Hypoglycemic and Anti-Inflammatory Effects of Triterpene Glycoside Fractions from Aeculus hippocastanum Seeds

Horse chestnut (Aesculus hippocastanum L.)-derived drugs have shown their potential in biomedical applications. The seed of A. hippocastanum contains various kinds of chemical compounds including phenolics, flavonoids, coumarins, and triterpene saponins. Here, we investigated the chemical components in A. hippocastanum L. grown in Uzbekistan, which has not yet been studied in detail. We identified 30 kinds of triterpene saponins in an extract of A. hippocastanum L. Classifying extracted saponins into eight fractions, we next studied the hypoglycemic and the anti-inflammatory activities of escin and its derivatives through in vivo experiments. We came by data indicating the highest (SF-1 and SF-2) and the lowest (SF-5 and SF-8) antidiabetic and anti-inflammatory effects of those eight fractions. These results imply the prospective use of A. hippocastanum L. grown in Uzbekistan in the production of pharmaceutical drugs to treat diabetes and inflammation.

Escin belongs to the triterpenoid saponin or glycoside group and is classified as αand β-escin [21]. While the latter has been used as an antiseptic, antioxidant, analgesic, and antiaging compound [22], several experiments have pointed to new putative therapeutic activities of β-escin, namely, venotonic, anti-inflammatory [23,24], and antiedematous [16,[25][26][27][28][29]. Disruption of proteolytic enzymes by escin results in activation of leukocytes, which is followed by decreased permeation of capillaries and veins, increasing the blood vessel wall's tone [30]. The tension in veins accelerates the blood flow and, in turn, enhances microcirculation and delivery of oxygen to the tissues. Extensive in vivo and in vitro experiments, as well as a proposed mechanism of action of escins against inflammation and edema, can be found in the literature [25].
The genus Aesculus includes 12 species which belong to the Hippocastanaceae family. The abovementioned species are mainly found in Northern Hemisphere, specifically in territories of eastern Asia, southeastern Europe, and eastern America [3,31]. Moreover, horse chestnut is indigenous in the southern part of the Balkan peninsula, and the similar species are also widely cultivated in urban areas. The geographic location of Uzbekistan also offers a favorable habitat for horse chestnut. Chemical components in A. hippocastanum L. grown in Uzbekistan have not yet been studied in detail. Here, we present the isolation and characterization of derivatives of escin in A. hippocastanum L. native to the country. Our in vivo studies showed high hypoglycemic and anti-inflammatory activities of escin and its derivatives.

Identification of Chemical Structures of Saponin Derivatives by HPLC-MS
Saponins derived from the seeds of A. hippocastanum L. and their derivatives were separated using preparative high-performance liquid chromatography (HPLC) and detected with single-ion monitoring mass spectrometry (SIM-MS) in negative ionic mode ( Figure 1).  The TICs for SF-1, SF-2, SF-3, SF-4, SF-5, SF-6, SF-7, and SF-8 (1 for SF-1, 2 for SF-2, and so on). All chromatograms were obtained using a Shimadzu liquid system equipped with a single-quadrupole Shimadzu 2020 LC-MS system in negative electrospray ion mode with total diode array detection (DAD) at 226 nm.
The natural chestnut seeds grown in Uzbekistan were analyzed (see Figure 1B) and found to contain 4.2% ± 1.3% triterpene saponins (see Figure 1A,B). The escin saponins (ES) were isolated (see Figure 1C) after gradual purification of the alcohol extract of chestnut seed (see Section 3.3). According to previous studies, multiple peaks occurred on escin's chromatogram [19,32,33]. The main cause is that escins have more than 30 isomers and derivatives. Thus, even the chromatogram obtained by Kimura et al. using a standard sample of escin was shown to have a record of several peaks [34]. The peaks in the chromatogram of the sample seeds correspond to those of the standard sample of escin (see Figure 1A,B). Furthermore, analysis by ESI-MS showed that compounds with m/z [M − H] − of 1129, 1099, 1131, 1113, 1089, 1117, 1087, and 1057 were detected ( Figure 1C). In order to identify the chemical structure of escins, the obtained results were compared with the 1 H-NMR and 13 C-NMR data present in the literature [34].

Hypoglycemic Activity
In order to determine the nutrient activity of escin and its derivatives from the seeds of A. hippocastanum L., inhibition of the increase in blood glucose levels was tested by glucose tolerance tests in mice. The saponins were administered orally to two separate groups of mice at two different doses: 100 and 200 mg/kg. The 200 mg/kg dose of saponins was found to be effective in preventing an increase of blood glucose levels on mice after oral intake of glucose at intervals of 30 min and 1 h (see Table 2). Each sample was orally administered to mice 30 min before oral administration of D-glucose (0.5 g/kg). Values in parenthesis show the percentage difference in plasma glucose concentration between control and each sample treatment. Data represent the mean ± SEM (n = 5). * p < 0.05, ** p < 0.01.
The lowest hypoglycemic activity of SF-5 and SF-8 might be explained by the absence of the acetyl group in the structure of saponins present in these fractions. Saponins that do not contain the acetyl group in SF-8, when used in combination with those that contain acetyl group SF-1 and SF-2, resulted in an increase in hypoglycemic activity. This idea was also supported by a previous study on the inhibitory effect of escins isolated from Japanese fake chestnut seeds (A. hippocastanum L.) on elevated plasma glucose levels [34]. Thus, in hypoglycemia, the combined use of escins might be more effective. Certainly, the proposed hypothesis still requires more detailed research in order to establish clear conclusions.

Anti-Inflammatory Activity
Inflammation can be induced by applying various compounds such as lipopolysaccharides [36,37] and carrageenan [38][39][40]. We performed in vivo experiments using carrageenan as a mediator of inflammation. Furthermore, we investigated the anti-inflammatory activity of escin in acute inflammatory models. In rats, escin and its derivatives presented dose-dependent inhibitory effects in the early stages of development of edema induced by carrageenan. SF-1, SF-2 (see Figure 4A), and ED ( Figure 4E) showed significant inhibition by carrageenan in the second stage of edema formation.  D) and ES (E) we found a tendency toward inhibitory activity in the inflammatory models. From these results, we inferred that escin saponins isolated from the seeds of horse chesnut have an anti-inflammatory effect against inflammation caused by carrageenan. In addition, the presence of the acetyl moiety and the oligosaccharide probably plays an important role in anti-inflammatory activity. This can be seen in case of SF-5 and SF-8, which lack the acetyl moiety and showed reduced anti-inflammatory effects in our experiments.

Materials
A. hippocastanum L. seeds were collected in the autumn of 2020 in the Gazalkent district of the Tashkent region (Uzbekistan). The seeds were air-dried (residual moisture 4-6%) and stored at room temperature until used.

Preparation of Horse A. hippocastanum L. Seeds
The air-dried seeds of horse chestnut were peeled manually, and seed kernels were pulverized using a grinder to a size of 0.05-0.1 mm. Seed powder was defatted by hexane using a Soxhlet extractor. After removal of lipids, defatted seeds were dried under vacuum and kept at −20 • C before extraction.

Extraction, Fractionation, and Isolation of Saponins from Defatted A. hippocastanum L. Seeds
Preparation of plant extract. The preparation of plant extract was performed as described previously in the literature [41]. Briefly, the powdered seeds (2 kg) were extracted with 65% ethanol (EtOH) (2 L) by intermittent stirring at room temperature for 2 days, and the ethanol extract was filtered and the solvent was evaporated under reduced pressure below 40 • C. This process was repeated several times to remove the extractable components. The dry extract was dissolved in 200 mL of methanol. Next, this extract was transferred drop-by-drop to a flask of ice-cold diethyl ether stirring at a rate of 300 spin/min. The white precipitate was then filtered using a vacuum system through Whatman filter paper. The resulting precipitate was dried in a vacuum oven at 30-35 • C (yield: 46.15 g).
The ethanol extracts were applied to absorption column chromatography (500 mm × 60 mm i.d.) using Diaion HP-20 with 0.5 mm diameter particles (synthetic adsorbent). After washing the column with 3 L of distilled water to remove sugars, the fraction containing saponins (56 g of dried materials) was obtained by eluting with 3 L of ethanol.
To purify individual components of saponins from A. hippocastanum L. seeds, HPLC analysis was carried out on Shimadzu LC-2020 system equipped with a preparative HPLC column YMC-Pack ODS (150 mm × 10 mm i.d.). The preparative column was eluted at a flow rate of 3 mL/min with a mobile phase of methanol/10 mM sodium phosphate buffer (pH 7) (62:38, v/v). The elution of saponins was detected by monitoring the optical absorbance at 226 nm. For the analytical purposes, YMC-Pack ODS AM (150 mm × 6 mm i.d.) was eluted at a flow rate of 0.8 mL/min with the same mobile phase as above.

Animals
Sixty female and male Wistar rats (seven weeks old) and 55 female (4 weeks old) mice of the BALB/c strain were purchased from the Tashkent Pharmaceutical Institute. Animals (five per cage) were housed under standard conditions (light-dark cycle of 12 h, 20-22 • C room temperature, 50-60% humidity) and fed on a commercial diet and water ad libitum. Before the beginning of the experiment, animals were first acclimatized for 1 week to the lab conditions. All animals were used with the approval of the Ministry of Health of the Republic of Uzbekistan, Laboratory Animal Use Ethics Committee under the Veterinary Service. All procedures in this study were approved by the Tashkent Pharmaceutical Institute for animal experimentation and performed under isoflurane anesthesia.

Glucose Tolerance Test in Mice
After the mice were kept in food deprivation conditions for 16 h, the blood was withdrawn from the tail vein and subjected to the assay of blood level of glucose. Next, saponin fractions from natural seeds of chestnut were individually suspended in 0.3 mL of physiological saline. The resulting suspension was administered orally into the stomach of mice before a single oral injection of glucose (0.5 g/kg mouse) dissolved in 0.1 mL of physiological saline. Thereafter, the blood was withdrawn at 0.5, 1, and 2 h for analysis of blood glucose levels using GlucoDr Auto AGM-4000 (Gyeonggi-do, Republic of Korea) in compliance with the manufacturer's instructions. The elevation of blood glucose levels was calculated by subtracting the blood glucose levels prior to the administration of glucose from those after the administration of glucose.

Experimental Design of the Study
All mice were randomly divided into eleven experimental groups and were treated as follows: (i) Blank control group: injected with olive oil used as the vehicle (2 mL/kg); (ii) SF-1-, SF-2-, SF-3-, SF-4-, SF-5-, SF-6-, SF-7-, and SF-8-treated groups: injected by a single oral administration of glucose at dose of 0.5 g/kg (n = 5); (iii) Escin-treated groups: injected with two different doses (100 and 200 mg/kg) by a single oral administration of glucose at dose of 0.5 g/kg (per group n = 5).

Evaluation of Anti-Inflammatory Activity
The male Wistar rats weighing 140-160 g were used for in vivo experiments. Inflammation was induced by carrageenin in these rats. Accordingly, 0.1 mL of 1% carrageenin was injected subcutaneously into the left hind paw l h after the administration of test samples [23]. The volume of the hind paw was measured by a plethysmometer (Animal research plethysmometer 37140, UgoBasile, Italy). The results were expressed as swelling (%), denoting the percentage increase in hind paw volume as compared with the initial volume.

Carrageenan-Induced Acute Inflammatory Model
Anti-inflammatory activity was measured using the carrageenan-induced rat paw edema assay [41,42]. Edema was induced by subplantar injection of 100 µL of 1% freshly prepared solution of carrageenan in distilled water into the right hind paws of each rat of all the groups. Then, 30 min prior to carrageenan injection, paw thickness was determined before (0 h) and 1, 2, 3, 4, and 24 h after carrageenan injection. The increase in paw thickness was measured as the difference in paw thickness at 0 h and paw thickness at respective hours.

Statistical Analysis of Data
Data obtained from animal experiments were expressed as the mean ± standard error of mean (± SEM). Statistical differences between the treatments and control group were evaluated by one-way ANOVA. A difference in the mean values of p < 0.05 was considered to be significant.

Conclusions
We enhanced the extraction and isolation of saponins from the seeds of A. hippocastanum L. grown in Uzbekistan. Processing of the plant extract resulted in a pharmacological substance with more than 90% of its content being escin. An HPLC/MS analysis of the extracts showed that escin and its derivatives accounted for 4.2% ± 1.3% of the chemical components isolated from the seeds of A. hippocastanum L. Furthermore, we divided isolated saponin compounds into eight fractions on the basis of their molecular mass. and we investigated the antidiabetic and anti-inflammatory effects of the fractions in vivo. The results of the in vivo experiments indicated higher hypoglycemic and anti-inflammatory activity of the first and second fractions in comparison with the other six fractions. The absence of acetyl moiety and the oligosaccharide might have been responsible for the reduced biological activity of saponin fractions 3-8.
Escin saponins also show other potential pharmacological activities, for instance, against thromboembolism, viruses, and even cancer. Our future investigations will be focused on studying such new biological activities of escin saponins in a wider scope.

Conflicts of Interest:
The authors declare no conflict of interest.
Sample Availability: Samples of the compounds are available from the authors.