Glycoside Hydrolases and Non-Enzymatic Glycation Inhibitory Potential of Viburnum opulus L. Fruit—In Vitro Studies

Phytochemicals of various origins are of great interest for their antidiabetic potential. In the present study, the inhibitory effects against carbohydrate digestive enzymes and non-enzymatic glycation, antioxidant capacity, and phenolic compounds composition of Viburnum opulus L. fruits have been studied. Crude extract (CE), purified extract (PE), and ethyl acetate (PEAF) and water (PEWF) fractions of PE were used in enzymatic assays to evaluate their inhibitory potential against α-amylase with potato and rice starch as substrate, α-glucosidase using maltose and sucrose as substrate, the antioxidant capacity (ABTS, ORAC and FRAP assays), antiglycation (BSA-fructose and BSA-glucose model) properties. Among four tested samples, PEAF not only had the highest content of total phenolics, but also possessed the strongest α-glucosidase inhibition, antiglycation and antioxidant activities. UPLC analysis revealed that this fraction contained mainly chlorogenic acid, proanthocyanidin oligomers and flavalignans. Contrary, the anti-amylase activity of V. opulus fruits probably occurs due to the presence of proanthocyanidin polymers and chlorogenic acids, especially dicaffeoylquinic acids present in PEWF. All V. opulus samples have an uncompetitive and mixed type inhibition against α-amylase and α-glucosidase, respectively. Considering strong anti-glucosidase, antioxidant and antiglycation activities, V. opulus fruits may find promising applications in nutraceuticals and functional foods with antidiabetic activity.


Introduction
Type 2 diabetes mellitus (T2DM) is a global problem because it concerns hundreds millions of people and has a significant impact on the health, quality of life and the health care system [1]. Evidence from epidemiological studies suggests that many cases of T2DM can be prevented by improving the main modifiable risk factors such as obesity, sedentary lifestyle and unhealthy diet contains large amounts of carbohydrates and fats [2,3]. Therefore, diet, weight control and exercise are essential and effective means of improving glucose homeostasis. Some medicinal plants, fruits, and vegetables have been reported in the literature as having been used to control diabetes [4][5][6][7][8][9]. In addition, the antidiabetic agents from the natural resources can be an alternative to hypoglycemic drugs that are not always satisfactory and show side effects [10,11]. The antidiabetic activity of plant-derived foods and extracts has been attributed, inter alia, to the naturally available phenolic compounds [6,[12][13][14][15][16][17][18][19]. These secondary plant metabolites comprise one (phenolic acids such as hydroxybenzoic and hydroxycinnamic acids) or more (polyphenols containing flavonoids, tannins, lignans) aromatic rings with attached hydroxyl group in their structures [20]. These secondary plant metabolites have the capacity to inhibit activity of carbohydrate hydrolyzing digestive enzymes, modulate the glucose transport, stimulate insulin secretion

Plant Material and Samples Preparation
The dried fruits of V. opulus were purchased from "Natura Wita Ltd." (Kopernia, Poland; https://natura-wita.pl/kontakt.html, accessed on 15 June 2021). Before the extraction process, dry fruits were ground in a coffee grinder and extracted with 70% acetone (1:20, w/v) on a magnetic stirrer at room temperature for 3 h. Then, the mixture was incubated at room temperature for 18 h followed by the extraction on a magnetic stirrer at room temperature for 3 h. After centrifugation at 5000 rpm for 10 min the supernatant was evaporated at 40 • C under reduced pressure in order to remove acetone and lyophilized to obtain the crude extract (CE). The part of CE was applied onto a Sep-Pak C18 cartridge (10g capacity, Waters Corp., Milford, MA, USA) that was previously activated with methanol and water. The column was washed with water in order to eliminate carbohydrates, proteins and other polar compounds. The phenolic compounds were then eluted with methanol. Methanol fraction was evaporated under reduced pressure (T < 40 • C), and the aqueous phase was lyophilised to afford purified extract (PE). The part of PE (300 mg) was suspended in 30 mL of water, and then, partitioned with 30 mL of ethyl acetate, repeated three times. The ethyl acetate fraction was evaporated to dryness, solubilized with water and lyophilised. This sample was named ethyl acetate fraction of purified extract (PEAF). Water fraction was concentrated, lyophilised and called water fraction of purified extract (PEWF). The yield and abbreviations for each of V. opulus samples are shown in Figure 1. For in vitro analysis CE, PE, PEWF and PEAF were dissolved in water at a concentration 10 mg/mL.

Quantification of Individual Phenolic Compounds
UPLC analysis was performed on an ultra-performance liquid chromatograph (Waters Acquity UPLC system, Milford, MA, USA) equipped with a binary pump, an autosampler, a column compartment and a diode array detector. Briefly, samples were eluted with a gradient of solvent A (4.5% formic acid in ultrapure water) and B (acetonitrile) on an Acquity UPLC HSS T3 C18 column (150 × 2.1 mm, 1.8 µm; Waters) operating at 30 °C as described in the previous work [38]. The gradient program was as follows: initial conditions 99% (A), 12 min 75% (A), 12.5 min 100% (B), 15.0 min 99% (A). The flow rate was 0.45 mL/min and the injection volume was 5 µL. The identification of phenolic compounds was described in detail previously [39]. The content of neochlorogenic acid, chlorogenic acid, cryptochlorogenic acid, caffeic acid, procyanidin C1, quercetin 3-glucoside, quercetin 3-rutinoside, and quercetin 3-rhamnoside were quantified using corresponding standard calibration curves. A quantificative analysis of other phenolics was based on the standards as follow: chlorogenic acid was used for the hydroxycinnamic acid derivatives, (+)-catechin for (epi)-catechin hexoside, procyanidin C1 for procyanidin tetramer, cinchonine for cinchonain Ix derivatives, quercetin 3-glucoside for quercetin 3sambubioside. The results were expressed as mg/g CE, PE, PEWF, and PEAF.

Total Proanthocyanidins Content
The content of total proanthocyanidins was determined after their acid depolymerization to the corresponding anthocyanidins as described by Rösch et al. [40] and calculated by the molar extinction coefficient of cyanidin (ε = 17,360 L/mol −1 ·cm −1 and molar mass 287 g/mol), and was expressed as mg of cyanidin equivalents (CYE)/g of V. opulus samples.

α-Amylase Inhibition Assay
The α-amylase inhibition assay was based on a previously-described spectrophotometric method [41] with some modifications. All reagents were prepared with 0.1 M phosphate buffer containing 6 mM CaCl2 (pH 6.9). Briefly, 20 µL of diluted V. opulus samples and 40 µL of potato starch (0.83 g/L) or starch from rice (1 g/L) solution were mixed with 20 µL of α-amylase (0.1 mg/mL for potato starch and 0.2 mg/mL for starch from rice) in a 96-well plate. After incubation at 37 °C for 10 min, the reaction was stopped by addition

Total Proanthocyanidins Content
The content of total proanthocyanidins was determined after their acid depolymerization to the corresponding anthocyanidins as described by Rösch et al. [40] and calculated by the molar extinction coefficient of cyanidin (ε = 17,360 L/mol −1 ·cm −1 and molar mass 287 g/mol), and was expressed as mg of cyanidin equivalents (CYE)/g of V. opulus samples.

α-Amylase Inhibition Assay
The α-amylase inhibition assay was based on a previously-described spectrophotometric method [41] with some modifications. All reagents were prepared with 0.1 M phosphate buffer containing 6 mM CaCl 2 (pH 6.9). Briefly, 20 µL of diluted V. opulus samples and 40 µL of potato starch (0.83 g/L) or starch from rice (1 g/L) solution were mixed with 20 µL of α-amylase (0.1 mg/mL for potato starch and 0.2 mg/mL for starch from rice) in a 96-well plate. After incubation at 37 • C for 10 min, the reaction was stopped by addition of 80 µL of 0.4 M HCl followed by 100 µL of 5 mM I 2 in 5 mM KI. The absorbance was read at 600 nm using a microplate reader (Synergy2, Biotek Instruments Inc., Winooski, VT, USA). Acarbose and chlorogenic acid were also used as a positive control and reference phenolic compound, respectively. Each sample was measured in triplicate. The IC 50 value was calculated by regression analysis as the mean concentration of tested sample required to demonstrating at 50% inhibition against α-amylase in this condition.

α-Glucosidase Inhibition Assay
The assessment of the α-glucosidase activity was slightly modified according to Adisakwattana et al. [42] and Gao et al. [43]. Briefly, 125 mg of rat intestinal acetone powder was mixed with 2.5 mL of 0.9% NaCl solution. Enzyme isolation was performed in an ultrasonic cleaner (for 30 sec) in an ice bath, then 30 sec without ultrasonic. This step was repeated 12 times. After that, the solution was centrifuged at 4 • C (3000 rpm, 30 min). 50 µL of diluted V. opulus samples were mixed with 50 µL of enzyme supernatant (diluted 2 times for reaction with maltose). After incubation at 37 • C for 10 min, 50 µL of substrate (0.1 M maltose or 0.5 M sucrose solution prepared in 0.1 M phosphate buffer, pH 6.9) was added and incubated in a 96-well microplate at 37 • C for 20 min for maltose and 60 min for sucrose. The reaction was stopped by adding 150 µL of 2 M Tris-HCl buffer (pH 7.0). The concentration of glucose released from the reaction mixture was determined by the commercial glucose test (BIOMAXIMA SA, Lublin, Poland). Acarbose and chlorogenic acid were also used as a positive control and reference phenolic compound, respectively. The concentration of the tested sample able to inhibit the α-glucosidase by 50% (IC 50 value) was calculated by regression analysis.

Mode of Enzyme Inhibition
The α-amylase and α-glucosidase inhibition modes of different V. opulus fruit samples (PE, PEWF, PEAF) and reference substances (chlorogenic acid and acarbose) were determined by Lineweaver-Burk model. The enzymes activities without and with tested samples were studied using the methods described in Sections 2.5 and 2.6. A double reciprocal graph (1/v versus 1/S, where "v" is the reaction rate and "S" the concentration of the substrate), was plotted. All samples were tested at three different doses using six concentrations of substrate.

Fluorescence Measurements
The effect of PE on fluorescence spectra of α-amylase and α-glucosidase at different concentrations of PE (from 0.014 to 0.136 mM) were performed using FluoroMax 4 (Jobin Yvon Spex, UK, Stanmore, Middlesex) spectrofluorometer (Horiba Scientific, Piscataway, NJ, USA) using a method described by Zhu et al. [44]. The fluorescence spectra of enzymes and their changes upon adding increasing amounts of PE of V. opulus fruits were recorded in the wavelength range of 315-450 nm upon excitation at 295 nm. Both excitation and emission slit widths were set at 2 nm. All measurements were performed in a standard quartz cuvette at 20 • C. The time-resolved fluorescence of α-amylase and α-glucosidase before and after adding increasing concentrations of PE was measured using the FL900CDT time-correlated single-photon counting fluorimeter (Edinburgh Analytical Instruments, Edinburgh, UK). The excitation and emission wavelengths were set to 295 and 350 nm, respectively. The instrumental response function was~1.18 ns (FWHM full width at half-maximum). The time-resolved measurements were performed at 20 • C in a standard quartz cuvette. Data acquisition and analysis of the obtained results were performed with the use of software provided by Edinburgh Analytical Instrumentation (F900). The goodness of fit was estimated by using χ2 values.

Protein Glycation Inhibition Assay
The bovine serum albumin (BSA) and glucose/fructose assay was carried according to the method of Błaszczak et al. [45] with some modifications. We prepared 25 mL of the test solution containing D-glucose or D-fructose (1.0 M), BSA (10 mg/mL) and sodium azide (0.1 mg/mL) in the phosphate buffer (0.1 M, pH 7.4). The test solution (1 mL) was incubated in the dark at 37 • C for 13 days for glucose and for 6 days for fructose with or without 0.5 mL of tested V. opulus samples. Aminoguanidine was used as positive control in the concentration from 41.67 to 166.67 µg/mL for glucose and from 22.22 to 133.33 µg/mL for fructose. Next, the formation of AGE was evaluated based on fluorescence intensity at λ = 330 nm (excitation wave) and λ = 460 nm (emission wave). Fluorescence intensity was measured in 96-well plates (black, OptiPlate-96 F, PerkinElmer, UK) in a microplate reader (Synergy2, BioTek Instruments Inc., Winooski, VT, USA).

In Vitro Antioxidant Assays
Antioxidant capacity of V. opulus fruit samples was determined by ABTS radical cation (ABTS), oxygen radical (ORAC) scavenging capacity, and as ferric reducing power (FRAP). Antioxidant capacity was expressed as mM Trolox equivalents (TE) per g of CE and PE extracts, and PEAF and PEWF fractions. ABTS and FRAP assays procedures have been detailed in our previous works [41,46]. The ABTS •+ monocation was generated by oxidation of ABTS with potassium persulphate. The tested sample (20 µL) was added to the ABTS •+ solution (1 mL) and after 6 min, the absorbance at 734 nm was measured. In the FRAP assay, 90 µL of tested sample was mixed with 0.27 mL of water and freshly prepared FRAP reagent (2.7 mL). After 10 min the absorbance at 593 nm was measured. ORAC assay was carried out using a Synergy TM 2 microplate reader (BioTek Instruments Inc., Winooski, VT, USA). Peroxyl radicals were generated by the spontaneous decomposition of AAPH at 37 • C. The loss of fluorescence of fluorescein was an indication of the extent of damage from its reaction with the peroxyl radicals. Fluorescence was read at 485 nm excitation and 520 nm emission every 2 min for 2 h.

Statistical Analysis
All samples were assayed in triplicate and results are given as the mean ± standard deviation using Microsoft Excel XP. Significance differences were calculated using one-way analysis of variance (ANOVA) using Statistica Ver. 6.0 (TIBCO Software Inc., Palo Alto, CA, USA) followed by post hoc Tukey test. Statistically significant differences were set at p < 0.05.

Effects of V. opulus Fruit Components on α-Amylase and α-Glucosidase Activity
The inhibitory effects of different samples separated from V. opulus dried fruits such as crude extract (CE), purified extract (PE), and two fractions separated from PE, called water fraction (PEWF) and ethyl acetate fraction (PEAF), were analyzed by the dose-effect plots (Supplementary Figure S1). In the present study, the α-amylase assay was performed in two systems using potato starch or starch from rice as substrates. Measurement of α-glucosidase activity was also carried out in two systems using sucrose or maltose as substrates. Inhibitory activity of extracts and fractions of V. opulus dried fruit had a direct linear relationship between concentration and percentage inhibitory activity.
The degree of inhibition of the tested enzymes depended on both the type of V. opulus sample and the type of substrate used. The V. opulus samples differed significantly (p < 0.05) in their IC 50 values (Table 1). PEWF was the strongest inhibitor of porcine pancreatic αamylase while PEAF against α-glucosidase as indicated by the lowest IC 50 values. CE and PEAF performed a weak inhibitory effect on α-amylase while CE on α-glucosidase. Purification of CE by solid-phase extraction on the C18 bed increased the inhibitory activity of PE against α-amylase 6.5-fold and almost 4-fold using potato starch or starch from rice, respectively. In the case of α-glucosidase, the inhibitory activity of PE exceeded almost 11fold and more than 7-fold the effectiveness of CE in the assay system containing maltose or sucrose as a substrate, respectively. However, the comparison of the IC 50 values for PEWF and PEAF, i.e., fractions separated from PE by liquid-liquid extraction with ethyl acetate, indicates a higher affinity of α-amylase inhibitors to water, and α-glucosidase inhibitors to ethyl acetate. All V. opulus fruit samples showed lower inhibitory activity against both enzymes compared to acarbose, with greater differences occurring with α-glucosidase. Chlorogenic acid-the reference phenolic compound-was a weaker inhibitor of both glycoside hydrolases compared to acarbose, and the weaker α-amylase inhibitor compared to all V. opulus fruit samples. In contrast, in the case of α-glucosidase, chlorogenic acid showed comparable inhibition efficiency to PEAF in the assay system with sucrose, but exhibited 1.7-fold lower inhibitory activity than PEAF in the system using maltose as a substrate.
A Lineweaver-Burk plot between 1/[substrate] and 1/V (reaction rate) was used to examine the action type of the most active V. opulus samples (PE and PEWF) against α-amylase (Supplementary Figure S2), or PE and PEAF samples against α-glucosidase (Supplementary Figure S3). For comparison, chlorogenic acid and acarbose were also tested.
In the study, the kinetic Michaelis-Menten parameters such as Michaelis constant (Km), maximum rate (Vmax), and inhibition constant (Ki) were calculated from Lineweaver-Burk graphs ( Supplementary Figures S2 and S3) and are summarized in Table 2. All tested V. opulus samples similar to acarbose have an uncompetitive type inhibition against α-amylase while chlorogenic acid inhibits this enzyme activity in competitive manner, regardless of Antioxidants 2021, 10, 989 7 of 18 the type of starch. Whereas, in the case of α-glucosidase, the tested V. opulus samples and acarbose have a mixed type of inhibition (competitive and noncompetitive), regardless of the type of disaccharide as substrate. The results also suggest that chlorogenic acid has a noncompetitive-type inhibition against this enzyme in both assay models.

Effect of PE on α-Amylase and α-Glucosidase Spectra
The fluorescence spectra of α-amylase and α-glucosidase at different concentrations of PE (0.014-0.136 mM as chlorogenic acid equivalents) are depicted in Figure 2. It was evident that with the increasing concentration of PE, the fluorescence intensity of both enzymes reduced progressively, which is indicative of molecular interactions between PE components and α-amylase as well as α-glucosidase. From Figure 2A,B it can be seen that addition of PE to α-amylase or α-glucosidase solution resulted in tryptophan fluorescence quenching. Moreover, a slight redshift of the tested enzymes emission spectra maximum may be observed. Due to the presence of this alteration to the fluorescence maximum, the Stern-Volmer plots ( Figure 2C  The bimolecular quenching constant of tryptophan α-amylase or α-glucosidase fluorescence by PE was determined from the following equation  The bimolecular quenching constant of tryptophan α-amylase or α-glucosidase fluorescence by PE was determined from the following equation F 0 and F are the fluorescence in the absence and presence of quencher, respectively, k q is the bimolecular quenching constant, τ 0 is the native lifetime of the fluorophore, and [Q] is the concentration of quencher. It is equal to 6 × 10 11 1/M × s.

Antiglycation Capacity of V. opulus Fruit Samples
The V. opulus fruit samples and reference substances (chlorogenic acid and aminoguanidine) were assessed in vitro for inhibitory effects on end glycation products (AGE) formation through fluorometric detection of fluorescent AGE. Inhibitory activity of extracts and fractions of V. opulus dried fruit had a direct linear relationship between concentration and percentage inhibitory activity of AGE formation (Supplementary Figure S4). The inhibitory effects of V. opulus fruit samples as well as chlorogenic acid and aminoguanidine (therapic agent for the inhibition of AGE formation) expressed as IC 50 value is summarized in Table 4. The antiglycation capacity of V. opulus fruit samples depended on both the type of sample and the type of sugar used. PEAF and PEWF with IC 50 value equal to 32.32 µg/mL and 33.96 µg/mL exhibited the strongest inhibition of AGE formation in the BSA-fructose in vitro model. In addition, PE, PEWF and PEAF demonstrated stronger antiglycation effect than aminoguanidine (IC 50 = 106.08 µg/mL) and chlorogenic acid (IC 50 = 58.87 µg/mL). Alternatively, testing the same group of samples for their antiglycation activity in the BSA-glucose model showed aminoguanidine as the strongest inhibitor (IC 50 = 137.80 µg/mL) followed by PEAF with IC 50 value equal to 577.50 µg/mL. Antiglycation effects of PE, PEWF and chlorogenic acid in the BSA-glucose assay did not differ significantly (p < 0.05)

Antioxidant Capacity of V. opulus Fruit Samples
Antioxidant properties of V. opulus samples were estimated by three different methods, as scavenging potential towards stable, synthetic ABTS •+ radical cation (ABTS) and towards peroxyl radical (ORAC), and as the potential to reduce ferric to ferrous ion (FRAP). Troloxa water-soluble analog of vitamin E-was used as an antioxidant standard to determine the Trolox Equivalent (TE). The results for antioxidant capacity of tested samples are presented in Table 5. Significant differences (p < 0.05) were found among the analyzed V. opulus samples in the antioxidant capacity. The TE values varied from 0.61 to 6.41 mM TE/g of extract or fraction in ABTS assay, from 0.40 to 5.18 mM/g in ORAC assay, and in the range 1.15-24.12 mM/g in FRAP method. The antioxidant capacity of V. opulus samples investigated was in the following order: PEAF > PEWF > PE > CE, regardless of the method used.

The Phenolic Compounds Content of V. opulus Fruit Samples
The results of quantitative analysis of V. opulus fruit extracts (CE, PE) and fractions (PEAF, PEWF) determined by UPLC method are presented in Tables 6 and 7. Figure 3 presents the chromatograms of V. opulus fruit samples at 280 nm. According to UPLC analysis, the CE, PE and PEAF showed the presence of four groups of phenolic compounds, such as flavanols, flavalignans, hydroxycinnamic acids, and flavonols. For comparison, no flavanols and flavalignans were found in PEWF fraction. On the basis of our results from the UPLC assay concerning low molecular weight phenolics, hydroxycinnamic acids dominated in all samples, while flavonols were the lowest (Table 6). Hydroxycinnamic acids accounted for approximately 84% of the total phenolics in CE, PE and PEAF, and almost 98% of the phenolic compounds determined in PEWF. The purification process on the Sep-Pak C18 column was effective in removing CE non-phenolic compounds as PE was characterized by an approximately 5-fold increase in flavanols and hydroxycinnamic acids, a 6-fold increase in flavonols and proanthocyanidins, and a 7.5-fold increase in flavalignans. As a result of fractionation of the aqueous solution of PE with ethyl acetate, the acetate fraction (PEAF) was 4.7-fold and 4-fold richer in total phenolics and hydroxycinnamic acids, respectively, than the aqueous residue. Moreover, flavanols and flavalignans were only present in PEAF, with the recovery relative to PE being 28.2% and 72.9%, respectively.
In the tested samples, there were 24, 23, 18, and 14 phenolic compounds detected in PE, CE, PEAF and PEWF, respectively (Table 7 and Figure 3). UPLC results revealed that the contents of selected phenolic compounds in different V. opulus samples showed significant differences (p < 0.05). As the main phenolic compound in all samples, the chlorogenic acid amount in CE was equal to 65.62 mg/g, whereas in PE, PEAF and PEWF it reached 300.23 mg/g, 795.72 mg/g and 152.05 mg/g, respectively. Quantitatively, (epi)-catechin hexoside was the second-most prominent component in all samples with the exception of PEWF in which dicaffeoylquinic acid derivatives were found next after chlorogenic acid. the contents of selected phenolic compounds in different V. opulus samples showed significant differences (p < 0.05). As the main phenolic compound in all samples, the chlorogenic acid amount in CE was equal to 65.62 mg/g, whereas in PE, PEAF and PEWF it reached 300.23 mg/g, 795.72 mg/g and 152.05 mg/g, respectively. Quantitatively, (epi)-catechin hexoside was the second-most prominent component in all samples with the exception of PEWF in which dicaffeoylquinic acid derivatives were found next after chlorogenic acid.  Table 7 for the identification of each numbered peak. CE-crude extract, PE-purified extract, PEAF-ethyl acetate fraction of PE, PEWF-water fraction of PE. Quercetin 3-rhamnoside 0.11 ± 0.00a 1.05 ± 0.03c 4.87 ± 0.01d 0.25 ± 0.02b "-"-not identified. The content expressed as equivalents of: a -procyanidin C1, b -(+)-catechin, c -cinchonine, d -chlorogenic acid, e -quercetin 3-glucoside. The means within a same raw with different letters are significantly different.

Discussion
In Poland, V. opulus L. is mainly cultivated as an ornamental shrub and its red and shiny fruits are rarely used as food. However, in Turkey, Russia and Ukraine, these fruits are used in traditional cuisine. V. opulus fruits are widely used in natural medicine for the relief of asthma, cold, fever, cough, nervousness, rheumatoid diseases, menstrual cramps, uterine and urinary tract infections, and to treat high blood pressure [36]. Our previous studies conducted both in cell-based and cell-free models, revealed that the compounds included in the V. opulus fresh fruits exhibited antidiabetic properties [33][34][35]. Contrary to the previous studies, we chose dried fruits as the research material due to their commercial availability. The present research is in line with the trend of searching safe and effective antidiabetic agents from natural sources. This is the first study of the effect of V. opulus dried fruit extracts on both the activity of carbohydrate digestive enzymes and the process of AGE formation. Inhibition of αglucosidase and α-amylase can prevent the postprandial surge of glucose and help to maintain its safe level in blood which is one of the primary targets for the treatment and management of the early stage of T2DM [47]. Additionally, inhibition of AGE formation is another approach in delaying or preventing the onset of diabetic complications [48]. Many studies conducted on fruits indicated that phenolic compounds were the functional components that played a major role in their hypoglycemic effect [7][8][9]12,27,31,[49][50][51][52]. Thus, in order to verify this hypothesis in our research, the V. opulus fruit samples varying in terms of degree of purification of phenolic compounds and thus in the presence of other non-phenolic components were used. In this study four samples were obtained from V. opulus dried fruits. CE was prepared using 70% acetone as a solvent, PE was obtained from CE by solid phase extraction method as a purification step, and next PE was fractionated into ethyl acetate and water phase (PEAF and PEWF). Two analyzed V. opulus fruit extracts were characterized by different content of phenolic compounds, from 87.42 mg/g of CE to 418.39 mg/g of PE (Table 6). Moreover, as a result of PE fractionation by the liquid-liquid extraction method, PEAF (the ethyl acetate fraction) contained as much as 977.08 mg of total phenolics per gram and PEWF (the water fraction) with 206.87 mg/g of fraction.
In addition, PE was the most abundant in proanthocyanidins, almost 95% converted to PEWF, which may indicate the presence of proanthocyanidins with a high degree of polymerization. According to Saucier et al. [53] the proanthocyanidin oligomers are soluble in ethyl acetate while the polymers remain in the aqueous phase. The presence of tannins in the form of polymers was also confirmed by the absence of peaks corresponding to the flavanols in the UPLC chromatograms of PEAF ( Figure 3 and Table 7). According to Perova et al. [54], proanthocyanidins are a quantitatively significant component of the fresh V. opulus fruits with content 201-528 mg/100 g fresh weight and they accounted over 49.9% of total phenolics. However, UPLC results showed hydroxycinnamic acids dominated in all V. opulus samples tested. This group of simple phenolic compounds constituted over 84% of the total phenolics in the analyzed samples, with the highest participation (over 97%) in PEWF. Chlorogenic acid was identified as the major phenolic compound in all V. opulus fruit samples with amount 65.62 mg/g of CE, whereas in PE, PEAF and PEWF it reached 300.23 mg/g, 795.72 mg/g, and 152.05 mg/g, respectively. Chlorogenic acid has also been indicated as the main phenolic compound of V. opulus fruit by other authors [46,[55][56][57].
According to our results, the composition of phenolic compounds of V. opulus fruit samples had a significant impact on both the activity of α-amylase and α-glucosidase, the process of AGE formation, as well as the antioxidant activity of the analyzed V. opulus fruit samples. In this study, the porcine pancreatic α-amylase and rat intestinal α-glucosidase models have been used because these enzymes play a crucial role in starch breakdown during digestion and glucose uptake. Pancreatic α-amylase is an endoglucosidase presents in pancreatic juice secreted into the intestinal lumen [7] and it digests the starch to maltose, maltotriose and dextrins, which are then broken down into glucose by other carbohydrases, e.g., α-glucosidase [19]. α-Glucosidase is a specific membrane-bound enzyme in the small intestine catalyzing the hydrolytic cleavage of disaccharides into monosaccharides [58]. In our study, crude extract (CE) showed the lowest inhibitory potential against carbohydrate digestive enzymes, regardless of both enzyme and substrate types ( Table 1). Removal of non-phenolic compounds from CE by the solid phase extraction method increased the inhibitory activity of PE on average 5-fold in the case of α-amylase and 9-fold in the case of α-glucosidase. The most active inhibitor of α-amylase was PEWF while PEAF exhibited the highest inhibitory activity against α-glucosidase. The reference of the obtained results to the phenolic composition of these fractions suggests that proanthocyanidin polymers, which remain in the aqueous phase, are mainly responsible for the anti-amylase activity of V. opulus fruits, and hydroxycinnamic acids, as well as flavanol oligomers and flavalignans for α-glucosidase inhibition. PEWF contained 2.6-fold more proanthocyanidins, but 4-fold less hydroxycinnamic acids and 1.7-fold less flavonols than PEAF (Table 6). Proanthocyanidins from grape seed and skin, acacia bark, persimmon peel and leaf, apple, and almond peel skin have been previously reported to inhibit α-amylase [59]. Additionally, proanthocyanidins with a high degree of polymerization compared to those with low degree had a greater effect on α-amylase activity in mice [60]. The higher activity of PEWF than PEAF on α-amylase may be due not only to the presence of proanthocyanidins but also to dicaffeoylquinic acids which were not generally present in PEAF. Narita and Inouye [61] showed that the decreasing rank of hydroxycinnamic acid derivatives on porcine pancreas α-amylase isozyme I was as follows: dicaffeoylquinic acids > caffeoylquinic acids > feruoylquinic acids. Our results demonstrated a very weak inhibitory activity of chlorogenic acid towards α-amylase (the highest IC 50 values, on average 2.7-fold and 14-fold higher than the IC 50 values for CE and PE, respectively).
The dominant role of hydroxycinnamic acids, including chlorogenic acid in inhibition of α-glucosidase is confirmed by similar IC 50 values for chlorogenic acid and PEAF-the most active V. opulus sample (Table 1). Oki et al. [62] demonstrated that the activity of rat α-glucosidase depends on disaccharide type. Thus, both maltose (a breakdown product of starch) and sucrose (the major disaccharide in food) were used as substrates for rat intestine α-glucosidase in the test systems used. The IC 50 values of all V. opulus fruit samples and acarbose (a well-known and standard antidiabetic drug) were lower in the maltose-containing system. These differences are due to the different affinity of the enzyme for the substrate as demonstrated by Oki et al. [62] because maltose had near 8-fold higher affinity for rat α-glucosidase than sucrose (Michaelis constants were equal to 3.0 and 22.8 mM, respectively). On the other hand, the inhibitory activity of chlorogenic acid (the dominant phenolic compound in V. opulus samples) was comparable in both assay systems. Study of Iwai and co-workers [31] confirmed inhibitory activity of chlorogenic acid against sucrase and maltase from rat small intestine with IC 50 equal to 0.50 and 8.81 mg/mL, respectively. This study also showed weaker inhibitory activity of this phenolic acid on αamylase porcine pancreas (IC 50 = 13.17 mg/mL) which is in line with our results. However, Pollini et al. [26] demonstrated that chlorogenic acid had higher inhibitory activity against α-amylase than other phenolics acids (e.g., caffeic acid, p-coumaric acid, sinapic acid, and syringic acid) and 5-fold lower activity compared to acarbose. In addition, the inhibitory activities of the V. opulus fruit samples against both α-amylase and α-glucosidase were lower than the acarbose, probably due to the complexity of the chemical composition of fruit matrices, and weak inhibitory activity of chlorogenic acid-quantitatively the main component of the tested samples.
The evaluation of the inhibition mode showed that PE and PEWF (the most active V. opulus samples against α-amylase) and acarbose inhibited the α-amylase uncompetitively (Supplementary Figure S2, Table 2). In this case, inhibitor binds only to the complex form between the enzyme and the substrate, i.e., potato or rice starch. On the other hand, PE and PEAF (the most active V. opulus samples against α-glucosidase) as well as acarbose displayed mixed mode of inhibition towards α-glucosidase (Supplementary Figure S3, Table 2). This means that the inhibitor may bind to the enzyme whether or not the enzyme has already bound the substrate, but has a greater affinity for one state or the other. Contrary, the Lineweaver-Burk plots showed that chlorogenic acid inhibits α-amylase noncompetitively and has a competitive mode towards α-glucosidase. Karim et al. [63] and Sun et al. [64] reported that chlorogenic acid has a mixed type inhibitory effect on porcine pancreatic α-amylase hydrolysis of potato and maize starch, respectively. In line with our results, a chlorogenic-rich fraction from Smilax aristolochiifolia root inhibited α-amylase activity in an uncompetitive manner and α-glucosidase in a noncompetitive mode [65]. To the best of our knowledge, the inhibitory mode of V. opulus fruit extracts and fractions are reported for the first time.
Interactions between both enzymes and V. opulus fruit components were also confirmed by fluorescence spectroscopy for PE. Under our measurement condition, PE reduced the fluorescence intensity of α-amylase and α-glucosidase, and the maximal absorption peak was slightly red-shifted. The same effects were found with the addition of chlorogenic acid and its derivatives to the α-glucosidase from Saccharomyces cerevisiae [66]. The authors suggested that these changes led to a polarity variation for tryptophan (Trp) residues in the enzyme and made the microenvironment change from hydrophilic to hydrophobic, which resulted in greater exposure of Trp residues and development of the protein structure. In the present study, the value of the bimolecular quenching constant is equal to 6 × 10 11 1/M × s, and is above the diffusion-limited rate constant of the biomolecule equal to 1 × 10 10 1/M × s [67]. It suggests the involvement of static mechanism in the fluorescence quenching of the α-amylase and α-glucosidase with PE. Thus, the binding affinity of PE to the studied enzymes may be estimated by the determined Stern-Volmer constants.
Chronic hyperglycemia promotes the formation of free radicals, that further increases the formation of reactive oxygen species, which in turn lead to a state of oxidative stress-a key factor in the development of diabetic complications [48]. There are increasing evidence that phenolic compounds due the antioxiadant properties may protect cell constituents against oxidative damage and, therefore, limit the risk of various degenerative diseases associated with oxidative stress [68,69]. In the current study, all V. opulus fruit samples demonstrated antioxidant activity, indicated by scavenging potential towards stable, synthetic ABTS •+ radical cation (ABTS) and towards peroxyl radical (ORAC), and as Fe 3+ reducing activity (FRAP). The antioxidant capacity of these samples followed this order: PEAF > PEWF > PE > CE, regardless of the method (Table 5). Chlorogenic acids, including mono-and dicaffeoylquinic acids as well as coumaroylquinic and feruoylquinic acid which were identified in V. opulus fruit samples by UPLC method (Table 7) are known to be potent antioxidants [70]. Dicaffeoylquinic acids possessed better antioxidant activities, mostly because they have more hydroxyl groups than caffeoylquinic acids. On the other hand, caffeoylquinic acid isomers showed quite similar antioxidant activities, indicating that the position of esterification on the quinic moiety of caffeoylquinic acid had no effect on its antioxidant activities.
At the early stage of T2DM, there is a relationship between hyperglycemia, increased oxidative stress, and excessive AGE formation [48]. The prevention of protein glycation is a viable strategy to prevent AGE formation since these molecules have ability to modify the chemical and functional properties of various biological structures [71]. All V. opulus fruit samples (extracts and fractions), as well as chlorogenic acid, showed inhibitory activity against AGE formation and had significantly (p < 0.05) lower IC 50 values than the positive control, aminoguanidine in the BSA-glucose model (Table 4). Other relationships were found in the BSA-fructose measurement system, where only crude extract (CE) of V. opulus fruit was a weaker inhibitor of AGE formation compared to aminoguanidine. The higher IC 50 values obtained with BSA-glucose compared to BSA-fructose are probably due to higher reactivity of ketose sugar than which has a more open-chain structure than aldose sugar [51]. Regardless of the test used, the ethyl acetate fraction (PEAF) showed the strongest inhibitory properties against the formation of fluorescent AGE. In studies by Justino et al. [52] ethyl acetate fraction from both ethanol and aqueous crude extracts of Eugenia dysenterica fruit pulp also showed the lowest IC 50 values in the BSA-fructose model. This fraction in both studies was characterized by the highest concentrations of phenolic compounds. The ethyl acetate fraction of V. opulus fruit contained different groups of phenolics with a predominance of chlorogenic acid and much less flavanols and flavalignans. This fraction of E. dysenterica fruit was composed of various phenolic compounds such as ferulic acid derivatives, flavonols and flavanols. These results may indicate that simple phenolic acids as well as more complex flavonoids can be able to inhibit the AGEs formation. In addition, some studies showed antiglycation activity of V. opulus fruit components such as 3,5-dicaffeoylquinic acid [72], chlorogenic and caffeic acid [73,74], catechins and procyanidins [75]. The antiglycation potential of V. opulus fruit had not been previously studied.

Conclusions
The potential ability of V. opulus dried fruit extracts, especially enriched with phenolic compounds to inhibit α-amylase and α-glucosidase, and AGE formation, and exhibition of antioxidative activities could be some of antidiabetic properties. The effects listed above are likely attributed to fruit phenolic compounds. The purification of crude extract resulted in extract enriched with phenolic compounds showing higher inhibitory activity against carbohydrate hydrolases and AGE formation, as well as antioxidant capacity. Further separation of the aqueous PE solution with ethyl acetate led to the obtaining of the acetate fraction rich in chlorogenic acid, proanthocyanidin oligomers and flavalignans with the highest inhibitory activity against α-glucosidase, formation of glycation products and antioxidant potential. The highest inhibitory activity of the water fraction of PE against α-amylase is related to proanthocyanidin polymers and chlorogenic acids, especially dicaffeoylquinic acids. This study suggested that the V. opulus fruits can serve as prospective source of natural phenolics, which could be developed as potential effective component of nutraceuticals and functional foods, especially after isolating a bioactive fraction, e.g., with ethyl acetate. The present results broaden knowledge regarding phytochemical and biological potential of non-conventional fruits, especially those related to the inhibition of AGE formation and enzymes activity that hydrolyze carbohydrates.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.