Kumquat fruit and leaves extracted with different solvents: phenolic content and antioxidant activity

study, we used various solvents to obtain extracts from fresh and lyophilized dried fruits and leaves of kumquat plant, as well as six mutants, to compare their total phenolic and flavonoid contents and antioxidant activities. Study objects and methods. The total phenolic and flavonoid content was determined by the Folin-Ciocalteu method and the colorimetric method, respectively. The antioxidant capacities of the extracts were determined by commonly used antioxidant tests, such as the DPPH radical scavenging activity, reducing power, and metal chelating activity. Results and discussion. The total phenolic content of the extracts was in the range of 3705–86 329 mg GAE/g extract. The total amount of flavonoid substance ranged from 5556 to 632 222 mg QUE/g extract. The highest free radical scavenging activity was observed in the kumquat leaves. We also found that the activity of dried fruit was lower than that of fresh fruit. According to our results, the differences in the phenolic contents of the studied plants affected their antioxidant properties. We determined that the extracts with a high phenolic content showed high antioxidant activity. No significant difference was detected between the rootstock kumquat type and its mutants. Finally, we found no chelating activity in the extracts obtained from fresh and lyophilized dried fruits. Conclusion. Kumquat fruit and its leaves can be considered as functional foods due to phenolic compounds, which are able to neutralize free radicals.


INTRODUCTION
Constantly developing technology, environmental pollution, ultraviolet radiation, and many other factors cause us to be exposed to various toxic substances. This results in more diseases caused by external environmental effects, including more pronounced genetic diseases. Preventing these diseases should become our priority. Since most of them occur in people with a weak immune system, we must focus on strengthening it. For this, we should consume foods with high antioxidant capacity, especially fruits and green leafy vegetables that contain antioxidative phytochemicals [1,2].
Phytochemicals, or "plant chemicals", are compounds of plant origin, mostly polyphenols, that are essential for human life. They work alongside macronutrients such as carbohydrates, fats, and proteins, as well as 13 essential vitamins and 17 minerals [3]. Antioxidant phytochemicals, especially in fruits and vegetables, combine with free radicals in the human body to protect cells from the attacks of harmful radicals [4]. Bioactive compounds in fruits contain ascorbic acid, organic and phenolic acids, flavonoids, anthocyanins, and carotenoid substances [5,6].
Citrus fruits come in different types, varieties, and flavors and have positive effects on health and nutrition. Although they have been known as the best sources of vitamin C for a long time, studies on their use as an antioxidant substance have recently gained momentum, due to their richness in phenolic compounds [7]. These bioactive components are responsible for various health benefits of citrus fruits, such as prevention of various diseases or protective effects to lower the risk of various cancers [8][9][10].
Citrus is a fruit group belonging to the genus Citrus, which is a member of the Aurantioideae subfamily of the Rutaceae family. The most common citrus varieties are orange (Citrus sinensis L.), mandarin (Citrus reticulata L.), lemon (Citrus limon L.), golden ball (Citrus paradisi L.), bitter orange (Citrus auranthium L.), and bergamot (Citrus bergami L.) [11]. In addition to fresh table consumption, citrus fruits are used as jam, marmalade or fruit juice, as well as raw material in the cosmetics sector [11].
Citrus fruits grow in subtropical climate areas. While mainland China, Southeast Asia, and India are major producers of citrus fruits due to suitable ecological conditions, they are also cultivated in the Mediterranean and Aegean coastal regions and partly in the Eastern Black Sea region of Turkey [12,13]. The distribution of species and varieties of citrus fruits has gained a regional identity. For example, Washington navel, as well as other navel oranges, and Jaffa are harvested in the Eastern Mediterranean region.
Orange is one of the most produced and consumed citrus fruits in Turkey due to its preference in the juice industry and its great potential in the oil industry [14]. Orange is followed by mandarin and lemon products, respectively. Apart from these species, kumquat, which is called the "little gem of the citrus family", has recently grown in popularity, as well as such species as Altıntop and citrus, which are lower in production but can be considered important [15].
Kumquat is also called "citrus fortunella", taking its name from the Scottish horticultural expert Robert Fortune (1812-1880). This species, referred to as "komquot" in some countries, is also called a "golden orange" [16]. It is like a tiny lemon in shape and orangish in color. However, while orange and lemon are consumed after they are peeled, kumquat is consumed with its peel. Its scent is reminiscent of bergamot. It tastes sweet and leaves a lasting scent when you hold it in your hand.
In addition to fresh consumption, kumquat can be used in products such as confectionery, marmalade, liquor, and wine [17,18]. Essential oil and bioactive ingredients obtained from its peel are used in the perfumery, pharmaceutical, and food industries [19]. Kumquat is an excellent source of nutrients containing minerals, ascorbic acid, carotenoids, flavonoids, and essential oils [20]. It contains remarkable antioxidant properties due to its flavonoid content [18]. However, there are very few studies about kumquat grown in Turkey.
In this study, we aimed to determine the antioxidant capacity and the total phenolic and flavonoid contents of the extracts obtained from fresh and lyophilized dried fruits and leaves of kumquat and six mutants from the Mersin Alata Horticultural Research Institute Directorate.

STUDY OBJECTS AND METHODS
Plant materials. Kumquat leaf and fruit samples were obtained from the Mersin Alata Horticultural Research Institute in November 2017 and January 2018, respectively. We used EP (Old Parcel) with rootstock species; EP.4, EP.29, EP.31 and YP (New Parcel); YP.117, YP.141, YP.188 mutants. The leaf samples were dried in room conditions and in the shade, and stored in a dry and cool environment for analysis. The fruit samples were freeze-dried, or lyophilized.
Chemicals and equipment. We used chemicals and solvents of analytical purity produced by Sigma, Aldrich, and Riedel-de Haen.
Extraction process. Phenolic compounds were extracted from kumquat fruits and leaves with a Soxhlet extraction device, using 260 mL of 99, 80, 60, and 50% methanol and pure water as solvents. In addition, 1 and 0.5% acidified ethanol and hexane solvents were used for kumquat leaves.
For extraction, 20 g of the samples were weighed into the cartridge and then placed in the Soxhlet extractor. The solvent(s) was added to the glass flask and kept in the Soxhlet device for 8 h. The solvent used for extraction was concentrated from the obtained phenolic extracts using a laboratory scale rotary evaporator under vacuum. The remaining part was removed by standing in the open air. The extracts were weighed gravimetrically and stored in dark vials at +4°C in the refrigerator until analysis.
Determination of free radical capture capacity (DPPH method). We used 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical to determine the free radical capture capacity according to the Blois method [21]. This method is based on the ability of the extracts to donate a proton or electron and to decolorize the purple colored DPPH solution (from violet to yellow). A decrease in the absorbance of the reaction mixture is indicative of high free radical scavenging activity.
All the extracts, BHA and BHT standards, and α-tocopherol were dissolved in ethanol at 1 mg/mL. After taking the samples and standards into 5 different volumes of 50, 100, 150, 250, and 500 µL, ethanol was added to a total volume of 3 mL. 1000 µL of 0.1 mM DPPH was added to the tubes and vortexed. The absorbance of the mixture, which was incubated for 30 min in the dark at room temperature, was measured in the UV-visible spectrophotometer at 517 nm. Calculations were made using the following formula:  where A C is the absorbance of the control reaction; A S/S is the absorbance of the sample or standard.
Determination of reducing capacity. The Oyaizu method was used to determine the reduction capacity [22]. According to this method, the reducing agent in the medium reduces Fe 3+ ions to Fe 2+ ions and a complex is formed by adding FeCl 3 . The absorbance of the resulting complex is measured in the UV-visible spectrophotometer at 700 nm. The increase in absorbance of the reaction mixture is directly proportional to the reducing power of the sample.
All the extracts, BHA and BHT standards, and α-tocopherol were dissolved in ethanol at 1 mg/mL. 100, 250, and 500 µL of the samples and standards were taken into test tubes in three different volumes, and 3400, 3250, and 3000 µL of pH 6.6 phosphate buffer was added to them, respectively, to a total volume of 3500 µL. Then, after adding 2500 µL of 1% K 3 [Fe (CN) 6 ] and vortexing, it was left to incubate for 20 min in a water bath at 50°C. After the incubation, 2500 µL of 10% trichloroacetic acid (TCA) was added to the test tubes and centrifuged at 3000 rpm for 10 min. 1250 µL of the resulting supernatant was taken into empty tubes and 1250 µL of distilled water and 500 µL of 0.1% FeCl 3 were added to them. The mixture was vortexed and its absorbance was measured at 700 nm in the UV-visible spectrophotometer.
Determination of iron (II) ions chelating activity. Antioxidants with metal chelating properties inactivate free iron by binding it and thus inhibit the formation of radicals such as hydroxyl and peroxide, which are formed as a result of Fenton reactions (Fe 2+ + H 2 O 2 → Fe 3+ + HO • + HO -) [23]. The Dinis method was used to determine the activity of chelating iron (II) ions [24]. All the extracts and EDTA used as control were dissolved in ethanol to 1 mg/mL. The samples and standards were taken into 50, 100, 150, 250, and 500 µL test tubes, and 3700, 3650, 3600, 3500, and 3250 µL of ethanol was added to them, respectively, to a total volume of 3750 µL. Then, 50 µL of 2mM FeCl 2 was added and vortexed to incubate at room temperature for 10 min. Then, 200 µL of 5mM ferrosine was added. The resulting purple color was measured in the UV-visible spectrophotometer at 562 nm after the mixture was kept at room temperature for 25 min.
Determination of total phenolic content. The Folin-Ciocalteu method was used to determine the total phenolic content [25]. The Folin-Ciocalteu reagent (FCR) used in this method is molybophosphotungstic heteropolyacid (3H 2 O·P 2 O 5 ·13WO 3 ·5MoO 3 ·10H 2 O). This method is based on the transfer of electrons from phenolic compounds and other reducing compounds to molybdenum. Phenolic compounds only react with the FCR in basic conditions (pH ~ 10) [26].

Mo(VI) + e -(antioxidant) → Mo(V)
Commercially available 2N Folin-Ciocalteu reagent was prepared daily by diluting it with purified water at a ratio of 1/1 (v/v). 500 µL of the extracts (1 mg/mL) was taken into test tubes and 500 µL of distilled water was added. After 250 µL of 1 N Folin reagent was added to the mixture, it was incubated for 5 min by vortexing. 1250 µL of 2% Na 2 CO 3 solution was added to it, vortexed, and then kept at room temperature for 2 h. The absorbance of the resulting mixture was measured at 765 nm in the UV-visible spectrophotometer. The phenolic content of the extracts was given as mg gallic acid equivalent (GAE)/g extract.
Determination of total flavonoid content. The total flavonoid content was measured by an aluminum chloride colorimetric test according to Zhishen et al. [27]. All the extracts and a quercetin solution used as a standard were dissolved in 1 mg/mL ethanol. 500 µL was taken from the extracts prepared in the test tubes and pure water was added to a total volume of 5000 µL. To this mixture, 300 µL of 5% NaNO 2 solution was added and left to incubate at room temperature for 5 min, and then 300 µL of 10% AlCl 3 solution was added. After waiting for 6 min, 2 mL of 1.0 M NaOH solution was added and the volume was completed to 20 mL with distilled water. The absorbance of the solution was measured at 510 nm in the UV-visible spectrophotometer. The total flavonoid content of the extracts was given as mg quercetin equivalent (QUE)/g extract.

RESULTS AND DISCUSSION
The solubility and distribution of phenolic compounds in the solvent depend on the polarity of their structure, so the choice of solvent and method is one of the most important steps. In our study, for fresh and lyophilized dried fruits, we preferred methanol and its aqueous solutions, as well as pure water. For leaves, we preferred methanol and aqueous solutions, distilled water, and ethanol acidified with hexane.
Three different methods (DPPH radical scavenging activity, reducing capacity, and iron (II) ions chelating activity) were used to determine the antioxidant capacity. We thought that the extracts could show activity through different mechanisms depending on the diversity of phenolic substances. In addition, we determined the total phenolic content and flavonoid amounts in all the extracts in order to show that the antioxidant effect was proportional to the plant content.
Free radical scavenging activity. The DPPH method is commonly used to evaluate the antioxidant activity of natural products, as it is easy and highly sensitive. DPPH (2,2-diphenyl-1-picrylhydrazyl) is a commercially available stable organic nitrogen radical. The antioxidant effect is proportional to the removal of the DPPH radical. The DPPH radical (DPPH • ) is purple in color due to the unpaired nitrogen atom. When the DPPH solution reacts with an oxygen atom of a substance (antioxidant chemical) that can give hydrogen atoms, the initial purple color disappears as the radical reduces, turning yellow [28]. The reaction takes place stoichiometrically according to the number of hydrogen atoms absorbed. Therefore, the antioxidant effect was easily determined by following the decrease in UV absorbance at 517 nm until it stabilized.
We observed that the highest free radical scavenging activity was in kumquat leaves, and the activity of kumquat fruit decreased when dried (Table 1). There was no significant difference between the rootstock kumquat type and its mutants. The free radical scavenging activities of the extracts were slightly below the standards (BHA, BHT, and α-Tocopherol). The highest activity (81.66%) was seen in the YP.188 hybrid leaf extract using 80% methanol solvent. As for the fruits, the highest activity (61.37%) was in the EP.4 hybrid extract using a pure methanol solvent.
When we examined all the samples, we associated high phenolic content with high antioxidant activity. We found that the total phenolic content was higher in the samples with high antioxidant activity. As a matter of fact, the leaf extract with high antioxidant activity also had a high phenolic content (85.651 ± 0.030 mg GAE/g extract).
However, when we carefully examined the results, we saw that having a high amount of phenolic substances did not give high results in all antio-xidant activity methods. For example, although the YP.188 Leaf 80% methanol and the YP.188 Leaf 50% methanol extracts contained almost the same amount of phenolic substances, the former had higher activity in the applied antioxidant activity methods. This could be explained by the differences between the phenolic substances they contained depending on the solvent used.
In fact, other studies have found that the antioxidant activity of methanol and ethanol extracts, which generally contained phenolic substances, was higher than in other solvent systems [29]. For example, Jayaprakasha et al. extracted powdered kumquat fruit in 5 different solvents and investigated the radical capture capacities of the extracts, their amounts in total phenolic matter, and their inhibitory properties for prostate cancer [30].
In this study, the extracts obtained from EtOAc and MeOH-water (4:1, v/v) solvents were found to have the highest and lowest total phenolics, respectively, according to the Folin-Ciocalteu method. It was also observed that the EtOAc and MeOH extracts exhibited the highest and lowest 1,1-diphenyl-2-picyrylhydrazyl (DPPH) radical scavenging activity, respectively [30]. The chelating activity of iron (II) ions. Antioxidants with metal chelating properties inactivate it by binding free iron and thus inhibit the formation of radicals such as hydroxyl and peroxide, which are formed as a result of Fenton reactions. Therefore, metal chelating plays an important role in determining antioxidant activity [31].
We evaluated the metal ion chelating activity according to the competition between plant extracts with ferrosine in order to bind Fe 2+ ions in the solution. We observed no chelating activity in the extracts obtained from moist and lyophilized dried fruits ( Table 2). The pure methanol extracts showed weak activity in kumquat leaves, while the extracts obtained from aqueous solvents showed no activity at all. In addition, weak chelating activity was detected in the 0.5 and 1% acidified ethanol extracts of kumquat leaves and the hexane solvent extracts. The highest activity (50.37%) was found in 62.5 µg/mL concentration of the extract obtained from kumquat leaves with a hexane solvent. We determined no correlation between the chelating activity of the extracts and their concentration. No significant difference was found between the rootstock kumquat type and its hybrids.
When we evaluated all the activities, we concluded that the extracts obtained from kumquat fruits and leaves were not good at chelating iron (II) ions. The most important feature that affects the metal chelating activity depends on the functional groups in the structure of phenolic compounds and the position and amount of these functional groups. For this reason, the difference in the chelating activity of the samples can be explained  6 ]) formed by adding FeCl 3 to the reduced product is measured at 700 nm [22]. The increase in absorbance of the reaction mixture is directly proportional to the reducing power of the sample.
We found that the capacity of kumquat leaves to reduce Fe 3+ ions was higher than that of lyophilized and wet kumquat fruits (Table 3). We observed that lyophilizing and drying of kumquat fruits did not cause a significant change in their reducing capacity. The reducing capacity of the fruit and leaf extracts was lower than the standards (BHA, BHT and α-tocopherol). The highest reducing capacity (0.307 ± 0.001) was observed at a concentration of 29.41 µg/mL of the EP.4 mutant leaf extract obtained with pure water. Among the fruits, the highest reducing capacity (0.199 ± 0.001) was found at a concentration of 29.41 µg/mL of the EP.4 hybrid wet fruit extract obtained with pure methanol. The reducing capacities of the standards were 1.984 ± 0.001, 1.290 ± 0.002, 0.872 ± 0.001 for BHA, BHT, and α-toceferol, respectively, at the highest concentration of 29.41 µg/mL.
No significant difference was observed between the rootstock kumquat plant and its mutants. Although the reducing power is an important factor of antioxidant activity, in our study, the reducing power was lower in the extracts with high antioxidant activity. Other studies also show that extracts with high antioxidant activity may have low reducing power [33,34]. This is because in the systems where free iron ions are present in trace amounts, the net oxidation rate increases with the Fenton reaction. Substances with high reducing power may cause further acceleration of oxidation by reducing Fe(III) to Fe(II). The presence of trace levels of iron ions in kumquat materials may have caused its low reducing power and ncreased antioxidant activity [35].
Phenolic and flavonoid content. Since phenolic and flavonoid compounds contain hydroxyl groups in their structures and can easily give a hydrogen radical in hydroxyl groups, they have free radical quenching properties. Therefore, it is important to know the total phenolic and flavonoid contents of the samples to determine their contribution to the antioxidant activity, including radical scavenging activity tests. For this, we used the Folin-Ciocalteu method, a standard method in antioxidant studies. The basis of the method is that phenolic compounds dissolved in water and other organic solvents form a colored complex with a Folin reagent in an alkaline medium. The total phenolic content of the extracts obtained by Soxhlet extraction with different solvents was calculated using the regression equation (y = 0.0292x + 0.0749 and  Table 3 R² = 0.9994) of the calibration line of the standard gallic acid solution prepared in the concentration range of 5-50 µg/mL and expressed as gallic acid equivalent (mg GAE/g extract). The gallic acid standard curve is shown in Fig. 1. We found that the kumquat leaf extracts had the highest total phenolic content (Table 4). In particular, the highest total phenolic content (86.329 ± 0.022 mg GAE/g extract) was in the YP.188 mutant extract obtained with 60% methanol. In the fruit samples, the highest total phenolic content (20.281 mg GAE/g extract) was found in the EP.4 mutant extract obtained with pure methanol. There was no significant difference in total phenolic contents between the fresh and dried fruit samples. Lou et al. compared total phenolic contents in fresh and dried kumquat fruits [36]. The scientists investigated changes in total phenolic matter by changing the drying degree and time. They found that the total amount of phenolic substances increased with drying, amounting to 15-17 mg GAE/g extract and 48-50 mg GAE/g extract in fresh and dried fruit (130°C), respectively [36].
In another study, Özcan et al. dried kumquat fruit in hot air, under vacuum, and in a microwave oven [27]. The authors found that the total phenolic content of hot air-dried fruit was approximately 5 mg GAE/g extract, but with other drying methods, it varied in the range of 25-30 mg GAE/g extract [37].
Yıldız Turgut et al. studied the functional quality parameters of the powder obtained from Fortunella margarita kumquat varieties grown in Turkey. They reported the total phenolic content of kumquat between 2.62 ± 0.051 -6.97 ± 0.053 mg GAE/g depending on the type of drying method [38].
Having determined the total phenolic content, we measured the total flavonoid content of the samples. Total flavonoid concentration was determined colorimetrically using a UV spectrophotometer according to the method applied by Zhishen et al. [27]. In our study, quercetin was used as a standard and the results were calculated as quercetin equivalent (mg QUE/g extract) from the quercetin standard calibration chart (y = 0.0185x -0.0019 and R² = 0.9666) (Fig. 2). The highest amount of total flavonoid substance was seen in kumquat leaves (Table 4). In particular, the highest flavonoid content was found in the EP.31 mutant extract (632.222 ± 0.033 mg QUE/g extract) obtained with 1% acidified ethanol.
Among the fruit samples, the highest amount (313.333 ± 0.023 mg QUE/g extract) was found in the YP.141 mutant extract obtained with pure methanol. There were no significant differences between the total flavonoid amounts in the fresh and dried fruits.
Lou et al. reported that the total amount of flavonoid substance in kumquat varied between 58.23-91.42 mg/g depending on the drying temperature [36]. In another study, Lou et al. found that the total phenolic and flavonoid contents were higher in the extracts from kumquat and calamondin peel compared to fruit pulp, and that they were higher in the extracts from unripe kumquat compared to those from ripe kumquat [39,40].

CONCLUSION
In antioxidant activity studies, it is common to use a different polarity solvent system in order to determine which compound types have the highest activity. There may be a relationship between phenolic or flavonoid amounts and antioxidant capacity determination methods. In particular, a relationship between methods such as the DPPH, which is based on radical capture, and total phenolic and flavonoid amounts may be important in some plant structures. Phenolic acids and flavonoids are soluble in polar solvents and show strong activity in polar systems.
In this study, we investigated the effect of different solvents and their concentrations on the bioactivity of kumquat fruit and leaf extracts. We found that the solvent type was extremely important for the extracts' bioactivity. In particular, the extraction performed with pure methanol in the fruits and 60 or 80% methanol in the leaves showed the highest total phenolic and flavonoid contents, the highest extraction efficiency (50.18-59.95%), and the highest antioxidant capacity.
We found no statistically significant difference between the total amount of phenolic/flavonoid substances and % inhibition value in the extraction performed with 60 and 80% methanol solutions. This shows that the amount of phenolic substances was affected by the polarity of the solvent, depending on the difference in phenolic compounds found in kumquat fruit and leaves. We concluded that phenolic components in the structure of a kumquat fruit could be extracted with a single solvent type, whereas those in the structure of a kumquat leaf could be extracted better with an aqueous solution of the relevant solvent, rather than a single solvent type.
We also observed that the aqueous solutions gave better results than the pure solutions in the production of phenolics from kumquat leaves, maximizing at certain water ratios and showing different distributions according to the solvents. These results can be explained by the fact that water increases diffusion by causing swelling in the leaf structure. In this context, methanol was the most effective solvent for bioactive component extraction from the kumquat fruits, whereas methanol + water was most effective for the leaves.
Having examined the effect of a solvent amount, we concluded that the extraction with 260 mL solvent ensured the highest total phenolic content, extraction efficiency, and antioxidant capacity. In addition, since methanol is a toxic solvent, it must be removed so that the obtained extract can be used in foods or consumed as a food supplement.
Plants are complex systems by nature and have multiple reaction characteristics and dissolution properties in different phases. Thus, it is not possible for a single method to reveal all of their radical sources or antioxidants [41][42][43]. For these reasons, we used a combination of methods, namely the DPPH, metal chelation, and iron reduction. In addition, we used the Folin-Ciocalteu method and the aluminum chloride method to determine the total phenol and flavonoid contents, respectively. The results clearly showed that the differences in the phenolic contents affected the plants' antioxidant properties.
We found that having a high phenolic content or high radical scavenging activity did not yield high results in all antioxidant activity studies. Thus, we concluded that determining the antioxidant activity with a single method was not the right approach and that it would be more accurate to simulate biochemical events in living systems by using a variety of methods. In summary, antioxidant structures can demonstrate their antioxidant activities by different mechanisms such as binding transition metal ions, breaking down peroxides, preventing hydrogen absorption, and removing radicals.
Our study revealed that the kumquat leaf extracts had a higher DPPH radical scavenging power than the fruit extracts. However, both the fruit and leaf extracts showed high levels of free radical scavenging activity with high antioxidant activity at a 125 μg/mL concentration. Due to high antioxidant activity, kumquat leaves can be recommended to be used as food, just as kumquat fruit, against many diseasesfrom gastrointestinal to infertility, from cardiovascular to respiratory and excretory disorders, especially to prevent cell damage caused by free radicals in human and animal bodies.