Lemongrass (Cymbopogon citratus) Ethanolic Extract Exhibited Activities That Inhibit -glucosidase Enzymes and Postprandial Blood Glucose Elevation

Lemongrass is a common ingredient in Indonesian traditional herbal medicine that potently inhibits carbohydrate hydrolysis. In this study, the in vitro α-glucosidase inhibitory (AGI) activity of lemongrass was compared with its in vivo activity to retard postprandial blood glucose elevation, and the bioactive compounds responsible for these activities were observed. Both water and ethanolic extracts of lemongrass (WLG and ELG, respectively) were tested in vitro for its inhibition of the sucrose and maltose hydrolyzing activities of rat intestinal glucosidase. ELG was observed to exert higher inhibitory activities (Sucrase IC50 = 8.74 mg/mL; Maltase IC50 = 18.93 mg/mL) than WLG. ELG was evaluated for its in vivo activity to retard blood glucose elevation in mice after sucrose, maltose, and glucose consumption. ELG was also fractionated using activity-guided chromatography, followed by liquid chromatography–mass spectrometry and high-performance liquid chromatography (HPLC). In vivo sugar tolerance test confirmed the AGI activity in a nondose-dependent manner and showed potential additional mechanisms that may prevent postprandial hyperglycemia. The active principles were acquired in methanol-soluble fraction and purified using preparative thin-layer chromatography. HPLC analysis with commercial standards identified caffeic acid and kaempferol as the compounds responsible for the bioactivity of ELG. Results showed that lemongrass has the potential as herbal medicine ingredient in the management of diabetes.


Introduction
The treatment of diabetes mellitus has been an increasing concern in Indonesia, as it became the fifth country with the highest prevalence of diabetes after China, India, Brazil, and the US [1]. The α-glucosidase inhibitor (AGI) inhibits the hydrolysis of polysaccharides and disaccharides in the small intestine, thereby delaying the spike of postprandial blood glucose after meals [2][3][4][5]. AGIs can be a reasonable option as the first-line drug in the treatment of patients with diabetes mellitus type 2 as it specifically targets postprandial hyperglycemia. AGIs are also expected to cause no hypoglycemic events or other life-threatening events, even at overdoses, and no weight gain [6].
Lemongrass (Cymbopogon citratus) is widely cultivated in tropical and subtropical countries. Lemongrass is a popular spice and flavoring agent in various traditional beverage and food recipes that are also known to have various health benefits [7,8]. Because of its strong roots that hold the soil, lemongrass plantations in Indonesia have also been known to prevent soil erosion in terrace farming. Commercialization of lemongrass as scientificbased herbal medicine, known as obat herbal terstandar (OHT) in Indonesia, is expected to increase both the value and the plantation of this plant [9]. The National Agency of Drug and Food Control of the Republic of Indonesia stated that herbal medicine registration as OHT must include the identification of active compounds as biomarkers, use of biomarker content for product quality control, and in vivo confirmatory tests of its activity and toxicity [10].
Ethnopharmacological and pharmacological investigations of lemongrass summarized the uses and potential functions of the essential oil and water extract of lemongrass (WLG) for various health benefits [8], in-December 2020  Vol. 24  No. 4 cluding the hypoglycemic effect of its fresh leaf aqueous extract [11] and in vitro antidiabetic activity of its essential oil [12]. Previous research on the exploration of Indonesian plant activities also reported the AGI activity of aqueous and methanolic extracts of lemongrass [13][14][15]. However, in vivo activity confirmation and evaluation of active compounds responsible for the AGI activity have not been conducted.
Common preparation methods for medicinal herbs in the traditional industry include water extraction and/or juice expression to produce ready-to-drink concoctions. Meanwhile, large herbal companies in Indonesia employ ethanolic extraction, followed by the drying process to produce an herbal powder with a long shelf life [16]. This indicates the need to evaluate the activities of both WLG and ethanolic extract of lemongrass (ELG). In this study, the in vitro AGI activities of both WLG and ELG were compared, and the extract with higher activity was further evaluated for its in vivo activity and subjected to a series of activity-guided fractionations to isolate and identify the active compound(s). Therefore, this study is expected to provide valuable insights into Indonesian herbal industries and the use of lemongrass as an herbal ingredient with AGI activity for the treatment of diabetes.
The materials used for in vivo evaluation include rat intestinal acetone powder (Sigma-Aldrich, Germany), sodium carboxymethyl cellulose (Brataco Chemika, Indonesia), blood glucose test strip (glucometer; GlucoDr® Auto, All Medicus Co., Ltd.), and mice feed (pellet BR2). All other materials were analytical grade and supplied by Merck, Germany. Adult male Swiss mice weighing 20-30 g were procured from the Imono Laboratory, Sanata Dharma University, Indonesia. The mice were housed in standard cages under standard laboratory conditions of 22 ± 2 °C and 12:12 h light/dark cycle. Standard pellet diet and water were given ad libitum. All procedures described were reviewed and approved with approval number KE/FK/1013/EC/2019 by the Medical and Health Research Ethics Committee, Faculty of Medicine, Gadjah Mada University-Dr. Sardjito General Hospital, Yogyakarta, Indonesia.

Lemongrass extraction.
Dried lemongrass was macerated using water and ethanol 96% for 24 h. The dried lemongrass-to-solvent ratio was 3:5 (v/v). Crude extracts were filtered using vacuum filtration and were evaporated using a vacuum rotary evaporator at 50°C. Then, both WLG and ELG were stored in a dark glass bottle at 4 °C.

In vitro rat intestinal glucosidase inhibitory activity.
Glucosidase inhibitory activity analysis was conducted using enzymatic reactions as previously described [17] with some modifications. To produce the crude rat intestinal glucosidase, 1 g rat intestinal acetone powder was crushed until homogenous and mixed with 20 mL of 0.1 M potassium phosphate buffer containing 5 mM EDTA in a cold mortar on an ice bath to prevent enzyme degradation. The mixture was placed in prechilled centrifuge tubes and centrifuged for 60 min (11,000 rpm, 4 °C). Any precipitation from centrifugation was solubilized and centrifuged for a second time in the same manner to increase efficiency.
The inhibitory activity against sucrose hydrolysis was measured using the following procedures: Two test tubes, as sample and control, containing 0.2 mL sucrose solution (56 mM) in potassium phosphate buffer (0.1 M, pH 7) and 2 test tubes, as sample and control blanks, containing 0.4 mL potassium phosphate buffer (0.1 M, pH 7) were preincubated at 37 °C for 5 min. The control and control blank were defined as 100% and 0% enzyme activity, respectively. Lemongrass extract diluted in 50% DMSO (0.1 mL) was added to the sample and sample blank test tubes; whereas 50% DMSO (0.1 mL) was added to the control and control blank test tubes. Then, crude rat intestinal glucosidase (0.2 mL) was added only to the test tubes containing the sucrose solution. The reaction was conducted at 37 ℃ for 15 min and terminated by adding Tris-HCl buffer (2 M, pH 6.3, 0.75 mL). Then, the reaction mixtures were passed through a short column of basic alumina (30 mm × 5 mm) to remove phenolics, which may interfere with the subsequent glucose quantification. The mixture (1)

Equation 1. Formula for AGI activity calculation,
where C is the absorbance value for the control, Cb is the absorbance value for the control blank, S is the absorbance value for the sample, and Sb is the absorbance value for the sample blank The procedures for inhibitory activity assay against maltose hydrolysis were the same as above, except for replacing sucrose solution (56 mM, 0.2 mL) with maltose solution (3.5 mM, 0.35 mL) and reducing the amount of enzyme solution from 0.2 mL to 0.05 mL. The experiments were conducted in two replications, and the results were presented as half-maximum inhibitory concentration (IC50). The slope test was employed to determine IC50. Analysis of variance (ANOVA) and Tukey's HSD were used (α value = 0.05) to determine the highest inhibitory activity between WLG and ELG. The extract with the highest AGI activity was selected for further tests.
In vivo sugar tolerance test. On the day of the experiment, 80 mice were fasted overnight and randomly divided into 16 groups, each consisting of 5 mice. One group was assigned as normal control (Group C), five groups (SC, SA, S1, S2, and S3) were assigned to the oral sucrose tolerance test (OSTT), five groups (MC, MA, M1, M2, and M3) were assigned to the oral maltose tolerance test (OMTT), and five groups (GC, GA, G1, G2, and G3) were assigned to the oral glucose tolerance test (OGTT).
Groups SC, MC, and GC served as sucrose control, whereas Groups SA, MA, and GA received 0.08 g acarbose/kg BW treatment. In OSTT, Groups S1, S2, and S3 received 0.22, 0.67, and 2.00 g ELG/kg BW treatments, respectively. Groups SC, SA, S1, S2, and S3 received 4 g sucrose/kg BW sugar induction (OSTT) 30 min after ELG or acarbose treatment. OMTT and OGTT were conducted in the same manner as above, except for replacing 4 g sucrose/kg BW with 3 g maltose/kg BW for OMTT and 2 g glucose/kg BW for OGTT (Table 1). Blood was collected from the tail vein of the mice at 0 min before sugar induction and at 15, 30, 60, 90, and 120 min after sugar induction. Postprandial glucose (PPG) levels were determined using a glucometer [18,19].
The results are expressed as the mean ± standard deviation. The trapezoidal rule was used to determine the area under the curve (AUC) blood glucose [20,21,22]. Data were analyzed using one-way ANOVA, followed by post hoc Scheffe tests using SPSS 22. A p value of <0.05 was considered statistically significant. AUC was estimated using the following formula for the trapezoidal rule (Equation 2): (2)

Results and Discussion
In vitro evaluation of -glucosidase inhibitory activity.
AGIs are expected to effectively inhibit postprandial hyperglycemia by suppressing carbohydrate digestion ( Figure 2). In this study, AGI activity against the digestion of sucrose and maltose was observed. In conjunction with solvents commonly used in Indonesian herbal industries, lemongrass was extracted using ethanol 96% (ELG) and water (WLG). The inhibitory activities of both extracts against the hydrolysis of sucrose and maltose by rat intestinal α-glucosidase were evaluated. The activities were expressed as IC50 to show the concentration required to inhibit 50% of enzymatic activity. ELG showed higher inhibitory activities (Sucrase IC50 = 8.74 mg/mL; Maltase IC50 = 18.93 mg/mL) than WLG (Sucrase IC50 = 132.89 mg/mL; Maltase IC50 = 302.27 mg/mL). Both extracts tend to have high inhibitory activities against sucrose hydrolysis.
In a previous study, lemongrass had been extracted using methanol, water, chloroform, and acetone. Evaluation of AGI activities of the extracts showed that methanol extract has the highest activity [15]. Due to the proximity of methanol and ethanol polarities, the previous study was in agreement with the better activity of ethanolic extraction in this study. This study confirmed the lemongrass activities to inhibit the hydrolysis of sucrose and maltose by intestinal mammalian glucosidase enzyme with higher activity extracted in ethanol as solvent. Due to its higher AGI activity, ELG was selected for further tests.

Figure 2. Schematic Mechanism of the Retardation of PPG Elevation Due to the Inhibition of the Hydrolysis of Disaccharides to Glucose by AGI
In vivo evaluation of antihyperglycemic activity. Sugar treatment increased the PPG level in mice 15 min after administration. The AUC of the PPG level 120 min after administration was compared between groups of mice to evaluate the significant changes caused by the treatments. Sugar treatments (SC, MC, and GC) significantly increased the total PPG level in comparison to the normal control group with no sugar treatment (C). Before glucose was transported to the cells and stored as a source of energy, glucose existed in the bloodstream [23]. Acarbose is a common antihyperglycemic agent that is used as a positive control in this study. Groups of mice treated with acarbose (SA, MA, and GA) before sugar administration have significantly lower total PPG levels in comparison to those that are only receiving sugar treatments (SC, MC, and GC), although the values are still higher than C. Acarbose was shown to suppress and decrease the spike of blood glucose as it inhibits the breakdown of carbohydrate [24] and the absorption of monosaccharides [25].
In OSTT, the total PPG elevation of mice treated with acarbose and ELG (SA, S1, S2, and S3) was significantly lower than that of mice in the sucrose control group (SC) (Figure 3), confirming its antihyperglycemic activity that inhibits the breakdown of carbohydrate [26]. At the concentrations used in this study (i.e., 0.22, 0.67, and 2.00 g/kg BW), ELG activity was significantly lower than acarbose activity and did not show a dose-dependent manner. Therefore, the optimum effective dose in this study was 0.22 g/kg BW ELG (S1) treatment as it has a significantly lower total PPG level compared with SC and a similar effect to higher ELG doses (S2 and S3).
The PPG levels of SC continued to increase up to 15.7 ± 3.1 mmol/L 30 min after sucrose administration. The PPG levels of SA, S1, S2, and S3 also increased but only up to 15 min after sucrose administration and returned to the basal level faster than SC. These results indicated that ELG might also facilitate glucose clearance from the blood vessel, adding another mechanism for its postprandial antihyperglycemic activity [25].
In OMTT, ELG treatments at the concentrations used in this study (M1, M2, and M3) were unable to reduce the total PPG elevation in comparison to those untreated maltose control (MC) (Figure 4). This result was consistent with the result of the in vitro evaluation, which showed that ELG has lower inhibitory activity against maltose hydrolysis than sucrose hydrolysis. Although ELG activity to inhibit maltose hydrolysis was not shown as the AUC of the total PPG elevation, M2 and M3 have significantly lower maximum PPG levels (13.85 ± 2.3 and 15.46 ± 1.9 mmol/L, respectively) than MC (17.4 ± 1.4 mmol/L) 15 min after maltose administration. Groups of mice treated with a low dose of ELG (M1) have a similar maximum PPG level (17.5 ± 2.9 mmol/L) to MC. However, M1 was able to delay the increase to 30 min after maltose administration in comparison to MC (15 min after maltose administration). ELG inhibits maltose digestion and retards the absorption of glucose, thereby preventing PPG spikes that are often associated with atherosclerosis [27] and other complications [4]. Although M1 was unable to reduce the total PPG elevation, the PPG level of mice treated with 0.67 and 2.00 g/kg BW (M2 and M3) returned to the basal state faster than MC, indicating its mild activity in facilitating glucose clearance from the blood vessel [25].
All treatments were done orally. The results are presented as the average ± SD; a: p < 0.05 vs normal; b: p < 0.05 vs sucrose; c: p < 0.05 vs acarbose.   ELG activities that prevent PPG spikes after sucrose and maltose consumption and total PPG elevation after sucrose consumption confirmed its in vivo AGI activity that inhibits digestive enzymes from breaking down disaccharides. The in vivo result was consistent with the in vitro result, indicating the stronger activity of ELG to inhibit sucrose hydrolysis than maltose hydrolysis ( Figure 3). The non-dose-dependent manner of the activity showed that ELG activity is also correlated with glucose counterregulatory responses to hypoglycemia [26,28]. The fast PPG level reduction to its basal state indicated that ELG might also have cell glucose uptake activity that is also preferable for the treatment of hyperglycemia [29][30]. Therefore, OGTT was further conducted to confirm this activity.
In OGTT, ELG at doses of 0.22 and 2.00 mg/kg BW (G1 and G3) were surprisingly able to reduce the total PPG elevation in comparison to glucose control (GC) with a similar level to acarbose (GA), whereas ELG at a dose of 0.67 mg/kg BW (G2) did not ( Figure 5). Again, the ELG activity did not show a dose-dependent manner, and the 0.22 mg/kg dose was once again shown to be a better treatment option because of its similar or even stronger effect on PPG level reduction compared with higher doses. Glucose counterregulatory responses to hypoglycemia at higher doses in healthy individuals might be responsible for the observed result in G2 mice [26,28]. Glucose, which is derived from the diet and synthesized by the body, needs a transporter, such as SGLT and GLUT, to be distributed to the bloodstream and cells [31]. Several mechanisms for the antihyperglycemic activity of lemongrass extract against glucose induction include the inhibition of the glucose transporter and the acceleration of the number of glucose transporters [32,33,34]. ELG activities that inhibit both PPG spike and total PPG elevation in OGTT showed that ELG not only inhibits disaccharides from hydrolyzing enzymes but also reduces the glucose transporter that transports glucose from the digestive tract to the blood vessel and/or increases cell glucose uptake that facilitates glucose clearance from the blood vessel.
The active fraction (ELG-Me) was subjected to PLC fractionation using ethyl acetate-methanol (4:1) as the mobile phase. Collected fractions were subjected to AGI activity assay. The active fractions with 42.39 ± 4.24% AGI activity (ELG-Me-V, Rf = 0.03-0.25, 890 mg) was further subjected to LC-MS evaluation, and the resulting compound with the mass spectrum of m/z 182.80 was identified as caffeic acid (C9H8O4) with the molecular mass of 180.16 and the mass spectrum of m/z 285.37 was identified as kaempferol (C15H10O6) with the molecular mass of 286.23 ( Figure 6). The RP-PLC purified fraction (ELG-Me-V-3) was further compared with the standard in HPLC to confirm the isolation of caffeic acid.
Kaempferol and caffeic acid were previously identified to be present in lemongrass [8]. Moreover, research conducted by Zang et al. in 2011 and Chandramohan et al. in 2015 showed that kaempferol exhibited an antidiabetic effect by increasing the insulin level and decreasing the plasma glucose of diabetic rats [35,36]. Another research showed that caffeic acid was also able to lower the blood glucose of diabetic rats [37]. The finding of this study indicated the potential use of ELG in the management of hyperglycemia.

Conclusion
The AGI principles in lemongrass were better extracted in the ethanolic solvent. ELG showed a non-dosedependent manner of inhibition of PPG elevation following all sugar challenges tested in this study, namely, sucrose, maltose, and glucose, with the lowest dose of 0.22 g/kg BW treatment showing the most optimum result. Further fractionation and identification of the bioactive compounds resulted in the identification of caffeic acid and kaempferol as the α-glucosidase inhibiting principles of ELG. This finding suggests the utilization of lemongrass in traditional and complementary medicine in the management of diabetes.