In Vitro Biological Activities of Fruits and Leaves of Elaeagnus multiflora Thunb. and Their Isoprenoids and Polyphenolics Profile

The objective of this study was in-depth identification of carotenoids and polyphenolic compounds in leaves and fruits of Elaeagnus multiflora Thunb. An additional aim was to assay their antioxidant and in vitro biological activities (the ability to inhibit pancreatic lipase, α-amylase, and α-glucosidase activity) of two cultivars: ‘Sweet Scarlet’ and ‘Jahidka’. Study results showed the presence of 70 bioactive compounds, including 20 isoprenoids and 50 polyphenols. The profile of identified bioactive compounds had not been examined in this respect until now. The total carotenoid, chlorophyll, and polyphenol levels and antioxidant activity of the foliar samples were virtually identical in both cultivars and clearly higher relative to those in the fruits. On the other hand, the ability to inhibit pancreatic lipase, α-amylase, and α-glucosidase activity of the fruits was clearly higher as compared to the leaves. The highest amount of phenolic acids, flavonols, and polymeric procyanidins was in the ‘Sweet Scarlet’ for fruit and leaves, while the highest amount of chlorophylls and carotenoids was in the ‘Jahidka’. The inhibition of α-amylase, α-glucosidase, and pancreatic lipase activities appeared to be better correlated with the carotenoid content, which warrants further studies of the possible anti-diabetic and anti-obesity actions of the major carotenoids found in the fruits (lycopene, phytoene, and lutein). In addition, strong correlation between antioxidant activity and phenols of E. multiflora Thunb. components can be effective in removing reactive oxygen species. The results of our study show that both the fruits and leaves of E. multiflora Thunb. can be important for health promotion through the diet and for innovating in the industry of functional food and (nutri)cosmetics.


Introduction
The study of biodiversity in the context of food security and health promotion through the diet has been and will continue to be an important research field, especially in the challenging current scenario Considering that there are very few works about the bioactive compounds' profile in selected morphological parts of E. multiflora Thunb. and their biological activity [2,3,5] this study was aimed at profiling the isoprenoid and polyphenolic contents of the leaves and fruits of two cultivars of E. multiflora Thunb, namely 'Jahidka' and 'Sweet Scarlet'. Additionally, the antioxidant as the ability to radical cation scavenging activity (ABTS), radical scavenging activity (DPPH) and reducing activity (FRAP) and in vitro biological activities as the ability to inhibit pancreatic lipase, α-amylase, and α-glucosidase activity of such materials was assessed. Given the versatility and importance of the compounds examined, the results of this study can be important to encourage the inclusion in the diet of E. multiflora Thunb. for health promotion and for the development of innovative products for health promotion or cosmetic purposes.

Materials
Fruits and leaves of Elaeagnus multiflora Thunb. cultivars, 'Jahidka' and 'Sweet Scarlet' (~5 kg per cultivar) were collected from Milanówek near Warsaw and the University of Warmia and Mazury in Olsztyn, Poland. The samples were collected at the optimum ripening time in 2019.

Determination of Polyphenols
For the extraction and determination of phenolic compounds, a protocol described before by Lachowicz et al. [17,18] was followed. The samples of fruits and leaves (1 g) were extracted with 10 mL of mixture grade methanol, ascorbic acid, and acetic acid. The extraction was performed twice by incubation under sonication (20 min). The samples were centrifuged (10 min. 19,000× g) and supernatants were filtered (Hydrophilic PTFE (polytetrafluoroethylene) 0.20 µm membrane). All extractions were carried out in triplicate.

Analysis of Proanthocyanidins by Phloroglucinolysis
Analysis of proanthocyanidins of samples was performed as described by Lachowicz et al. [18]. Fruits and leaves (5 mg) were mixed methanol solution with methanolic HCl and were incubated (at 50 • C, 30 min). After that the vials were put in an ice bath and 0.6 mL of the reaction medium was added, diluting with 1.0 mL of sodium acetate buffer. The samples were centrifuged (10 min, 20,000× g at 4 • C).
Phloroglucinolysis was analyzed using the liquid chromatograph Waters (Waters, Milford, MA, USA), consisting of a diode array and scanning fluorescence detectors, column manager. The separation was carried out using Cadenza CD C18 column (75 mm × 4.6 mm, 3 µm) kept at 15 • C. For phloroglucinolysis investigation the following solvent system (mobile phase A (25 mL acetic acid and 975 mL water) and mobile phase B (acetonitrile)) was applied. The detection of fluorescence was recorded at emission wavelength 360 nm and excitation wavelength 278 nm. The calibration curves and quantification were evaluated using standards: (−)-epicatechin, (+)-catechin, (−)-epicatechin-phloroglucinol, and (+)-catechins-phloroglucinol. The degree of polymerization was analyzed by evaluating the molar ratio of all the flavan-3-ol units. All results were evaluated in triplicate and administered as mg/100 g dry matter (d.m.).

Isoprenoids' Analysis
The extraction of isoprenoids (carotenoids and chlorophylls) was evaluated, according to Stinco et al. [30]. Approximately 50 mg of powder from the samples were used for the extractions. Subsequently, 1 mL of extracting solvent (dichloromethane/methanol/acetone, 50:25:25 v/v/v) was added. The mixture was vortexed 5 min at 3500 rpm and ultrasonicated (Ultrasons, JP Selecta, Barcelona, Spain) for 2 min. To promote phase separation, the sample was centrifuged (Microfuge 22R, Beckman Coulter, Madrid, Spain) at 18,000× g for 5 min at 4 • C. After recovering the colored fraction, the residue was extracted again with another aliquot of 1 mL dichloromethane. This procedure was repeated until the color exhaustion.
The Rapid Resolution Liquid Chromatography (RRLC) and Ultra Performance Liquid Chromatography (UPLC) analyses of isoprenoids compounds were carried out according to the validated method described by Stinco et al. [16].
2.6. In Vitro Assessment of Health-Promoting Capacities 2.6.1. Extraction Procedure The material of fruit and leaves (~1 g) were mixed with 20 mL of PBS (phosphate-buffered saline) buffer (pH 7.4) and incubated (1 h). After that, the material of fruit and leaves was centrifuged per 15 min and the supernatants and supernatants were filtered.
2.6.2. Antiradical activity ABTS radical activity (ABTS•+) method was carried out according to an earlier protocol described by Re et al. [31]. Briefly, after 6 min of reaction of sample (0.03 mL + 3 mL of ABTS solution) the material was analyzed at 734 nm. All material was evaluated in triplicate and administrated in mmol Trolox(TE)/g d.m.

Antiradical Activity
The free radical (DPPH) method was carried out according to an earlier protocol described [32]. A diluted extract was mixed with DPPH• dissolved in ethanol and water, and after 10 min of reaction the absorbance was measured at 517 nm. All samples were obtained in triplicate and expressed in mmol TE/g d.m.

Reducing Activity
The ferric reducing antioxidant activity (FRAP) method was carried out following the method described by Benzie and Strain [33]. In brief, 0.1 mL of sample was mixed with 0.9 mL of distilled water and 3 mL of ferric complex. After 10 min of reaction, the absorbance was measured at 593 nm. All samples were obtained in triplicate and expressed in mmol TE/g d.m.
2.6.5. The α-Amylase Inhibitor Activity Assay The α-amylase inhibitor (αA) activity was measured according to the method described by Jakubczyk et al. [34]. The αA activity was expressed as AIU (the amount of inhibitor that completely inhibited one enzyme unit)/mg of sample.
2.6.6. The α-Glucosidase Inhibitor Activity Assay The α-glucosidase inhibitor (αG) activity was measured according to the method described by Jakubczyk et al. [34]. The αG was administrated as AIU (the amount of inhibitor that completely inhibited one enzyme unit)/mg of sample.

Pancreatic Lipase Inhibitor Activity Assay
Pancreatic lipase inhibitory activity was determined with p-nitrophenyl acetate (pNPA) as the substrate according to the test described by Jakubczyk et al. [34]. Pancreatic lipase inhibitor activity was expressed as LIU (the amount of inhibitor that completely inhibited one enzyme unit)/mg of sample.

Polyphenolics' Profile
Qualitative and quantitative data about the phenolic profile of the two cultivars studied are summarized in Tables 1 and 2. Tentative identification revealed a total of 50 phenolic compounds in the examined material, including three phenolic acids, one hydrolyzable tannin, one stilbene, and 45 flavonols; 16 compounds were detected in fruits and 38 compounds in leaves of Elaeagnus multiflora Thunb. To date, very scarce information was available on the polyphenolic profile of fruits and leaves of these cultivars. A few compounds, usually phenolic acids, that were not detected in this study (Tables 1 and 2) have been previously reported [2,3]. Values are expressed as the mean (n = 3) ± standard deviation and different letters (between cultivars) within the same row indicate statistically significant differences (p < 0.05). Values are expressed as the mean (n = 3) ± standard deviation and different letters (between cultivars) within the same row indicate statistically significant differences (p < 0.05).
Among the phenolic acids, two compounds with numbers 1 and 2 were identified, based on fragmentation patterns and comparison with literature data [23,35], as quinic acid (m/z (m stands for mass and z stands for charge number of ions) 191) and 3-p-coumaroylquinic acid (m/z 337). These compounds were identified only in the leaves of E. multiflora Thunb. In turn, the compound detected in leaves (no. 11) and in fruits (no. 2) was tentatively identified as sinapic acid-O-hexoside based on findings reported by Spínola et al. [26] for E. umbellata Thunb.
However, of the 43 flavonols identified, 27 had not been previously determined in any plant material. In leaves, tentative identification revealed 21 compounds, including five compounds belonging to quercetins and 15 to kaempferol derivatives (or glycons). Compound no. 4 was tentatively identified based on the primary ion at m/z 757 and MS/MS fragmentation giving ions at m/z 595, 463, and 301 through the loss of two glycoside residues and one pentose residue, respectively. A deprotonated molecule at m/z 901 was identified for compound no. 6 based on the loss of a rhamnose residue (146 Da) and fragmentation peaks at m/z 755, 609, 447, and 285, and based on the rhamnose residue and two glycol residues. These compounds were tentatively identified as quercetin glycoside-pentoside-glycoside and kaempferol di-rhamnoside-di-glucoside, respectively. In turn, compound no. 7 was classified as quercetin pentoside-rutinoside based on the loss of a pentose residue (132 Da) at the primary peak of m/z 741, and losses of rhamnose and glycol residues for ions at m/z 609, 463, and 301. Compound nos. 12 and 14 were identified at the primary m/z ion of 887, however, they had various primary ions, which for the first compound was determined at m/z 285 and allowed identifying it as kaempferol pentoside-rhamnoside-rutinoside, and for the second compound at m/z 301, which allowed identifying it as quercetin rhamnoside-pentoside-rhamnoside. Compound no. 16 with the primary ion of m/z 739 lost the rhamnose residue, while its fragmentation revealed peaks at m/z 593 and 285 and the loss of rutinoside residue, which enabled its tentative identification as kaempferol rhamnoside-rutinoside. In turn, compound no. 17  Compound no. 26 was tentatively identified as kaempferol pentoside-rhamnoside-glucuronide based on its fragmentation at the primary ion of m/z 739 into ions: At m/z 563 through the loss of a glucuronide residue, at m/z 417 through the loss of an additional rhamnose residue, and at m/z 285 through the loss of a pentose residue. In the case of compound no. 28, the primary peak occurred at m/z 709 due to the loss of a pentose residue (132 Da) and its fragmentation revealed peaks at m/z 577, 431, and 285 due to the loss of two additional rhamnose residues. Therefore, this compound was tentatively identified as kaempferol pentoside-di-rhamnoside. Based on the primary ion at m/z 577 and fragmentation ions at m/z 431 and 285, compound no. 29 was tentatively identified as kaempferol di-rhamnose due to the loss of two rhamnose residues. Compound no. 31 was tentatively presented as kaempferol glucoside-glucuronide based on its primary ion at m/z 623 and MS/MS fragmentation ions at m/z 447 and 285. In addition, compound no. 34 with the primary peak at m/z 891, respectively, had an acyl group (42 Da) and was tentatively identified as acetylated kaempferol derivatives based on the primary peak at m/z 285.
In addition, seven compounds, including four quercetin derivatives, two kaempferol derivatives, and one isorhamnetin derivative were tentatively identified in fruit, based on the presence of primary peaks at m/z 301, 285, and 315. For this reason, compound no. 1 was tentatively presented as quercetin-rhamnoside-pentoside-rutinoside based on its primary ion appearing at m/z 887 due to the loss of a rhamnose residue (146 Da) and fragmentation peaks at m/z 609, 579, and 301 due to the loss of pentose, rhamnose, and hexose residues, respectively. Compound no. 3 showed the main peak at m/z 725 and fragmentation ions at m/z 597, 417, and 285, and was tentatively identified as kaempferol-pentoside-rutinoside. Compound no. 5 was tentatively identified as quercetin-pentoside-rutinoside based on the primary ion at m/z 741 and its MS/MS fragmentation giving peaks at m/z 609 and 433. In turn, compound no. 6 was tentatively presented as quercetin-3-O-rhamnose-7-O-pentoside based on its primary ion at m/z 595 and fragmentation ions at m/z 433 and 301. Compound no. 7 was classified based on the pseudomolecular ion at m/z 739 due to the loss of rhamnose residue, and fragmentation of this compound revealing peaks at m/z 539 and 285 due to the loss of deoxyhexose residues. In turn, compound no. 8 was fragmented into peaks at m/z 593, 447, and 301, and due to the loss of three rhamnose residues was tentatively identified as quercetin-tri-rhamnoside (at m/z 739). In addition, compound no. 13 was tentatively identified as isorhamnetin 3-O-(6"malonyl)-glucuronide-rhamnoside based on its primary ion at m/z 723 and its MS/MS fragmentation ions at m/z 491 and 315.
The polyphenolic content in fruits of 'Sweet Scarlet' cultivars reached 1388.63 mg/100 g d.m., and was 1.4 times higher than in the fruits of 'Jahidka' cultivars. In turn, in leaves, it reached 1501.46 mg/100 g d.m. in the case of 'Sweet Scarlet' cultivars and 1080.84 mg/100 g d.m. in the 'Jahidka' cultivars, which was ca. 1.3 times higher compared to the fruits of the same cultivars. Leaves turned out to be richer in phenolics compared to fruits due to the complexity of the biosynthesis process in plants, which is dependent on various factors, including environmental signals related to the protection against biotic and abiotic stress or developmental signals [17,36,37]. On the other hand, it is important to consider that the levels of secondary metabolites in plants are dependent on diverse factors (agronomic, climatic, genotypic) [38], hence the difficulty of meaningfully comparing them across different studies. As reported by Jaakola et al. [39], a higher content of polyphenolic compounds was identified in leaves of plants growing under intensive sunlight, which results in the enhanced gene expression coupled with phenolic biosynthesis. In turn, according to Lee et al. [3], the content of phenols in fruits of E. multiflora Thunb. from Korea reached 315 mg/100 g d.m. and was 2.9 and 4.0 times lower than the fruits of 'Jahidka' and 'Sweet Scarlet' cultivars growing in Poland. As shown by Spínola et al., [26], the content of polyphenols in fruits of E. umbellata Thunb. was at 556 mg/100 g d.m. and was 1.6 and 2.3 lower compared to fruits of 'Jahidka' and 'Sweet Scarlet' cultivars of E. multiflora Thunb. belonging to the same family. In turn, the content of phenols in leaves of 'Jahidka' and 'Sweet Scarlet' cultivars of E. multiflora Thunb. was 3.6 and 2.6 times lower compared to their content in the same morphological part of E. umbellata Thunb. In turn, fruits of E. angustifolia L. were, on average, 10.8 times poorer in polyphenolics than fruits of E. multiflora Thunb., whereas in leaves their content reached 10.28 mg gallic acid equivalents (GAE)/100 g fresh mass [40]. From this information it can be stated that, considering the published data, the fruits of E. multiflora Thunb. growing in Poland are richer in polyphenolic compounds compared to other fruits of the same family.
In the analyzed fruits, procyanidin polymers accounted for 94.8% > flavonols for 5% > phenolic acids for 0.2% of total phenols, whereas in leaves they were also the major group of compounds and accounted for 81.2% > flavonols for 18.5% > phenolic acids 0.2% of total phenols. However, according to literature data, these were the flavonols that represented the major group of phenols in fruits of E. umbellata Thunb. with their content reaching 78.8%. In contrast, ellagitannins turned out to be the prevailing class in leaves accounting for 56.8% of total phenols [26]. In the fruits of E. multiflora Thunb. growing in China, the predominating class of compounds was represented by flavan-3-ols (75% of total compounds) [3]. In addition, the major flavonol of fruits turned out to be kaempferol-deoxyhexoside-pentoside (30%), whereas in leaves, kaempferol di-rhamnoside-glucoside (59%). Among the acids, sinapic acid-O-hexoside predominated in both fruits and leaves and accounted for 100% and 53% of total acids, respectively. In turn, kaempferol-O-(coumaroyl)hexoside and bis-HHDP-O-hexoside were the major compounds in fruits and leaves of E. umbellata Thunb., with their contents reaching 37.2% and 26.7%, respectively. Procyanidin polymers are characterized by a higher degree of polymerization compared to flavan-3-ol dimers, while these compounds have earlier been detected in, e.g., fruits and leaves of cranberry, with the HPLC method [17]. It is noteworthy that the content of procyanidin polymers has never been determined in fruits from the family E. multiflora Thunb. before. According to Pei et al. [41], the procyanidin polymers' content in fruits of E. umbellata Thunb. was 255 mg catechin equivalents (CE)/100 g d.m., and was on average 4 times lower than in the fruits of E. multiflora Thunb.

Isoprenoids' Profile: Chlorophylls and Carotenoids
Carotenoids are associated to chlorophylls in the photosynthetic apparatuses of plants and other photosynthetic organisms, where they enhance the efficiency of light harvesting and electron transfer, protect against photooxidation, and intervene in their assembly and stabilization, with the amount of chlorophylls in these associations being higher [42,43]. Thus, the levels of chlorophylls were considerably higher relative to those of carotenoids. The levels of total chlorophylls were significantly different between the two cultivars (Table 3) [44]. The carotenoid profile of photosynthetic tissues is very well conserved. Normally, the major carotenoids are lutein, β-carotene, violaxanthin, and neoxanthin, in this order (the latter two at very similar levels). Lutein and the provitamin A carotenoid β-carotene are bioavailable in humans and among the most extensive carotenoids in relation to health promotion [45]. Minor carotenoids include α-carotene, β-cryptoxanthin, zeaxanthin, antheraxanthin, and lutein 5,6-epoxide. Lettuce photosynthetic tissues are a special case as they accumulate an unusual xanthophyll named lactucaxanthin [46]. Considering this paradigm, it can be readily inferred that the green leaves of the samples analyzed contained unusually elevated levels of α-carotene, which was, in fact, the second predominant carotenoid to lutein (Table 3). This fact can also be important from a nutritional point of view as α-carotene is a major provitamin A carotenoid, which is essential to fight vitamin-A deficiency in many regions [11]. Interestingly, the nutritional importance of the little-studied α-carotene is being pinpointed in studies carried out in the last decade. In relation to the major dietary provitamin A carotenoids, both α-caroteneand β-cryptoxanthin-rich foods are thought to exhibit greater apparent bioavailability than β-carotene-rich foods in Western diets [47]. The α-carotene circulating levels have been reported to be positively correlated with bone mineral density in Chinese individuals [48]. Furthermore, α-carotene has been shown to inhibit metastasis in Lewis lung carcinoma in vitro and to suppress lung metastasis and tumor growth (in association with taxol) in tumor-xenografted mice [49]. The intake of this carotenoid has also been shown to be associated to a reduced risk of prostate malignancy in Japanese [50].
Regarding the carotenoid profile of fruits, lycopene was the predominant carotenoid. This has also been extensively studied in relation to health promotion [45]. Lutein and phytoene were also detected. The colorless carotenoid, phytoene, has been largely neglected in food science and technology, nutrition, and health, although it is gaining momentum in research due to accumulating evidence indicating that it is a major dietary carotenoid, readily bioavailable in humans and that can intervene in biological actions that can be positive for health and even beauty promotion [12,51]. The total carotenoid levels of the foliar samples (calculated as the sum of the individual contents) were virtually identical, although significantly different. In the case of the fruits, there were clear differences as the levels in the 'Jahidka' cultivar were~2.4-fold higher relative to those of the 'Sweet Scarlet' cultivar sample (971.48 and 400.93 mg/100 g d.m., respectively, data not shown). Statistically significant differences in the levels of lutein and lycopene were obtained. The levels of carotenoids in plants depend on factors of a different nature, including genotype, climatic conditions of the production area, or agronomic factors, among others [52]. Table 3. Quantification of carotenoids and chlorophylls (mg/100 g dry matter (d.m.)) in leaves and fruits of Elaeagnus multiflora Thunb. Values are expressed as the mean (n = 3) ± standard deviation and different letters (between cultivars) within the same row indicate statistically significant differences (p < 0.05).

In Vitro Antioxidant Capacity
The antiradical activity assayed in leaves of 'Sweet Scarlet' and 'Jahidka' cultivars of E. multiflora Thunb. reached 5.56 and 5.08 mmol Trolox (TE)/g d.m. for ABTS assay and 2.69 and 2.51 mmol TE/g d.m. for DPPH assay, respectively (Table 4). These values were 9.7 and 6.2 times and 9.8 and 5.9 times higher than in the fruits of the respective cultivars. In turn, the reducing potential (FRAP assay) determined in leaves of 'Sweet Scarlet' and 'Jahidka' cultivars reached 0.30 and 0.33 mmol TE/g d.m., respectively, and was 3.7 and 2.6 times higher compared to the fruits of the same cultivars of cherry silverberry. As indicated by literature data and results of the present study, the value of the antioxidant capacity is strongly Pearson correlated with the content of bioactive compounds, including phenols (r 2 = 0.912) and carotenoids (r 2 = 0.539), which was confirmed by strong correlations [17,26]. In addition, as reported by Spínola et al. [26], leaves of E. umbellate Thunb. exhibited 6.9 times antiradical activity compared to fruits, which was consistent with our observations. Furthermore, leaves and fruits of E. umbellate showed 1.2 and 1.4 times lower activity, respectively, compared to the same morphological parts of E. multiflora Thunb. The values of antiradical activity obtained for fruits and leaves of cranberry, bilberry, and quince [36] were 7.0 and 5.3 times, 1.7 and 6.6 times, and 8.9 and 4.4 times lower, respectively, compared to fruits and leaves of E. multiflora Thunb. In turn, the reducing potential determined for fruits and leaves of cranberry [36] was similar to the results obtained for E. multiflora Thunb., whereas its value assayed for fruits and leaves of bilberry and quince [36] was, respectively, 2.8 and 1.8 times as well as 2.0 and 2.2 times higher compared to the same morphological parts of E. multiflora Thunb. It is worthy of mention that the antiradical activity of E. multiflora Thunb. leaves was similar to the activity determined for green tea [53], which indicates a high potential of not only fruits but also leaves of fruit trees, including leaves of E. multiflora Thunb., as ingredients in developing tea formulas or as additives in designing functional foods safe to consumers. In their study into the toxicity of cranberry leaves, Booth et al. [54] demonstrated that their extract could be used in the production of functional food safe to consumers. Values are expressed as the mean (n = 3) ± standard deviation and different letters (between cultivars) within the same row indicate statistically significant differences (p < 0.05).
3.3.2. The Ability to Inhibit α-Amylase, α-Glucosidase, and Pancreatic Lipase Activity Fruits and leaves of Elaeagnus multiflora Thunb. were also analyzed as potential sources of substances inhibiting activities of enzymes involved in the pathogenesis of the metabolic syndrome ( Table 4). The against inhibition of α-amylase (αA) and α-glucosidase (αG) activities in cherry silverberry fruits of 'Jahidka' and 'Sweet Scarlet' cultivars showed no statistically significant differences, whereas leaves of 'Jahidka' cultivars were 20.6 and 1.7 times, and leaves of 'Sweet Scarlet' cultivars were 7.7 and 2.1 times less active than the fruits of the same cultivars. The aforementioned enzymes take part in the degradation of complex saccharides to become absorbable during digestion, whereas inhibition of their activity is used in the prevention of type 2 diabetes [50,52]. A similar extent of αG activity inhibition was obtained in the case of leaves of E. angustifolia L. [53]. The study conducted by Spínola et al. [26] demonstrated that the inhibition of αA and αG activities determined in E. umbellata Thunb. reached the biochemical half maximal inhibitory concentration (IC 50 ) = 4.01 and 6.02 mg/mL in fruits as well as IC 50 = 2.18 and 4.76 mg/mL in leaves, respectively, whereas in Sambucus lanceolata it reached IC 50 = 6.01 and 7.55 mg/mL in fruits as well as IC 50 = 7.71 and 4.97 mg/mL in leaves, respectively. In that study, however, results were provided as IC 50 values, which makes their comparison with our data impossible. Nevertheless, the above authors noted that the leaves exhibited more inferior ability to inhibit activities of the analyzed enzymes, which was consistent with our observations. In addition, they showed a weak Pearson's correlation with the total content of phenols.
In turn, the inhibition of pancreatic lipase activity by fruits of 'Jahidka' cultivars reached 1.39 U/g dry matter and was 1.9 times higher compared to 'Sweet Scarlet' cultivars, whereas no pancreatic lipase-inhibiting activity was demonstrated in leaves (Table 4). This enzyme is responsible for the breakdown of lipids to become absorbable in the intestinal lumen [55]. Inhibition of its activity is exploited in obesity prevention [26]. As reported by Spínola et al. [26], the inhibition of pancreatic lipase activity determined in E. umbellate Thunb. and S. lanceolate reached IC 50 = 9.68 and 7.75 mg/mL in fruits, respectively, and turned out to be 1.9 and 1.2 times lower in leaves, respectively. These authors demonstrated also that the higher capability for inhibiting its activity in S. lanceolate was strongly correlated with the phenolic profile, including mainly anthocyanins, claimed to be its inhibitors, which were however not identified in E. umbellate Thunb. Other reported inhibitors of pancreatic lipase activity include ellagitannins and 5-O-coumarylquinic acid [26,56]. This can be the cause of the incapability for inhibiting this enzyme activity in leaves of Elaeagnus multiflora Thunb., characterized by a diverse phenolic profile. What is more, the results showed a strong Pearson's correlation with the total content of carotenoids with the inhibition of pancreatic lipase activity r 2 = 0.938, while the correlation between the inhibition of pancreatic lipase activity and sum of polyphenols was positive r 2 = 0.307. The higher correlation with carotenoids is due to the fact that adipose tissue is a main organ in obesity etiology and site of storage for carotenoids [57].
According to literature data [26,58,59], hyperglycemia enhances the generation of reactive oxygen species that may lead to the development of inflammatory conditions and carcinogenic lesions. However, it turns out that both leaves and fruits exhibit high antioxidant activities in in vitro biological activities due to the abundance of their bioactive compounds [60], which are capable of not only inducing hypoglycemia but also removing reactive oxygen species [26].
As far as the inhibition of the three enzymes' activities tested was concerned, it was observed that the correlations between the levels of phenolics and the ability to inhibit αA, αG, and pancreatic lipase were r 2 = 0.647, r 2 = 0.350, and r 2 = 0.307, respectively. In the case of carotenoids, the correlations were r 2 = 0.996, r 2 = 0.959, and r 2 = 0.938, respectively. These results suggested that the inhibiting activities were more related to the carotenoid levels than to the phenolic levels. Furthermore, our results suggest that the total phenolics' content was less important in the inhibition of the analyzed enzymes compared to contents of particular classes of phenols [25,61]. A large diversity of structures of phenols or carotenoids within a class or between various classes of these compounds determines their capability to bind with digestive enzymes [26,59,62]. This may be the reason for differences in the ability to inhibit enzymatic activity by leaves and fruits of E.s multiflora Thunb. Besides, phenols in plant extracts can exhibit the antagonistic and/or synergistic activity towards digestive enzymes (the main enzymes digestive of carbohydrates are α-amylase and α-glucosidase, and of fats it is pancreatic lipase) and influence the inhibitory activities [26,62,63]. Reported inhibitors of αG activity of E. umbellata Thunb. include coumaroylquinic acid and proanthocyanidins, while proanthocyanidins, rutin, and quercetin were the main compounds of E. umbellata Thunb. [26] and also occurred in E. multiflora Thunb. against to inhibit α-amylase.

Conclusions
The leaves were found to contain 13 isoprenoids with chlorophyll found as the major compound, and 38 phenolic compounds with kaempferol di-rhamnoside-glucoside and sinapic acid-O-hexoside as the predominating ones. In fruits, eight isoprenoids were detected, including (all-E)-lycopene as the major compound, and also 16 phenolics including kaempferol-deoxyhexoside-pentoside and sinapic acid-O-hexoside as the major compounds. However, among the identified classes of phenolic compounds, the procyanidin polymers were found to prevail as they accounted for 94.8% and 81.2% of total phenols in fruits and leaves of E multiflora Thunb., respectively.
The total carotenoid levels of the foliar samples were virtually identical in both cultivars and clearly higher relative to those in the fruits. Apart from the health-promoting carotenoids, lutein and β-carotene, the leaves of the E multiflora Thunb. samples analyzed contained unusually high levels of α-carotene, a provitamin A carotenoid that may be important to promote bone health and to decrease the risk of cancer. The fruits contain lycopene as the predominant carotenoid as well as low levels of lutein and phytoene, all being of interest in terms of health promotion. On the other hand, the study of the unusually high levels of α-carotene in leaves in the context of photosynthesis appears as an interesting topic for further studies, as the carotenoid profile of plant photosynthetic tissues is normally very constant and well conserved. Leaves also turned out to be richer in phenolics compared to fruits. The polyphenolic content in leaves and fruits of 'Sweet Scarlet' were~1.5-and 1.4-fold higher relative to those of the 'Jahidka' cultivar. In addition, the antiradical capacity and reduction potential were found to be higher in the leaves than in the fruits and also showed strong correlation with polyphenols that may be important to remove reactive oxygen species, and thus prevent the promotion of cancer.
The against inhibition of αA and αG activities of the fruits was similar in both cultivars, and clearly higher as compared to the leaves. The inhibition of pancreatic lipase activity was only detected in fruits and it was~1.9-fold higher in the 'Jahidka' cultivars. The inhibition of the analyzed enzymes' capacities appeared to be better correlated with the carotenoid content, which warrants further studies of the possible anti-diabetic and anti-obesity actions of lycopene, phytoene, and lutein.
In summary, the results of this study indicate that the inclusion of fruits and leaves on E. multifolia Thunb. in the diet is desirable in order to increase the intake of vitamin A (in the form of αand β-carotene) and polyphenolic compounds that could contribute to promote health by diverse mechanisms and also may be effective in the prevention of chronic noncommunicable diseases. Such materials also appear as promising in order to innovate in the context of functional foods and other industries, including (nutri)cosmetics.