Experimental and chemometric study of antioxidant capacity of basil (Ocimum basilicum) extracts

Industrial Crops and Products 100 (2017) 176–182

Contents lists available at ScienceDirect

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

Experimental and chemometric study of antioxidant capacity of basil (Ocimum basilicum) extracts ´ Branislava Teofilovic´ a,b , Nevena Grujic-Leti c´ a , Svetlana Goloˇcorbin-Kon a , b c – Srdan Stojanovic´ , Gyöngyi Vastag , Slobodan Gadˇzuric´ c,∗ a

University of Novi Sad, Faculty of Medicine, Hajduk Veljkova 3, Novi Sad, Serbia University Business Academy, Faculty of Pharmacy, Trg Mladenaca 5, Novi Sad, Serbia c University of Novi Sad, Faculty of Science, Trg Dositeja Obradovica 3, Serbia b

a r t i c l e

i n f o

Article history: Received 15 December 2016 Received in revised form 6 February 2017 Accepted 26 February 2017 Keywords: Antioxidant activity Basil Free radical scavenging Phenols Flavonoids

a b s t r a c t The interest in a natural and healthy lifestyle has moved the plant crops under the spotlight. The aim of the work was to test the antioxidant activity of basil (Ocimum basilicum L.) extracts obtained by extraction with water (in presence and absence of light), methanol (95%), ethanol (30, 40, 50, 60, 96%), chloroform, dichloromethane and hexane. Fragmentation of plant material was 0.3 and 2 mm and the extraction was performed during 10 and 30 min. The total phenolic content ranged from (5.17 ± 0.15 to 65.25 ± 2.19) mg of gallic acid equivalents per gram of a dry weight of extract, and the content of the total flavonoids from (0.11 ± 0.01 to 40.63 ± 2.14) mg of quercetin per gram of a dry weight of extract. All the extracts showed an antioxidant activity with an IC50 values of 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical inhibition in the range from (0.22 ± 0.01 to 20.49 ± 1.54) ␮g/ml. The evaluation of experimental data for 44 basil extracts was performed by applying hierarchical cluster analysis (HCA) and principal component analysis (PCA). It was found that the increased time of extraction, solvent polarity and plant fragmentation increase the quality of the extracts in terms of the content of phenolic components and antioxidant effects. Extracts with the strongest antioxidant capacity were obtained by concentrated ethanol and methanol maceration. The chemometric analysis showed good correlation between the yield and total phenolic composition, and between the flavonoid content and antioxidant activity, predicting thus, basil extract quality. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Looking for a new approach to nutrition, more and more people believe that plant crops contribute directly to their health. Functional food plays an important role considering the fact that it can offer an excellent benefit (Urala and Lahteenmaki, 2007). The healthy lifestyle has a great impact on the well-being by preventing nutrition-related diseases, scavenging of free radicals and blocking chain reactions (Menrad, 2003). Free radicals are the normal product of human body metabolic reactions (Salla et al., 2016). They

Abbreviations: DPPH, 2,2–diphenyl–1–picrylhydrazyl radical; HCA, hierarchical cluster analysis; PCA, principal component analysis; FC, Folin-Ciocalteu; GAE, gallic acid equivalents; DE, dried extract; QE, quercetin equivalents; RSC, radical scavenger capacity; IC50 , inhibitory concentration; TPC, total phenolic compound; FLV, flavonoid compound. ∗ Corresponding author at: Department of Chemistry, Biochemistry and Environmental Protection, Faculty of Sciences, University of Novi Sad, Trg Dositeja ´ 3, 21000, Novi Sad, Serbia. Obradovica ´ E-mail address: [email protected] (S. Gadˇzuric). http://dx.doi.org/10.1016/j.indcrop.2017.02.039 0926-6690/© 2017 Elsevier B.V. All rights reserved.

are usually defined as any molecular species with unpaired electron in an atomic orbital capable of independent existence. The presence of an unpaired electron is responsible for many common properties shared by most radicals. Many radicals are unstable and highly reactive and can behave as oxidants or reductants due to ability to donate or to accept an electron from other molecules. The most important oxygen-containing free radicals in many diseases are hydroxyl radical, superoxide anion radical, hydrogen peroxide, oxygen singlet, hypochlorite, nitric oxide radical, and peroxynitrite radical (Lobo et al., 2010). They can be products of tobacco smoke, pollutants, radiation, organic solvents, pesticides, etc. Free radicals are highly reactive oxygen species capable of attacking unsaturated fatty acids of the membrane system and cause lipid peroxidation, which is one of the reactions that lead to oxidative stress (Kaurinovic et al., 2011). Overproduction of these free radicals can cause the oxidative damage of biomolecules (lipids, proteins, DNA) and finally induce the occurrence of many chronic diseases such as atherosclerosis, cancer, diabetes, stroke, heart diseases, gastric ulcer, aging and other degenerative diseases in humans (Masuoka et al., 2012; Cai et al., 2004; Gülc¸in et al., 2007).

B. Teofilovi´c et al. / Industrial Crops and Products 100 (2017) 176–182

Antioxidants are substances which can scavenge free radicals and prevent those disorders. They have redox properties due to their ability to reduce agents, hydrogen donors and singlet oxygen structure (Hakkim et al., 2007). The most commonly used synthetic antioxidants in food are butyl-hydroxytoluene (BHT) and butylhydroxyanisole (BHA) which are very effective as antioxidants. However, their usage in nutrition is avoided due to their structural instability and potential carcinogenic effect. Nowadays, the tendency in pharmaceutical and food industries is to replace synthetic antioxidants with the natural ones. For these reasons there is a growing interest in analyzing natural, healthy and non-toxic additives as potential antioxidants (Politeo et al., 2007). Some plants which contain many phenolic compounds are increasingly of interest in food industry due to their ability to scavenge free radicals (Gülc¸in et al., 2007). Basil (Ocimum basilicum L.) is one of the most important industrial and pharmaceutical crop species from Lamiaceae family having a major application in the food, pharmaceutical and cosmetic industries. It contains many antioxidant substances which contribute to its intense antiradical activity (Kwee and Niemeyer, 2011) and could have potential human health benefits (Flanigan and Niemeyer, 2014; Lee and Scagel, 2009). The main antioxidant compounds in basil extracts are chlorogenic, p-hydroxybenzoic, caffeic, vanillic and rosmarinic acids, as well as apigenin, quercetin and rutin. Basil is an ornamental, medicinal and aromatic plant native to Asia, Africa and India, but is widely cultivated in many countries under variety conditions (Makri and Kintzios, 2007). It is used for pharmaceutical and cosmetic preparations because of the rich phenolic and flavonoid content. Due to the strong antioxidant activity, basil acts as a protector to prevent heart diseases, reduce inflammation, lower the incidence of cancers and diabetes (Mastaneh et al., 2014). Antioxidant compounds from natural plants can be obtained by different procedures under different process conditions (time, lightening) and using various solvents. Maceration with organic solvents provides a great amount of phenolic acids and flavonoids which are holders of an antioxidant activity (Vidovic´ et al., 2012). There are many studies about the antioxidant activity of the basil extract, but there is no evidence how different extraction conditions affect yield and total phenolic compounds of obtained extracts. Factors which have an impact on the phenolic composition are important because an antioxidant capacity, bioavailability and bioefficacy strongly depend on phenolic compounds (Nguyen et al., 2010; Manach et al., 2005). Thus, the main objectives of this study were: analysis of the influence of operational conditions and extraction methods on antioxidant activity, correlation between antioxidant activity and total phenolic/flavonoid content and optimization of its application in the industry. Although basil is popular and widely consumed, there is no data of optimization of extraction procedure performed by different chemometric techniques in order to obtain extracts with the highest amount of bioactive compounds. In this work large amounts of obtained experimental data were analyzed by applying different chemometric tools – hierarchical cluster analysis (HCA) and principal component analysis (PCA). The HCA and PCA procedures were applied in order to find similarities between the analyzed data and they have been already described elsewhere (Miller and Miller, 2004). HCA searches for objects which are close together in the variable spaces and puts them into the same cluster. Successive stages of clustering can be shown on a dendrogram where the vertical axis is a measure of similarity between two objects in obtained clusters. Principal component analysis is an effective method which is used for reducing the amount of data without much loss of information finding new variables – principal components (PCs) which are linear combinations of the original variables. The principal components are formed in the way that, unlike the original variables, they are not correlated with each other. However, the principal compo-

177

nents are also chosen so that the first principal component (PC1) accounts for most of the variation in the data set, while the second (PC2) accounts for the next largest variation and so on. Hence, when the significant correlation occurs, the number of useful PCs is much smaller than the number of the original variables. Usually, two PCs are adequate to describe the most of the data variations in the specified data analysis (Vastag et al., 2014). 2. Material and methods 2.1. Chemicals Gallic acid, aluminium chloride, dichloromethane, hexane and chloroform were obtained from Sigma Aldrich (St. Louis, MO, USA), methanol and sodium-carbonate from POCH (Gliwice, Poland) and quercetin from Extra synthese (Genay Cedex, France). Ethanol was purchased from J.T. Baker (Nederland) and 2, 2-diphenyl-1-picrilhydrazil (DPPH) reagent from Alfa Aesar (Karlsruhe, Germany). The Folin-Ciocalteu’s reagent was obtained from Merck (Darmstadt, Germany). Ultra-pure water was used for the preparation of all solutions. All solvents and reagents were of an analytical grade unless indicated otherwise. 2.2. Plant material and preparation Voucher specimens (Ocimum basilicum L. 1753, Institute for Medicinal Plant Research “Dr Josif Pancic”, Belgrade, Serbia, 2-1518) were confirmed and deposited at the Herbarium BUNS of the University of Novi Sad, at the Department of Biology and Ecology, Faculty of Sciences. Ground parts of basil (Ocimum basilicum L.) were run through different sieves and standardized on 0.3 and 2 mm. The extraction was performed with ethanol-water mixtures (30, 40, 50, 60, 96%, v/v), concentrated methanol (95%, v/v), water (in presence and absence of light), dichloromethane, chloroform and hexane during different periods of time (10 and 30 min). Different concentrations of ethanol and presence/absence of light have been applied in order to find the most appropriate solvent and condition for the extraction that can be applied in any of the industrial processes. For that purpose, 1 g of dry plant was overflowed with 5 ml of solvent and shaken on magnetic stirrer. After 10 or 30 min samples were filtered and rinsed with another 5 ml of specific solvent and evaporated on rotary evaporator. Dry extracts were used for further analysis. Total extraction yield, total phenolic content, flavonoides and the inhibition of DPPH radical were determined in 44 obtained dry extracts and were performed in triplicate. 2.3. Analysis of total phenolic compounds The amount of total phenolic compounds in the extracts was determined colorimetrically with the Folin-Ciocalteu (FC) reagent (Boˇzin et al., 2008). The reaction mixture contained 0.1% dilution of a dry extract (0.1 ml), a freshly prepared 0.2 M FC reagent (0.5 ml) and a 7.5% sodium carbonate solution (0.4 ml) and it was kept in the dark under ambient conditions for 30 min to complete the reaction. The absorbance of the resulting solution was measured at 760 nm in a UV–vis spectrophotometer (model 8453 Hewlett Packard, Agilent Technologies, USA). The concentration of the total phenolic compounds was expressed as a milligram of gallic acid equivalents (GAE) per gram of a dried extract (d.e.), using the standard curve of gallic acid. All the measurements were carried out in three replicates. 2.4. Estimation of total flavonoid content The measurement of total flavonoid content in the investigated extracts was determined spectrophotometrically (Boˇzin et al.,

178

B. Teofilovi´c et al. / Industrial Crops and Products 100 (2017) 176–182

2008), using a method based on the formation of the complex flavonoid-aluminium with the absorption maximum at 430 nm. The aqueous dilutions of samples, in the amount of 1 ml, were separately mixed with 1 ml of 2% AlCl3 ·6H2 O. After the incubation at room temperature for 15 min, the absorbance of the reaction mixtures was measured at 430 nm. The flavonoids content was expressed as a milligram of quercetin equivalents (QE) per gram of a dried extract (de), by using the standard graph. All measurements were carried out in three repetitions.

2012). With the reduction of particle size of plant a supercritical area can be increased and is available for the mass transport which increases the yield extraction (Spigno et al., 2007). The extension of the extraction period provides a higher yield of extraction due to the prolongation of the contact between the plant material and the solvent. Other authors also confirmed that the yield increased with the duration of maceration (Spigno et al., 2007).

2.5. Antioxidant activity

3.2. Total phenolic content

An antioxidant activity of extracts was determined using the DPPH method (Grujic´ et al., 2012). Herbal extracts were dissolved in methanol. Different volumes (10–50 ␮l) of herbal extracts were mixed in the test tubes with 1 ml of DPPH solution (0.1 mM in methanol) and made up to the final volume of 4 ml with methanol. A control sample was prepared containing the same volume without test compounds. Methanol was used as a blank. Absorbances were measured at 515 nm. The experiment was performed in triplicate and the average absorption was noted for each concentration. Radical Scavenger Capacity (%RSC) was calculated using the following formula:

Fig. 1b shows the results of the total phenolic compounds after the extraction of basil leaves. In this study, the total phenolic content (TPC) was in the range from (5.17 ± 0.15 to 65.25 ± 2.19) mg GAE/g d.e. for polar solvents and from (6.56 ± 0.17 to 28.91 ± 0.65) mg GAE/g d.e. for non-polar solvents. These values are higher in comparison with Kaurinovic et al., where obtained values ranged from (8.45 ± 0.02 to 11.88 ± 0.02) mg GAE/g d.e. for polar and (4.21 ± 0.01 to 9.76 ± 0.03) mg GAE/g d.e. for non-polar solvents, but this can be explained by the fact that different solvents were used, namely acetic anhydride, ethyl-acetate and n-butanol. However, the same observation is reached – higher values for TPC after using more polar solvents. The highest values for the total phenolic compounds were obtained using the concentrated ethanol (96%, v/v) and concentrated methanol during 30 min, but the application of ethanol may be more adequate considering the food application of the extracts (Spigno et al., 2007). The lowest values were obtained after maceration with dichloromethane during 10 min. This is reasonable due to the fact that the majority of phenolic compounds possesses hydroxyl groups and has polar characteristics. However ethanol and methanol were more selective in comparison with water due to the fact that those alcoholic solvents are less polar and more efficient in the cell wall degradation (Lapornik et al., 2005). The examination of total phenolic content indicated that with the increase of the plant fragmentation, the phenolic composition is also increased, which was consistent with the literature (Kaurinovic et al., 2011). Time was also an important variable for the investigation of the total phenolics. The yield of TPC increased during a longer period of extraction, which is in line with other publications (Lapornik et al., 2005). Maceration during a longer period was impossible due to the enormous volatility of non-polar solvents.

%RSC = 100 × (Acontrol − Asample )/Acontrol The IC50 value, defined as the concentration of the test sample leading to 50% reduction of the free radical concentration, was calculated graphically and expressed as ␮g of the extract/mL of the final solution in the measuring cell. 2.6. Chemometric analysis The HCA and PCA were performed by Statistica v.12 software (StatSoft Inc., Tulsa) where applied solvents represent the columns (variables), while the obtained values for flavonoid content, the total phenolic compound, DPPH radical and the yield of extraction represent the rows (cases). Before calculating, data were standardized subtracting the column values from each matrix element and dividing each matrix element by the standard deviation of each column. The measure of dissimilarity between objects in cluster analysis the Euclidean distance was applied, while the Ward’s linkage method was used for testing the linkage measure. 3. Results and discussion 3.1. Extraction yield

3.3. Flavonoid content

Fig. 1a represents an influence of the polarity of solvent, fragmentation of plant material and time on the total yield extraction of basil. After the extraction with different parameters different results were obtained. Higher values were obtained for polar solvents in comparison with non-polar. Among all the samples the greatest yield of extraction was obtained by using diluted ethanol (30%). Just by observing the non-polar solvents the highest yield was achieved during the maceration with chloroform and the lowest by the extraction with hexane. This means that with the increase of solvent polarity, the extraction efficiency also increased, which was confirmed by the similar studies (Kaurinovic et al., 2011). The extraction of basil with the greater plant fragmentation (sieve 0.3 mm) gave higher values of the yield than the extraction with the smaller plant fragmentation (sieve 2 mm). This kind of results support the fact that the increase of the plant fragmentation increases the specific surface area of a plant material and consequently leads to a better contact with the solvent (Grujic´ et al.,

Results in Fig. 1c shows a flavonoid content in samples obtained using polar and non-polar solvents during a different period of time. The flavonoid content (FLV) ranged from (3.73 ± 0.14 to 40.63 ± 2.14) mg QE/g of ɑ dry extract for non-polar samples and from (0.11 ± 0.01 to 22.9 ± 1.05) mg QE/g of ɑ dry extract for polar samples. These values were slightly different from the similar results (Kaurinovic et al., 2011) where ɑ flavonoid content was from (12.98 to 23.12) mg of rutin per gram of a dry extract for non-polar and (19.28–26.42) mg of rutin per gram of a dry extract for polar solvents. These circumstances could be explained by the fact that the authors did not use completely the same solvents. Also, dissimilarities in the results might be attributed to the different ways of preparing extracts and the time of extraction. The minimum content of flavonoids was obtained by maceration with water in darkness during 10 min and the highest content was obtained by the extraction with chloroform. That phenomenon could happen due to the possibility of flavonoid substances to form glycosides with aglycons which are less polar.

B. Teofilovi´c et al. / Industrial Crops and Products 100 (2017) 176–182

179

Fig. 1. Influence of the different polarity solvents applied in dark/light, fragmentation of plant material (sieves 0.3 and 2 mm) and time (10 and 30 min) on: (a) Extraction yield; (b) Total phenolic content; (c) Flavonoid content; (d) 50% inhibition of DPPH radical (␮g/ml).

The impact of plant fragmentation (sieves 0.3 and 2 mm) on flavonoid content was also investigated. Samples with higher plant fragmentation (sieve 0.3 mm) contained more flavonoids. The extraction time was not a crucial parameter in the extraction of flavonoid substances. With the different time of extraction approximately the same results were obtained. Other experimental results confirmed that the extraction time had an influence on the total phenolic compounds and the ability of DPPH radical inhibition, but not on the flavonoid content (Thoo et al., 2010).

3.4. DPPH test Fig. 1d represents the concentrations required for inhibition of 50% of DPPH radical and showed that all the extracts of basil have certain antioxidant activity. The less concentration implies better antioxidant activity of extracts. The results for IC50 values were ranged from (0.22 ± 0.01 to 12.99 ± 0.87) ␮g/ml for polar and from (12.12 ± 0.54 to 20.49 ± 1.54) ␮g/ml for non-polar solvents. These results are similar to the ones of the survey conducted with chloroform, ether, ethyl acetate, water and n-butanol extracts of basil in which the IC50 value varied in the range from (8.17 to 24.91) ␮g/ml (Kaurinovic et al., 2011). The lowest antioxidant activity was obtained using hexane and chloroform after 10 min (sieve 2 mm). The lowest IC50 value, or the highest antioxidant activity was found in ethanol extract of basil (60%, v/v). Free radical scavenging capacity of ethanol extracts was increased with the increase of a concentration which is in line with other analysis (Thoo et al., 2010). The extraction of substances which are responsible for an antioxidant activity (chlorogenic, p-hydroxybenzoic, caffeic, ferulic, vanillic, cinnamic, rosmarinic acids, apigenin, quercetin, naringenin and rutin,) depends on the compatibility of the compounds with the solvents, according to the system “like dissolves like” principle (Zhang et al., 2007). Obviously, our extracts contained different antioxidant compounds with a different polarity. While changing the polarity of a solvent, an antioxidant activity can be changed.

The extraction time had an influence on the antioxidant activity (Thoo et al., 2010), however these differences between 10 and 30 min are not very high. Free radical scavenging capacity was slightly better during the longer extraction time. An antioxidant activity varied with the extraction time, which means that the scavenging capacity was not dependent on only one group of antioxidant substances but on many compounds present in extracts (Prior et al., 2005). Differences in the plant fragmentation had notable effects on the antioxidant activity especially when water and non-polar solvents were used. This appearance can be explained by the fact that ethanolic solvents more easily destroy cell walls in comparison with water and non-polar solvents. 3.5. Chemometric analysis The dendrogram obtained using the cluster analysis for different solvents is presented in Fig. 2a and another one for the investigated parameters in Fig. 2b. In Fig. 2a, the separation of analyzed data into two clusters was evident. The first cluster included ethanol in different concentrations (except E 96), while the second cluster, which can be divided in two sub-clusters contained the remaining solvents. It is important to emphasize that one sub-cluster involved nonpolar solvents and the second contained water, methanol and concentrated ethanol. This suggests that diluted ethanol shows the different influence on the extraction process. Applying the cluster analysis on the examined parameters (Fig. 2b) indicated that two clearly defined clusters were obtained. The first cluster contained the total flavonoid content and the activity of DPPH radical and the second cluster included the total phenolic content and the total yield extraction regardless of the plant fragmentation. This clustering was expected, because DPPH and FLV are strongly dependent, as well as TPC and the yield of the extraction as discussed earlier. Both clusters were divided into two sub-clusters. Each of these sub-clusters contains one of the investigated parameters, and within this cluster there was an additional

180

B. Teofilovi´c et al. / Industrial Crops and Products 100 (2017) 176–182

Fig. 2. (a) Dendrogram for different solvents (HEX – hexane, MH – dichlormethane; HL – chloroform; WD – water in dark; WL – water in light; M – methanol; E96–96% solution of ethanol; E60–60% solution of ethanol; E50–50% solution of ethanol; E40–40% solution of ethanol; E30–30% solution of ethanol; (b) Dendrogram of examined parameters with different fragmentation degree (sieve 0.3 and 2 mm).

Eigenvalues of correlation matrix Active variables only 6 47,27% 5

Eigenvalue

4

33,11%

3

2 10,36%

1

4,53% 2,51% 1,04% ,73% ,25% ,15% ,04% ,02%

0

-1

-2

0

2

4

6

8

10

12

14

Eigenvalue number

Fig. 3. Obtained PCs for examined parameters.

grouping according to the fragmentation of the plant material. This suggests the high influence of the plant fragmentation on the total extraction. By PCA the original data matrix was decomposed into loading (applied solvents) and score (tested parameters) vectors, whereby new variables – the principal components were obtained (Fig. 3). In our case, four principal components described 95% of the total variance in the data (PC1 47.27% PC2 33.11%, PC3 10.36 and PC4 4.53%) as presented in Fig. 3. Comparing the obtained PCs of cases, the score plot was obtained (Fig. 4a). It was evident that the first two PCs resulted in the almost

similar classification of the investigated parameters as the cluster analysis. The total flavonoid content (FLV) and the activity of DPPH radical were described with the negative values of PC1, while in the case of the total phenolic content and total yield extraction a positive PC1 was registered. Values of PC2 showed the potential for the further separation of studied parameters (FLV and DPPH) based on the values of the plant fragmentation. The PC2 values separate parameters more or less based on the plant fragmentation (sieve 03 and 2 mm). Better separation of these parameters based on the plant fragmentation (except Yield) can be observed on the basis of PC3 values (Fig. 4b).

B. Teofilovi´c et al. / Industrial Crops and Products 100 (2017) 176–182

181

Fig. 4. PCA plots of FLV (total flavonoide content) DPPH, TPC (total phenolic content) and yield obtained after 10 and 30 min using the sieves of 0.3 and 2 mm: (a) Score plot as a result of PC1 variation with PC2; (b) Score plot as a result of PC1 versus PC3.

0,2

4. Conclusion E30-E60

0,0

PC2 (33.11%)

-0,2

-0,4

MH

HL

-0,6

WL

HEX

WD

M

-0,8

E96

-1,0 -0,8

-0,6

-0,4

-0,2

0,0

0,2

0,4

0,6

0,8

1,0

1,2

PC1(47.27%) Fig. 5. Loading plot as a result of PCA (HEX – hexane, MH – dichlormethane; HL – chloroform; WD – water in dark; WL – water in light; M – methanol; E96–96% solution of ethanol; E30-60–aqueous solutions of ethanol of different concentrations: 30, 40, 50 and 60%).

The total extraction yield was increased with increasing the time of extraction, solvent polarity and plant fragmentation. The highest extraction yield was obtained using diluted ethanol (30%, v/v) during 30 min. The highest total phenolic content was detected in extract obtained by extraction with 96% ethanol during 30 min (sieve 2 mm), while the highest content of flavonoids was achieved using chloroform during 30 min. All the obtained extracts of basil showed an antioxidant activity. The strongest free radical scavenging action of the extract was obtained using diluted ethanol (60%) as solvent during 30 min (sieve 2 mm). The chemometric analysis classified the yield and total phenolic compounds together on one side and the flavonoid content and antioxidant activity on the other side which indicated the correlation between the yield and total phenolic compound, and also between the flavonoid content and antioxidant activity. It can be concluded that applied chemometric techniques could be a helpful tool for choosing the most promising extraction method that can provide basil extract with the highest content of health promoting components which can be explored in different market applications.

Acknowledgments The variation of PC1 with PC4 did not give any additional new information. Also, the grouping of the analyzed parameters based on the extraction time was not registered applying PCA. Fig. 5 shows the loading plot of the PC1 and PC2 against each other showing the similarities and differences between the analyzed variables. Three specific groups of solvents can be registered in Fig. 5. The first group included non-polar solvents which have the negative values of both PCs (chloroform, dichloromethane, hexane), polar solvents (except diluted ethanol) form the second group with the positive values of PC1 and the negative values of PC2, and different diluted ethanol solutions are grouped separately, designated with the positive values of both PCs. This classification was very similar to those obtained by cluster analysis.

The work was financially supported by the Ministry of Science and Technological Development of the Republic of Serbia (Project No. 172021 and ON172012).

References ´ N., Samojlik, I., Anaˇckov, G., Igic, ´ R., 2008. Phenolics as Boˇzin, B., Mimica-Dukic, antioxidants in garlic (Allium sativum L., Alliaceae). Food Chem. 111, 925–929. Cai, Y., Luo, Q., Sun, M., Corke, H., 2004. Antioxidant activity and phenolic compounds of 112 traditional Chinese medicinal plants associated with anticancer. Life Sci. 74, 2157–2184. Flanigan, P., Niemeyer, E., 2014. Effect of cultivar on phenolic levels, anthocyanin composition, and antioxidant properties in purple basil (Ocimumbasilicum L.). Food Chem. 164, 518–526.

182

B. Teofilovi´c et al. / Industrial Crops and Products 100 (2017) 176–182

Gülc¸in, I., Elmastat, M., Aboul-Enein, H.Y., 2007. Determination of antioxidant and radical scavenging activity of basil (Ocimumbasilicum L. family Lamiaceae) assayed by different methodologies. Phytother. Res. 21, 354–361. – – J., 2012. Effects of different ´ N., Lepojevic, ´ Zˇ ., Srdenovi ´ B., Vladic, ´ J., Sudi, Grujic, c, extraction methods and conditions on the phenolic composition on mate tea extracts. Molecules 17 (3), 2518–2528. Hakkim, F.L., Shankar, C.G., Girija, S., 2007. Chemical composition and antioxidant property of holy basil (Ocimum sanctum L.) leaves, stems, and inflorescence and their in vitro callus cultures. J. Agric.Food Chem. 55 (22), 9109–9117. Kaurinovic, B., Popovic, M., Vlaisavljevic, S., Trivic, S., 2011. Antioxidant capacity of ocimumbasilicum L. and Origanum vulgare L. extracts. Molecules 16, 7401–7414. Kwee, E., Niemeyer, E., 2011. Variations in phenolic composition and antioxidant properties among 15 basil (Ocimumbasilicum L.) cultivars. Food Chem. 128, 1044–1050. Lapornik, B., Proˇsek, M., Wondra, A.G., 2005. Comparison of extracts prepared form plant by-products using different solvents and extraction time. J. Food Eng. 71, 214–222. Lee, J., Scagel, C., 2009. Chicoric acid found in basil (Ocimum basilicum L.) leaves. Food Chem. 115, 650–656. Lobo, V., Patil, A., Phatak, A., Chandra, N., 2010. Free radicals, antioxidants and functional foods: impact on human health. Pharmacogn. Rev. 4 (8), 118–126. Makri, O., Kintzios, S., 2007. Ocimumsp (basil): Botany, cultivation, pharmaceutical properties, and biotechnology. J. Herbs Spices Med. Plants 13, 123–150. Manach, C., Williamson, G., Monard, C., Scalbert, A., Remesy, C., 2005. Bioavailability and bioefficacy of polyphenols in humans. I. Review of 97 bioavailability studies. Am. J. Clin. Nutr. 81, 230S–242S. Mastaneh, M., Ahmad, M., Taher, N., Mehrdad, H., 2014. Antioxidant effect of Ocimum basilicum cv. dark opal (Lamiaceae) phenolics. Orient. J. Chem. 30 (4), 1965–1969. Masuoka, N., Matsuda, M., Kubo, I., 2012. Characterization of the antioxidant activity of flavonoids. Food Chem. 131, 541–545. Menrad, K., 2003. Market and marketing of functional food in Europe. J. Food Eng. 56, 181–188. Miller, J.N., Miller, J.C., 2004. Statistics and Chemometrics for Analytical Chemistry, 5th ed. Pearson Education Limited, Edinburgh Gate Harlow, England.

Nguyen, P., Kwee, E., Niemeyer, E., 2010. Potassium rate alters the antioxidant capacity and phenolic concentration of basil (Ocimum basilicum L.) leaves. Food Chem. 123, 1235–1241. Politeo, O., Jukic, M., Milos, M., 2007. Chemical composition and antioxidant capacity of free volatile aglycones from basil (Ocimum basilicum L.) with its essential oil. Food Chem. 101, 379–385. Prior, R.L., Wu, X., Schaich, K., 2005. Standardized methods for the determination of antioxidant capacity and phenolics in food and dietary supplements. J. Agric. Food Chem. 53 (10), 4290–4302. Salla, S., Sunkara, R., Ogutu, S., Walker, L.T., Verghese, M., 2016. Antioxidant activity of papaya seed extracts against H2 O2 induced oxidative stress in HepG2 cells. LWT Food Sci. Technol. 66, 293–297. Spigno, G., Tramelli, L., De Faveri, D.M., 2007. Effects of extraction time: temperature and solvent on concentraction and antioxidant activity of grape marc phenolics. J. Food Eng. 81, 200–208. Thoo, Y.Y., Ho, S.K., Liang, J.Y., Ho, C.W., Tan, C.P., 2010. Effects of binary solvent extraction system, extraction time and extraction temperatire on phenolic antioxidants and antioxidant capacity from mengkudu (Morindacitrifolia). Food Chem. 120, 290–295. Urala, N., Lahteenmaki, L., 2007. Consumers’ changing attitudes towards functional foods. Food Qual. Preference 18, 1–12. ´ B., Marinkovic, ´ A., 2014. Vastag, G.Y., Apostolov, S., Matijevic, Chemometricapproach instudyingof the retentionbehavior andlipophilicityof potentiallybiologically active N-substituted-2-phenylacetamide derivatives. J. Braz. Chem. Soc. 25, 1948–1955. ´ S., Zekovic, ´ Z., Lepojevic, ´ Zˇ ., Radojkovic, ´ M., Jokic, ´ S., Anaˇckov, G., 2012. Vidovic, Optimization of the Ocimum basilicum L. extraction process regarding the antioxidant activity. APTEFF 43, 315–323. Zhang, Z.S., Li, D., Wang, L.J., Ozkan, N., Chen, X.D., Mao, Z.H., 2007. Optimisation of ethanol-water extraction of lignans from flaxseed. Sep. Purif. 57 (1), 17–24.