Investigation of the Safety and Antihyperglycemic Effect of Apios americana Flower Intake as a Food Material in Normal and Diabetic Mice

Abstract

In order to utilize Apios americana flower as a functional food material, its safety and antihyperglycemic effects were investigated in mice. Normal mice fed dried A. americana flower or its methanolic extract (AFE) for 4 weeks did not show abnormalities in growth or plasma biochemistry profiles. However, ingestion of AFE lowered plasma glucose levels in ICR and diabetic KK-A^y mice. AFS60, fractionated from AFE by Sephadex LH-20 column chromatography, inhibited maltase; and suppressed the elevation of plasma glucose levels by 15% compared with control after oral administration of maltose in KK-A^y mice. Caffeoyl β-d-glucopyranoside was identified as the strongest maltase inhibitor in AFS60. While further toxicological studies are required, these results suggest that A. americana flower has maltase inhibitory activity and antihyperglycemic effects, and is a safe and useful novel food material for the prevention of diabetes.

Introduction

The increased incidence of non-insulin dependent diabetes mellitus, a typical metabolic disease, is a serious problem in Japan. Accordingly, α-glucosidase inhibitors are often prescribed to control blood glucose levels in diabetic patients. To date, a number of natural compounds that act as α-glucosidase inhibitors have been found in food (Peyrieras et al., 1983; Patel and Mishra, 2012), and many researches have elucidated the preventive effects of food ingredients on hyperglycemia (Shipra et al., 2010; Oe and Ozaki, 2008). Therefore, the identification of natural α-glucosidase inhibitors in food materials is of value in the development of new functional foods for the prevention of diabetes development and to reduce its incidence.

Apios americana    Medikus is cultivated in northern Japan for its edible tuber. Several studies have focused on the cultivation and nutritional components of A. americana tuber (Hoshikawa and Juliarni, 1995; Wilson et al., 1986; Wilson et al., 1987). In our previous study, diet supplemented with dried A. americana tuber exhibited an antihypertensive effect in spontaneously hypertensive rats (Iwai and Matsue, 2007). Moreover, angiotensin-converting enzyme inhibitory peptides from the tuber hydrolyzate were identified as the active compound in the antihypertensive effect (Kuramoto et al., 2013). Meanwhile, to cultivate high-quality A. americana tubers, the red-purple butterfly-shaped flowers are removed and discarded. While it is anticipated that the flower can also be used as a functional food material, the toxicological and physiological effects of A. americana flower and its constituents have not yet been studied.

We employed a number of in vitro bioactivity screening assays, and potential α-glucosidase inhibitory activity was identified in A. americana flower extracts in the preliminary experiments. Therefore, in vivo and in vitro experiments were carried out to clarify the physiological and toxicological effects of A. americana flower. In this study, the safety and antihyperglycemic effects of A. americana flower were investigated in normal and type 2 diabetic model (KK - A^y) mice fed diets containing A. americana flower dried powder (AFDP) or its extract (AFE). The suppressive effect of the α-glucosidase inhibitory fraction isolated from AFE on postprandial plasma glucose levels was investigated in KK-A^y mice. Moreover, the component responsible for the α-glucosidase inhibitory activity of A. americana flower was investigated in vitro.

Materials and Methods

Reagents and Preparation of A. americana Flower    Rat small intestinal acetone powder and methanol-d₄ were purchased from Sigma Chemical Co. (St. Louis, MO, USA). All other reagents and solvents used in this study were of the highest available or high-performance liquid chromatography (HPLC) grade.

A. americana flowers were harvested in full bloom at an Apios farm in Gonohe-machi, Aomori, Japan, in August. After drying at 50 – 60°C in warm air, AFDP was prepared by grinding the dried petals using a Wonder Blender WB-1 (Osaka Chemical Co., Osaka, Japan). The yield of dried petals from the raw flower was 9.69%. For the preparation of AFE, 200 g of AFDP was homogenized in 400 mL of 80% methanol containing 1% formic acid using a SKS-A700 Mixer (Tiger Co., Osaka, Japan). After stirring the mixture at 4°C for 2 h, the supernatant was collected by centrifugation at 8000 rpm for 30 min; the precipitate was then suspended in 400 mL of 80% methanol containing 1% formic acid and stirred for 2 h. This procedure was repeated 3 times, and all supernatants were collected and concentrated using a rotary evaporator. After freeze-drying the supernatants, 42.8 g of AFE was obtained from 200 g of AFDP. Similarly, 60.3 g of an aqueous extract of A. americana flower was prepared with distilled water from 200 g of AFDP.

The AFE was applied onto a Sephadex LH-20 column (28 mm i.d. × 980 mm L; GE Healthcare Bio-Sciences AB, Uppsala, Sweden), and then subjected to stepwise elution with water and 20, 60, and 100% methanol containing 1% formic acid. These eluates were obtained as AFS0 (water), AFS20 (20% methanol), AFS60 (60% methanol), and AFS100 (100% methanol) fractions, and were evaporated by a rotary evaporator.

Animals    Male ICR and KK-A^y mice were purchased from Clea Japan, Inc (Tokyo, Japan), and they were used after acclimation for 1 week. The study was approved by the Ethics Committee of Aomori University of Health and Welfare, and all animal experiments were conducted according to the Guidelines for Animal Experiments of the Committee. Animals were individually housed in wire cages at 22 ± 2°C with 65% humidity and a 12-h light/dark cycle.

Dietary Experiment in Normal Mice    The composition of the experimental diets, which were based on the AIN-93 diet (Reeves et al., 1993), are shown in Table 1. The normal control (NC), low AFDP (AFDP-L), high AFDP (AFDP-H), and AFE (AFE) diets contained 0, 1, and 5% of AFDP, and 1% AFE respectively. The nutrient contents of the AFDP diets were adjusted in reference to the nutritional composition of AFDP as follows: 7.0% moisture, 16.1% crude protein, 7.4% crude fat, 37.1% carbohydrate, 27.2% fiber, and 5.2% ash. In the AFE diet, the addition of AFE was adjusted using corn starch content. Solid diets were prepared.

Table 1. Nutritional composition of the experimental diets

	Normal mice				Diabetic mice
	NC	AFDP-L	AFDP-H	AFE	DC	AFE-L	AFE-H
Casein	20.0	19.8	19.1	20.0	20.0	20.0	20.0
Corn starch	39.8	39.6	38.5	38.8	39.8	39.6	38.8
α-Corn starch	13.5	13.4	13.1	13.5	13.5	13.5	13.5
Sucrose	10.0	10.0	9.8	10.0	10.0	10.0	10.0
Soybean oil	7.0	6.9	6.6	7.0	7.0	7.0	7.0
Cellulose	5.0	4.7	3.5	5.0	5.0	5.0	5.0
Mineral mixture*	3.5	3.4	3.2	3.5	3.5	3.5	3.5
Vitamin mixture*	1.0	1.0	1.0	1.0	1.0	1.0	1.0
Choline bitartrate	0.2	0.2	0.2	0.2	0.2	0.2	0.2
AF	0	1.0	5.0	0	0	0	0
AFE	0	0	0	1.0	0	0.2	1.0

*  AIN-93G composition.

Seven-week-old male ICR mice were divided into 4 groups of 6 animals each and given free access to the experimental diets for 4 weeks. At the end of the experimental period, fasted mice for 20 h were sacrificed under ether anesthesia, and blood was collected from the abdominal artery. Then, the heart, lung, liver, stomach, pancreas, spleen, kidneys, adrenal glands, small intestine, cecum, large intestine, testes, and prostate were removed and weighed. The plasma was immediately separated by centrifugation at 3000 rpm for 15 min, and biochemical parameters such as glucose (GLU), total cholesterol (TCHO), triglyceride (TG), total protein (TP), aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), lactate dehydrogenase (LDH), blood urea nitrogen (BUN), uric acid (UA), and creatinine (CRE) were measured using a Fuji Dri-Chem 3500 (Fuji Medical System Co., Tokyo, Japan).

Dietary Experiment in Diabetic Mice    All diets were based on AIN-93 (Reeves et al., 1993), and the diabetic control (DC), low AFE (AFE-L), and high AFE (AFE-H) diets contained 0, 0.2, and 1% of AFE, respectively (Table 1). The nutritional content of the AFE diets was adjusted using corn starch content, and solid diets were prepared.

Seven-week-old male KK-A^y mice were divided into 3 groups of 6 animals each and given free access to the diets for 4 weeks. At 0, 2, and 4 weeks of the experiment, blood was collected from the tail artery of non-fasted mice using a heparinized capillary tube, and the plasma was immediately separated by centrifugation at 12000 rpm for 5 min. The plasma glucose concentration was measured using a Fuji Dri-Chem 3500 after dilution of the plasma with a 0.15 M NaCl solution. Plasma insulin concentrations of non-fasted mice were measured using a Mouse-Insulin ELISA kit (Mercodia AB, Uppsala, Sweden).

At the end of the experimental period, the fasted mice were sacrificed under ether anesthesia, and blood was collected from the abdominal artery. Next, the liver and kidneys were removed and weighed, and the plasma was immediately separated by centrifugation at 3000 rpm for 15 min. Plasma biochemical parameters such as GLU, TCHO, TG, TP, AST, ALT, BUN, UA, and CRE were measured using a Fuji Dri-Chem 3500.

Oral Maltose Tolerance Test    Eight-week-old male KK-A^y mice were fasted for 20 h before the tolerance test. Mice were divided into control (C), low dose AFS60 (AFS-L), and high dose AFS60 (AFS-H) groups of 6 mice each. AFS60 was suspended in a 0.5% carboxymethyl cellulose solution containing 100 mg/mL of maltose. The dose of maltose in all groups was 1 g/kg body weight, and the dose of AFS60 in the C, AFS-L, and AFS-H groups was 0, 60, and 180 mg/kg, respectively. After oral administration of the mixture of maltose and AFS60 by gavage, blood was collected from the tail artery at 0, 0.5, 1, and 2 h, and the plasma was immediately separated by centrifugation at 12000 rpm for 5 min. The plasma glucose concentration was measured using a Glucose C-II Test kit (Wako Pure Chemical Industries Ltd., Osaka, Japan) after dilution with 0.15 M NaCl solution.

Measurement of α-Glucosidase Inhibitory Activity    The α-glucosidase inhibitory activity was measured according to the previously reported method (Iwai, 2008) as follows. The crude enzyme solution was prepared from rat small intestinal acetone powder by homogenization in 56 mM maleate buffer (pH 6.0). Maltose, sucrose, soluble starch, and palatinose were used as substrates for the maltase, sucrase, glucoamylase, and isomaltase assays, respectively, and the substrates and samples were also dissolved in 56 mM maleate buffer (pH 6.0).

In brief, 20 µL of sample was added to 100 µL of the 20 mg/mL substrate solution and mixed. After incubation of the mixture at 37°C for 5 min, 20 µL of the enzyme solution was immediately added into the mixture and incubated at 37°C. At 0, 15, 30, 60, and 90 min of incubation, 15 µL of the reaction mixture was transferred into other tubes and immediately boiled for 10 min to stop the enzymatic reaction; the glucose concentration in the mixture was measured using a Glucose C-II Test kit. The enzymatic reactions were conducted in triplicate, and the kinetics were calculated from the production of glucose. The inhibition rate (%) of the sample was standardized at 100% of the kinetic responses without the sample.

Isolation and Identification of Maltase Inhibitory Component    The maltase inhibitory component was isolated and purified from AFS60 by preparative HPLC using an Alliance 2695 separation module with 2998 photodiode array detector (Waters Co., MA, USA). Acetonitrile and 10% formic acid solution were eluted through an Inertsil ODS-3 preparative column (20 mm i.d. × 300 mm L; GL Science Inc., Tokyo, Japan) by a linear gradient of 0 to 70% acetonitrile for 70 min at a flow rate of 5 mL/min. The eluate was monitored at a wavelength range of 200 to 800 nm, and the peaks were collected. In the purification step using the same HPLC, the ingredients were eluted with acetonitrile and 1% formic acid solution by a linear gradient of 0 to 45% acetonitrile for 180 min and 45 to 95% of acetonitrile for 210 min at a flow rate of 5 mL/min. The eluate was detected at a range of 200 to 800 nm, and the individual peaks were collected. Analytical HPLC using the same instruments was performed on an Inertsil ODS-3 analytical column (4.6 mm i.d. × 250 mm L; GL Science Inc.) with acetonitrile and 1% formic acid solution at a flow rate of 1.0 mL/min as follows: 0 to 45% of acetonitrile for 60 min and 45 to 90% of acetonitrile for 70 min, and the eluate was detected at a range of 200 to 800 nm.

For chemical structure analysis, a JNM-EX270 instrument (JEOL Ltd., Akishima, Japan) was used to determine the nuclear magnetic resonance (NMR) spectrum of the compound dissolved in 99.9% methanol-d₄, which contained tetramethylsilane as an internal standard.

Liquid chromatography-tandem mass spectrometry (LC/MS/MS) analysis of the compound was also performed on an API 3000 LC/MS/MS system using an electrospray ionization source (Applied Biosystems Japan, Tokyo, Japan) and 1100 HPLC system (Agilent Technologies Japan, Ltd., Hachioji, Japan). The mobile phase was eluted through an Inertsil ODS-3 analytical column (2.1 mm i.d. × 150 mm L; GL Science Inc.) at a flow rate of 0.2 mL/min under a linear gradient of 8 to 95% of acetonitrile in 1% formic acid solution for 33 min. The eluate was detected by absorption and ESI-MS in the positive ion mode under the following conditions: nebulizer gas, 4 units; curtain gas (N₂), 6 units; collision gas, 7 units; ion spray voltage, 5500 V; temperature, 300°C; declustering potential, 25 V; focusing potential, 200 V; collision cell rod offset, 20 V; collision cell exit potential, 20 V; and mass scanning, m/z 100 to 3000. In the MS/MS analyses, samples were ionized under the above conditions, and mass was measured under the following conditions: declustering potential, 20 V; entrance potential, 10 V; collision energy, 10 V; and collision cell exit potential, 15 V.

Statistical Analysis    Data were presented as the mean ± standard deviation (SD) and were analyzed using the Tukey-Kramer test for multiple comparisons after one-way analysis of variance using the StatView System (SAS Institute Inc., NC, USA). Differences were considered significant at P < 0.05.

Results and Discussion

Effects of Dietary A. americana Flower on Normal Mice    Table 2 shows body weights, body weight gain, dietary intake and tissue weights of normal ICR mice fed diets containing AFDP and AFE for 4 weeks. All groups showed good changes of body weight during the experimental period. After 4 weeks, the AFDP-H group tended to have the lowest body weight and body weight gain; however, there were no significant differences in the 4 groups, and the dietary intake of all groups was equal. At dissection, the fasted body weights of NC, AFDP-L, AFDP-H, and AFE were not significantly different at 36.12 ± 2.50 g, 37.82 ± 2.64 g, 35.84 ± 1.90 g, and 38.35 ± 1.66 g, respectively. There were also no significant differences in any tissue weights and the ratio of tissue weight to body weight in the 4 groups (data not shown). From these results, it was clear that the ingestion of A. americana flower for 4 weeks did not negatively affect growth parameters such as body weight, tissue weight and food consumption.

Table 2. Body weights, body weight gain, dietary intake and tissue weights of male ICR mice fed AFDP and AFE diets for 4 weeks

	NC	AFDP-L	AFDP-H	AFE
Initial body weight (g)	28.38 ± 0.61	28.31 ± 0.84	28.41 ± 1.05	28.83 ± 1.00
Final body weight (g)	39.55 ± 2.75	40.89 ± 2.13	38.61 ± 2.51	41.62 ± 0.85
Body weight gain (g)	11.17 ± 2.56	12.59 ± 2.12	10.19 ± 2.09	12.79 ± 1.03
Dietary intake (g/d)	5.32 ± 0.31	5.47 ± 0.13	5.33 ± 0.31	5.50 ± 0.13
Heart (g)	0.160 ± 0.025	0.157 ± 0.017	0.154 ± 0.007	0.168 ± 0.019
Lung (g)	0.196 ± 0.031	0.192 ± 0.020	0.179 ± 0.009	0.194 ± 0.014
Liver (g)	1.453 ± 0.151	1.388 ± 0.166	1.358 ± 0.078	1.422 ± 0.074
Stomach (g)	0.227 ± 0.018	0.229 ± 0.047	0.232 ± 0.019	0.240 ± 0.018
Pancreas (g)	0.215 ± 0.098	0.247 ± 0.071	0.245 ± 0.086	0.260 ± 0.061
Spleen (g)	0.075 ± 0.015	0.078 ± 0.020	0.080 ± 0.007	0.084 ± 0.017
Kidneys (g)	0.527 ± 0.049	0.543 ± 0.070	0.537 ± 0.044	0.544 ± 0.071
Adrenals (g)	0.006 ± 0.003	0.005 ± 0.002	0.007 ± 0.002	0.006 ± 0.001
Small intestine (g)	1.080 ± 0.066	1.134 ± 0.174	1.177 ± 0.160	1.137 ± 0.061
Cecum (g)	0.238 ± 0.048	0.267 ± 0.093	0.200 ± 0.059	0.234 ± 0.057
Large intestine (g)	0.149 ± 0.019	0.169 ± 0.008	0.156 ± 0.015	0.159 ± 0.025
Testes (g)	0.261 ± 0.030	0.291 ± 0.051	0.270 ± 0.032	0.266 ± 0.012
Prostate (g)	0.119 ± 0.019	0.098 ± 0.028	0.095 ± 0.029	0.102 ± 0.026

NC, control diet; AFDP-L, 1% AFDP-containing diet; AFDP-H, 5% AFDP-containing diet; AFE, 1% AFE-containing diet. Body weights of mice were measured under a non-fasted condition, and tissues were removed under a fasted condition. Data are presented as the mean ± SD of 6 mice.

Plasma biochemistry profiles of ICR mice fed AFDP and AFE diets for 4 weeks are shown in Table 3. The plasma GLU of both the AFDP-H and AFE groups and BUN of AFE group were significantly lower than those of the NC group; the AFDP-H group also tended to have lower TG. There were no significant differences in the other parameters among the 4 groups, and abnormal values were not found in any of the groups. The GLU level of the AFE group was similar to the normal plasma glucose level of fasted ICR mice (approximately 100 mg/dL). While the NC group tended to show higher than normal levels, the reason for this was unclear. Although it is necessary to further investigate the effects of A. americana flower on normal plasma glucose levels and on hormones involved in plasma glucose homeostasis, it is proposed that A. americana flower has a lowering effect on the plasma glucose concentration.

Table 3. Plasma biochemistry profiles of male ICR mice fed AFDP and AFE diets for 4 weeks

	NC	AFDP-L	AFDP-H	AFE
GLU (mg/dL)	147.6 ± 14.4	128.6 ± 33.1	100.8 ± 29.4*	93.8 ± 24.2*
TCHO (mg/dL)	130.3 ± 28.8	140.7 ± 37.9	155.3 ± 32.9	139.5 ± 45.7
TG (mg/dL)	125.8 ± 21.7	85.5 ± 36.8	78.0 ± 32.9	93.8 ± 30.9
TP (g/dL)	4.850 ± 0.176	4.883 ± 0.306	4.900 ± 0.253	5.017 ± 0.431
AST (U/L)	42.00 ± 6.96	37.83 ± 3.54	48.83 ± 13.00	43.50 ± 12.32
ALT (U/L)	18.50 ± 3.27	14.50 ± 2.43	20.00 ± 7.09	18.00 ± 9.89
ALP (U/L)	250.8 ± 145.5	242.5 ± 79.4	244.2 ± 54.0	210.8 ± 60.0
LDH (U/L)	215.8 ± 90.9	233.3 ± 66.2	252.5 ± 96.9	233.3 ± 56.7
BUN (mg/dL)	23.78 ± 4.81	22.38 ± 2.23	20.60 ± 2.87	18.28 ± 2.20*
UA (mg/dL)	1.033 ± 0.294	1.100 ± 0.297	1.083 ± 0.264	0.950 ± 0.302
CRE (mg/dL)	0.103 ± 0.010	0.103 ± 0.016	0.103 ± 0.019	0.100 ± 0.011

NC, control diet; AFDP-L, 1% AFDP-containing diet; AFDP-H, 5% AFDP-containing diet; AFE, 1% AFE-containing diet. GLU, glucose; TCHO, total cholesterol; TG, triglyceride; TP, total protein; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ALP, alkaline phosphatase; LDH, lactate dehydrogenase; BUN, blood urea nitrogen; UA, uric acid; CRE, creatinine. Plasma was obtained under a fasted condition. Data are presented as the mean ± SD of 6 mice.

*  Significant difference is indicated from NC group at P < 0.05.

On the other hand, the amounts of dried A. americana flower ingested in the AFDP-L, AFDP-H, and AFE groups were calculated as 1.45, 7.43, and 6.53 g/kg/d, respectively, from their dietary intakes. These values in mice correspond to the range of 87 to 446 g of dried A. americana flower in humans (60 kg body weight). Even if a significant amount of A. americana flower was ingested over 4 weeks, the nutritional status would not be affected, and we considered that the risk of a negative influence on hepatic and renal function was low. In this study, the 4-week dietary experiment was equivalent to a subacute toxicological experiment and was performed as an initial investigation of the safety of A. americana flower. Furthermore, it is likely that a long-term ingestion experiment will be necessary for further toxicological studies including genotoxicity.

Effects of Dietary A. americana Flower Extract on Diabetic KK-A^y Mice    Since A. americana flower demonstrated a possible plasma glucose lowering effect, the antihyperglycemic effect of AFE was investigated in type 2 diabetic mice. Table 4 shows the body weights, body weight gain, liver and kidney weights, dietary intake, and plasma biochemistry profiles in male KK-A^y mice fed diets containing AFE for 4 weeks. There were no significant differences in body weights, liver and kidney weights, and dietary intake among the 3 groups. At dissection, the fasted body weights of the DC, AFE-L, and AFE-H groups were not significantly different at 35.08 ± 1.76 g, 33.94 ± 1.07 g, and 36.44 ± 1.90 g, respectively. The AFE-L and AFE-H groups ingested 15.5 mg/day (0.40 g/kg) and 82.1 mg/day (2.07 g/kg) AFE as calculated from their dietary intakes.

Table 4. Body weights, body weight gain, liver and kidney weights, dietary intake and plasma biochemistry profiles of male KK-A^y mice fed AFE diets for 4 weeks

	DC	AFE-L	AFE-H
Initial body weight (g)	31.24 ± 2.34	31.32 ± 0.95	31.54 ± 2.65
Final body weight (g)	39.51 ± 2.26	38.75 ± 1.06	39.68 ± 1.55
Body weight gain (g)	8.27 ± 0.98	7.43 ± 1.15	8.14 ± 2.25
Dietary intake (g/d)	8.19 ± 0.55	7.75 ± 0.35	8.21 ± 0.77
Liver weight (g)	1.368 ± 0.042	1.343 ± 0.079	1.370 ± 0.079
Kidney weight (g)	0.448 ± 0.033	0.457 ± 0.034	0.436 ± 0.025
Plasma biochemistry
GLU (mg/dL)	305.5 ± 56.5	275.7 ± 37.9	224.2 ± 35.2*
TCHO (mg/dL)	161.7 ± 33.3	155.5 ± 37.4	128.7 ± 12.2
TG (mg/dL)	149.1 ± 50.2	141.5 ± 63.5	144.1 ± 16.8
TP (g/dL)	5.420 ± 0.545	5.080 ± 0.370	5.220 ± 0.286
AST (U/L)	60.93 ± 18.49	64.07 ± 22.81	44.55 ± 3.01
ALT (U/L)	6.800 ± 1.483	6.480 ± 2.086	5.000 ± 2.000
ALP (U/L)	365.8 ± 66.6	269.5 ± 66.7	274.5 ± 32.3
LDH (U/L)	232.9 ± 33.9	268.4 ± 68.8	193.2 ± 38.9
BUN (mg/dL)	29.36 ± 2.50	32.38 ± 2.76	30.20 ± 5.58
UA (mg/dL)	3.020 ± 0.432	2.580 ± 0.278	2.980 ± 0.602
CRE (mg/dL)	0.560 ± 0.089	0.502 ± 0.005	0.506 ± 0.009

DC, control diet; AFE-L, 0.2% AFE-containing diet; AFE-H, 1% AFE-containing diet. GLU, glucose; TCHO, total cholesterol; TG, triglyceride; TP, total protein; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ALP, alkaline phosphatase; LDH, lactate dehydrogenase; BUN, blood urea nitrogen; UA, uric acid; CRE, creatinine. Body weights of mice were measured under a non-fasted condition. Under a fasted condition, blood was collected, and tissues were removed and weighed. Data are presented as the mean ± SD of 6 mice.

*  Significant difference is indicated from DC group at P < 0.05.

The GLU level of the AFE-H group was significantly lower than that of the DC group. No abnormal values and no significant differences in TCHO, TG, and TP levels were observed in the 3 groups. Moreover, there were no significant differences in the parameters of hepatic and renal functions among the 3 groups, and the AFE-H group showed lower AST, ALT, ALP, and LDH levels. There have been no previous reports on the toxicity of A. americana flower. These results suggested that A. americana flower did not negatively affect the growth and biochemical parameters of mice; further, no abnormalities were observed upon visual inspection.

Figure 1 shows the changes in plasma glucose and insulin levels in non-fasted KK-A^y mice during the experimental period. All groups showed the same high plasma glucose level at the start of the experiment, and this level increased 1.5- and 1.7-fold in the DC group at 2 and 4 weeks, respectively. The plasma glucose levels in the AFE-L and AFE-H groups increased during the experimental period, but these increases tended to be lower than that in the DC group. Moreover, after 4 weeks, the AFE-H group showed a significantly lower glucose level than the DC group (Fig. 1-A). The increase in the plasma glucose concentration during 4 weeks in the AFE-H group was significantly lower than that in the DC group.

Fig. 1.

Changes in plasma glucose (A) and insulin (B) levels in male KK-A^y mice fed AFE diets for 4 weeks

DC, control diet; AFE-L, 0.2% AFE-containing diet; AFE-H, 1% AFE-containing diet. Blood was collected under a non-fasted condition. Data are presented as the mean ± SD of 6 mice. *Significant difference is indicated from DC group at P < 0.05.

Although the plasma insulin level in all groups increased during the experimental period, the AFE-L group showed a lower plasma insulin level than the DC group, and the plasma insulin level of the AFE-H group changed less than the AFE-L group (Fig. 1-B).

This animal experiment was designed to investigate the antidiabetic effects of AFE, since the plasma glucose concentration was lowered by the ingestion of AFE in normal mice. KK-A^y mice, a non-insulin dependent diabetes model, have genetic obesity and diabetic syndromes such as hyperglycemia and hyperinsulinemia (Iwatsuka et al., 1970). However, in the KK-A^y mice fed AFE diets, the increase in the non-fasted plasma glucose level with aging was depressed in an AFE-dose-dependent manner. Corresponding to the changes in the plasma glucose concentration, the increase in the plasma insulin level was also suppressed in relation to the AFE dose. Since the KK-A^y mice had hyperinsulinemia, it was thought that the reduction in plasma insulin level was caused by suppression of the increase in plasma glucose. Therefore, it is expected that the reduction in plasma glucose level by ingestion of AFE was induced by the insulin-like action of AFE or its inhibitory action on the hydrolysis and absorption of carbohydrates in the digestive tract. Since cinnamic acid derivatives have been reported to promote insulin secretion (Adisakwattana et al., 2008), the possibility that AFE affects insulin secretion is an interesting concept. There have been several studies on the hypoglycemic activity and antidiabetic effects of flowers and their constituents (Bhaskar et al., 2011; Dias et al., 2010; Nakamura et al., 2011); however, there have been no reports on these effects in a type 2 diabetes model. Although the influence of AFE on insulin secretion in normal mice was not elucidated in this study, plasma glucose levels were lowered in a dose-dependent manner with AFE in the ICR mice that had a normal insulin response. Therefore, the suppressive effect of AFE on the increase of non-fasted plasma glucose levels was presumed to be dependent on the inhibition of α-glucosidase in the digestive tract.

α-Glucosidase Inhibition and Antihyperglycemic Effects of A. americana Flower    In order to elucidate the suppressive effect of AFE on the increase of plasma glucose levels, α-glucosidase inhibitory activities of extracts and fractions were investigated. The inhibitory activities of an aqueous extract of A. americana flower on maltase, sucrase, glucoamylase, and isomaltase, all originating from the rat small intestine, were 13.9%, 4.9%, 11.1%, and 9.1%, and those of AFE were 41.8%, 4.4%, 22.6%, and 16.3%, respectively, at 10 mg/mL of each extract concentration. AFE as the 80% methanolic extract showed stronger activities than the aqueous extract and strongly inhibited maltase. AFE was then fractionated to AFS0, AFS20, AFS60, and AFS100, with yields of 65.0%, 2.8%, 3.3%, and 25.2%, respectively, by Sephadex LH-20 column chromatography. These fractions had maltase inhibition rates of 8.2%, 35.6%, 67.1%, and 5.5%, respectively, at 2 mg/mL of each fraction concentration, and the maltase inhibitory activity of AFS60 was stronger than that of AFE.

The antihyperglycemic effect of AFS60, as the maltase inhibitory fraction of A. americana flower, was investigated to elucidate the suppressive effect of AFE on the increase of plasma glucose levels. Figure 2 shows the temporal changes of plasma glucose levels in KK-A^y mice after oral administration of maltose and AFS60 mixtures. The initial plasma glucose concentrations of fasted mice were high, and there were no significant differences in levels among the 3 groups. The plasma glucose level of the C group increased immediately, and reached a maximum level that corresponded to 2.3-fold the initial level at 0.5 h after administration. Although the glucose level decreased after 2 h, it remained at 1.5-fold the initial level. The plasma glucose levels in the AFS-L and AFS-H groups were elevated after maltose administration, but the AFS-H group level tended to be lower than that in the C group. Moreover, the AFS-H group showed lower glucose levels than the AFS-L group, and its concentration at 1 h was significantly lower than those in the C and AFS-L groups.

Fig. 2.

Temporal changes of plasma glucose levels in male KK-A^y mice after administration of maltose and AFS60 mixture

C, 0 mg/kg AFS60 with 1 g/kg maltose; AFS-L, 60 mg/kg AFS60 with 1 g/kg maltose; AFS-H, 180 mg/kg AFS60 with 1 g/kg maltose. Mice were fasted for 20 h before administration. Data represent the mean ± SD of 6 mice. A significant difference is indicated from ^aC and ^bAFS-L groups at P < 0.05.

This test was designed to investigate the maltase inhibitory activity of AFS60 in KK-A^y mice; the area under the curve (AUC_{0– 2h}) of plasma glucose levels from 0 to 2 h after maltose administration was estimated as an index of glucose exposure. The AUC_0–2h in the C, AFS-L, and AFS-H groups were 8.854 ± 0.359 mg/mL·h, 8.322 ± 0.682 mg/mL · h, and 7.492 ± 0.861 mg/mL · h, respectively. The AUC_0–2h of the C and AFS-L groups did not differ; however, the AFS-H group showed a significantly lower AUC_0–2h than the C group. This result suggests that AFS60 can inhibit the elevation of postprandial plasma glucose levels in diabetic mice.

There are a number of reports about the antihyperglycemic effects of natural resources such as Morus alba, Rosa damascena, Lonicera japonica, and Ganoderma lucidum (Yatsunami et al., 2003; Naowaboot et al., 2009; Gholamhoseinian et al., 2009; Zhang et al., 2013; Usui et al., 2007). AFS60 showed 15% inhibition of the postprandial plasma glucose level at a dose of 180 mg/kg, and it is suggested that the antihyperglycemic effect of A. americana flower is similar to that observed with the above natural resources. The flower of Inula japonica showed an antidiabetic effect in alloxan-induced diabetic mice, which was presumed to be based on the recovery of insulin secretion and the inhibition of gluconeogenesis (Shan et al., 2006). The ingestion of 500 mg/kg of an aqueous extract of Cleistocalyx operculatus flower lowered postprandial plasma glucose levels after maltose administration in streptozotocin-induced diabetic mice; the effect was thought to be caused by the glucosidase inhibitory effects of polyphenolic compounds (Mai and Chuyen, 2007). Since similar efficiency to that of C. operculatus was found in our experiments, these findings suggested that A. americana flower could inhibit maltase and had a suppressive effect on postprandial plasma glucose elevation. Moreover, it may suggest that the suppression of non-fasted plasma glucose levels in KK-A^y mice fed the AFE diet was caused by AFS60 inhibition of α-glucosidase.

Identification of Maltase Inhibitory Component from A. americana Flower    In order to isolate the maltase inhibitors, 70 fractions were isolated from AFS60 by preparative HPLC; a strong inhibitory fraction, whose yield was 14.8 mg from 239 mg of AFS60, was obtained through this process. Figure 3 shows the analytical HPLC chromatogram of this inhibitory fraction. The major peak A (AFS60A) was detected at a retention time of 41 min and showed maximum absorption at 329 nm. As a result of preparative HPLC, 3.7 mg of AFS60A was isolated from 14.8 mg of this fraction. The maltase inhibition of 0.1 mg/mL AFS60A was 41.7%, an activity stronger than that of AFS60. Because the contents and inhibitory activities of the other peaks were low, this result showed that AFS60A contributed strongly to the maltase inhibition of AFS60.

Fig. 3.

Analytical HPLC chromatogram of maltase inhibitory fraction from AFS60 and UV spectrum of peak A The analytical conditions are described in the text.

Since AFS60A had a lambda maximum at 329 nm in the absorption spectrum, it was expected to be a phenolic compound such as a caffeic acid derivative. The LC/MS spectrum of AFS60A showed the mass intensity at m/z 343.1; the MS/MS spectrum showed the major mass numbers of m/z 179.1 and 163.2, a typical fragment of chlorogenic acid. These results suggested that AFS60A is a derivative of chlorogenic acid with a molecular weight of 342.

The NMR data of AFS60A is summarized in Table 5. In the ¹³C-NMR spectrum of AFS60A, 15 carbon atoms were detected. The signals at δ_C 127.63, 114.43, 149.96, 148.39, 115.28, and 123.26 ppm were assigned to the benzene ring in the caffeoyl group, and the signals at δ_C 146.93 and 116.56 ppm were identified as carbons of unsaturated α- and β-carbonyls. The signal at δ_C 95.82 ppm was assigned to C-1′ of anomer carbonyl in the β-d-glucopyranoside group, and the δ_C 78.86 − 62.38 ppm signals were identified as carbons in a pyranose ring and a hydroxymethyl group. In the ¹H-NMR spectrum, the doublet signal at δ_H 5.55 ppm (J = 7.8 Hz) was assigned to the H-1′ proton of β-d-glucopyranose, and the doublet signals at δ_H 6.29 and 7.65 ppm were assigned to the H-8 and H-7 protons of the caffeoyl group. The double doublet signal at δ_H 6.96 ppm (J = 1.98 and 8.1 Hz) and the doublet signals at δ_H 6.78 ppm (J = 8.1 Hz) and 7.05 ppm (J = 1.9 Hz) were assigned to protons in an aromatic ring (H-6, H-5, and H-2). From these spectroscopic data, the structure of AFS60A was identified as caffeoyl β-d-glucopyranoside (CBG; Fig. 4). The concentration of CBG in dried A. americana flower was determined as 127.0 µg/g by analytical HPLC, and it was maximum at full bloom (data not shown).

Table 5. ¹H-(270 MHz) and ¹³C-(67.5 MHz) NMR spectroscopic data for AFS60A

Group	Atom	δC	δH	(J in Hz)
caffeoyl	1	127.63
	2	114.43	7.05	d (J = 1.9)
	3	149.96
	4	148.39
	5	115.28	6.78	d (J = 8.1)
	6	123.26	6.96	dd (J = 1.9, 8.1)
	7	146.93	7.65	d (J = 15.9)
	8	116.56	6.29	d (J = 15.9)
	9	167.56
β-d-glucopyranose	1′	95.82	5.55	d (J = 7.8)
	2′	78.86
	3′	74.09
	4′	71.15
	5′	78.09
	6′	62.38

AFS60A was dissolved in CD₃OD. Abbreviations: d, doublet; dd, double doublet.

Fig. 4.

Chemical structure of AFS60A (caffeoyl β-d-glucopyranoside)

In this study, the HPLC-PDA and LC/MS/MS analyses clarified that AFS60 contained anthocyanins and flavonoids; however, the identification is likely incomplete. Moreover, the antihyperglycemic effect of A. americana flower might not be fully explained by AFS60A alone. Nevertheless, the activities of the other ingredients were weaker than that of AFS60A. Future study will focus on the structural analyses of the compounds that participate in α-glucosidase inhibition.

These results suggest that AFS60, which was fractionated from A. americana flower, has an antihyperglycemic effect in diabetic mice, and that CBG might contribute to the effect of AFS60 by maltase inhibition. As CBG has already been identified from L. japonica (Qian et al., 2008) and yacon (Terada et al., 2006), it cannot be considered as a unique compound. However, there are few reports on the identification of caffeic acid derivatives from petals (Gholamhoseinian et al., 2009; Li et al., 2005), and CBG has only been reported from Spiraea cantoniensis flowers (Yoshida et al., 2008). Furthermore, there are no reports on the antihyperglycemic effect of these ingredients in diabetic animals. Therefore, the identification of CBG as a maltase inhibitor from A. americana flower, which also has suppressive effects on the elevation of postprandial plasma glucose levels, is a novel finding.

While the aim was the effective usage of A. americana flowers, which are removed and discarded to promote high-quality tuber development, the consumption of flowers is uncommon. Therefore, investigations of the dietary ingestion and antihyperglycemic effect of A. americana flower were conducted to elucidate its safety and beneficial effects. Consequently, this study demonstrated that A. americana flower can be safely consumed in the daily diet for 4 weeks and that it had a suppressive effect on postprandial plasma glucose elevation. Moreover, CBG was identified as a maltase inhibitor from A. americana flower. As a first step, these results provide useful information regarding the applicability of A. americana flower as a new functional food material. In future, it will be important to identify other active ingredients in the flower as well as demonstrate both chronic and genetic safety.

Acknowledgements    This research was financially supported by the Center for Promotion of Research and Intellectual Property, Aomori University of Health and Welfare, Aomori, Japan. We thank Dr. S. Yamaguchi (Aomori Prefectural Industrial Technology Research Center, Aomori, Japan) for support in the NMR measurements and Mr. S. Matsuo (Aomori Pharmaceutical Association, Sanitary Inspection Center) for operation of the LC/MS measurements.

References

•  Adisakwattana,  S.,  Moonsan,  P., and  Yibchok-Anun,  S. (2008). Insulin-releasing properties of a series of cinnamic acid derivatives in vitro and in vivo. J. Agric. Food Chem., 56, 7838-7844.
•  Bhaskar,  J.J.,  Shobha,  M.S.,  Sambaiah,  K., and  Salimath,  P.V. (2011). Beneficial effects of banana (Musa sp. var. elakki bale) flower and pseudostem on hyperglycemia and advanced glycation end-products (AGEs) in streptozotocin-induced diabetic rats. J. Physiol. Biochem., 67, 415-425.
•  Dias,  T.,  Bronze,  M.R.,  Houghton,  P.J.,  Mota-Filipe,  H., and  Paulo,  A. (2010). The flavonoid-rich fraction of Coreopsis tinctoria promotes glucose tolerance regain through pancreatic function recovery in streptozotocin-induced glucose-intolerant rats. J. Ethnopharmacol., 132, 483-490.
•  Gholamhoseinian,  A.,  Fallah,  H., and  Sharifi,  F. (2009). Inhibitory effect of methanol extract of Rosa damascena Mill. flowers on α-glucosidase activity and postprandial hyperglycemia in normal and diabetic rats. Phytomedicine, 16, 935-941.
•  Hoshikawa,  K. and  Juliarni. (1995). The growth of Apios (Apios americana Medikus), a new crop, under field conditions. Jpn. J. Crop Sci., 64, 323-327.
•  Iwai,  K. (2008). Antidiabetic and antioxidant effects of polyphenols in brown alga Ecklonia stolonifera in genetically diabetic KK-A^y mice. Plant Food Hum. Nutr., 63, 163-169.
•  Iwai,  K. and  Matsue,  H. (2007). Ingestion of Apios americana Medikus tuber suppresses blood pressure and improves plasma lipids in spontaneously hypertensive rats. Nutr. Res., 27, 218-224.
•  Iwatsuka,  H.,  Shino,  A., and  Suzuoki,  Z. (1970). General survey of diabetic features of yellow KK mice. Endocrinologia Japonica, 17, 23-35.
•  Kuramoto,  S.,  Kaneyoshi,  G.,  Morinaga,  Y.,  Matsue,  H., and  Iwai,  K. (2013). Angiotensin-converting enzyme-inhibitory peptides isolated from pepsin hydrolyzate of Apios americana tuber and their hypotensive effects in spontaneously hypertensive rats. Food Sci. Technol. Res., 19, 399-407.
•  Li,  Y.,  Wen,  S.,  Kota,  B.P.,  Peng,  G.,  Li,  G.Q.,  Yamahara,  J., and  Roufogalis,  B.D. (2005). Punica granatum flower extract, a potent α-glucosidase inhibitor, improves postprandial hyperglycemia in Zucker diabetic fatty rats. J. Ethnopharmacol., 99, 239-244.
•  Mai,  T.T. and  Chuyen,  N.V. (2007). Anti-hyperglycemic activity of an aqueous extract from flower buds of Cleistocalyx operculatus (Roxb.) Merr and Perry. Biosci. Biotechnol. Biochem., 71, 69-76.
•  Nakamura,  S.,  Matsuda,  H., and  Yoshikawa,  M. (2011). Search for antidiabetic constituents of medicinal food. Yakugaku Zasshi, 131, 909-915.
•  Naowaboot,  J.,  Pannangpetch,  P.,  Kukongviriyapan,  V.,  Kongyingyoes,  B., and  Kukongviriyapan,  U. (2009). Antihyperglycemic, antioxidant and antiglycation activities of mulberry leaf extract in streptozotocin-induced chronic diabetic rats. Plant Food Hum. Nutr., 64, 116-121.
•  Oe,  H. and  Ozaki,  S. (2008). Hypoglycemic effect of 13-membered ring thiocyclitol, a novel α-glucosidase inhibitor from kothala-himbutu (Salacia reticulata). Biosci. Biotechnol. Biochem., 72, 1962-1964.
•  Patel,  M.B. and  Mishra,  S.M. (2012). Magnoflorine from Tinospora cordifolia stem inhibits α-glucosidase and is antiglycemic in rats. J. Func. Food, 4, 79-86.
•  Peyrieras,  N.,  Bause,  E.,  Legler,  G.,  Vasilov,  R.,  Claesson,  L.,  Peterson,  P., and  Ploegh,  H. (1983). Effects of the glucosidase inhibitors nojirimycin and deoxynojirimycin on the biosynthesis of membrane and secretory glycoproteins. EMBO J., 2, 823-832.
•  Qian,  Z.M.,  Wen,  X.D.,  Li,  H.J.,  Liu,  Y.,  Qin,  S.J., and  Li,  P. (2008). Analysis of interaction property of bioactive components in Flos Lonicerae Japonicae with protein by microdialysis coupled with HPLC-DAD-MS. Biol. Pharm. Bull., 31, 126-130.
•  Reeves,  P.G.,  Nielsen,  F.H., and  Fahey,  G.C. Jr. (1993). AIN-93 purified diets for laboratory rodents: final report of the American Institute of Nutrition ad hoc writing committee on the reformulation of the AIN-76A rodent diet. J. Nutr., 123, 1939-1951.
•  Shan,  J.J.,  Yang,  M., and  Ren,  J.W. (2006). Anti-diabetic and hypolipidemic effects of aqueous-extract from the flower of Inula japonica in alloxan-induced diabetic mice. Biol. Pharm. Bull., 29, 455-459.
•  Shipra,  G.,  Suman,  B.S.,  Usha,  R.S.,  Surendra,  K.B., and  Krishna,  M.P. (2010). Elucidation of mechanism of action of Cassia auriculata leaf extract for its antidiabetic activity in streptozotocin-induced diabetic rats. J. Med. Food, 13, 528-534.
•  Terada,  S.,  Ito,  K.,  Yoshimura,  A.,  Noguchi,  N., and  Ishida,  T. (2006). The constituents relate to anti-oxidative and α-glucosidase inhibitory activities in yacon aerial part extract. Yakugaku Zasshi, 126, 665-669.
•  Usui,  T.,  Okazaki,  M.,  Kamiuchi,  S.,  Suzuki,  F.,  Iizuka,  H., and  Hibino,  Y. (2007). Inhibitory effects of a water-soluble extract from culture medium of Ganoderma lucidum (Rei-shi) mycelia on postprandial blood glucose elevation in mice and additional effect with α-glucosidase inhibitor. J. Jpn. Soc. Nutr. Food Sci., 60, 249-255.
•  Wilson,  P.W.,  Gorny,  J.R.,  Blackmon,  W.J., and  Reynolds,  B.D. (1986). Fatty acids in the American groundnut (Apios americana). J. Food Sci., 51, 1387-1388.
•  Wilson,  P.W.,  Pichardo,  F.J.,  Liuzzo,  J.A.,  Blackmon,  W.J., and  Reynolds,  B.D. (1987). Amino acids in the American groundnut (Apios americana). J. Food Sci., 52, 224-225.
•  Yoshida,  K.,  Hishida,  A.,  Iida,  O.,  Hosokawa,  K., and  Kawabata,  J. (2008). Flavonol caffeoylglycosides as α-glucosidase inhibitors from Spiraea cantoniensis flower. J. Agric. Food Chem., 56, 4367-4371.
•  Yatsunami,  K.,  Saito,  Y.,  Fukuda,  E.,  Onodera,  S., and  Oshigane  K. (2003). α-Glucosidase inhibitory activity in leaves of some mulberry varieties. Food Sci. Technol. Res., 9, 392-394.
•  Zhang,  Z.,  Luo,  A.,  Zhong,  K.,  Huang,  Y.,  Gao,  Y.,  Zhang,  J.,  Gao,  H.,  Xu,  Z., and  Gao,  X. (2013). α-Glucosidase inhibitory activity by the flower buds of Lonicera japonica Thunb. J. Func. Food, 5, 1253-1259.