Hypoglycemic and hypolipidemic effect of Meyna spinosa leaves in high fat diet-alloxan induced type 2 diabetic rats
Abstract
This study investigated the antidiabetic and hypolipedemic action of Meyna spinosa leaf fractions. Effect of methanol, ethyl acetate and petroleum ether fraction (75 and 150 mg/kg p.o.) from M. spinosa leaf methanol extract was evaluated in high fat diet-alloxan induced type 2 diabetic rats after 21 days treatment. Glibenclimide (5 mg/kg), methanol and ethyl acetate fraction at 150 mg/kg exhibited 57.7, 63.4 and 53.8% reduction in serum glucose level after 21 days. Fractions demonstrated significant decrease in triglycerides, total cholesterol, low density lipid, very low density lipid, α-amylase level, and increase in body weight, high density lipid level in diabetic rats. Fractions showed significant (p<0.05) hypoglycemic effect in glucose loaded animals but not in normal rats. Petroleum ether fraction did not produced significant effect. Result validated the claim made by folk medicinal uses and confirmed the antidiabetic potential of methanol and ethyl acetate fraction of M. spinosa.
Introduction
Type 2 or non insulin dependent diabetes account for about 95% of all diagnosed cases of diabetes mellitus, and has a significant impact on the health, excellence and expectancy of life (Dewanjee et al., 2009). Though several oral hypoglycemic agents are available, but most of these drugs possess different side effects. Thus, the hunt for a new therapeutic agent devoid of adverse effect originating from plants would be of interest (Kannur et al., 2006). Meyna spinosa Roxb. (Rubiaceae) has been used in the treatment of hepatic disorder, gastrointestinal problems, skin infection, and as refrigerant, and abortifacient (Bora and Kumar, 2003; Dea et al., 2009; Mitra and Mukherjee, 2009). The boiled fruit extract is used to treat diabetes in Manipur, while fruit and leaf juice are recommended to cure diabetes by the tribes of Tripura, India (Khan and Yadava, 2010; Sen et al., 2011). Scientific literature suggests that M. spinosa fruits extracts possess antimicrobial activity (Buragohain, 2008). Methanol extract of M. spinosa also possess antifungal activity (Goswami et al., 2006). The present study aim to investigate the in vivo antidiabetic, hypolipedemic potential of fractions of Meyna spinosa leaf against high fat diet-alloxan induced type 2 diabetes.
Materials and Methods
Plant materials
Leaves of Meyna spinosa Roxb. were collected from Tripura, India and identified by Dr. BK Datta, Department of Botany, Tripura University, Tripura, India (Voucher specimenno: TU/BOT/HEB/SS23072011a).
Preparation of extract and fractionation
The fresh leaves were air dried under shade, pulverized into coarse powder (500 g) and extracted with methanol (2.5 L) using Soxhlet apparatus for 18 hours. The extract was concentrated to dryness under reduced pressure to obtain the methanol extract (15.2% w/w). The methanol extract (20 g) was pre-absorbed onto silica gel (60-120 mesh, SD Fine Ltd., Mumbai) and fractionated with column chromatographic technique using the following solvents in the order of increasing polarity viz petroleum ether, ethyl acetate and methanol. The yields afforded petroleum ether (PFMS, 8.3% w/w), ethyl acetate (EFMS, 34.5% w/w), and methanol (MFMS, 49.0% w/w) fractions.
Experimental animals
Healthy Albino mice (20-30 g) were used for acute toxicity study and Wistar rats (150-200 g) were used for the antidiabetic activity. Animals were maintained under controlled environmental conditions. The study was approved by the Institutional Animal Ethical Committee (Reg. No. 1305/ac/09/CPCSEA).
Acute toxicity study
The fixed dose (OCED Guideline no. 423, Annexure 2d) method of Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA) was used for toxicity studies. The tested fractions were administrated orally.
Effect of fractions on normoglycemic plus glucose-hyperglycemic model (NG-OGTT)
To access the effect of fractions on combined NG-OGTT model, blood samples were collected from overnight fasted rats, followed by the administration of vehicle, glibenclimide (5 mg/kg), MFMS, EFMS, and PFMS at a dose of 75 and 150 mg/kg, p.o. Serum glucose level was determined after ½, 1, and 2 hours after drug administration. The rats were orally loaded with 2 g/kg glucose after the last measurement (at 2 hours) and the serum glucose levels were continued to determine after 2½, 3, 4 and 6 hours by using a commercially available kit (Agapee Diagnostic Ltd, Kerala) (Orhan et al., 2005).
Induction of type 2 diabetes
High fat diet was administered continuously for 10 days to healthy rats. On 11th day freshly prepared alloxan solution in physiological saline (120 mg/kg) injected into the caudal vein of overnight fasted rats. Fifteen minute later 0.4 U insulin was injected (i.p.); 2.5 and 5.0 hours later, 25% glucose (10 mL/kg, p.o.) administered to decrease animal mortality rate caused by the hypoglycemia and hyperglycemia in rats with injection of alloxan. Twenty four hour after first alloxan injection, alloxan (100 mg/kg body weight) was injected again into the caudal vein of rats. Insulin (0.4 U) was intraperitoneally injected 15 min later to rats, then 25% glucose (10 mL/kg) given orally 2.5 and 5.0 hours later respectively. The blood glucose was tested and animals with fasting blood glucose more than 200 mg/dL were used for the experiments (Shu et al., 2009).
Antidiabetic and hypolipedemic evaluation of fractions
Healthy and diabetic animals were divided into nine groups in fix sex ratio (3 males and 3 females). Groups I and II served as normal control and diabetic untreated control respectively. Group III, diabetic rat treated orally with 5 mg/kg/day glibenclamide. Group IV-VII, diabetic rat treated with the MFMS (75 and 150 mg/kg/day, p.o.), EFMS (75 and 150 mg/kg/day, p.o.) and PFMS (75 and 150 mg/kg/day, p.o.) for 21 days respectively. Serum glucose levels and body weight were measured on day 0, 7, 14 and 21 of the study.
Initial and final total cholesterol (TC), triglyceride (TG), high density lipids (HDL), low density lipids (LDL), very low density lipids (VLDL), and α–amylase level in serum were determined using a commercially available kit (Agapee Diagnostic Ltd, Kerala).
Statistical analysis
Results are expressed as mean ± SEM, (n = 6). Statistical difference was tested by using one-way analysis of variance (ANOVA) followed by Tukey tests. A level of p<0.05 was used as the criterion for statistical significance.
Results
Fractions did not show any mortality at the dose of 2,000 mg/kg. Therefore, 2000 mg/kg dose was considered as ALD50 cut off the dose under Globally Harmonised Classification System (GHS) category 5 (safe dose).
Fractions did not show any significant change in serum glucose level, while glibenclamide decreases serum glucose levels in normoglycemic rats (between 14.3 and 44%). However, methanol and ethyl acetate fraction prevented drastic increase of serum glucose after the glucose loading, and decrease the serum glucose level below normal levels after 360 min (Table I).
Table I: Effects of the M. spinosa fractions on blood glucose levels in normal and glucose-loaded hyperglycaemic rats
Test sample |
Dose (mg/kg) | Serum glucose concentration (mg/dL)[%inhibition] | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 0 min | 30 min | 60 min | 120 min (glucose load) | 150 min | 180 min | 240 min | 360 min | ||
| Control | - | 66.1 (2.9) |
70.2 (3.2) |
72.4 (3.3) |
75.1 (2.7) |
150.8 (5.0) |
137.4 (4.4) |
98.6 (4.2) |
77.9 (3.2) |
| Glibenclimide | 5 | 68.3 (3.1) |
60.2 (2.1) [14.3] |
56.1 (3.1) [22.5]a |
42.1 (2.0) [44.0]†|
106.9 (4.0) [29.2]a |
86.8 (2.5) [36.8]‡ |
60.2 (3.1) [38.9]‡ |
53.6 (2.9) [31.2]‡ |
| Methanol fraction | 75 | 68.5 (2.7) |
72.2 (3.0) |
69.9 (3.2) |
75.6 (3.2) |
137.0 (4.1) |
117.5 (3.9) [14.5]a |
88.8 (3.7) |
80.7 (3.0) |
| Methanol fraction | 150 | 64.2 (2.4) |
68.3 (3.9) |
70.0 (4.5) |
69.2 (2.5) |
119.5 (5.4) [20.8]a |
89.3 (2.9) [35.0]‡ |
65.8 (4.1) [33.3]‡ |
51.0 (2.2) [34.5]‡ |
| Ethyl acetate fraction | 75 | 69.7 (2.3) |
71.2 (3.3) |
67.5 (3.0) |
74.0 (3.5) |
144.4 (5.0) |
120.0 (3.4) [12.6] |
95.1 (3.9) |
87.6 (3.3) |
| Ethyl acetate fraction | 150 | 69.1 (2.1) |
63.8 (1.9) |
70.1 (3.8) |
65.6 (2.7) |
126.3 (5.4) [24.6]†|
102.3 (2.6) [25.5]a |
77.3 (4.2) [21.6]†|
62.3 (2.2) [20.0]a |
| Petroleum ether fraction | 75 | 71.2 (2.4) |
72.1 (3.4) |
68.7 (3.2) |
73.8 (3.1) |
148.0 (5.0) |
135.0 (3.0) |
107.0 (3.0) |
92.0 (2.8) |
| Petroleum ether fraction | 150 | 67.1 (2.8) |
71.3 (2.9) |
64.4 (3.0) |
67.9 (2.2) |
144.3 (3.1) |
131.2 (2.4) |
105.3 (3.4) |
87.7 (2.5) |
| Values are given as mean ± (SEM). n = 6; Only inhibitory effect on blood glucose concentration above 10% is shown in bracket [] ap<0.05, †p< 0.01, ‡p<0.001 when test drug treated group compared with control animals | |||||||||
MFMS and EFMS exhibited significant hypoglycaemic activity in diabetic rats, while no significant effect observed for PFMS (Table II). After 21 days of treatment with MFMS (75 and 150 mg/kg/day), there was 34.9 and 63.4% decrease (p<0.001) in serum glucose levels, while EFMS at a dose of 75 and 150 mg/kg results 30.7 and 53.8% inhibition of serum glucose level. At the end of 21 days treatment standard drug, MFMS, and EFMS significantly increases body weight (p<0.05, 0.01) (Table II).
Table II:Effects of the M. spinosa fractions on body weight and serum glucose level in NIDDM rats
| Test sample | Dose (mg/kg) | Body weight (g) | Serum glucose concentration (mg/dL)[%inhibition] | ||||||
|---|---|---|---|---|---|---|---|---|---|
| 0 day | 7 day | 14 day | 21 day | 0 day | 7 day | 14 day | 21 day | ||
| Normal control | - | 180.2 (12.1) |
185.3 (13.8) |
189.9 (14.2) |
193.2 (16.1) |
65.8 (3.9) |
70.5 (4.0) |
67.3 (3.9) |
73.6 (4.2) |
| Diabetic control | - | 170.4 (10.2) |
156.9 (13.3) |
143.3 (11.4)†|
145.4 (14.9)†|
230.1 (7.4)‡ |
242.1 (8.9)‡ |
252.8 (10.1)‡ |
258.0 (10.1)‡ |
| Glibenclimide | 5 | 166.3 (11.1) |
164.3 (10.4) |
170.1 (13.3)†|
175.3 (10.8)†|
251.3 (8.1) |
182.0 (8.0)†[24.8] |
131.4 (7.2)‡ [48.0] |
109.0 (4.9)‡ [57.7] |
| Methanol fraction | 75 | 172.1 (13.9) |
163.4 (10.5) |
165.2 (16.0)†|
167.3 (14.1)a |
240.3 (9.4) |
210.2 (8.4)a [13.2] |
198.1 (9.2)†[21.7] |
168.1 (8.3)‡ [34.9] |
| Methanol fraction | 150 | 167.9 (14.1) |
163.4 (13.0) |
168.2 (14.1)a |
172.2 (15.1)†|
244.9 (9.1) |
173.0 (7.0)†[28.5] |
127.1 (8.2)‡ [49.7] |
94.6 (3.1)‡ [63.4] |
| Ethyl acetate fraction | 75 | 168.3 (10.1) |
166.3 (12.5) |
162.6 (12.7)†|
161.3 (12.2)a |
238.1 (8.8) |
222.0 (7.8) |
208.8 (8.0)a [17.4] |
178.8 (7.0)†[30.7] |
| Ethyl acetate fraction | 150 | 164.6 (11.7) |
160.6 (9.9) |
165.7 (12.8)a |
169.3 (13.3)†|
246.9 (9.4) |
172.0 (6.1)†[29.0] |
138.3 (7.0)‡ [45.3] |
119.2 (3.5)‡ [53.8] |
| Petroleum ether fraction | 75 | 163.2 (12.0) |
150.6 (11.5) |
145.3 (11.2) |
140.5 (13.3) |
245.8 (8.2) |
254.9 (11.7) |
260.3 (7.1) |
267.2 (11.1) |
| Petroleum ether fraction | 150 | 170.3 (13.7) |
157.1 (15.3) |
150.1 (13.9) |
143.3 (12.9) |
231.0 (7.2) |
227.4 (8.9) |
224.8 (9.9) [11.1] |
215.7 (9.9)a [16.4] |
| Values are given as mean ± (SEM); n = 6; Only inhibitory effect on blood glucose concentration above 10% is shown in the bracket [ ] ap<0.05, †p<0.01, ‡p<0.001 when test drug treated group compared with diabetic control group and diabetic control group compared with normal control group | |||||||||
A significant decrease in serum TG, TC, LDL and VLDL levels, and increase in HDL level were observed after the treatment with glibenclamide, MFMS, and EFMS.
Higher dose of ethyl acetate fraction showed similar hypolipedemic activity compare to standard, but methanol fraction (150 mg/kg) produced better effect than glibenclamide (Table III). Diabetic control rats had significantly higher serum amylase activity compared to normal control. The enzyme level was reduced (p<0.05, 0.01) following the treatment with glibenclamide, MFMS and EFMS (Table III).
Table III:Effect of fraction of M. spinosa leaves on serum lipid profile and α-amylase in type 2 diabetic rats
| Test sample |
Dose (mg/kg) | TG (mg/dL) | TC (mg/dL) | LDL (mg/dL) | VLDL (mg/dL) | HDL (mg/dL) | α-amylase (U/L) | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Initial | Final | Initial | Final | Initial | Final | Initial | Final | Initial | Final | Initial | Final | ||
| Normal control | 95.6 (6.1) |
98.0 (6.0) |
48.3 (4.1) |
52.1 (4.9) |
14.5 (2.2) |
16.0 (2.6) |
19.1 (2.0) |
19.6 (2.2) |
26.4 (3.3) |
28.0 (3.6) |
206.0 (8.9) |
211.4 (7.9) |
|
| Diabetic control | 150.4 (8.0)‡ |
159.4 (8.0)‡ |
179.4 (10.3)‡ |
186.6 (11.0)‡ |
75.3 (3.8)‡ |
91.0 (6.6)‡ |
30.1 (3.2)†|
31.9 (3.8)‡ |
22.3 (3.3)a |
17.9 (3.2)†|
288.6 (10.0)‡ |
296.3 (11.6)‡ |
|
| Glibenclimide | 5 | 147.4 (8.8) |
109.8 (5.3)‡ |
170.0 (8.8) |
79.4 (7.0)‡ |
74.2 (3.9) |
35.3 (3.2)‡ |
29.5 (3.4) |
22.0 (1.7)†|
21.4 (3.0) |
27.1 (4.0)‡ |
290.3 (10.0) |
222.6 (9.0)‡ |
| Methanol fraction | 75 | 155.3 (9.9) |
128.0 (6.4)†|
181.2 (12.7) |
121.5 (7.1)†|
78.1 (4.0) |
52.1 (4.0)‡ |
31.1 (3.0) |
25.6 (2.3)†|
22.8 (3.0) |
25.3 (3.6)†|
275.8 (9.4) |
250.8 (8.3)a |
| 150 | 142.6 (7.1) |
103.5 (5.9)‡ |
164.7 (8.6) |
66.4 (6.8)‡ |
72.1 (5.0) |
32.1 (3.3)‡ |
28.5 (2.9) |
20.7 (1.9)†|
20.4 (1.9) |
27.0 (3.9)‡ |
282.7 (9.0) |
228.7 (7.8)†|
|
| Ethyl acetate fraction | 75 | 144.9 (9.2) |
131.0 (8.0)†|
169.5 (9.6) |
118.3 (7.0)†|
74.8 (3.9) |
50.3 (4.1)‡ |
29.0 (2.6) |
26.2 (3.2)†|
21.1 (2.8) |
25.9 (3.0)†|
280.8 (8.7) |
261.3 (8.1)a |
| 150 | 148.7 (8.3) |
107.7 (5.9)‡ |
174.3 (10.9) |
83.7 (6.1)‡ |
76.9 (4.1) |
38.7 (3.0)‡ |
29.7 (3.3) |
21.5 (3.6) |
23.4 (3.3) |
27.1 (3.3)†|
272.1 (9.9) |
134.0 (6.3)†|
|
| Petroleum ether fraction | 75 | 151.6 (7.8) |
153.2 (8.5) |
175.7 (9.7) |
182.3 (10.4) |
79.0 (3.3) |
83.2 (6.3) |
30.3 (4.3) |
30.7 (4.1) |
19.4 (2.0) |
16.8 (1.9) |
281.8 (8.5) |
179.0 (8.0) |
| 150 | 147.0 (8.4) |
138.7 (7.5)a |
167.4 (8.2) |
140.5 (7.0)a |
72.0 (4.6) |
70.4 (6.0)a |
29.4 (3.6) |
27.7 (3.0)a |
20.0 (2.7) |
19.6 (2.3) |
286.4 (10.2) |
172.9 (8.5) |
|
| Values are given as mean ± (SEM); n = 6; ap<0.05, †p<0.01, ‡p<0.001 when test drug treated group compared with diabetic control group and diabetic control group compared with normal control group | |||||||||||||
Discussion
Despite the fact that substantial progress has been made in the management of diabetes mellitus, but the search for plant based products mainly from folk/traditional medicinal system for the control of diabetes mellitus continues with great expectation (Kannur et al., 2006). Oral glucose tolerance test is acknowledged for indirect assessment of in vivo peripheral insulin action and insulin resistance (Liou et al., 2002). Fractions did not have any effect on normal animal, but prevent drastic increase in serum glucose level after glucose administered. Hence the fraction may have the potential to improve insulin sensitivity.
Alloxan has been observed to cause necrosis of β-cells and induce hyperglycemia. Intraperitoneal injection of alloxan after administration of high fat diet causes type 2 diabetes (Shu et al., 2009). MFMS, EFMS significantly decreased the fasting serum glucose in diabetic animals. Alloxan causes characteristic loss in body weight in diabetic rats which is due to improved muscle wasting and due to loss of tissue protein (Nammi et al., 2003; Jain et al., 2010), and a similar phenomenon was observed in the present study also. Treatment with MFMS and EFMS was found effective, which may due to the protective effect of fractions in controlling muscle wasting, i.e. reversal of gluconeogenesis and may also be due to proper glycemic control of fractions may responsible for this effect.
The sulphonylureas cause hypoglycaemia by increasing the secretion of insulin and mainly active in mild alloxan-induced diabetes (Al-Shamaony et al., 1994). Fractions of M. spinosa had similar activity like that of glibenclamide in diabetic rats which could be due to improving the metabolism of glucose, and increasing insulin secretion from the β-cells.
Marked increase in TG, TC, LDL, and VLDL in diabetic rats was observed. The most frequent lipid abnormalities in diabetes are hypertriglyceridemia and hypercholesterolemia (Al-Shamaony et al., 1994; Shirwaikar et al., 2005). Hypertriglyceridemia is coupled with metabolic consequences of hypercoagulability, hyperinsulinemia, insulin resistance and insulin intolerance (Shirwaikar et al., 2005). In our study, administration of MFMS and EFMS significantly (p<0.05) improved these parameters. Decreased of cholesterogenesis and fatty acid synthesis may responsible for the observed hypolipidemic effect. Significant lowering of TC, LDL and raise in HDL is a very enviable biochemical state for avoidance of atherosclerosis and ischemic conditions (Shirwaikar et al., 2005; Akpan et al., 2012). The result indicated fractions have the potential to prevent or treat the complications of diabetes. The acini produce α-amylase enzyme that involved in the digestion of polysaccharide in diet. Diabetic cause damage to acini cells results leakage of the enzyme into blood (Akpan et al., 2012). MFMS and EFMS may prevent cellular damage thus decrease the enzyme level in serum. Preliminary phytochemical analysis suggested that methanol extract contain alkaloids, glycosides, triterpenoids, tannins and flavonoids. Therefore, presence of any of these compounds in fraction may be responsible for observed biological activities. These statistical data indicated MFMS possess highest hypoglycemic effect is likely due to the increase in insulin secretion or ameliorating insulin sensitivity or due to the inhibition of α amylase activity.
Conclusion
MFMS, EFMS has beneficial effects, in reducing the elevated blood glucose level, alpha amylase activity and lipid profile of high fat diet-alloxan induced-diabetic rats. Thus, justifying the claim made by folk medicinal uses and can be viewed as potential source of new antidiabetic molecule.
References
Akpan HD, Ekaidem IS, Usoh IF, Ebong PE, Isong NB. Effect of aqueous extract of Azadirachta indica (Neem) leaves on some indices of pancreatic function in alloxan-induced diabetic Wistar rats. Pharmacologia 2012; 3: 420-25.
Al-Shamaony L, Al-Khazraji SM, Twaij HAA. Hypoglycemic effect of Artemisiab herbaalba II. Effect of a valuable extract on some blood parameters in diabetic animals. J Ethnopharmacol. 1994; 43: 167-71.
Bora PJ, Kumar Y. Floristic diversity of Assam: Study of Pabitora wild life sanctuary. Delhi, Daya Publishing House, 2003.
Buragohain J. Phytochemical and antimicrobial investigation on the fruits of Meyna spinosa. Res J Biotechnol. 2008; 3: 372-74.
Das HB, Majumdar K, Datta BK, Ray D. Ethanobotanical uses of some plants by Tripuri and Reang tribes of Tripura. Nat Prod Radiance. 2009; 8: 172-80.
Dewanjee S, Maiti A, Das AK, Mandal SC, Dey SP. Swietenine: a potential oral hypoglycemic from Swietenia macrophylla seed. Fitoterapia 2009; 80: 249-51.
Goswami S, Bora L, Das J, Begam M. In vitro evaluation of some medicinal plants against Candida albicans. J Cell Tissue Res. 2006; 6: 837-39.
Jain S, Bhatia G, Barik R, Kumar P, Jain A, Dixit VK. Antidiabetic activity of Paspalum scrobiculatum Linn. in alloxan induced diabetic rats. J Ethnopharmacol. 2010; 127: 325-28.
Kannur DM, Hukkeri VI, Akki KS. Antidiabetic activity of Caesalpinia bonducella seed extracts in rats. Fitoterapia 2006; 77: 546-49.
Khan MD, Yadava PS. Antidiabetic plants used in Thoubal district of Manipur, Northeast India. Indian J Trad Knowl. 2010; 9: 510-14.
Liou SS, Liu IM, Hsu JH, Wu YC, Hsu SF, Chen JT. Release of acetylcholine by Die–Huang–Wan to enhance insulin secretion for lowering plasma glucose in Wistar rats. Auton Neurosci. 2002; 100: 21-26.
Mitra S, Mukherjee SK. Some arbortifacient plants used by the tribal people of West Bangal. Nat Prod Radiance. 2009; 8: 167-71.
Nammi S, Boini MK, Lodagala SD, Behara RBS. The juice of fresh leaves of Catharanthus roseus Linn. reduces blood glucose in normal and alloxan diabetic rabbits. BMC Complement Altern Med. 2003; 3: 1-4.
Orhan DD, Aslan M, Sendogdu N, Ergun F, Yesilada E. Evaluation of the hypoglycemic effect and antioxidant activity of three Viscum album subspecies (European mistletoe) in streptozotocin-diabetic rats. J Ethnopharmacol. 2005; 98: 95-102.
Sen S, Chakraborty R, De B, Devanna N. An ethnobotanical survey of medicinal plants used by ethnic people in West and South district of Tripura, India. J For Res. 2011; 22: 417-26.
Shirwaikar A, Rajendran K, Punitha ISR. Antidiabetic activity of alcoholic stem extract of Coscinium fenestratum in streptozotocin-nicotinamide induced type 2 diabetic rats. J Ethnopharmacol. 2005; 97: 369-74.
Shu XS, Lv JH, Tao J, Li GM, Li HD, Ma N. Antihyperglycemic effects of total flavonoids from Polygonatum odoratum in STZ and alloxan induced diabetic rats. J Ethnopharmacol. 2009; 124: 539-43.
