Evaluation of Ethnomedical Claim III: Anti-hyperglycemic activities of Gongronema latifolium root and stem (民族药物需求评估III： Gongronema latifolium 根与茎的降糖活性)

ORIGINAL ARTICLE

Evaluation of Ethnomedical Claim III: Anti-hyperglycemicactivities of Gongronema latifolium root and stemAdeleke Clement ADEBAJO,1,2,3 Marcus Durojaye AYOOLA,1 Samuel Akintunde ODEDIRAN,1

Adetunji Joseph ALADESANMI,1 Thomas Jürgen SCHMIDT2 and Eugene Joseph VERSPOHL3

1Department of Pharmacognosy, Faculty of Pharmacy, Obafemi Awolowo University, Ile Ife, Nigeria, 2Institute of Pharmaceutical Biologyand Phytochemistry, and 3Department of Pharmacology, Institute of Pharmaceutical and Medicinal Chemistry, WestfälischeWilhelms-Universität, Munster, Germany

Correspondence

Adeleke Clement Adebajo, Department ofPharmacognosy, Faculty of Pharmacy,Obafemi Awolowo University, Ile Ife,Nigeria.Tel: +234 8033679390Email: [email protected]

Received 6 February 2012; revised 23November 2012; accepted 3 December2012.

doi: 10.1111/1753-0407.12019

Abstract

Background: The insulinotropic activity of the combined root and stem ofGongronema latifolium (Asclepiadaceae) was evaluated to justify its Africanethnomedicinal use in the management of diabetes.Methods: A methanolic extract and its chromatographic fractions (A1–A6)were tested for glucose-reducing and in vitro insulin-stimulating abilities usingglucose-loaded rats and INS-1 cells, respectively. In vivo insulin-releasingactivities for the significantly (P < 0.05) active antihyperglycemic A5 and A6

and in vitro insulinotropic activity of the C1 fraction and its isolated constitu-ents were also similarly determined.Results: The extract (100 mg/kg) had higher in vivo antihyperglycemic activ-ity than the individual A1–A6, indicating a synergistic effect of the plantconstituents. Higher in vivo insulin release in response to A5 (100 mg/kg) thanA6, agreed with their in vivo antihyperglycemic activities and confirmedinsulin release as a mechanism of action of the plant. Compared with5.6 mmol/L glucose (negative control), the extract and the A3, A6, and C1 (allat 100.0 mg/mL) elicited significantly higher in vitro insulin release that wassimilar to that of glibenclamide (1.0 mg/mL; P > 0.05). Fraction C1 yielded a1:1 mixture of a-amyrin and b-amyrin cinnamates (1a:1b), lupenyl cinnamate(2), lupenyl acetate (3), and two unidentified triterpenoids, Y and Z. The 1a:1bmixture (100.0 mg/mL) demonstrated the highest insulinotropic activity thatwas comparable (P > 0.05) to that of glibenclamide (1.0 mg/mL).Conclusions: The results confirm pancreatic activity as a mechanism under-lying the antidiabetic action of G. latifolium and justify its ethnomedical use.

Keywords: antihyperglycemic, Gongronema latifolium, insulin release,a-amyrin cinnamate, b-amyrin cinnamate.

Significant findings of the study: The results demonstrated synergistic antihyperglycemic activities of the differentfractions of Gongronema latifolium. A 1:1 mixture of a-amyrin : b-amyrin cinnamates was the main active con-stituent, with similar in vitro insulinotropic activity as glibenclamide. The most active antihyperglycemic fractions(A5 and A6) gave high plasma and in vitro insulin release.What this study adds: The study identifies insulin release as a mechanism of action and justifies the ethnomedicinaluse of the plant as an antidiabetic. It is the first report of 1a/1b in this plant and demonstrates the contribution of2, 3 and Z to the insulinotropic activity of the plant.

bs_bs_banner

Journal of Diabetes 5 (2013) 336–343

336 © 2013 Wiley Publishing Asia Pty Ltd and Ruijin Hospital, Shanghai Jiaotong University School of Medicine

Introduction

Diabetes mellitus is a disease of chronic disorders ofcarbohydrate, protein. and fat metabolism due tovarying degrees of absence of insulin and/or its resistanceor the absence of insulin altogether.1 The high cost ofanti-diabetic drugs, severe long-term complications ofdiabetes mellitus and serious side effects of synthetichypoglycemic agents compel continued pharmacologicalinvestigations into plants used ethnomedicinally asantidiabetic agents.1–3

Managing diabetes is one of the ethnomedicinal uses ofthe combined root and stem of Gongronema latifoliumBenth & Hook (Asclepiadaceae) in Africa, especiallyin south-eastern Nigeria, where it is called Utazi.4,5

A hypoglycemic action of a methanolic leaf extract(100 mg/kg) has been reported in alloxan-diabetic mice.6

In addition, the ethanolic leaf extract (100 mg/kg) hasbeen reported to have various enzyme activities instreptozotocin-diabetic rats, representing extrapancre-atic activity.3,5,7 b-Sitosterol, lupeol acetate, and lupeolcinnamate have been isolated from the stem,8 whereaspregnane esters have been isolated from the leaves.9,10

In the present study, we tested the methanolic extract ofG. latifolium stem and root, its vacuum liquid chromato-graphic (VLC) and column chromatographic (CC)fractions for in vivo antihyperglycemic and in vitroinsulin-stimulating activities. The compounds isolatedwere investigated only for their ability to stimulate insulinrelease in vitro to determine whether a pancreatic actionmay explain the folkloric antidiabetic use of G. latifolium.

Methods

Chemicals, equipment and instrumentation

Samples were evaluated using vacuum liquid chroma-tography (dimensions 9 ¥ 12 cm, silica gel HR60,size 10–40 mm), column chromatography (dimensions60 ¥ 4 cm, silica gel, mesh No. 70–230), and preparativethin layer chromatography (PTLC; silica gel 60 F254 0.25,0.5, 1, 2 mm). Nuclear magnetic resonance (NMR)spectra (400, 500 and 600 MHz) were obtained usingBruker AMX 400, Varian Nova 500 and Varian UnityPlus 600 NMR instruments (Bruker Daltonics, Bremen,Germany). High-resolution electrospray ionization massspectra (HRESIMS) were obtained using loop injection,electrospray ionization (ESI), and determination of the[M + Na]+ and [2M + Na]+ ions using a MicroTof massspectrophotometer (Bruker Daltonics). Electrosprayionization mass spectra (ESIMS) were also obtainedusing a GCQ mass spectrometer (Finnigan MATGmbH, Bremen, Germany), spectrometer (70 eV, +veElectron Impact (EI)) using the Electrospray ionization

(ESI) technique.1,20,21 The 96-well flat-bottomed microc-ulture plates used in the present study and the Titertekmotorized hand diluter were obtained from Costar(Corning, New York, NY, USA) and Flow Laboratories(Uxbridge, UK), respectively. Biological chemicals andmedia, as well as the equipment used for studies in therat insulinoma and insulin-secreting (INS-1; generouslyprovided by Prof. Dr. Claes B. Wollheim, Department ofCell Physiology and Metabolism, University MedicalCenter, University of Geneva, Geneva, Switzerland) celllines were as reported previously.1,11,12 Blood glucoselevels were determined using a Finetest glucometer(model IGM-0005A) with Finetest strips (Infopia Co.Ltd., Anyang Kyunggi, Korea). All solvents used were ofanalytical grade, whereas materials and equipment usedfor the in vitro tests were sterile.

Animals

Healthy Wistar albino rats of either sex (400 g averageweight) were maintained at 22°C under natural 12 h day–night conditions in the Animal House of the Institute ofPharmaceutical and Medicinal Chemistry (Münster,Germany) and the Animal House of the Faculty of Phar-macy, Obafemi Awolowo University (Ile-Ife, Nigeria).Rats were acclimatized for at least 5 days before experi-ments commenced. Rats were fed a standard pellet diet(Altromin, Lage, Germany; or Bendel Feeds, Ewu,Nigeria) and water was available ad libitum. Groups ofsix to seven normal rats were fasted for 18 h beforeadministration of either the extract, fractions, drugs orvehicle (negative control).1,12 All animal experiments con-formed to the Guide for the Care and Use of LaboratoryAnimals published by the National Academies Press.13

Sample preparation

Prior to use, all extract, fractions, and compounds werefirst solubilized by 0.1 mL of 100% dimethylsulfoxide(DMSO, for in vitro experiments) or 100% Ethanol(EtOH, for in vivo experiments) followed by subsequentdilution in sterile water or 0.9% (w/v) saline to 10 mL.This method of dissolution was used to ensure uniformdosage of the drugs or extract or fraction while a � 1%final concentration of DMSO and EtOH was chosen as ithas been reported to have no solvent effect on the testsystems.1,11,12

Culture of rat insulinoma or INS-1 cell lines

The INS-1 cells were grown in 24-multiwells plates for5–6 days (half confluence: 1–2 ¥ 106 cells/mL) in RoswellPark Memorial Institute (RPMI) medium supplementedwith 10% (v/v) fetal calf serum, 100 U/mL penicillin and

A.C. ADEBAJO et al. Antidiabetic constituents of Utazi

© 2013 Wiley Publishing Asia Pty Ltd and Ruijin Hospital, Shanghai Jiaotong University School of Medicine 337

0.1 mg/mL streptomycin. The cells in the multiwells werewashed three times with Krebs’–Ringer–bicarbonatebuffer (KRBH), containing 20 mmol/L HEPES and0.5% bovine serum albumin, and thereafter incubated for90 min at 37°C in the same buffer containing 5.6 mmol/Lglucose.1,11,12

In vivo antihyperglycemic potential inglucose-loaded rats

Ten groups of five fasted rats were used. An oral glucosetolerance test (OGTT) was performed administering10 g/kg glucose solution to normal rats. Thirty minuteslater (T0), rats with blood glucose levels �7.0 mmol/Lwere administered 100 mg/kg, p.o., methanolic extract(GLSR), VLC (A1–A6) or CC (C1) fractions, or 5 mg/kg,p.o., glibenclamide. A drop of blood from the tail of eachrat was dropped onto strips that were inserted into theglucometer and the values recorded. Blood glucose levelsat T0 were taken as 100%, with values obtained othertimes in each group expressed as a percentage of T0.1

In vitro insulin determination

Insulin released into the medium was determined byradioimmunoassay (RIA) using rat insulin as the stan-dard, (mono-125I-Tyr A14)-porcine insulin as the labeledcompound, and anti-insulin antibodies. Results obtainedfor the effects of 5.6 mmol/L glucose on insulin releasefrom INS-1 cells were taken as 100%, with the results forall other agents tested expressed as a percentage of theglucose results.1,11,12 Experiments were replicated six toeight times.

In vivo determinations in glucose-loaded rats

In vivo experiments were performed in three groups ofsix to seven fasted rats in each. A glucose tolerance testwas performed by the concurrent administration of0.6 mL glucose solution (10 g/kg, i.p.) and 100 mg/kg,p.o., of the most active VLC fractions (i.e. A5 or A6) orglibenclamide (5 mg/kg, p.o.) at Time 0 (T0) to normalrats. Blood was drawn by the retrobulbar techniqueusing heparinised capillary tubes and the samples (15drops) were collected within 20 s into chilled tubes at 0,0.5, 1.0 and 2.0 h after drug administration. Bloodglucose was determined in duplicate by the glucoseoxidase method.1,11,12 After the addition of 150 IUheparin-Na to the blood samples, the samples were cen-trifuged at 2106 g for 10 min at 10°C and the sera col-lected and kept frozen at –20°C until use in the insulinRIA. Insulin levels at T0 were taken as 100%, with thoseat other times within each of the groups expressed as apercentage of these values at T0.1,11,12

Plant material

Whole G. latifolium plants were collected in May 2003from the Obafemi Awolowo University (Ile-Ife, Nigeria)after authentication by Dr C.O. Illoh (Botany Depart-ment, Obafemi Awolowo University). A voucher speci-men (No. UHI 15362) has been deposited in theuniversity herbarium.

Extraction

Powdered stem and root (1.367 kg) were soxhletextracted with methanol (5 L). The extract was concen-trated in vacuo to give a 178.4 g (13.05% w/w yield)methanolic extract (Fig. 1).

Antihyperglycemic and insulin-stimulating actions of themethanolic extract

The in vivo antihyperglycemic activity of the methanolicextract was evaluated using the glucose tolerance testdescribed above.1 In addition, the ability of the extract toelicit insulin release in vitro was determined as describedabove.1,11,12

Root and stem powder (1.4 kg)Methanol

GLSR (178 G)

A1 A2 A3 A4 A5 A6

B1 B2 B3 B4 B5 B6

C1 C2 C3 C4 C5

N-Hexane wash

MARC

VLC

CC

CC

B8B7

D1D2

F4F1 F5F2

PTLC

PTLC

G1 G3 G4G22 *

PTLC

I4

K1 K2 K3 K4 K5 K6

L1 L2 L3 L4 L5

PTLC

M1 M2 M3 M4 M5

PTLC

J1 J2 J3 J4 3 *

PTLC

1a/1b*

Z *

I3I2

HI

E1 E2 E3

I1

Y*

PTLC

PTLC

F3

PTLC

Figure 1 Flow chart showing the isolation of constituents of Gon-gronema latifolium stem and root. GLSR, methanolic extract; VLC,vacuum liquid chromatography; CC, column chromatography; PTLC,preparative thin layer chromatography; A1–A6, VLC fractions; B1–B8,C1–C5, CC fractions; E1–M5, PTLC fractions; *, pure isolated com-pound; 1a, a-amyrin cinnamate; 1b, b-amyrin cinnamate; 2, lupenylcinnamate; 3, lupenyl acetate; Y, Z, unidentified triterpenes.

Antidiabetic constituents of Utazi A.C. ADEBAJO et al.


Fractionation of the methanolic extract

Vacuum liquid chromatography of the active methanolicextract (178.0 g) yielded fractions A1–A6 (Fig. 1).

Antihyperglycemic and insulin-stimulating actions offractions A1–A6

The in vivo antihyperglycemic and in vitro insulin-stimulating actions of fractions A1-A6 were evaluatedusing the glucose tolerance test and INS-1 cells, respec-tively.1,11,12 In addition, the in vivo insulinotropic effectsof the most active antihyperglycemic fractions (i.e. A5

and A6), which produced high in vitro insulin release,were determined in glucose-loaded rats.15

Purification of fraction A1

Column chromatography of fraction A1 (4.85 g) pro-vided gradient fractions B1–B8. Thin layer chromatogra-phy revealed that the components of fraction A1 weremore concentrated in fraction B1. Hence, further columnchromatography purification of B1 (1.60 g) yielded frac-tions C1–C5 (Fig. 1).

Antihyperglycemic and insulin-stimulating actions offraction C1

Fraction C1 was tested for its in vivo anti-hyperglycemicand in vitro insulin-stimulating actions using the glucosetolerance test and INS-1 cells, respectively.1,11,12

Isolation of the components of fraction C1

An N-hexane solvent wash of fraction C1 (750 mg)yielded D1 (precipitate, 300 mg) and a 443 mg filtrate(D2). Fraction D1 was purified by repeated PTLC(cyclohexane : CH2Cl2 6:4) to obtain pure G4 (30.2 mg).Further PTLC (0.25 mm, cyclohexane : CH2Cl2 6:4)purification of combined G2 and G3 (40 mg) gave pure I2

(17.3 mg). Fractions E2, E3, F4, and F5 were combined(35 mg) and resubjected to PTLC (0.25 mm,cyclohexane : CH2Cl2 6:4) to give pure J3 (21.7 mg).Further PTLC purification of G1 and D2 gave pure H1

(2.5 mg) and M5 (3.8 mg), respectively. Detailed frac-tionation of the methanolic extract and the isolation ofthese compounds are shown in Fig. 1 and are available asSupplementary Material to this paper.

Identification of the compounds isolated

The melting points, infrared, 1H–1H-NMR (Homo-nuclear Correlated Spectroscopy (COSY), Homo-nuclear Total Correlated Spectroscopy (TOCSY)),13C-NMR (Attached Proton Test (APT), DistortionlessEnhancement Polarisation Transfer (DEPT)),

1H-13C-NMR (Heteronuclear multiple bond correlation(HMBC), Heteronuclear Single Quantum Correlation(HSQC)), ESIMS (positive, negative), and HRESIMSdata of the isolates I2, G4 and J3, were comparedwith information in the literature,14–17 enabling theiridentifications.

In vitro insulinotropic effects of the constituents offraction C1

The ability of isolates I2, G4, J3, H1 and M5 toelicit insulin release in vitro were assessed as describedpreviously.1,11,12

Statistical analysis

Data are given as the mean � SEM. Differences betweengroups were evaluated by one-way analysis of variance(ANOVA) followed by Bonferroni’s t-test and theStudent–Newman–Keuls’ test as post hoc tests. P < 0.05was considered significant.1,11,12

Results and discussion

Figure 1 shows the isolation of compounds 1a/1b, 2, 3, Y,and Z, whereas the structures of the compounds identi-fied are given in Fig. 2. The in vivo antihyperglycemiceffects of the methanolic extract, its VLC fractions, andglibenclamide in glucose-induced hyperglycemic rats arepresented in Table 1. The data show that blood glucoselevels in rats given only glucose (10 g/kg) were signifi-cantly (P < 0.05) higher at all times (0.5–4.0 h) thanblood glucose levels in rats administered the GLSRextract, VLC fractions A1–A6 (100 mg/kg), or glibencla-mide (5 mg/kg). Fractions A1–A4 exhibited hyperglyce-mic actions throughout the duration of the experiment.Glibenclamide and polar A5 and A6 (as deduced fromtheir VLC elutions), had the highest antihyperglycemicactivities at 4 h (Table 1), suggesting an insulinotropicaction. A synergistic effect of the constituents of G. lati-folium stem and root may be suggested by the fact thatnone of the VLC fractions (A1–A6) had activity compa-rable to that of the extract (P > 0.05; Table 1).

The in vitro effects of the methanolic extract, its VLCfractions, and glibenclamide on glucose-mediated insulinrelease from INS-1 cells are given in (Table 2). The invivo antihyperglycemic GLSR extract elicited dose-dependent in vitro insulin release, with values at100 mg/mL that were significantly higher than those for5.6 mmol/L glucose (control) and comparable to theeffects of 1.0 mg/mL glibenclamide (Table 2). Theseobservations suggest that the extract has insulinotropicactivity. Dose-dependent effects of A2–A6 on insulinrelease were observed, whereas only the effects of A3 and



A6 were comparable to that of glibenclamide. The induc-tion of plasma insulin in vivo by A5, A6, and glibencla-mide is shown in Fig. 3. Due to the differences observedin the relative in vivo and in vitro potencies of A5 and A6,the two most active partition fractions, their in vivo

insulinotropic activities were ascertained. Results of anOGTT with a supplemental i.p. glucose load revealedcomparable plasma insulin release by glibenclamide, A5,and A6 at 0.5 h (Fig. 3), whereas at 1 and 2 h the degreeof plasma insulin released by A5 was significantly less

O

H

H

H

H

O

H

H

H

O

O

H

H

H

H

O

O

H

1a: α-amyrin cinnamate

1b: β-amyrin cinnamate

2: lupenyl cinnamate

3: lupenyl acetate

Figure 2 Structures of the isolated constituents of Gongronema latifolium stem and root.

Table 1 In vivo antihyperglycemic activities of the extract and fractions of Gongronema latifolia stem and root

Blood glucose levels as a percentage of T0 (% reduction in blood glucose relative to negative control at Tt)

0 h 0.5 h 1 h 2 h 4 h

Glucose (10 g/kg;negative control)

100.0 � 0.1 83.3 � 0.3c 85.7 � 0.1d 76.2 � 0.2c 73.8 � 0.2e

GLSR (100 mg/kg) 100.0 � 0.7 72.7 � 0.3a (12.8%) 70.1 � 0.3b (18.4%) 68.4 � 0.3b (10.3%) 53.0 � 0.2b (28.2%)A1 (100 mg/kg) 100.0 � 0.4 89.8 � 0.1e (-7.7%) 88.8 � 0.2e (-3.6%) 81.8 � 0.2d (-7.3%) 74.0 � 0.4e (-0.3%)A2 (100 mg/kg) 100.0 � 0.4 93.8 � 0.1f (-12.6%) 93.8 � 0.2g (-9.4%) 83.8 � 0.2e (-9.9%) 80.0 � 0.4g (-8.4%)A3 (100 mg/kg) 100.0 � 0.3 97.6 � 0.3h (-17.2%) 95.2 � 0.3h (-11.1%) 88.1 � 0.2f (-15.6%) 76.2 � 0.2f (-3.2%)A4 (100 mg/kg) 100.0 � 0.2 98.8 � 0.3i (-18.6%) 97.6 � 0.1i (-13.9%) 96.4 � 0.3g (-26.5%) 86.8 � 0.3h (-17.6%)A5 (100 mg/kg) 100.0 � 0.6 87.5 � 0.5d (-5.0%) 77.3 � 0.8c (9.8%) 67.1 � 0.8b (12.0%) 61.4 � 0.8c (16.9%)A6 (100 mg/kg) 100.0 � 0.3 96.2 � 0.2g (-15.5%) 92.4 � 0.2f (-7.8%) 77.2 � 0.4c (-1.3%) 64.6 � 0.7d (12.5%)Glibenclamide (5 mg/kg;

positive control)100.0 � 1.1 77.2 � 0.4b (13.3%) 67.0 � 0.2a (21.8%) 55.6 � 0.2a (27.0%) 43.3 � 0.3a (41.3%)

Data show the mean � SEM blood glucose levels at the different time points expressed as a percentage of levels at 0 h (T0; n = 5). Values inparentheses are the reduction in blood glucose values as a percentage of values in the negative control group at the same time point. Withincolumns, values with different superscript letters differ significantly (P < 0.05; one-way ANOVA followed by Bonferroni’s t-test and theStudent–Newman–Keuls’ test).GLSR, methanolic extract of Gongronema latifolium stem and root; A1–A6, vacuum liquid chromatographic fractions of GLSR.



than those of A6 and glibenclamide (Fig. 3). These obser-vations are in agreement with the in vitro insulin values(Table 2) and suggest that the antihyperglycemic activityof A5 may reflect extrapancreatic activity,3,5,7 as well as atransient early insulinotropic activity. Conversely, the in

vivo insulinotropic actions of A6 and glibenclamide werecomparable (Fig. 3) and sustained throughout the periodof the experiment. In addition, these two agents gave thehighest in vitro insulin release (Table 2). Therefore, A6

may elicit its antihyperglycemic effect principally viainsulin release (pancreatic action).

Comparisons of the in vivo antihyperglycemic activi-ties of the extract and its VLC A1 and CC C1 fractions areshown in Fig. 4. The in vivo antihyperglycemic activityof C1 was significantly higher at 4 h than that of itsmother fraction A1 (Fig. 4). In addition, its in vitroinsulin releasing ability, especially at 100 mg/mL, wascomparable to that of glibenclamide and significantlyhigher than that of A1 (Fig. 5). These two effects of C1

(Figs 4,5) were comparable to those of A6 (Tables 1,2)and suggest the presence of insulinotropic constituents inC1. Purification of C1 yielded five isolates (Fig. 1), ofwhich I2, G4 and J3 with purity >97% were conclusivelyidentified as 1:1 mixture of a-amyrin and b-amyrin cin-namates (1a/1b), lupenyl cinnamate (2), and lupenylacetate (3), respectively (Fig. 2). The purity of the iso-lates H1 and M5, coded Y and Z, respectively, was >94%and they were partly identified as unknown triterpenes(A.C. Adebajo et al., unpubl. data, 2013). The com-pound 1a/1b has not been reported previously for thisplant. With the exception of Y, all the isolates exhibiteddose-dependent insulinotropic activity. Their in vitroinsulin-stimulating ability was in the order 1a/1b =glibenclamide > Z = 2 = 3 > Y (Fig. 5), indicating thatthe most active constituent was 1a/1b. The possiblecontribution of 2, 3, and Z to the insulinotropic and

Table 2 Effects of extract and fractions of Gongronema latifoliumstem and root on glucose-mediated insulin release from INS-1 cells

Concentration of thetest agents (extract,fractions)

Insulin release (% of 5.6 mmol/L glucose)

10.0 mg/mL 100.0 mg/mL

Glucose (mmol/L)3.0 (substimulatory

concentration)62.8 � 5.9a 62.8 � 5.9a

5.6 (negative control) 100.0 � 0.0b,c,d 100.0 � 0.0b,c,d

5.6 mmol/L glucose+GLSR 117.6 � 6.9c,d,e 150.7 � 9.5f,g,h

+A1 119.9 � 9.0c,d,e 113.7 � 12.6b,c,d,e,f

+A2 114.4 � 6.5c,d,e 133.7 � 5.4d,e,f,g

+A3 120.5 � 5.4c,d,e 175.9 � 6.8h,i

+A4 84.8 � 7.6a,b,c 106.9 � 6.4b,c,d,e

+A5 112.0 � 11.3c,d,e 127.1 � 4.6c,d,e,f,g

+A6 130.2 � 3.6d,e 173.8 � 7.4g,h,i

+Glibenclamide(1.0 mg/mL;positive control)

173.2 � 6.5f 173.2 � 6.5g,h,i

Data show the mean � SEM (n = 6–8). Insulin release in response to5.6 mmol/L glucose was taken as 100%, with all other valuesexpressed as a percentage of this value. Within columns, values withdifferent superscript letters differ significantly (P < 0.05; one-wayANOVA followed by Bonferroni’s t-test and the Student–Newman–Keuls’ test).GLSR, methanolic extract of Gongronema latifolium stem and root;A1–A6, vacuum liquid chromatographic fractions of GLSR.

250

200

150

100

50

0

0 30 60

Time (min)

120

Pla

sm

a insulin

(%

of T

0)

‡†

† †

†‡

†‡

*†‡

***

*

*

Figure 3 Insulin-releasing effects of glibenclamide (5 mg/kg, p.o.)(•), fractions A5 (�) and A6 (�) (100 mg/kg, p.o.) of Gongronemalatifolium stem and root in glucose-induced hyperglycemic rats.(�). Insulin levels at T0 were taken as 100%, with those at othertimes within the groups expressed as a percentage of these values.Experiments were replicated six to seven times. Data are themean � SEM. *P < 0.05 compared with the negative control at thesame time point; †P < 0.05 compared with glibenclamide at the sametime point; ‡P < 0.05 compared with A5 at the same time point.

120

100

80

60

40

20

00 0.5 1 2

Time (h)4

Pla

sm

a insulin

(%

of T

0)

*†

*†

*†

*†*†

*†

*†

*†

*†

*†

*†

**

**

† †

† † †

Figure 4 In vivo antihyperglycemic activities of the extract and frac-tions of Gongronema latifolia stem and root. (�), 10 g/kg glucose(GL; negative control); ( ), 100 mg/kg methanolic extract of Gon-gronema latifolium stem and root (GLSR); ( ), A1 (vacuum liquidchromatographic fraction of GLSR); (�), C1 (column chromatographicsubfraction of A1); ( ), 5 mg/kg glibenclamide (positive control).Blood glucose levels at T0 were taken as 100%, with those at othertimes within the groups expressed as a percentage of these valuesData are the mean � SEM (n = 5). *P < 0.05 compared with GL atthe same time point; †P < 0.05 compared with glibenclamide at thesame time point.



ultimately antihyperglycemic activity of the plant,extract, and fractions A1 and C1 may confirm the syner-gistic nature of the plant constituents. This has alreadybeen observed with the fractions (Tables 1,2).

There are no prior reports regarding the insulin-stimulating properties of G. latifolium. The glucose-loaded model used in the present study enables detailedcomparisons of the actions of glibenclamide, the extract,and its fractions and may give an insight into theirmechanism(s) of action.18,19

A lack of insulin-stimulating action by A1 and A4

(Table 2) may explain their in vivo hyperglycemic activ-ity (Table 1). The VLC fractions of leaves of Stachytar-pheta cayennensis (Verbenaceae)1 and Bauhiniamonandra (Fabaceae)20 have similarly been reported tohave hyperglycemic activity and to inhibit insulin releasein vitro. Using the same INS-1 cell method, quercetin-3-O-rutinoside (rutin) and carbazole alkaloids have beenidentified as the insulin-inhibitory agents in B. monan-dra20 and Murraya koenigii (Rutaceae)11 leaves, respec-tively. Hyperglycemic A2 (Table 1) also exhibited mildinsulinotropic activity (Table 2). However, it is notknown why the high in vitro insulin activity demon-strated by A3 (Table 2) did not produce corresponding invivo antihyperglycemic activity (Table 1). Conversely,significant in vivo antihyperglycemic and weak and/or noin vitro insulinotropic activities have been reported forrutin, isolated as the major compound of the butanolicfraction of B. monandra leaves.20,21 It is believed that rutinbehaves as a prodrug. Its metabolism by gastric enzymes

produces quercetin, which has these in vivo and in vitroactivities.21 Therefore, quercetin has been identified asthe active antihyperglycemic and insulinotropic com-pound of B. monandra leaves.21

A mixture of sitosteryl and stigmasteryl glucosides(charantin), oleanolic acid 3-O-monodesmoside and ole-anolic acid 3-O-glucuronide, isolated from various partsof Momordica charantia (Cucurbitaceae), demonstrateshypoglycemic activity.22–24 Furthermore, antihyperglyce-mic activity and increased plasma insulin levels havebeen reported for a triterpene glycoside (gymnemic acidIV) in streptozotocin-diabetic rats, with the actions ofgymnemic acid IV comparable to those of glibencla-mide.25 This compound was reported to be the mostactive of the five triterpene glycosides in the antidiabeticsaponin fraction of Gymnema sylvestre (Apocynaceae)leaves.25 Hence, similar identification of the insulinotro-pic constituents of G. latifolium as 1:1 mixture ofa-amyrin and b-amyrin cinnamates (1a/1b), lupenylcinnamate (2), lupenyl acetate (3), and the unidentifiedtriterpenoid Z conclusively justify the antidiabeticethnomedical claims of the plant and established stimu-lation of insulin release as its hitherto unreported mecha-nism of action.

Acknowledgements

The authors thank Sanofi-Aventis (Charlotte, NC, USA)for supplying the radiolabeled insulin and the UniversityResearch Committee of Obafemi Awolowo Universityfor funding part of this work (grants 11-813-A and11-812-ARN). The authors are also grateful to the Deut-scher Akademischer Austauchdienst (Bonn, Germany)for a senior fellowship for ACA.

Disclosure

The authors declare no conflicts of interest.

References

1. Adebajo AC, Olawode EO, Omobuwajo OR et al.Hypoglycaemic constituents of Stachytarpheta cayennen-sis leaf. Planta Med. 2007; 73: 241–50.

2. Pattanayak S, Nayak SS, Panda D, Shende V. Hypogly-cemic of Cajanus scarabaeoides in glucose overloaded andstreptozotocin-induced diabetic rats. Bangladesh J Phar-macol. 2009; 4: 131–35.

3. Ugochukwu NH, Babady NE. Antihyperglycemic effectof aqueous and ethanolic extracts of Gongronema latifo-lium leaves on glucose and glycogen metabolism in liversof normal and streptozotocin-induced diabetic rats. LifeSci. 2003; 73: 1925–38.

Figure 5 Effects of extract, fractions and isolates of Gongronemalatifolium stem and root on glucose-mediated insulin release fromINS-1 cells. Glu, Glucose (sub-stimulatory and negative controls);GLSR, Methanolic extract of Gongronema latifolium stem and root;A1, Vacuum Liquid Chromatographic fraction of GLSR; C1, ColumnChromatographic subfraction of A1; Y, Z, unidentified triterpenes;1a/1b, 1:1 mixture of a- (1a) and b-amyrin (1b) cinnamates; 2, lupenylcinnamate; 3, lupenyl acetate; Glib, Glibenclamide. *P < 0.05 vs5.6 mmol/L of glucose; †P < 0.05 for values vs that of glibenclamide(positive control). Insulin level of glucose (5.6 mmol/L) was taken as100% and those of the extract/fractions/drug were expressed aspercentage of this value. N = 6–8.



4. Morebise O, Fafunso MA, Makinde JM, Olajide OA,Awe EO. Anti-inflammatory property of the leaves ofGongronema latifolium. Phytother Res. 2002; 16 (Suppl.1): S75–77.

5. Ugochukwu NH, Babady NE, Cobourne MK, GassetSR. The effect of Gongronema latifolium extracts onserum lipid profile and oxidative stress in hepatocytes ofdiabetic rats. J Biosci. 2003; 28: 1–5.

6. Ogundipe OO, Moody JO, Akinyemi TO, Raman A.Hypoglycemic potentials of methanol extracts of selectedplant foods in alloxanized mice. Plant Foods Hum Nutr.2003; 58: 1–7.

7. Ugochukwu NH, Cobourne MK. Modification of renaloxidative stress and lipid peroxidation in streptozotocin-induced diabetic rats treated with extracts from Gon-gronema latifolium leaves. Clin Chim Acta. 2003; 336:73–81.

8. Ekundayo O. Constituents of Gongronema latifoliumBenth & Hook. (Asclepiadiaceae). Q J Crude Drug Res.1980; 18: 127–29.

9. Schneider C, Rotscheidt K, Braimaier E. Zwei neue Preg-nanesterglykoside aus Gongronema latifolium (Asclepia-daceae). J Prakt Chem/Chemiker-Zeitung. 1993; 335:532–36.

10. Schneider C, Rotscheidt K, Braimaier E. Vier neue Preg-nanglycoside aus Gongronema latifolium (Asclepia-daceae). Liebigs Ann Chem. 1993; 10: 1057–62.

11. Adebajo AC, Olayiwola G, Verspohl JE et al. Evaluationof the ethnomedical claims of Murraya koenigii. PharmBiol. 2004; 42: 610–20.

12. Adebajo AC, Iwalewa EO, Obuotor EM et al. Pharma-cological properties of the extract and some isolated com-pounds of Clausena lansium stem bark: anti-trichomonal,anti-diabetic, anti-inflammatory, hepatoprotective andantioxidant effects. J Ethnopharmacol. 2009; 122: 10–9.

13. Committee for the Update of the Guide for the Care andUse of Laboratory Animals, Institute for LaboratoryAnimal Research, Division on Earth and Life Studies,National Research Council of the National Academies.Guide for the Care and Use of Laboratory Animals, 8thedn. The National Academies Press, Washington, DC,2011.

14. Kuo YH, Chang CI, Li SY et al. Cytotoxic constituentsfrom the stems of Diospyros maritime. Planta Med. 1997;63: 363–65.

15. Menezes FS, Borsatto AS, Pereira NA, Matos FJA,Kaplan MAC. Chamaedrydiol, an ursane triperpenefrom Marsypianthes chamaecdrys. Phytochemistry. 1998;48: 323–25.

16. Chiang YM, Kuo YH. New peroxy triterpenes from theaerial roots of Ficus microcarpa. J Nat Prod. 2001; 64:436–39.

17. Humberto MMS, Nascimento MCBS, Sant’Ana AEG.Pentacyclic triterpene 5-phenylpenta-2,4-dienoyl estersfrom Peltastes peltatus (Vell) Woodson. PhytochemAnal. 2004; 15: 339–44.

18. Luzi L, Pozza G. Glibenclamide: an old drug with a novelmechanism of action? Acta Diabetol. 1997; 34: 239–44.

19. Murray RK, Granner DK, Rodwell VW. Harper’sIllustrated Biochemistry, 27th International Edition.McGraw-Hill Education (Asia), Singapore, 2006.

20. Alade GO, Omobuwajo OR, Adebajo CA, Verspohl EJ.Evaluation of the hypoglycaemic activity of Bauhiniamonandra leaf in Alloxan- diabetic rats and INS-1 insulincells. J Chem Pharm Res. 2011; 3: 506–21.

21. Alade GO, Adebajo AC, Omobuwajo OR, Proksch P,Verspohl EO. Quercetin, a minor constituent of the anti-hyperglycaemic fraction of Bauhinia monandra leaf.J Diabetes. 2012; 4: 439–41.

22. Lotlikar MM, Rao MR. Pharmacology of a hypoglyce-mic principle isolated from the fruits of Momordica cha-rantia Linn. Indian J Pharm. 1966; 28: 129–33.

23. Matsuda H, Li Y, Murakami T, Matsumura N,Yamahara J, Yoshikawa M. Antidiabetic principles ofnatural medicines. Part III. Structure-related inhibitoryactivity and action mode of oleanolic acid glycosideson hypoglycemic activity. Chem Pharm Bull. 1998; 46:1399–403.

24. Haque ME, Alam MB, Hossain MS. The efficacy ofCucurbitane type triterpenoids, glycosides and phenoliccompounds isolated from Momordica charantia: a review.Int J Pharm Sci Res. 2011; 2: 1135–46.

25. Sugihara Y, Nojima H, Matsuda H, Murakami T,Yoshikawa M, Kimura I. Antihyperglycemic effects ofgymnemic acid IV, a compound derived from Gymnemasylvestre leaves in streptozotocin-diabetic mice. J AsianNat Prod Res. 2000; 2: 321–27.

Supporting information

Additional Supporting Information may be found in theonline version of this article:

Supplementary Material Detailed fractionation andisolation procedure of the compounds of Gongronemalatifolium.



Documents

Evaluation of Ethnomedical Claim III: Anti-hyperglycemic activities of Gongronema latifolium root and stem (民族药物需求评估III： Gongronema latifolium 根与茎的降糖活性)