In vitro and in vivo antioxidant and antihyperglycemic potentials of phenolic fractions of Syzygium zeylanicum (L.) DC trunk‐bark

Abstract Syzygium zeylanicum L. (DC) (SZL) has been used in antidiabetes treatment for ages. However, the scientific evidence of active agents that have antidiabetic activity and response against biological activities is limited. In this study, the active components of SZL trunk‐bark extract (SZL extract) were identified using principal component analysis (PCA), and their antidiabetic activities were assessed. The results indicated that the ethyl acetate fraction (EAF) had the highest concentration of phenolic compounds, antioxidants, and antihyperglycemic activities in the postprandial zebrafish model. The major antioxidant contributors were gallic acid, catechin, epicatechin, ellagic acid, quercetin, caffeine, and apigenin, and their concentrated levels reduced α‐amylase inhibitory activity, whereas rutin and ethyl gallate influenced the α‐glucosidase inhibitory activity. This study showed the bio‐functional properties of active phenolic compounds present in the SZL extract, potentially serving as a functional food to control hyperglycemia.

as natural antioxidants, which provide effective therapy for diabetes treatment by influencing human physiological function (Shori, 2015).
Syzygium zeylanicum (L.) DC. (SZL) is a valuable medicinal plant found in secondary populations across Vietnam, from Kon Tum to the South. According to recent reports, SZL has the potential to be used in diabetic treatment due to its ability to resist oxidation and inhibit starch hydrolyzing enzymes. SZL leaf extract was discovered as a novel promising material that has proven outstanding performance in terms of α-glucosidase inhibition and antioxidant activities (Mai et al., 2007).
Our preliminary findings showed that the methanolic extract of SZL trunk-bark had greater α-glucosidase inhibitory activity than acarbose and effectively decreased blood glucose levels in streptozotocininduced diabetic rat models (Nguyen et al., 2019). However, the active components and their characterization are still unknown. Moreover, several recent investigations discovered that the fractionation approach profoundly influenced the bioactive component contents and their related biological activities (Franco et al., 2020;Nguyen Quang et al., 2021;Yadav et al., 2011). However, the influence of various solvent fractionation on the bioactive substances from SZL and their related bioactivities has not yet been investigated. Therefore, this study aimed to evaluate the antioxidant and antidiabetic activities of the fractionated extracts of SZL. In addition, the correlation between the phenolic compounds as the main components present in the fractions of the extract and their biological activities was identified by PCA.
Ltd. supplied ethanol, methanol, and acetic acid. All chemicals and reagents were of analytical grade.

| Plant sample
The trunk-bark of SZL collected from Easo Nature Reserve, Ea Kar District, Dak Lak Province, Vietnam was dried at room temperature in a well-ventilated environment, powdered, sealed in PE bags, and stored at −30°C until further use.

| Preparation of crude extract and fractions
The crude extract (CE) was prepared by immersion method as reported by Nguyen, Van Chuyen, et al. (2022), with slight modifications for using ultrasound-assisted extraction. In brief, the powdered trunk-bark of SZL was steeped in 50% aqueous ethanol solution at a ratio of 1:10 w:v for 2 min before being sonicated for 15 min at 28 kHz using a micro-tipped probe 10 mm in diameter (Vietsonic, VS28H). The mixture was entirely extracted for 24 h at room temperature before being filtered using filter paper (No.1). The residues were extracted a second time using the same procedure. The extracts were then combined and concentrated in a rotary evaporator (IKA) at 60°C until half of the volume was recorded. The extract solution was further fractionated with different polar solvents using the procedure described by Nguyen Quang et al. (2021). The crude extract (CE), n-hexane (HF), chloroform (CF), ethyl acetate (EAF), n-butanol (BF), and water fractions (WF) were fully dried by an Operon freeze dryer (Operon FBD-5503, Korea) at -56°C and 0.001 Mbar pressure and stored at -30°C until further use.

| Determination of the total polyphenol and flavonoid contents
The total polyphenol and flavonoid contents (TPC and TFC) were estimated using methods described in a previous publication (Nguyen et al., 2016).

| Determination of antioxidant capacity
The capacity of DPPH • and ABTS •+ radical scavenging was evaluated using the approach outlined in a previous study (Nguyen, Huyen, et al., 2022).

| Determination of the capacity to inhibit αamylase and α-glucosidase activities
The inhibition of α-amylase and α-glucosidase activities was measured using the methods reported in a previous publication (Nguyen, Huyen, et al., 2022).  (Westerfield, 2000). The dechlorinated tap water was filtered through a filtration system and the Ultraviolet filter was used to minimize contaminants. The culture medium was kept at a pH of 6.8-7.0, a temperature of 28 ± 1°C, a light cycle of 14 h (light)/10 h (dark), and adequate oxygen was supplied by an air pump. The fish was placed at the rate of 15 adults in each 4 L tank (Westerfield, 2000).

| Postprandial blood glucose level assay
Adult zebrafish (300-400 mg BW/fish) fasted for 12 h before the experiment. The fish after the fasting period was put in freezing water at 4°C to induce hypothermia. The anesthetic stage was detected when the fish lost balance, and operculum movements without response. The anesthetic fishes were then force-fed following the standard method with slight modifications (Collymore et al., 2013).
The fishes were divided into six groups (10 fishes/group) and treated with maltodextrin (MD), acarbose, SZL extract, and its fractions, as described below: Group 1: Maltodextrin was administered at 1 mg/g BW (negative control).
After feeding for 30, 60, 120, and 180 min, the blood was collected by the standard method (Zang et al., 2015). Postprandial blood glucose level was measured using a glucometer (Multicare-In, Biochemical Systems International).

| Ultra-performance liquid chromatography (UPLC) analysis
The phenolic constituents in the extract and its fractions were figured out using the Thermo-Ultimate 3000UPLC system (Thermo Scientific) with the gradient method as described by Nguyen, Huyen, et al. (2022).

| Data analysis
The mean values of each treatment were compared using an ANOVA test followed by a post hoc test (STATGRAPHICS Centurion XV). The LSD test identified significant variations in means (p < .05). The results were reported as the mean ± standard deviation (SD). The PCA was performed using R software version 4.1.2 (R Core Team, 2021) with the package FactoMineR (Lê et al., 2008).

| The total polyphenol and flavonoid contents
Polyphenol and flavonoids were found to be a rich source of Syzygium sp. which exhibits several valuable bioactivities such as anti-inflammatory, cardiometabolic, antioxidant, and antidiabetic activities (Adefegha & Oboh, 2012;Chagas et al., 2018;Gajera et al., 2017;Priya et al., 2017;Thomas et al., 2022). In our current work, the n-hexane and chloroform fractions were collected in extremely insignificant amounts that were not enough quantity for further characterization and bioactivity tests. As an observation, SZL extract contains high quantities of polyphenols and flavonoids. The total polyphenol and flavonoid contents of CE were 531.64 mg GAE/g and 473.44 mg QE/g dry weight (DW), respectively (Table 1). These results were in good agreement with previous studies on Syzygium sp. (Priya et al., 2017;Ruan et al., 2008).
According to Mai et al. (2007), the total polyphenol content in SZL leaf is the highest among 11 species of culinary plants that grow naturally in Vietnam, with values of 176.9 mg/g DW and 251.7 mg/g DW for crude and methanolic extract, respectively. Among all fractions, the EAF had the highest total polyphenol and flavonoid contents, followed by the BF, while the WF had the lowest contents. In comparison with WF, the EAF and BF showed a TA B L E 1 Total polyphenol, flavonoid contents, antioxidant activity, starch-hydrolyzing enzyme inhibitory activity of SZL extract and its fractions. superior performance to extract polyphenols and flavonoids from the CE (Table 1). These results showed that the polyphenols and flavonoids in the trunk-bark of SZL have similar polarity to ethyl acetate and n-butanol, allowing them to be fractionated readily using these solvents.

| DPPH • radical scavenging activity
The DPPH • radical scavenging activity results were expressed in gram Trolox equivalent (TE) per gram DW of extract/fraction which was observed to be significantly different amongst extract and fractions (p < .05; Table 1

| ABTS •+ radical scavenging activity
Like DPPH • radical scavenging activity, EAF was determined to be relatively high in the ABTS •+ radical scavenging activity compared to all other fractions. The activity of this fraction was found to be greatest at 3.27 g TE/g DW, while the BF and WF showed a value of 1.4 g TE/g DW and 1.7 g TE/g DW, respectively (Table 1).
Especially, all extracts and fractions exhibited greater ABTS •+ radical scavenging activity than that DPPH • radical scavenging activity. The findings demonstrate that the EAF obtained most of the antioxidant compounds, the electron transfer compounds being the most abundant. In addition, the WF showed stronger ABTS •+ radical scavenging activity than that of BF and CE, suggesting that highly polar compounds have significantly higher antioxidative potential than nonpolar compounds. In literature, oxidative stress and inflammation engage in the onset and evolution of diabetes, as well as their repercussions (Kanwugu et al., 2022).
Previous research highlighted the correlation between the antioxidant, anti-inflammation, and antidiabetic potentials of Syzygium sp. (Arumugam et al., 2014). Thus, further investigation should be conducted to figure out the correlation between antioxidant and antidiabetic SZL bioactive components.

| α-Amylase inhibitory activity
The IC 50 value of α-amylase inhibitory activity revealed a significant difference between the extract to the fractions (Table 1). CE inhibited α-amylase activity at the highest level, followed by WF (IC 50 : 4.19 and 4.52 μg/mL, respectively), and both had higher activity than BF and EAF. BF and acarbose exhibited no significant difference in activity (p > .05). Interestingly, EAF had the lowest activity despite having the greatest TPC, TFC, and antioxidant activity levels.
The α-amylase inhibitory activity of SZL extract and its fractions was markedly higher as compared to S. cumini leaves (IC 50 : 39.9 μg/ mL; Poongunran et al., 2017)

| α-Glucosidase inhibitory activity
The efficiency of α-glucosidase inhibition, as represented by IC 50 values, was shown to differ considerably between extract and frac-  (Mai et al., 2007). The methanol SZL extract was found to have the highest α-glucosidase inhibitory activity, with the IC 50 value of 109 μg/mL (Nguyen et al., 2019). The SZL extract and its fraction could be an excellent source of α-glucosidase inhibition activity as associated with the selection of an appropriate solvent. The most effective solvent and extraction processes for the active ingredients associated with α-glucosidase inhibitors should be explored further.
According to the literature, an effective therapeutic strategy for treating diabetes by inhibiting starch-hydrolyzing enzymes, such as α-amylase and α-glucosidase. When an inhibitor derived from medicinal plants was treated, it may reduce negative consequences by reducing aberrant fermentation or breakdown of polysaccharides in the colon due to α-amylase activity (Wang et al., 2012). Our findings in this study revealed that the EAF and BF of SZL extracts could be an alternative treatment method for type 2 diabetes due to their substantial inhibition of α-glucosidase and moderate inhibition of α-amylase. The PBGL of all zebrafish groups reached the highest peak after 60 min of feeding, and the strongest hypoglycemic activity of SZL extract and its fractions were also observed at this time. better after 180 min. These findings revealed that the mechanism of action may be distinct and that the fractionation procedure altered the quantities of bioactive components, influencing plasma blood glucose levels over time.

| Postprandial antihyperglycemic activity on zebrafish model
In the previous reports, we found that methanolic SZL extract had significant hypoglycemic activity in mice at a dosage ranging from 100 to 200 mg/kg BW (Nguyen et al., 2019). In this work, we firstly reported the postprandial glycemic reduction of SZL extract and its fractions in a zebrafish model with a lower dosage and longer lasting activity.

| UPLC analysis and phenolic compound properties
The most prevalent secondary metabolites found in plants are phenolic compounds which supplied a wide variety of bioactivity in general . Gallic acid, catechin, epicatechin, caffeine, ethyl gallate, rutin, ellagic acid, quercetin, quercitrin, apigenin, and chlorogenic acid were among 11 recognized phenolic compounds discovered in the SZL extract and its fractions. Their concentration altered from the extract to the fractions (  Table 2). The concentration of individual phenolic compounds and their ratio in the SZL extract and its fractions may influence their bioactivities. However, the primary compound that has the greatest impact on bioactivities must be investigated further.

F I G U R E 1
The effect of SZL extract, its fraction, and acarbose on blood glucose levels after oral maltodextrin loading in Zebrafish. SZL extract and its fraction (75 mg/kg BW), acarbose (75 mg/kg BW), and maltodextrin (1 mg/g BW) were given to zebrafish (n = 10) orally. Blood was drawn at 0, 30, 60, 120, and 180 min after feeding. Different labels (a-c) indicate a significant difference within the group of times (p < .05), Aca, acarbose; BF, nbutanol fraction; CE, crude extract; EAF, ethyl acetate fraction; MD, maltodextrin; WF, water fraction.

| Correlation analysis and the principal component analysis
Pearson's correlations (r) between TPC, TFC, DPPH • , and ABTS •+ radical scavenging activities, and phenolic components along with α-glucosidase and α-amylase inhibitory activities were observed and arranged from lowest to highest (Table 3) Table 3). These results suggested that the higher levels of ellagic acid, quercetin, and apigenin in CE may support their DPPH • radical scavenging activity better than that of BF, and WF, which had lower levels of ellagic acid, quercetin, and absence of apigenin (Table 2).
Similarly, the higher levels of catechin, epicatechin, and caffeine in WF may be responsible for its ABTS •+ radical scavenging activity better than that of CE and BF.
Interestingly, a strong positive correlation between phenolic compounds with the IC 50 value of α-amylase inhibitory activity was observed (Table 3). This situation implied that the increasing concentration of these compounds is suggested to reduce the αamylase inhibitory activity. In literature, the phenolic interaction with each other affects their α-amylase inhibitory activity due to decreasing their interaction with enzymes (Sultana et al., 2020).
A previous study reported that an elevated level of phenolic compounds in yellow raspberries did not increase α-amylase inhibition (Grussu et al., 2011). Thus, these findings suggested the behavior of α-amylase inhibitory activity of SZL extract and its fraction may affect by the interaction of specific types of phenolic compounds than their total concentration level.
In terms of α-glucosidase inhibition, only ethyl gallate and rutin showed a significant association with r = −.68 (p < .016) and r = −.79 (p < .0025), respectively, whereas other compounds showed an insignificant correlation (Table 3). Furthermore, ethyl gallate and rutin may differ from other phenolic compounds in their activity. They have a strong association with each other (r = .97, p < .001), but only a moderate link with the other compounds (excluding quercitrin in the case of ethyl gallate; Table 3). This discovery demonstrated that ethyl gallate and rutin may become intertwined during fractionation and that changes in concentrated levels over fractions may impact the α-glucosidase inhibitory activity.
The PCA of SZL extract and fractions is shown in Figure 2

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.