee M. K.
2013
Nat. Prod. Sci.
19
166 
172
Ferreira S. L.
,
Bruns R. E.
,
Ferreira H.S.
,
Matos G. D.
,
David J. M.
,
Brand ão G. C.
,
da Silva E. G.
,
Portugal L. A.
,
doe Reis P. S.
,
Souza A. S.
,
dos Santos W. N.
2007
Anal. Chim. Acta
597
179 
186
DOI : 10.1016/j.aca.2007.07.011