Received: 4 June 2020 | Revised: 4 July 2020 | Accepted: 17 July 2020 DOI: 10.1111/jfpp.14813 ORIGINAL ARTICLE Plum (Prunus domestica L.) leaves extract as a natural antioxidant: Extraction process optimization and sunflower oil oxidative stability evaluation Nesren Elsayed | Karima Said Mohamed Hammad Department of Food Science, Faculty of Agriculture, Cairo University, Egypt Correspondence Ekram Abd El-Salam Abd El-Salam, Department of Food Science, Faculty of Agriculture, Cairo University, Giza 12613, Egypt. Email: eaam2000@agr.cu.edu.eg | Ekram Abd El-Salam Abd El-Salam Abstract The current work aimed to maximize the yield of total phenolic content (TPC) and associated antioxidant activity of plum leaf extract. The feasibility of using the optimal extract as a natural antioxidant for retarding the oxidation process in sunflower oil using rancimat method was also investigated. One-factor-at-a-time (OFAT) method and 32 full factorial design were implemented in sequence to optimize the studied factors (extraction time [ET] (10–60 min) and ethanol percentage [EtOH%] (0%– 75%)). DPPH scavenging ability of the obtained extracts were positively (r ≥ 0.7493) and significantly (p < .0001) correlated to TPC yield. The optimal conditions for extraction plum leaves were 64.37% EtOH% and 60 min with desirability of 0.856. The optimal plum leaves extract was subjected to HPLC analysis. At the same concentration (200 ppm), the protection factor of BHT and optimal plum leaves extract against sunflower oxidation were insignificantly (p > .05) differed. Practical applications Plum (Prunus domestica L.) leaves are characterized by a high content of phenolic compounds which varied between 66.50 to 143.7 mg GAE/g powder. Plum leaf extracts exhibited high scavenging activity against DPPH radicals. The protection factor of the freeze-dried optimal leaves extract was insignificantly differed than that of BHT at the concentration of 200 ppm. These findings indicate that plum leaves extract could efficiently retard the oxidation process of edible oils; and consequently, improve their quality and extend their shelf life. 1 | I NTRO D U C TI O N attributed to their phenolic compounds content (Moure et al., 2001; Pokorný, 2007). Plant leaves are a great source of bioactive materials that are uti- Phenolic compounds present in all plants and comprise a group lized in various food applications due to their functional properties of greater than 8,000 identified compounds. These molecules are (Bernhoft, 2010). In this context, the antiradical activities of leaf considered as a secondary metabolite in the plant. They contribute extract of several plants such as olive leaves (Andrikopoulos, Salta, to plant growth, pollination; and protection against ultraviolet radia- Mylona, Chiou, & Boskou, 2007; Farag, Mahmoud, & Basuny, 2007), tion, pathogens, and environmental stresses. Polyphenols molecules murta leaves (Rubilar et al., 2006, 2012), Ginkgo leaves (Kobus are characterized by the existence of hydroxyl groups that are at- et al., 2009), and thyme leaves (Beddows, Jagait, & Kelly, 2000) tached to a benzene ring. The number of phenolic hydroxyl groups as were evaluated and their abilities to retard edible oils oxidation well as their location determine the antioxidant activity of phenolic were determined. The antioxidant activity of plant leaf extracts was compounds (Serra, Almeida, & Dinis, 2018). J Food Process Preserv. 2020;00:e14813. https://doi.org/10.1111/jfpp.14813 wileyonlinelibrary.com/journal/jfpp © 2020 Wiley Periodicals LLC. | 1 of 11 2 of 11 | ELSAYED Et AL. Fatty acids composition in addition to the antioxidants that are nat- and stirred for 10–60 min using a benchtop lab stirrer (Heidolph, urally found in edible oils affect their oxidative stability, which is an Germany) at maximum speed. All extraction experiments were con- important parameter to evaluate the shelf life and quality of oils during ducted at room temperature and repeated three times. The extract processing and storage (Abril et al., 2019). Polyphenols are natural anti- was filtered and kept at 8°C for further analysis. The significant ef- oxidants that have a valuable role for retarding unsaturated fatty acids fects of studied factors (extraction time [ET] and ethanol percent- oxidation and provide an efficient protection against oxidative stress in age [EtOH%]) were estimated using One-way ANOVA followed by the human body (Farhoosh, Khodaparast, & Sharif, 2009). Tukey's test at p < .05. Degree of association between variables Plum (Prunus domestica L.) is a deciduous tree and belongs to the was calculated as Pearson correlation coefficient using XLSTAT Rosaceae family. There are more than 2,000 varieties of plum and the 2014.5.03 software (Addinsoft, USA). The three levels of studied major species are found in Europe and Australia (Mocan et al., 2018). factors that significantly resulted in the highest polyphenol yield Several studies have been conducted to investigate the total phe- and radical scavenging activity were estimated and used for further nolic content of plum leaves and its associated antioxidant activity experimental design. (Gougoulias, 2015; Mocan et al., 2018). However, based on our knowledge, there is no single study has been conducted to optimize the ex- Full factorial design (FFD) traction process of plum leaves and evaluate the ability of the obtained RSM using 32FFD was further implemented to maximize the yield extract to extend oxidative stability of edible oils. Therefore, response of extracted polyphenols and its corresponding antioxidant activity. surface methodology (RSM) using 32 full factorial design (FFD) was The chosen three levels of ET (X1) an EtOH% (X 2) that result in the implemented to maximize the TPC yield of plum leaves extract and highest polyphenol yield and radical scavenging activity (Figure 1) its scavenging activity against DPPH radicals. Furthermore, rancimat were 40, 50, and 60 min and 25%, 50%, and 75%, respectively. method was used to investigate the ability of optimal plum leaves ex- Experimental data were subjected to Browne–Forsythe and One- tract to extend the oxidative stability of sunflower oil. way ANOVA test with post hoc Tukey's test (p < .05) to assess responses variance homogeneity and significant effects among various treatments (combinations), respectively. Design-Expert version 11 2 | M ATE R I A L S A N D M E TH O DS (Stat-Ease, Inc., USA) was used to implement 32 FFD. Reduced cubic model including linear, squared, and interaction terms (Equation 1) 2.1 | Materials 2.1.1 | Plant material and chemicals was used to fit experimental data. Y = b0 + k ∑ j=1 b j Xj + k ∑ j=1 b𝑗𝑗 X2j + k ∑ b𝑖𝑗 Xi Xj + i<j k ∑ b𝑖𝑗 X2i Xj + i<j k ∑ b𝑖𝑗 Xi X2j + ei i<j (1) Leaves (5 months old) of plum tree (Prunus domestica L. SSP Hollywood) where, Y is the predicted response; b0 is the intercept; bj, bjj, and bij were detached from 5 years old trees grafted on “Marianna” plum are coefficients of linear, quadratic, and interaction effects of coded root stock and grown in sandy soil at the Faculty of Agriculture farm independent variables (Xi and Xj ), respectively; and ei is the error. [Latitude 30° 1′ 38.0208″ (N), Longitude 31° 11′ 39.5628″(E)], Cairo Statistical significance of the model and their various terms were de- University, Egypt in the summer of 2019. Plum trees were received the termined using analysis of variance (ANOVA). Lack of fit test in ad- normal agriculture practices and cultivated under flood irrigation sys- dition to R 2, adjusted R 2 and predicted R 2 values were used to check tem. Plum leaves were identified by Dr. Ibrahim Hmmam, Pomology the adequacy of generated models. Department, Faculty of Agriculture, Cairo University. The leaves were dried at 40°C in a forced air oven (Shel-lab, USA) until constant weight, then they were grinded with analytical mill (Cole-Parmer, USA), sieved 2.2.2 | Total phenolic compounds up to 50 mesh and stored in a dark place at room temperature till analysis. Free antioxidant sunflower oil was obtained from Cairo Oil Folin–Ciocâlteu assay (Hosseini, Bolourian, Yaghoubi Hamgini, & and Soap Company (Egypt). Folin–Ciocalteu reagent, Gallic acid, and Ghanuni Mahababadi, 2018) was used to determine Total phenolic 2,2-diphenyl-1-picrylhydrazyl (DPPH) were purchased from Sigma compounds (TPC) of dried plum leaves extract and expressed as mg Chemical Co., Ltd (St. Louis, MO, USA). Gallic acid equivalents (GAE)/g powder. 2.2 | Methods 2.2.3 | Determination of antioxidant activity 2.2.1 | Experimental design and statistical analysis The antiradical activity of the obtained extracts against DPPH radicals was determined according to the methodology described Preliminary extraction experiments by Fang et al. (2014). DPPH inhibition percentage (DPPH IN%) To extract polyphenols, 5 gm Prunus domestica dried leaves pow- was Spectrophotometrically (Unico UV-2000, USA) determined at der were added to 100 ml aqueous ethanol solution (0%–75%) 517 nm and calculated using Equation (2): | ELSAYED Et AL. 3 of 11 F I G U R E 1 Effect of extraction time and ethanol percentage on total polyphenols content [TPC (mg GAE/g dried powder] of plum leaves and its associated DPPH radical scavenging activity [DPPH IN%] DPPH IN% = ( A517nm of DPPH solution − A517nm of DPPH and extract solution A517nm of DPPH solution ) × 100 (2) 2.2.4 | HPLC analysis 2.2.5 | Physical and chemical properties of sunflower oil Refractive index, acid value, and peroxide value of sunflower oil were determined according to AOCS official methods (AOCS, 2009). The optimal plum leaves extract was analyzed using an Agilent 1260 series HPLC system (Agilent technologies Inc. CA, USA). 2.2.6 | GC analysis of fatty acids The separation was carried out using C18 column (100 mm × 4.6 mm i.d., 5 μm). The mobile phase consisted of (A) water 0.2% Fatty acids methyl esters (FAMEs) preparation H3PO 4, (B) methanol, and (C) acetonitrile at a flow rate 0.6 ml/min. The FAMEs were prepared using a cold saponification method ac- Gradient elution was according to the following scheme: 0–11 min cording to ISO standard No. 12966-2 (ISO, 2011). (96% A, 2% B); 11–13 min (50% A, 25% B); 13–17 min (40% A, 30% B); 17–20.5 min (50% B, 50% C), and 20.5–30 min (96% A, 2% B). Identification of FAMEs Detection wavelength was set at 284 nm. The injection volume The FAMEs were analyzed by an Agilent 6,890 series gas chro- was 20 μl and the column temperature was maintained at 30°C. matography equipped by a flam ionization detector and DB23 Compounds were identified by comparing their retention time (60 m × 0.32 mm × 0.25 µm) capillary column (Agilent technologies with those from authentic standards. Calibration curves were Inc. CA, USA). The carrier gas was N2 at a flow rate of 1.6 ml/min and used to calculate the compound amounts. split ratio of 50:1. The column oven temperature was programed at 4 of 11 | ELSAYED Et AL. initial temperature of 150°C for 1 min; rising at 10°C/min to 170°C for TPC of quince and cranberry leaves, which were found to be and held for 5 min; rising at 5°C/min to 220°C and held for 3 min. varied from 89.57 to 175.36 and from 89.81 to 127.64 mg GAE/g The injector and detector temperature were set at 250 and 270°C, dry matter, respectively (Teleszko & Wojdyło, 2015). Also, TPC of respectively. Gases flow rates for the detector were 450, 40, and black tea was ranged between 80.5 to 134.9 mg GAE/g dry mat- 25 ml/min for air, H2, and N2, respectively. Fatty acid standards were ter (Khokhar & Magnusdottir, 2002). Moreover, Gougoulias (2015) used to identify the peaks. found that TPC of plum leaves was 9.381 mg GAE/g dry matter, which is lower than our results. These variations in TPC could be attributed to several factors, such as plum leaves variety, climate, 2.2.7 | Rancimat test cultivation conditions, harvesting time, extraction methods, and solvent type (Brahmi, Mechri, Dhibi, & Hammami, 2013; Nashwa The optimal plum leaves extract was first concentrated using ro- & Abdel-Aziz, 2014). tary evaporator (EYELA rotary evaporator N-1000, Japan) at 40°C Several analytical methods have been adopted to evaluate the and then, freeze-dried (Edward freeze dryer (3,983), England). The antioxidant activity of different materials. Among these analytical lyophilized extract at concentration of 100, 200, and 400 ppm methods, DPPH method is extensively used to evaluate in vitro scav- GAE and BHT at 200 ppm were individually added to a free anti- enging ability of plant extracts against free DPPH radicals due to its oxidant sunflower oil. The tested oil samples (2.5 g) were placed in efficiency, simplicity, and cheapness (Kandi & Charles, 2019). DPPH rancimat tubes (Metrohm's 743, Switzerland), which were exposed IN% was increased from 60.07 ± 0.26 to 80.51 ± 0.89% as ET in- to air flow rate of 20 L/h and elevated temperature of 110°C. The creased from 10 to 40 min, respectively (Figure 1b); however, further organic acids, which were produced as a result of thermal de- increase in ET significantly decreased DPPH IN% with low decre- composition of oil, were absorbed in deionized water filling the ment to 77.01 ± 1.41%. The correlation between TPC and DPPH measuring vessel. Conductivity measuring cell was used to con- IN% of the tested extracts was significant (p < .001) and positive tinuously monitor the formed organic acids. The induction periods (r = 0.7853). Tohidi, Rahimmalek, and Arzani (2017) reported that of different samples were used to calculate the protection factor polyphenols have a capability to scavenge reactive oxygen interme- using Equation (3) diate compounds without any further support of oxidative reactions. Protection factor = induction period of sample with antioxidant induction period of sample without antioxidant (control) 3.1.2 | Effect of ethanol percentage (EtOH%) (3) At ET of 50 min, Data illustrated in Figure 1c show that increas- 3 | R E S U LT S A N D D I S CU S S I O N ing EtOH% to 50% significantly (p < .05) increased TPC yield to 119.12 ± 1.75 mg GAE/g powder; however, further increase in 3.1 | Preliminary extraction experiments EtOH% insignificantly (p > .05) decreased the yield of TPC to 116.23 ± 1.53 mg GAE/g powder. Similar trend was observed during Preliminary extraction experiments were conducted to investigate extraction polyphenols from Myrtuscommunis L. leaves using differ- the individual effect of studied factors (ET and EtOH%) on the yield ent EtOH% (Dahmoune, Nayak, Moussi, Remini, & Madani, 2015). of polyphenols and its corresponding antioxidant activity. Thus, The reason behind this decrement in TPC yield at higher EtOH% through all preliminary experiments only one factor was changed could be attributed to the binding affinity of polyphenols for pro- while the other extraction conditions were kept constant. The ob- tein (Papadopoulou & Frazier, 2004). Higher EtOH% induce protein tained results are outlined in Figure 1. denaturation which in turn decreased the dissolution of polyphenols leading to low TPC yield (Dahmoune et al., 2015). Furthermore, the low TPC yield at high EtOH% might be ascribed to the effect of 3.1.1 | Effect of extraction time (ET) ethanol on barrier properties of the plant cell membrane (Frontuto et al., 2019). Different ET (10–60 min) were used to extract polyphenols Moreover, the highest TPC yield at EtOH% of 50% could be from dried plum leaves powder at EtOH% of 50% (Figure 1a). returned to the polarity of ethanol water mixture. The polarity of The highest and lowest yields of TPC were 119.40 ± 1.50 and ethanol is lower than that of water. Thus, adding water to etha- 94.63 ± 4.72 mg GAE/g powder which were obtained at ET of nol increases its polarity regarding the proportion of added water. 60 and 10 min, respectively. In general, it could be noted that Polyphenols are polar compounds, which are more soluble in eth- extending ET increased TPC yield until it reached equilibrium at anol concentration of 50% (high polarity) other than high ethanol 50 min and there was no change in TPC yield after that. Similar concentration of 75% (low polarity) according to the principle of “like behavior was observed during extraction polyphenols from olive dissolve like” (Dahmoune et al., 2015; Zhang et al., 2007). leaves (Mkaouar, Gelicus, Bahloul, Allaf, & Kechaou, 2016). The Similar to TPC recovery pattern, the DPPH IN% of the obtained obtained results are consistent with previously reported data extracts was significantly (p < .05) increased as EtOH% increased to | ELSAYED Et AL. 5 of 11 YTPC(mg GAE∕g powder) = 119.10 + 2.32X1 + 15.52X2 + 6.25X1 X2 − 1.98X21 50% and significantly decreased after that. The correlation between TPC and DPPH IN% of the tested extracts was significant (p < .001) −18.35X22 + 11.68X12 X2 + 9.02X1 X22 and positive (r = 0.9638). (4) YDPPHIN% = 75.95 − 1.75X1 + 6.27X2 + 0.524X1 X2 3.2 | Full factorial design +3.34X21 − 13.40X22 − 0.563X21 X2 + 1.63X1 X22 (5) Three level FFD was found to be more adequate to fit experi- Multiple regression analysis in addition to analysis of variance mental data comparing with other experimental designs (Rakić, (ANOVA) was implemented to assess the adequacy of the obtained Kasagić-Vujanović, Jovanović, Jančić-Stojanović, & Ivanović, 2014). models to represent the variability of response variables (Table 2). Therefore, in the present study three level FFD was chosen. Mean Low probability values (p < .0001) of the obtained models indicate values of the actual and predicated values of TPC yield of plum their high significances. Despite, regression coefficients of X1 X2 , X21 X2 leaves extracts and their DPPH IN% are listed in Table 1. Probability and X1 X22 for DPPH IN% which were insignificant (p > .05), all other values of Brown–Forsythe's test (p ≥ .2486) and one-way ANOVA regression coefficients were significant (p ≤ .0320). Adjusted R 2 (p < .0001) indicate that all dependent variables are homoscedas- values of the obtained models were 0.9921 and 0.9647 for TPC tic and significantly differed. Implementing the preceding tests and DPPH IN% and their lack of fit were insignificant (p = .9368 are essential for performing FFD (Granato, de Araújo Calado, & and 0.0800, respectively), which show the ability of the obtained Jarvis, 2014). models to explain 99.21% and 96.47% of the variance, respectively TPC yield of various extracts were significantly (p < .0001) (Pedro, Granato, & Rosso, 2016). According to Maran, Priya, and varied between 66.50 ± 0.10 mg GAE/g powder (run 1) and Manikandan (2014), the desirable value of signal-to-noise ratio (ad- 143.57 ± 2.52 mg GAE/g powder (run 9), while the DPPH IN% were equate precision) is greater than 4. In the present study, adequate significantly (p < .0001) varied between 55.74 ± 0.77% (run 4) and precision values were greater than 28 which indicate the adequacy 80.51 ± 0.90% (run 2). The correlation between TPC and DPPH IN% of the signals. CV% values of the obtained models were lower than under applied extraction conditions was significant (p < .0001) and 2.3 which indicates high degree of experimental values precisions positive (r = 0.7493). and adequately of the obtained models (Maran et al., 2014). Results of fitting experimental data using various models are outlined in Table 2. Despite the cubic model which was confounding or aliased, the reduced cubic model was the most suitable model to 3.2.1 | Effect of extraction process variables fit experimental data. As, the reduced cubic model was significant (p < .0001), lack of fit was insignificant (p ≥ .0800) and it has the Equations (4) and (5) were used to draw 3D plots (Figure 2c,d) which highest values of R2, adjusted R 2 and predicted R 2. were used in addition to perturbation plots (Figure 2a,b) to illustrate In terms of coded factors, reduced cubic models (Equations 4 and 5) were adopted to the following forms. the interactive and individual effects of studied factors on response variables. Perturbation plots were drawn at midpoint of studied TA B L E 1 Full factorial design (FFD), experimental means, and predicted values of plum leaves extract total polyphenols content (TPC) and their radical scavenging activity against DPPH Run Extraction time (min) X1 Ethanol conc. (%) X2 TPC yield (mg GAE/gm powder) DPPH inhibition % Experimental Experimental Predicted f c Predicted 1 40 25 66.50 ± 0.10 66.48 61.09 ± 0.69 60.82 2 40 50 114.77b ± 5.01 114.80 80.51a ± 0.90 81.04 c b 3 40 75 108.40 ± 1.50 108.38 71.46 ± 0.38 71.19 4 50 25 85.20 d ± 0.60 85.23 55.74d ± 0.77 56.28 b a 5 50 50 119.17 ± 1.75 119.10 77.01 ± 2.52 75.95 6 50 75 116.23b ± 1.53 116.27 68.29b ± 1.22 68.82 e 7 60 25 76.67 ± 0.42 76.65 8 60 50 119.40 b ± 1.50 119.43 75 a 9 60 143.57 ± 2.52 P (Browne–Forsythe) P (ANOVA) Note: Values are expressed as means ± standard deviations of three replicates. Different letters in the same column indicate significant differences at p < .05. 143.55 59.80 cd ± 2.40 59.54 77.01a ± 1.41 77.54 b 72.27 ± 1.53 0.2486 0.9298 <0.0001 <0.0001 72.01 | 6 of 11 ELSAYED Et AL. TA B L E 2 Adequacy of the tested models and ANOVA analysis of the reduced cubic-order models in addition to their statistical parameters Std. Dev. R2 Adjusted R2 Predicted R2 PRESS Prob < F Prob (Lack of Fit) Linear 12.02 0.7607 0.7408 0.6997 4,351.53 <0.0001 <0.0001 2FI 11.42 0.7931 0.7661 0.7257 3,975.40 0.0706 <0.0001 Model TPC Quadratic 6.74 0.9341 0.9184 0.8931 1549.95 <0.0001 <0.0001 Reduced Cubic 2.10 0.9942 0.9921 0.9884 168.81 <0.0001 0.9368 2.10 0.9942 0.9921 0.9884 168.81 <0.0001 0.9368 * Cubic DPPH IN% Linear 7.09 0.3443 0.2896 0.2056 1,462.62 0.0063 <0.0001 2FI 7.24 0.3461 0.2608 0.1848 1,500.86 0.8041 <0.0001 Quadratic 1.68 0.9677 0.9601 0.9470 97.53 <0.0001 0.0611 Reduced Cubic 1.58 0.9742 0.9647 0.9507 90.76 <0.0001 0.0800 1.58 0.9742 0.9647 0.9507 90.76 0.1194 0.0800 * Cubic ANOVA analysis and statistical parameters of reduced cubic-order models Source Model DPPH IN% TPC RC SS p value RC S p value 119.10 14,408.90 <.0001 75.95 1793.63 <.0001 X1 2.32 32.20 .0141 −1.75 18.36 .0139 X2 15.52 1,444.60 <.0001 6.27 236.07 <.0001 6.25 468.75 <.0001 0.5241 3.30 .2651 X 21 −1.98 23.60 .0320 67.05 <.0001 X 22 1,077.53 <.0001 1.27 .4848 10.64 .0530 X1X2 3.34 −18.35 2020.34 <.0001 −13.40 X 21 X 2 11.68 546.00 <.0001 −0.5632 X 1 X 22 9.02 325.20 <.0001 Residual 83.68 Lack of fit 0.0300 Pure error C.V.% Adeq precision 47.48 .9368 7.62 83.65 Cor total Mean 1.63 14,492.59 105.54 1.99 67.4622 .0800 39.86 1841.11 69.24 2.28 28.7802 *Aliased model. variables (coded 0.0 (50 min, 50%)) to compare the influences of these variables on different responses. The results in Figure 2c show that the highest TPC yield was observed at the highest levels of ET and EtOH%. All regression coefficients for TPC yield were statistically significant (p ≤ .0320) which Effect of extraction process variables on TPC yield result in curvilinear change of TPC yield for all investigated factors. Data in Figure 2a illustrate that TPC yield was more sensitive to Two distinct effects of EtOH% on TPC yield were observed. At low the change in EtOH% than the change in ET. As, the line (B) that ET, it could be noted that increasing EtOH% increased the yield of represent the change of TPC yield with changing EtOH% at con- TPC till maximum which was observed at the range of 55%–65% and stant ET is highly curved, while the line (A) that represent the further increase in EtOH% decreased the yield of TPC. At high ET, change of TPC yield with changing ET at constant EtOH% is ap- increase EtOH% resulted in progressive increment of TPC yield. proximately flat line. Furthermore, regression coefficients of linear terms for ET and EtOH% were 2.32 and 15.52 (Table 2), Effect of extraction process variables on DPPH IN% respectively, which show greater effect of EtOH% than ET on Data in Figure 2b indicate that DPPH IN% was sensitive to the change TPC yield. of both EtOH% and ET. However, its sensitivity against the change in | ELSAYED Et AL. 7 of 11 F I G U R E 2 Perturbation and response surface plots of total polyphenols content [TPC (mg GAE/ g dried powder] of plum leaves (a & c) and its associated DPPH radical scavenging activity [DPPH I%] (b & d) EtOH% is higher. The values of regression coefficient of linear terms activity were 60 min and EtOH% of 64.37% with desirability of of both ET and EtOH% sustained the preceding observation. As, 0.856. To validate the obtained model, extraction of plum leaves their values were −1.75 and 6.27, respectively. Significances of linear powder using aqueous ethanol solution (65%) for 60 min were car- and quadratic terms for DPPH IN% resulted in a curvilinear change ried out. The predicted and validated values of TPC yield and DPPH of DPPH IN% for all investigated variables (Figure 2d). Two different IN% were 135.57 and 136.67 ± 2.52 mg GAE/g powder and 77.24 trends were observed as a result of increasing EtOH%. For EtOH% and 76.54 ± 0.86%, respectively. The predicted and validated values lower than 62%, DPPH IN% was increased as EtOH% increased. are very close indicating the adequacy of the obtained models to However, for percentages higher than 62%, increasing EtOH% re- predict experimental data. sulted in a decrease of DPPH IN%. 3.2.2 | Optimal extraction process conditions determination and model validation 3.3 | HPLC analysis Identification and quantification of phenolic components of the plum leaves extract obtained at optimal conditions were determined Desirability function was implemented to maximize the yield of TPC using HPLC (Figure 3 and Table 3). The total identified components and its associated radical scavenging activity. The optimum condi- were 18 compounds. The major identified phenolic compounds tions for extracting TPC with the highest DPPH radical scavenging were o-Coumaric acid, Rosmarinic acid, Resveratrol, Quercetin, 8 of 11 | FIGURE 3 ELSAYED Et AL. HPLC profile of phenolic compounds in plum leaves extract obtained at optimal conditions TA B L E 3 Phenolic compounds of plum leaves extract obtained at optimal conditions Compounds Retention time (min) Concentration (µg/g leaf) Guitard, Paul, Nardello-Rataj, & Aubry, 2016; Gülçin, 2010; Lesjak et al., 2018; Peñalvo et al., 2016), which showed high scavenging activities against DPPH radicals. Thus, high antioxidant activity of the obtained extracts could be related to its content of phenolic com- Gallic acid 3.71 24 ± 1.41 pounds and its synergistic effect with other ingredients present in Catechin 8.956 18 ± 0.80 the same extract (Xu et al., 2017). Chlorogenic acid 9.38 81 ± 1.50 Vanillic acid 9.693 499 ± 25.00 164 ± 10.00 3.4 | Rancimat test Caffeic acid 10.14 Syringic acid 10.402 p-Coumaric acid 12.933 61 ± 4.50 Benzoic acid 14.075 2,991 ± 225.00 fatty acids composition of the tested oil sample are in accordance Ferulic acid 15.016 1888 ± 112.01 with those specified values for sunflower oil (Codex Alimentarius Rutin 16.318 1,359 ± 51.33 Commission, 2005) which indicates the authenticity of the oil sam- Ellagic acid 16.956 264 ± 7.50 ple. Also, data in Table 4 and Figure 4 indicates that the sunflower oil o-Coumaric acid 17.517 6,836 ± 570.01 is rich with polyunsaturated fatty acids (greater than 88%) especially Resveratrol 19.465 4,060 ± 152.00 essential fatty acids (ω6 and ω3) (Delplanque, 2000). Quercetin 21.332 3,839 ± 113.00 3,300 ± 80.00 Rosmarinic acid 21.805 4,465 ± 51.99 Naringenin 22.286 1844 ± 66.00 Myricetin 23.301 3,180 ± 21.00 Kaempferol 24.595 2,678 ± 75.00 Note: Values are expressed as means ± standard deviations of two replicates. Physical and chemical characteristics of sunflower oil are listed in Table 4. Refractive index value, acid value, peroxide value, and The influence of optimal plum leaves extract on oxidative stability of sunflower oil was investigated using rancimat method and the obtained results are listed in Table 5. Adding antioxidants to sunflower oil significantly (p < .05) increased its induction period. The induction period of sunflower oil containing plum leaves extract at 200 ppm GAE was 4.730 ± 0.007 hr, which was close to the induction period of sunflower oil containing BHT at the same concentration. Moreover, data in Table 5 reveal that increasing plum extract concentration from 100 to 400 (ppm GAE) had insig- Syringic acid, Myricetin, Benzoic acid, and Kaempferol. Several nificant (p > .05) effect on extending the induction period of sun- studies have been conducted to examine the antioxidant activity flower. Despite the highest protection factor was recorded for BHT of identified phenolic compounds (Erkan, Ayranci, & Ayranci, 2008; (1.187 ± 0.012), it was insignificantly (p > .05) differed than that | ELSAYED Et AL. TA B L E 4 9 of 11 Physical and chemical properties of sunflower oil Parameter Sunflower oil Refractive index at 20°C/20°C 1.4750 ± 0.00 Acid value (mg KOH/g oil) 0.27 ± 0.01 Peroxide value (m. equiv./kg oil) 0.67 ± 0.02 Fatty acids Relative area percentage Myristic acid 0.072 ± 0.005 Palmitic acid 6.506 ± 0.425 Palmitoleic acid 0.106 ± 0.003 Stearic acid 3.574 ± 0.162 Oleic acid 28.617 ± 1.723 Linoleic acid 59.527 ± 3.124 Alpha-linolenic acid 0.291 ± 0.012 Arachidic acid 0.271 ± 0.015 Behenic acid 0.729 ± 0.041 Total unsaturated fatty acids 88.541 Total saturated fatty acids 11.152 Unknowns 0.307 Note: Values are expressed as means ± standard deviations of two replicates. FIGURE 4 Typical GC chromatogram of sunflower fatty acid methyl esters for plum extract (200 ppm GAE). Ghosh, Upadhyay, Mahato, and They ascribed the oxidative inhibition effect of Ginkgo leaves ex- Mishra (2019) found that the induction periods of antioxidant free tract to bioactive components of the extract and its interactions. sunflower oil under the same air flow rate were 6.13 and 2.74 hr at Additionally, Beddows et al. (2000) found that induction period of 100 and 110°C, respectively. Kobus et al. (2009) found that add- sunflower oil at 105°C was increased from 8.3 to 9.9 and 11.4 hr ing ethanolic Ginkgo leaves extract at concentration of 200 ppm as a result of adding thyme and turmeric extract to the oil, respec- to stripped triacylglycerols of rape seed oil increased its induc- tively. They also found that adding mixture of both extracts to the tion period from 8.14 hr for antioxidant free sample to 10.04 hr. oil increased its induction period to 13.8 hr. 10 of 11 | ELSAYED Et AL. TA B L E 5 Effect of the concentration of plum leaves extract on oxidative stability of sun flower oil according to Rancimat analysis Sample Induction period (hours) Protection factor Control 4.230 c ± 0.078 – a 1.187a ± 0.012 b 5.020 ± 0.042 BHT (200 ppm) Plum extract (100 ppm GAE) 4.650 ± 0.007 1.099b ± 0.022 Plum extract (200 ppm GAE) 4.730 b ± 0.007 1.118ab ± 0.022 Plum extract (400 ppm GAE) 4.720 b ± 0.051 1.116b ± 0.009 Note: Values are expressed as means ± standard deviations of two replicates. Different letters in the same column indicate significant differences at p < .05. 4 | CO N C LU S I O N The results of the present work showed that the antioxidant activity of ethanolic plum leaves extracts was highly dependent on their phenolic compounds content. TPC yield and DPPH IN% of the obtained extracts were more sensitive to EtOH% change than changing ET. Reduced cubic model was found to be the best model to represent experimental data. RSM using 32 FFD successively optimized the extraction parameters (64.37% EtOH%, 60 min) and TPC yield and DPPH IN% of optimal extract were 136.67 ± 2.52 mg GEA/g powder and 76.54 ± 0.86%, respectively. The major identified phenolic components were o-Coumaric acid, Rosmarinic acid, Resveratrol, Quercetin, Syringic acid, Myricetin, Benzoic acid, and Kaempferol. Optimal plum leaves extract showed superior inhibitory characteristics for sunflower oxidation process. At concentration of 200 ppm GAE, the protection factor of optimal plum leaves extract was non statistically differed than that of BHT. C O N FL I C T O F I N T E R E S T The authors have declared no conflicts of interest for this article. ORCID Nesren Elsayed https://orcid.org/0000-0002-7040-3142 Karima Said Mohamed Hammad https://orcid. org/0000-0003-4983-7920 Ekram Abd El-Salam Abd El-Salam https://orcid. org/0000-0003-1144-1672 REFERENCES Abril, D., Mirabal-Gallardo, Y., González, A., Marican, A., Durán-Lara, E. F., Santos, L. S., & Valdés, O. (2019). Comparison of the oxidative stability and antioxidant activity of extra-virgin olive oil and oils extracted from seeds of Colliguaya integerrima and Cynara cardunculus under normal conditions and after thermal treatment. Antioxidants, 8(10), 470. https://doi.org/10.3390/antiox8100470 Andrikopoulos, N. K., Salta, F. N., Mylona, A., Chiou, A., & Boskou, G. (2007). Oxidative stability of edible vegetable oils enriched in polyphenols with olive leaf extract. Food Science and Technology International, 13(6), 413– 421. https://doi.org/10.1177/1082013208089563 AOCS. (2009). Official and tentative methods of the American Oil Chemists’ Society (2nd ed.). Chicago, IL: Author. Beddows, C. G., Jagait, C., & Kelly, M. J. (2000). Preservation of α-tocopherol in sunflower oil by herbs and spices. International Journal of Food Sciences and Nutrition, 51(5), 327–339. https://doi. org/10.1080/096374800426920 Bernhoft, A. (2010). Bioactive compounds in plants—benefits and risks for man and animals. In Proceedings from a symposium held at The Norwegian Academy of Science and Letters, Oslo, Norway. Brahmi, F., Mechri, B., Dhibi, M., & Hammami, M. (2013). Variations in phenolic compounds and antiradical scavenging activity of Olea europaea leaves and fruits extracts collected in two different seasons. Industrial Crops and Products, 49, 256–264. https://doi.org/10.1016/j.indcrop.2013.04.042 Codex Alimentarius Commission. (2005). Codex Standard for Named Vegetable Oils, Codex Standard Series No. 210—2005, Rome. Dahmoune, F., Nayak, B., Moussi, K., Remini, H., & Madani, K. (2015). Optimization of microwave-assisted extraction of polyphenols from Myrtus communis L. leaves. Food Chemistry, 166, 585–595. https:// doi.org/10.1016/j.foodchem.2014.06.066 Delplanque, B. (2000). The nutritional value of sunflower oils: Linoleic sunflower seeds and seeds with high oleic content. Ocl—oilseeds and Fats Crops and Lipids, 7(6), 467–472. Erkan, N., Ayranci, G., & Ayranci, E. (2008). Antioxidant activities of rosemary (Rosmarinus officinalis L.) extract, blackseed (Nigella sativa L.) essential oil, carnosic acid, rosmarinic acid and sesamol. Food Chemistry, 110(1), 76–82. https://doi.org/10.1016/j.foodchem.2008.01.058 Fang, X., Wang, J., Wang, Y., Li, X., Zhou, H., & Zhu, L. (2014). Optimization of ultrasonic-assisted extraction of wedelolactone and antioxidant polyphenols from Eclipta prostrate L using response surface methodology. Separation and Purification Technology, 138, 55–64. https://doi. org/10.1016/j.seppur.2014.10.007 Farag, R. S., Mahmoud, E. A., & Basuny, A. M. (2007). Use crude olive leaf juice as a natural antioxidant for the stability of sunflower oil during heating. International Journal of Food Science and Technology, 42(1), 107–115. https://doi.org/10.1111/j.1365-2621.2006.01374.x Farhoosh, R., Khodaparast, M. H. H., & Sharif, A. (2009). Bene hull oil as a highly stable and antioxidative vegetable oil. European Journal of Lipid Science and Technology, 111(12), 1259–1265. https://doi. org/10.1002/ejlt.20090 0081 Frontuto, D., Carullo, D., Harrison, S. M., Brunton, N. P., Ferrari, G., Lyng, J. G., & Pataro, G. (2019). Optimization of pulsed electric fields-assisted extraction of polyphenols from potato peels using response surface methodology. Food and Bioprocess Technology, 12(10), 1708– 1720. https://doi.org/10.1007/s11947-019-02320 -z Ghosh, M., Upadhyay, R., Mahato, D. K., & Mishra, H. N. (2019). Kinetics of lipid oxidation in omega fatty acids rich blends of sunflower and sesame oils using Rancimat. Food Chemistry, 272, 471–477. https:// doi.org/10.1016/j.foodchem.2018.08.072 Gougoulias, N. (2015). Evaluation of antioxidant activity and polyphenol content of leaves from some fruit species. Oxidation Communications, 38(1), 35–45. Granato, D., de Araújo Calado, V. Ô. M., & Jarvis, B. (2014). Observations on the use of statistical methods in Food Science and Technology. Food Research International, 55, 137–149. https://doi.org/10.1016/j. foodres.2013.10.024 Guitard, R., Paul, J. F., Nardello-Rataj, V., & Aubry, J. M. (2016). Myricetin, rosmarinic and carnosic acids as superior natural antioxidant alternatives to α-tocopherol for the preservation of omega-3 oils. Food Chemistry, 213, 284–295. https://doi.org/10.1016/j.foodc hem.2016.06.038 Gülçin, I. (2010). Antioxidant properties of resveratrol: A structure-activity insight. Innovative Food Science and Emerging Technologies, 11(1), 210–218. https://doi.org/10.1016/j.ifset.2009.07.002 | ELSAYED Et AL. Hosseini, H., Bolourian, S., Yaghoubi Hamgini, E., & Ghanuni Mahababadi, E. (2018). Optimization of heat- and ultrasound-assisted extraction of polyphenols from dried rosemary leaves using response surface methodology. Journal of Food Processing and Preservation, 42(11), 1–15. https://doi.org/10.1111/jfpp.13778 ISO. (2011). ISO 12966-2:2011. Animal and vegetable fats and oils — Gas chromatography of fatty acid methyl esters — Part 2: Preparation of methyl esters of fatty acids. Kandi, S., & Charles, A. L. (2019). Statistical comparative study between the conventional DPPH [rad] spectrophotometric and dropping DPPH [rad] analytical method without spectrophotometer: Evaluation for the advancement of antioxidant activity analysis. Food Chemistry, 287, 338–345. 10.1016/j.foodchem.2019.02.110 Khokhar, S., & Magnusdottir, S. G. M. (2002). Total phenol, catechin, and caffeine contents of teas commonly consumed in the United Kingdom. Journal of Agricultural and Food Chemistry, 50(3), 565–570. https://doi.org/10.1021/jf010153l Kobus, J., Flaczyk, E., Siger, A., Nogala-Kałucka, M., Korczak, J., & Pegg, R. B. (2009). Phenolic compounds and antioxidant activity of extracts of Ginkgo leaves. European Journal of Lipid Science and Technology, 111(11), 1150–1160. https://doi.org/10.1002/ejlt.20080 0299 Lesjak, M., Beara, I., Simin, N., Pintać, D., Majkić, T., Bekvalac, K., … Mimica-Dukić, N. (2018). Antioxidant and anti-inflammatory activities of quercetin and its derivatives. Journal of Functional Foods, 40, 68–75. https://doi.org/10.1016/j.jff.2017.10.047 Maran, J. P., Priya, B., & Manikandan, S. (2014). Modeling and optimization of supercritical fluid extraction of anthocyanin and phenolic compounds from Syzygium cumini fruit pulp. Journal of Food Science and Technology, 51(9), 1938–1946. https://doi.org/10.1007/s1319 7-013-1237-y Mkaouar, S., Gelicus, A., Bahloul, N., Allaf, K., & Kechaou, N. (2016). Kinetic study of polyphenols extraction from olive (Olea europaea L.) leaves using instant controlled pressure drop texturing. Separation and Purification Technology, 161, 165–171. https://doi.org/10.1016/j. seppur.2016.02.002 Mocan, A., Diuzheva, A., Carradori, S., Andruch, V., Massafra, C., Moldovan, C., … Locatelli, M. (2018). Development of novel techniques to extract phenolic compounds from Romanian cultivars of Prunus domestica L. and their biological properties. Food and Chemical Toxicology, 119, 189–198. https://doi.org/10.1016/j.fct.2018.04.045 Moure, A., Cruz, J. M., Franco, D., Manuel Domínguez, J., Sineiro, J., Domínguez, H., … Carlos Parajó, J. (2001). Natural antioxidants from residual sources. Food Chemistry, 72(2), 145–171. https://doi. org/10.1016/S0308-8146(00)00223-5 Nashwa, F. S. M., & Abdel-Aziz, M. E. (2014). Efficiency of olive (Olea europaea L.) leaf extract as antioxidant and anticancer agents. Journal of Agroalimentary Processes and Technologies, 20, 46–53. Papadopoulou, A., & Frazier, R. A. (2004). Characterization of protein-polyphenol interactions. Trends in Food Science and Technology, 15(3–4), 186–190. https://doi.org/10.1016/j.tifs.2003.09.017 Pedro, A. C., Granato, D., & Rosso, N. D. (2016). Extraction of anthocyanins and polyphenols from black rice (Oryza sativa L.) by modeling and assessing their reversibility and stability. Food Chemistry, 191, 12–20. https://doi.org/10.1016/j.foodchem.2015.02.045 11 of 11 Peñalvo, G. C., Robledo, V. R., Callado, C. S. C., Santander-Ortega, M. J., Castro-Vázquez, L., Victoria Lozano, M., & Arroyo-Jiménez, M. M. (2016). Improving green enrichment of virgin olive oil by oregano. Effects on Antioxidants. Food Chemistry, 197, 509–515. https://doi. org/10.1016/j.foodchem.2015.11.002 Pokorný, J. (2007). Are natural antioxidants better—and safer—Than synthetic antioxidants? European Journal of Lipid Science and Technology, 109(6), 629–642. https://doi.org/10.1002/ejlt.20070 0064 Rakić, T., Kasagić-Vujanović, I., Jovanović, M., Jančić-Stojanović, B., & Ivanović, D. (2014). Comparison of full factorial design, central composite design, and Box-Behnken design in chromatographic method development for the determination of fluconazole and its impurities. Analytical Letters, 47(8), 1334–1347. https://doi.org/10.1080/00032 719.2013.867503 Rubilar, M., Morales, E., Sáez, R., Acevedo, F., Palma, B., Villarroel, M., & Shene, C. (2012). Polyphenolic fractions improve the oxidative stability of microencapsulated linseed oil. European Journal of Lipid Science and Technology, 114(7), 760–771. https://doi.org/10.1002/ ejlt.20110 0230 Rubilar, M., Pinelo, M., Ihl, M., Scheuermann, E., Sineiro, J., & Nuñez, M. J. (2006). Murta leaves (Ugni molinae Turcz) as a source of antioxidant polyphenols. Journal of Agricultural and Food Chemistry, 54(1), 59–64. https://doi.org/10.1021/jf051571j Serra, D., Almeida, L. M., & Dinis, T. C. P. (2018). Dietary polyphenols: A novel strategy to modulate microbiota-gut-brain axis. Trends in Food Science and Technology, 78, 224–233. https://doi.org/10.1016/j. tifs.2018.06.007 Teleszko, M., & Wojdyło, A. (2015). Comparison of phenolic compounds and antioxidant potential between selected edible fruits and their leaves. Journal of Functional Foods, 14, 736–746. https://doi. org/10.1016/j.jff.2015.02.041 Tohidi, B., Rahimmalek, M., & Arzani, A. (2017). Essential oil composition, total phenolic, flavonoid contents, and antioxidant activity of Thymus species collected from different regions of Iran. Food Chemistry, 220, 153–161. https://doi.org/10.1016/j.foodchem.2016.09.203 Xu, D. P., Zheng, J., Zhou, Y., Li, Y., Li, S., & Li, H. B. (2017). Ultrasoundassisted extraction of natural antioxidants from the flower of Limonium sinuatum: Optimization and comparison with conventional methods. Food Chemistry, 217, 552–559. https://doi.org/10.1016/j. foodchem.2016.09.013 Zhang, Z. S., Li, D., Wang, L. J., Ozkan, N., Chen, X. D., Mao, Z. H., & Yang, H. Z. (2007). Optimization of ethanol—Water extraction of lignans from flaxseed. Separation and Purification Technology, 57(1), 17–24. https://doi.org/10.1016/j.seppur.2007.03.006 How to cite this article: Elsayed N, Hammad KSM, Abd El-Salam EAE-S. Plum (Prunus domestica L.) leaves extract as a natural antioxidant: Extraction process optimization and sunflower oil oxidative stability evaluation. J Food Process Preserv. 2020;00:e14813. https://doi.org/10.1111/jfpp.14813