Phytochemicals from Phillyrea latifolia L. leaves and fruit extracted with various solvents: Their identification and quantification by LC-MS and antihyperglycemic effects

Mock privet leaves and fruit were collected from Antalya, Turkey, in December. Fruits were at the mature stage. The plant material was identified by Agriculturist Yavuz Çetin, TropikHal. Both the leaves and fruit were dried at 50 °C using an oven (Azaizeh et al., 2013). Next, the dried fruits and leaves were ground and tightly packed in polyethene bags and were stored at –20 °C for further experiments.

According to Sarikurkcu et al. (2018), phenolics were extracted from the dried mock privet leaves and fruit using ethyl acetate, methanol and water. Twenty grams of air-dried samples were extracted using ethyl acetate and methanol through a Soxhlet extractor for a duration of 5 hr. For water extract, 20 g of air-dried sample were mixed with 400 mL of deionized water and boiled for 15 min. Following, the ethyl acetate and methanol were removed using a rotary evaporator. The resulting water extract was subsequently freeze-dried. The extracts were stored at 4 °C.

Chromatographic analysis of mock privet leaves and fruit compounds was conducted using the LCMS-2020 quadrupole mass spectrometer (linked with Shimadzu LC-2010 HT) fitted with an electrospray ionization source (Milton Keys, UK). It was performed for the identification of phenolics in the extracts and confirmed with reference standards. A Kinetex C18 analytical column (2.10 mm × 150 mm, 2.6 μm, 35 °C) was used for chromatographic separation. The conditions were determined as a flow rate of 0.25 mL · min^–1, an injection volume of 5 μL, mobile phase A (0.1% formic acid, 95% demineralized water and 5% acetonitrile) and mobile phase B (95% acetonitrile, 5% demineralized water and 0.1% formic acid). Elution was started at 0% solvent B and increased to 10% at 5 min, afterwards solvent B rose to 25% and 35% at 10 min and 20 min, respectively. Following, solvent B reached 50% at 25 min and held up until 30.5 min. Finally, this gradient increased to 100% for a further 5 min and reduced to 0% at 36 min. The column was reequilibrated for 41 min. The mass-to-charge ratio in negative mode (m/z(–)) was used to identify each compound, and it was compared with their respective true standard. Cyanidin 3-O-glucoside, rutin, myricetin, quercetin-3-O-glucuronide, quercetin-7-O-glucoside, luteolin 7-O-glucoside and cyanidin 3-O-rutinoside were purchased from Extrasynthese (Lyon, France). 5-Caffeoylquinic acid was purchased from Acros Organics (Geel, Belgium). Ferulic acid, caffeic acid kaempferol aglycone, apigenin aglycone, rosmarinic acid and oleuropein were purchased from Sigma-Aldrich (St. Louis, MO, USA).

The total phenolic content (TPC) was determined using the Folin-Ciocalteu colorimetric method (T70 + UV/VIS spectrophotometer, PG Instruments, UK; Li et al., 2006). TPC was expressed as mg gallic acid equivalents (GAE) per 100 g of extract. In addition, the TPC of GFE and FDM was detected using the above procedure. Three replicates were performed for each sample.

α-diphenyl-β-picrylhydrazyl (DPPH) was analysed according to the method developed by Dorman et al. (2003). The mixture of 50 μL extract, 450 μL tris-HCl buffer (50 mM, pH 7.4) and 1.00 mL of fresh methanolic solution DPPH (0.10 mM) was well-shaken and kept in dark condition at room temperature for 30 min. The percentage of DPPH was calculated using the following equation after the samples’ absorbance was read (517 nm): Inhibition % (DPPH) = [(Abs_Control – Abs_Sample)/Abs_Control] × 100. The half maximal inhibitory concentration (IC₅₀) value of the extract concentration, to inhibit 50% of the free radicals, was determined based on the graph produced from DPPH (%) activity, and the results were shown as IC₅₀ = mg · mL^–1. All analyses were performed in triplicate.

A buffer solution of 20 mM sodium phosphate and 6.7 mM sodium chloride (pH 6.9) was used to dissolve human salivary α-amylase (S/3160/53 from Fisher Scientific, Loughborough, UK). On the other hand, 10 mM sodium phosphate buffer with pH 7.0 was used to dissolve intestinal acetone rat powder (I1630). Following, both enzyme solutions were vortexed for 30 s and then centrifuged at 17,000 g for 10 min. The supernatants were removed to conduct analysis, and these enzyme solutions were prepared freshly before each experiment (Aydin, 2015).

The developed method by Nyambe-Silavwe et al. (2015) was used. Twelve grams of sodium tartrate were added to 8 mL of 2M sodium hydroxide and heated until dissolved. It was included to stabilize the colour and to protect the product from oxidizing. The DNS solution was produced by combining the DNS powder with deionized water (20 mL) before placing the mixture immediately on a heating plate to dissolve. The colorant employed in the α-amylase reaction was DNS. The reducing sugars are produced as a by-product of the hydrolysis of starch by human salivary α-amylase. In an alkaline environment, DNS reacts with the free carbonyl group of the reducing sugars to form 3-amino-5-nitrosalicylic acid, which can be detected at 540 nm. As the reducing sugars are released, DNS altered the colour. DNS and prepared sodium tartrate were combined with deionized water (40 mL) to create the colour reagent solution.

A total of 200 mL of the sucrose solution, 50 mL of buffer (pH 7.0–10 mM), 50 mL of plant material and 200 mL of enzyme solution at various concentrations were combined for the experiment. Following, the mixture was vortexed (10 s) and incubated at 37 °C (10 min). As it was mentioned by Nyambe-Silavwe et al. (2015) study, a cartridge (Waters Oasis MAX-003036349A) was applied to eliminate the polyphenol’s potential to interfere with colour development, the samples were left for 10 min incubation in a boiling water bath to stop the enzyme activity before the addition of colour reagent solution. Following the incubation, to measure the production of reduced sugar amount, a colour reagent solution was added to each sample (1 mL). After that, the samples were immediately placed on ice for cooling down to RT after being placed in a boiling water bath (10 min) to prevent enzyme activity. Following this, the samples were transferred to a boiling water bath for 10 min again. Then they were placed in vials for electrochemical detection using HPAE-PAD.

A previously reported technique by Gao et al. (2007) was modified to detect the activities of sucrase in an acetone extract of rat intestinal tissues by analysing sugars [sucrose and its products (glucose and fructose)] via a Dionex system running Chromeleon 6.5. Ion-exchange-LC combined with electrochemical detection permits the direct quantification of low-level (pM) carbohydrates <10,000 Da without the need for derivatisation or intensive sample preparation. To optimise the assay conditions, first Michaelis Constant (K_m/V_max) was detected and K_m was found as 18 mM and V_max as 0.09 μmoL sucrose hydrolysis/min. The amount of enzyme at different concentrations and the hydrolysis of sucrose were carried out at 37 °C for different time intervals to determine the optimum incubation time and the amount of enzyme. Based on preliminary studies, 18 mM of sucrose (200 μL) and 15 mg · mL^–1 of acetone rat intestinal powder (200 μL) were mixed with 100 μL of sodium phosphate buffer as the control sample. To analyse the antidiabetic activity of the extract, as a test sample 100 μL of sodium phosphate buffer was changed with 100 μL of extract/solution. To analyse the antidiabetic activity of leaf and fruit extracts, as a test sample 100 μL of sodium phosphate buffer was changed with 100 μL of samples. After that, the samples were incubated at 37 °C for 15 min. To stop enzyme activity, 750 μL of acetone was added to mixtures and immediately vortexed for 10 s and hold in ice to cool down at RT. The acetone was removed under nitrogen gas and centrifugation was repeated. Finally, the supernatant was filtered, and the amount of sucrose and its products (glucose and fructose) were determined using HPAE-PAD.

Carbohydrate molecules are often separated using anion-exchange chromatography but considering they are weak acids, it may be more sensitive to identify them using amperometric detection, due to its speciality to depend on the oxidation of carbohydrates in the presence of sodium hydroxide at the gold electrode. HPAE-PAD (Dionex DX500, Sunnyvale, CA) comprised of a GP40 gradient pump, PAD system, an LC20 column oven and electrochemical detectors [ED 40, e.g., gold working, titanium and silver (reference) electrode]. The analytical column for α-glucosidase assay was used as Carbopac PA20 [Dionex, 3 mm × 150 mm and guard column (3 mm × 30 mm)] and for human salivary α-amylase assay, Carbopac PA200 (Dionex, 3 mm × 250 mm) with guard (3 mm × 50 mm). In addition, the mobile phase was 200 mM NaOH (flow rate: 0.4 mL × min^–1 and injection volume: 10 μL). Finally, the elution was achieved using a gradient from 0% to 30% 200 mM NaOH in 10 min, 50% 200 mM NaOH from 10 min to 15 min and re-equilibration at 30%, 200 mM NaOH for 15 min.

The data analysis process was carried out using IBM SPSS Statistics 22. The homogeneity of the mean groups was assessed using Levene’s test. The one-way analysis of variance (ANOVA) was followed by the Dunnett C test unless the condition was achieved, in which case the Tukey HSD post hoc test was used. If p ≤ 0.05, differences were regarded as statistically significant unless otherwise stated.

The type of solvents, such as ethyl acetate, chloroform, methanol, water, hexane, ethanol and acetone, is considerable for the extraction of bioactive compounds and these solvents have been widely used in the literature (Altemimi et al., 2017). Based on the type of solvent used, mock privet leaves and fruit extracts showed different efficiency rates (Table 1). For the leaves, the aqueous extract (18.02%) had the highest yield, followed by the methanol extract (16.98%) and the ethyl acetate extract (2.05%). Similar to leaves, for the extraction yield of the fruit, the aqueous extract had the maximum efficiency (19.02%), followed by the methanol extract (13.83%) and the ethyl acetate extract (1.47%). According to Altemimi et al. (2017), the extraction yield is influenced by the solvents, besides the temperature and time, the chemical characteristics of the samples have effects. Considering current findings, different studies indicated that water showed better extraction efficiency (EE) than methanol, thus these findings collaborated with what was reported in the literature (Kuo et al., 2015; Sarikurkcu et al., 2020). In addition, another recent study also reported that the polarity of solvents had a direct impact on the extraction yield for most of the plants, and water, methanol and ethyl acetate had polarity indexes of 9.0, 6.6, and 4.3, respectively (Ng et al., 2020). Therefore, water may show better EE than methanol and ethyl acetate.

Table 1 also demonstrated that the solvents that were most effective at extracting had the greatest phenolic compounds. The methanol extract (1,974 mg GAE · 100 g^–1 extract), ethyl acetate extract (1,497 mg GAE · 100 g^–1 extract) and aqueous extract (2,531 mg GAE · 100 g^–1 extract) had the maximum phenolic content of leaves, respectively. Similar to leaves’ TPC, fruit’s ethyl acetate extract (1,004 mg GAE · 100 g^–1 extract) has the lowest phenolic content whereas aqueous extract (2,260 mg GAE · 100 g^–1 extract) has the highest. Statistical analyses showed significant differences between TPC values for both samples (p < 0.05). In addition, due to the presence of polar phenolic hydroxyl groups, phenolics, including numerous flavonoids, exhibit a strong tendency to be extracted into methanol and water. Similarly, the hexanic extract with lowest TPC can also be elucidated based on the same principles. The significant role of retrieving polyphenols from diverse sources has been suggested to be influenced by the varying polarity of solvents (Teruel et al., 2015). The data obtained in our study showed that the total phenolic of aqueous extracts was higher than that of methanol and ethyl acetate extracts. Similarly, Romero-Diez et al. (2018) and Wang et al. (2019) observed that the aqueous extract of the sample plant exhibited superior extraction efficacy for phenolic compounds when compared with the methanol extract. The results obtained regardless of the solvent revealed that mock privet leaves and fruit contain a significant amount of phenolic compounds, with leaves having a higher phenolic content than fruit. Teleszko and Wojdylo (2015) also informed that the leaves of different plants (chokeberry, quince, cranberry or apple) contain significantly more polyphenols than the fruit.

For the first time, analysis of mock privet leaf and fruit phenolics was revealed using LC-MS and demonstrated a great diversity of compounds in both samples for all solvents used during extraction. A total of 14 phenolic compounds for the extractions of leaves and fruit of mock privet were identified and quantified, including 7 flavonols, 2 anthocyanidins, 4 hydroxycinnamic acids and 1 secoiridoid by using LC-MS (Table 2). The results demonstrated that the fruit and leaves of the plant have different phenolic component profiles and concentrations. The individual constituents within the different extracts for both fruit and leaves were detected as the luteolin-7-O glucoside in the ethyl acetate (854 μg · g^–1 and 1,098 μg · g^–1), methanol (1,241 μg · g^–1 and 2,136.43 μg · g^–1) and aqueous (509 μg · g^–1 and 898.23 μg · g^–1) extracts, respectively. Similar to this study, Yazici Tutuinis et al. (2016) also reported the major phenolic of mock privet leaves as luteolin-7-O glucoside. Followed by caffeic acid was investigated in the ethyl acetate (202.75 μg · g^–1 and 562 μg · g^–1), methanol (1,094.04 μg · g^–1 and 1,596 μg · g^–1) and 615.33 μg · g^–1 and 1,367 μg · g^–1 in order of aqueous extracts of fruit and leaves. Rutin was found in fruit and leaves extracts of the ethyl acetate (430.84 μg · g^–1 and 24.90 μg · g^–1), methanol (632.16 μg · g^–1 and 254.34 μg · g^–1) and aqueous extracts (822.09 μg · g^–1 and 361.94 μg · g^–1), respectively. Apparently, both leaves and fruit of mock privet comprised of oleuropein content. In addition, ethyl acetate extraction of both parts indicated the highest concentration of oleuropein, which was followed by methanol and aqueous extracts. Also, the oleuropein content of the leaves was higher than the fruit. Only one study reported the oleuropein content of the ethanolic extraction of mock privet leaves with an amount of 160 mg · L^–1 (Azaizeh et al., 2013). This is the first report about the oleuropein content of mock privet fruit. The results of phenolic composition indicated that leaves of mock privet contain more polyphenols than fruit. While apigenin aglycones were only detected for all extraction types of the leaves, the fruit did not have it. Ayranci and Erkan (2013) analysed the phenolic composition of methanolic extraction of the mock privet fruit using HPLC-DAD. It was reported that the fruit extracts included cyanidin, rosmarinic acid, cyanidin 3-O-glucoside, ferulic acid, cyanidin-3-O-rutinoside and caffeic acid at amounts of 191.4 mg · 100 g^–1, 190.1 mg · 100 g^–1, 90.4 mg · 100 g^–1, 289.1 mg · 100 g^–1, 225.2 mg · 100 g^–1 and 221.2 mg · 100 g^–1 fresh weight of fruit, respectively. Chlorogenic and p-coumaric acids were also detected in lower amounts (Ayranci and Erkan, 2013). Longo et al. (2007) investigated the phenolic composition of mock privet fruit, and it was found that cyanidin 3-O-rutinoside was the major polyphenol of the fruit whereas cyanidin 3-O-glucoside have found in low quantities. In this study, the aqueous extract of the fruit contains a higher amount of cyanidin 3-O-rutinoside than cyanidin 3-O-glucoside. In addition, fruit contains higher cyanidin 3-O-rutinoside and cyanidin 3-O-glucoside than the leaves extracts. Recently another study demonstrated that leaves of mock privet contain luteolin and quercetin derivatives as major polyphenols followed by caffeic acid, rutin and kaempferol derivatives (Gori et al., 2020). In addition to this study, Gori et al. (2021) analysed the phytochemical content of mock privet leaves’ different tissues (e.g., adaxial and abaxial epidermis, mesophyll parenchyma) during different seasons. The HPLC-DAD analyses showed that different leaf tissues have similar phytochemical content and the identified peaks were quercetin, kaempferol, luteolin-7-O-glycosides, hydroxycinnamic acid, apigenin and luteolin-4-O-glycoside derivatives. It was concluded that the evergreen habit of mock privet may affect flavonoids’ enhanced biosynthesis from spring to summer and the summer season contains higher phenolic content compared with other seasons (Gori et al., 2021).

Antidiabetic activity and DPPH radical scavenging of fruit and leaves of mock privet were presented in Table 3. The leaves for all extraction methods showed higher antioxidant and antidiabetic activities compared with fruit. This may be due to the higher amount of TPC and phenolic composition and their synergist effects on the leaves.

The extracts’ antioxidant capacity for both leaves and fruit were ranked using the IC₅₀ values (mg · mL^–1) as follows: aqueous extract, methanol extract and ethyl acetate. Table 3 shows that the aqueous extract was more effective than the methanol and ethyl acetate extracts for the DPPH radical scavenging activity whereas lowest for the antidiabetic activities. The phenolic and flavonoid content of the extracts, as well as individual phenolic compounds found in the extracts, such as luteolin-7-O glucosidase, which was effective in antioxidant and antidiabetic activities, may be related to the extracts’ functional potential (Park et al., 2016; Caporali et al., 2022). These findings are in agreement with the literature demonstrating that the antioxidant and antidiabetic activities are correlated to the phenolic content, which is also related to the solvent applied (Mokrani and Madani, 2016; Ismail et al., 2019; Venkatachalam et al., 2020). It is challenging to compare antioxidant assays due to the differences between the synergy of the antioxidant compounds in the mixture (e.g., some phenolic compounds respond slowly and take longer to react). Also, some compounds’ antioxidant activity responses may effect by the factors such as structure and concentration (Sarikurkcu et al., 2020). Ethyl acetate and methanol extractions indicated better inhibitory activity compared with α-amylase than the aqueous extract for both parts of the mock privet. Similarly, ethyl acetate extraction of leaves and fruit of the mock privet indicated significantly better inhibitory activity than the methanol and aqueous extracts, respectively, for the inhibition of α-glucosidase activity. A positive control (acarbose, 1 mM) inhibited both α-amylase and α-glucosidase activities with IC₅₀ values of 2.2 mg · mL^–1 and 1.3 mg · mL^–1, respectively. The higher ethyl acetate and methanol extracts activity in the results can be explained for various reasons. First, these solvents extracted the substance(s) more successfully with higher α-amylase inhibitory activity than water. Secondly, interactions are possible between the substance(s) in the extracts and the enzymes (Zengin et al., 2018). Finally, the observed effects may also have been caused by the synergistic interaction of phenolics with one another and with other extract constituents (Abdillahi et al., 2011). The antioxidant and antidiabetic outcomes of this study also revealed the mock privet’s leaves and fruit as a potential natural source for the food and pharmaceutical industries, regardless of solvent. Additionally, some studies reported important antioxidant capacities of mock privet leaves and fruit (Ayranci and Erkan, 2013; Gori et al., 2020, 2021). Therefore, both leaves and fruit of the mock privet may be used in the management of diabetes.