Journal of Ethnopharmacology,

Elsevier Scientific Publishers

30 (1990) 265-279
Ireland Ltd.

265

POSSIBLE
REGENERATION
OF THE ISLETS OF LANGERHANS
IN STREPTOZOTOCIN-DIABETIC
RATS GIVEN GYMAEMA
SYLVESZRE LEAF EXTRACTS

E.R.B. SHANMUGASUNDARAM,
K. LEELA
GOPINATH,
K. RADHA SHANMUGASUNDARAM
and V.M. RAJENDRAN
Department
of Biochemistry,
Postgraduate
Institute
Madras, Taramani Campus Madras 600-119 (Indial
(Accepted

of Basic Medical

Sciences,

University

of

June 25, 1990)

Summary
Two water soluble extracts, GS, and GS,, obtained from the leaves of
Gymnema sylvestre,
were tested in streptozotocin
treated rats for their
effects on blood glucose homeostasis and pancreatic endocrine tissue. In the
diabetic rats, fasting blood glucose levels returned to normal after 60 days of
GS, and after 20 days of GS, oral administration. Blood collected during the
conduct of oral glucose tolerance tests was used to assay for serum insulin.
GS, and GS, therapy led to a rise in serum insulin to levels closer to normal
fasting levels. In diabetic rat pancreas, both GS, and GS, were able to double the islet number and beta cell number. This herbal therapy appears to
bring about blood glucose homeostasis through increased serum insulin levels provided by repair/regeneration
of the endocrine pancreas.

Introduction
Gymnema sylvestre R.Br., a herb belonging to the Asclepiadaceae family,
has been employed to control diabetes by traditional medical practitioners of
India for several centuries. The hypoglycaemic effect of this herb was scientifically tested for the first time by Mhasker and Caius (1930). In alloxanobserved
a
induced diabetes in rabbits and dogs, these investigators
lowering of blood glucose levels, while in pancreatectomised
animals the
leaves of G. sylvestre did not have any hypoglycaemic effect. They concluded
that some residual pancreatic endocrine tissue is needed for this herb to be
effective. In a couple of diabetic patients, administration of the dried leaves
for several weeks led to reduction in glucose in the urine collected for 24 h,
Correspondence

to: Prof. K. Radha Shanmugasundaram.

0378.8741/$03.50
0 1990 Elsevier
Published and Printed in Ireland

Scientific

Publishers

Ireland

Ltd.

266

confirming its action. Prolonged administration (24 weeks) of the dried leaves
of the herb to alloxan-diabetic rabbits brought about increased serum insulin
levels both at fasting and after a glucose load (Shanmugasundaram
et al.,
19811.
In alloxan-induced diabetes, the insulin producing cells are damaged and,
to bring about blood glucose and serum insulin homeostasis, either regeneration or repair of the beta cells in the islets of Langerhans would appear to
be necessary. The data presented here provide evidence that G. sylvestre
extracts do increase islets and beta cells in experimentally induced diabetes
in albino rats. Streptozotocin
was used to induce diabetes, rather than
alloxan, since with this agent there is (i) no incidence of spontaneous reversion and (ii1 greater specificity for islets resulting in >90%
of the rats
becoming diabetic at the dosage used.
Two concentrates of the leaf extract were tested. GS, was obtained from
GS, during purification and is a component of the latter.
Materials

and Methods

Drug materials
Collagenase (Type Vl and streptozotocin were obtained from Sigma Chemicals Co., St. Louis, U.S.A. and radioimmunoassay
reagents for insulin assay
was from Bhabha Atomic Research Centre, Bombay, India.
Crushed dried leaves of authentic G. sylvestre
were steeped in 50%
ethanol and steam passed into the mixture for 3- 4 h. The resultant alcoholic extract was filtered before being concentrated under vacuum and acidified to pH 3.0. The residue obtained (GS,, 7% yield w/w relative to starting
material) was then further purified by dissolving in 0.1 N sodium hydroxide
and reprecipitation with acid to give GS, (3% yield w/w relative to starting
material). Both GS, and GS, were tested for their pharmacological effects in
diabetic rats in separate experiments.
Animal stock
Male albino rats (derived from the Wistar strain) weighing 120-150
g
were used. They were obtained from the inbred colonies maintained in the
Kings Institute for Preventive Medicine, Madras. All animals were screened
for detection of abnormalities in blood glucose homeostasis by subjecting
them to an oral glucose tolerance test. Only those animals meeting the following three criteria were used: (i) fasting blood glucose levels below 85 mgl
dl, (ii) 60 min after glucose load, a blood glucose level below 150 mgldl and
(iii) a return to fasting levels in 150 min.
Experimental
diabetes
Diabetes was induced in overnight fasted rats by intravenous (i.v.1 injection of 55 mg/kg streptozotocin
using a 5% solution of freshly prepared
streptozotocin
in 0.1 M citrate buffer (pH 4.51. Control rats received citrate

267

buffer only. Fasting

blood glucose was measured
and glucosuria
was
detected in all the animals seven days after streptozotocin
administration.
All animals were given powdered food (Gold Mohur rat feed, Lipton Co.,
Bombay) and water ad libitum.
Five weeks (35 days) after the injection of citrate buffer (normal) or streptozotocin (diabetic), half of the normal and diabetic animals were administered the extracts, while the other half remained untreated for comparison.
The dosage used in both cases (GS, and GS,) was 20 mgldaylrat. GS, and GS,
in powder form was mixed with 3 g of powdered food and moistened with
water to make a bolus which the rats ate in the mornings. Food trays were
filled only after the drug-mixed rations were fully consumed. The presence
of reducing sugar in the urine was tested using Benedicts solution daily and
fasting blood glucose was estimated every 5 days by the o-toluidine colour
reaction (Cooper and McDaniel, 19701. In the case of GS,-administered
diabetic rats (DGS, group), the fasting blood glucose reached the normal range
after 60 days (95 days after induction of diabetes), while with GS, therapy
(DGS, group) the fasting glucose level was normal on the 20th day and GS,
administration
discontinued forthwith. The rats were maintained for a further period of 40 days on food ad libitum. On the 95th day a glucose tolerance test (GTT) was performed on both treated groups (DGS, and DGS,) and
their respective controls (N, NGS, and NGS,). Group N is the control group,
while NGS, and NGS, are groups of non-diabetic rats under GS, and GS,
therapy, respectively.
Serum insulin levels were assayed in all the blood
samples. Two days later, animals were sacrificed and pancreatic tissue was
harvested.
Glucose tolerance

test and insulin assay

Oral glucose tolerance tests (OGTT) were carried out according to the
method of Cole and Harned (19381 as modified by Wexler and Fischer (1963).
After an overnight fast, blood samples were collected and then 200 mg glucose in 2 ml solution administered
at 30, 60, 120 and 180 min. Blood glucose
levels were estimated by the o-toluidine method of Cooper and McDaniel
(19701. Serum insulin levels were assayed by radioimmunoassay
(Herbert et
al., 19651 in all blood samples collected during OGTT.
Histologic

and morphometric

studies

on the pancreas

Each animal was sacrificed by decapitation and the whole pancreas perfused with formalin and removed immediately together with the spleen. The
total pancreatic weight was recorded. The three regions of pancreas, viz.
duodenal (head), gastric (body) and spleen (tail), were dissected according to
the specifications of Jaffe (19511, cut into smaller fragments and fixed separately in Bouins fluid for 24 h. The segments were dehydrated with ethanol
and embedded in paraffin-wax (56OC). Serial sections (5 pm1 were taken and
stained with chrome-haematoxylin
and phloxin according to Gomori (1941).

268

One hundred serial sections were studied for the number of islets and beta
cell content in each one of the three regions of the pancreas for each rat.
Beta cells stain deep purple and photographs
were taken with green filter to
obtain maximum contrast. Number of rats for which histological studies
were made was 10 in each group.
Morphometric studies were made using stereological procedures according
to Wiebel and Elias (1967) using a reticule. The reticule was mounted along
the focal plane of the eye piece, and consisted of 100 squares. At 400 x magnification, the area covered by the 100 squares was 0.0625 mmz. The area
covered by the section and diameters of the islets and beta cells were measured by counting the number of squares in the reticule occupied by the
structures. With the thickness of the sections at 5 pm, the volume of the section and islets was arrived at to express the percentage endocrine tissue.
In vitro studies

on isolated islets of Langerhans

From one set of normal and experimental

rats, islets were isolated and
insulin release under stimulation with glucose and extract GS, was studied.
The rats were sacrificed by decapitation
and the pancreas was removed
through an incision made in the ventral abdominal duct and placed in 3.0 ml
Hanks solution. The tissue taken from two rats was minced with a pair of
scissors and incubated with (35 - 50 mg) collagenase (Type V) in 5.0 ml Hanks
solution at 37OC for 6- 12 min with constant stirring according to the
method of Lacy and Kostianovsky
(1967) as modified by James and Keith
(1968). Aliquots of the mixture were examined periodically under a dissecting
microscope to determine the point at which optimum separation of islets had
occurred.
The solution was diluted to about 30 ml with Hanks solution and the
supernatant
containing predominantly
acinar tissue was removed. The residual islets were washed eight times with Hanks solution and the sedimentation period was progressively
reduced
from 4 to 1 min. The islets
sedimented under gravity. The washed islet suspension was poured into a
Petri dish surrounded by crushed ice and examined under a dissecting microscope. The islets appeared as free, round or ovoid structures
with greyish
white to brownish red colour.
They were transferred
in groups of 10 to the incubation mixture containing 1.0 ml of ice cold Krebs-Hensleit
buffer (KHB) with 3.3 mM glucose,
capped and gassed with oxygen/carbon
dioxide mix (95:5 v/v) for 3 min and
incubated for 30 min in a Dubnoff metabolic shaker at 37OC. The buffer was
removed and the islets were washed with physiological saline. They were
used in the insulin secretion studies described below.
To study GS,-mediated
release of insulin from the islets, two different
levels of the crude drug (80 mg and 160 mg/lOO ml) were tested with glucose
at 30, 150 and 300 mg/lOO ml at 37OC. In each case the reaction mixture contained 1.0 ml Krebs-Hensleit
buffer containing the test substance and 10
islets. A O.l-ml sample of the fluid was removed at 0, 30 and 60 min and insu-

269

lin content was assayed by radioimmunoassay.

The net insulin released was
corrected for dilution and is expressed here as the amount of immunoreactive insulin (IRI) released per 10 islets. The in vitro studies on isolated islets
were preferred over pancreatic perfusion, so as to assess the efficiency of a
definite number of islets (in releasing insulin) from the normal and experimental animals before and after therapy. However, the possibility of missing
small regenerating
islets during isolation cannot be ruled out.
In a third set of rats (normal, NGS, and DGS, groups), pancreas tissue was
excised at the end of the experimental
period and homogenised with cold
distilled water (4OCl. Amylase and lipase activities were assayed in the homogenate according to the method of Bernfield (1955) and Cherry and Crandale (19321, respectively.
For statistical analysis, the diabetic rats (D group) were compared with
the normals (N group), while those after GS, and GS, therapy were compared with their respective
controls without the herbal therapy, i.e. the
NGS, and NGS, groups were compared with group N, while the DGS,
and DGS, groups were compared with group D. Students t-test was performed.
Results

It can be seen from Table 1 that both GS, and GS, are able to lower fasting blood glucose levels in the diabetic rats (DGS, and DGS, groups). With
GS, administration,
fasting glucose level was controlled at the end of 60 days
(95th day), while almost the same degree of response was observed at the
end of 20 days with GS, therapy. Mean blood glucose remained around 90
mg/dl, even after discontinuing GS, therapy. Pancreatic amylase and lipase
activities lowered in the diabetic group (Dl are brought back to normal levels
after GS, therapy. Glucosuria disappeared
in the DGS, group on 75th day
and in the DGS, group on 55th day.
From Table 2 it can be seen that the diabetic rats after GS, or GS, therapy had a near normal glucose tolerance curve. While the GTT was performed on GS, treated diabetics 40 days after termination
of GS, therapy,
both the fasting and postprandial
blood glucose levels were under control
and insulin levels in serum were near normal.
Such complete glucose homeostasis
in streptozotocin
diabetes suggests
the repair or regeneration
of the insulin secreting beta cells in the islets of
Langerhans. When the herbal concentrates
are administered
to healthy rats
(NGS, and NGSJ insulin secretion is clearly augmented (Table 2).
In control rats (NJ, the islets of Langerhans
were distributed throughout
the three regions of the pancreas, viz. duodenal (head), gastric (body) and
splenic (tail) (Table 31. The islets of the duodenal and gastric regions were
more or less of the same diameter, but in the splenic region the islets were
larger and double the size. In the streptozotocin
diabetic rats, the gastric
region of the pancreas did not show any islets (Table 31 indicating that this is
the major target for streptozotocin.

79
81
76
75
80
685
29.4

15
35
55
75
95
96
96

0.9

* 1.3
f 1.6
-I- 2.0
f 1.6
= 1.8
zk 43
k 1.4

r
190
197
200
216
220
104
13.8

76

-t
f
2
f
2
r
-t

6.8
6.6
5.8
6.4c
6.2
5
1.0

2 0.9
79
80
74
70
68
730
38.5

76

NGS,

+
-c
f
-
2
+-c

k
1.3
1.6
1.5
1.6
0.9
37
2.1

0.9
190
199
150
103
76
641
39.3

76

DGS,

f
f
2
-rk
f
-

6.8
5.5
5.5
3.3
1.5
46
1.1

0.9

79
79
71
72
71

f 1.3
f 1.1
f 1.1
+ 1.5
f 0.9
N.D.
N.D.

76 f 0.9

NGS,

GS, was administered from 36th to 95th days.

bGS, was administered from 36th to 55th day only, although the animals were under observation through the 95th day.
cGlucosuria present.
dAmylase (mIU/mg protein) and lipase (mIU/mg protein) were assayed in pancreatic homogenates.

AmyIas&
Lipased

76

Initial

Blaod
glucose

Number
of days

Test

190 f 6.8
192 -+ 6.8
93 f 2.0
90 f 1.6
92 f 2.0
N.D.
N.D.

76 + 0.9

DGS,

Animals of groups N, NGS, and NGS, formed one group until the 35th day when they were divided. Similarly the diabetic animals (streptozototin injected) were divided into three groups D, DGS, and DGS, on the 35th day. Each tabular value is the mean + S.E.M. obtained in more than
30 rats in each group studied in five batches. ND = not determined.

FASTING BLOOD GLUCOSE (mg/lOO ml) VALUES DURING THE EXPERIMENT

TABLE 1

LEVELS OF NORMAL AND DIABETIC ANIMALS (BEFORE AND AFTER

TESTS CARRIED OUT AT THE END OF THE EXPERIMENTAL
PERIOD

GS, TREATMENT)

Glucose
Insulin
Glucose
Insulin
Glucose
Insulin
Glucose
Insulin

Glucose
Insulin

NGS,

DGS,

DGS,

NGS,

Glucose
Insulin

Group

f
f
f
+
f
+
f
f

1.6
0.6
1.3
1.6
4.1
0.6
2.2
0.5

93
+ 3.5
14.9 + 1.0

68
18.6
71
19.9
217
6.3
77
20.0

77
* 1.3
15.9 + 0.9

Fasting

99
33.5
122
35.2
295
9.4
112
29.0
156
25.2

f
+
+
-c
+
+
f
f
f

137
f
28.6 f

+ 30 min

4.1
0.7
2.9
0.7
5.1
0.7
5.1
0.5
4.1
0.7

1.9
0.8
+
f
f
f
f
f
f
-c

2.9
0.6
2.2
1.2
3.8
0.7
3.8
0.6
133
f 2.9
23.6 -c 3.0

97
25.5
88
26.4
353
9.6
98
27.0

99
2 2.5
21.1 f 0.8

+ 60 min

2
f
f
+
f
f
+
+
123
f
20.2 f

77
23.8
69
20
366
9.6
95
25.2

3.1
0.8

1.3
0.5
1.9
1.3
5.7
0.5
6.0
0.5

83
2 2.2
16.9 2 0.6

+ 120 min

+
f
f
+
f
+
f
+

1.3
0.8
1.9
1.1
5.1
0.5
2.5
0.7
99
f 2.2
19.9 + 0.8

69
19.1
67
16.1
396
9.6
86
22.1

80
+ 1.3
15.3 + 0.7

+ 180 min

The values of glucose and insulin are expressed as mg/lOO ml blood and pIU/ml serum respectively (mean f S.E.M. for 10 animals in each
group). NGS, and NGS, indicate normal healthy animals on GS, and GS, therapy, respectively. DGS, and DGS, indicate diabetic animals after
treatment with GS, and GS,, respectively.

BLOOD SUGAR AND SERUM INSULIN

DURINGORALGLUCOSETOLERANCE

TABLE 2

Significance
Significance

relative
relative

0.07
0.07
0.05
0.06

1.26
0.75
0.99
1.03

NGS,
D
DGS,
DGS,

2
f
2
k

1.12 + 0.15
1.13 5 0.60

f
2
f
f

3.5
2.5++
2.2-e
4.4**

58
65
62
41
53
33

f
f
f
2
+
r

4.4
5.7
5.4
4.4
4.7
2.9
28
36

304 * 29
77 2 12+++
144 k 18**
74k
7

247 f
341 f

cell/

110 f 11.1
120 -r- 6.0
109 k 6.0
0
28 2
1.9
36 2
4.1

of

to Group N: +P< 0.05;

++P< 0.01;
+++P
< 0.001.
to corresponding
untreated
group: *P < 0.05;
**P < 0.01;

55
28
57
44

48 f 4.4
53 2 2.2

(N)

Beta
islet

Number
islets

(pm)

Diameter

Number
islets

of

AND

cell/

113 2 16
101 f 14
116 + 14
0
82 2 10
52 f 13

w)

Beta
islet

***P< 0.001.

52 + 8.0
59 2 4.7
55 2 6.0
0
38 f 4.1
27 f 4.1

(pm)

87
34
73
82

k
f
2
+

5.7
4.1+++
4.7-e
6.6***

85 2 5.7
93 f 6.0

Number
islets

of

138
138
133
68
98
92

(pm)

AND

+ 11.4
2 11.7
-+ 14.2
+ 1.3+++
f
5.4**
-+ 6.6**

DIA-

635
719
778
156
295
272

vv)
f 83
k
69
f 115
2 lo+++
2 42*
2
50*

Beta cell/
islet

for each region.

IN NORMAL

Diameter

in each animal

CELLS/ISLET

Splenic

sections

OF BETA

and 100 serial

NUMBER

Diameter

in each group

(g)

on 10 animals
Gastric

are based

Duodenal

per mm3. Analysis

Pancreas
weight

is expressed

NGS,

Group

The number

MEAN NUMBER
f-(_ S.E.M.) AND SIZE OF ISLETS
OF LANGERHANS
BETIC, WITH AND WITHOUT
GS, AND GS, THERAPY

TABLE

273

GS, and GS, treatment helped to maintain the weight of pancreas, number
of islets and number of beta cells in the diabetic animals (Table 3).
It is interesting to note that with GS, and GS, therapy, the total pancreatic weight increased by about 30% in diabetic animals, while the number
of islets and beta cells were more than doubled. At the same time, the
increases in number of islets and beta cells of the normal animals treated
with GS, and GS, were not statistically significant, suggesting a regulated
response of the pancreatic tissue to stimulation by the G. sylvestre
leaf
extracts.
The streptozotocin-diabetic
animals had damage over the pancreatic acinar
region as can be seen by reduced pancreatic amylase and lipases and these
decreases are corrected after GS, therapy (Table 1).
Figure 1 is a section of the pancreas showing the normal architecture of
the islets of Langerhans with the granulated beta cells appearing dark.
Small and shrunken islets and destruction of beta cells are observed in the
diabetic condition (Fig. 21. Figures 3 and 4 show typical pancreatic sections
after GS, and GS, therapy. Well formed islets and increased granulation of
beta cells are observed in diabetic rats after therapy. The pancreatic sections studied are from the splenic region of the pancreas.
Table 4 and Figs. 5 and 6 describe immunoreactive
insulin (IRI) released
into the medium when islets (obtained from the splenic region of the pancreas) are exposed to varying glucose load and GS, in the incubation
medium. Insulin release from the normal islets increased with glucose con-

Fig. 1. Pancreatic
section from normal rat. The islet of Langerhans
lated dark beta cells. (chrome-haematoxylin
and phloxin stain; x 450).

shows

well-defined

granu-

274

Fig. 2. Diabetic rat pancreas,

are seen ( x 450).

Fig. 3. Pancreatic

( x 450).

showing imperfections.

The islet is atrophied

section of a diabetic rat after GS, therapy.

and hydropic ch,a

The islet is large and organised

275

Fig. 4. Islet from a diabetic

and granulated ( x 1000).

rat after

GS, therapy.

The beta cells appear

dark, well developed

centration (Table 41. With islets obtained from control rats at the severely
hyperglycaemic
state of 300 mg glucoseldl, the maximum insulin released
was 397 pIUnits/lO islets (i.e. 40 $Units/isletl.
By increasing glucose level
from 30 to 300 mg/dl (Table 41, insulin release was increased eight-fold and
this can be taken as the reserve capacity of the pancreas available for secretion. Under similar conditions, diabetic rat islets released a maximum of 100
$Units/lO islets, while in the DGS, and DGS, groups (diabetes controlled by
GS, and GS, therapy, respectively),
glucose mediated insulin release was
nearly 120 $Units/lO islets or 12 PIUnitslislet.
GS, stimulated insulin release from the normal islets to a very significant
degree even with the hypoglycaemic level of 30 mgldl, but in hyperglycaemic
conditions (150 and 300 mgldll, insulin secretory capacity and the maximum
release was about 100 I.tIUnits/lO islets or 10 pIU/islet. In diabetic rats controlled by GS, and GS, therapy (DGS, and DGS, groups), insulin release as a
response to glucose load was marginally higher than the diabetics at 30 min.
When the islets were incubated with glucose at hyperglycaemic
levels (300
mgldll, GS,-stimulated insulin release was 202 and 194 pIUnits/lO islets (or 20
$Units/isletl
after GS, and GS, therapy, respectively.
After therapy, the
diabetic islets appear to have a reserve capacity for insulin release which is
double that of the untreated diabetics, but only one half of the insulin
release observed in the non-diabetic controls. However, at 300 mg/dl glucose,
GS, induced insulin released from the islets appears to be reduced at the end

AND GS ON IN VITRO INSULIN

(DGS, AND DGS,) RATS

RELEASE

FROM

ISLETS

OF NORMAL

(N), UNTREATED

DIABETIC

(D), AND

300

160

N: +P < 0.001.
D: *P < 0.05; **p < 0.01; ***p < 0.001

150

160

to Group
to Group

30

160

relative
relative

30
60

300

Significance
Significance

235 + 16.0
190 2 9.5

30
60
30
60
30
60
30
60

150

1.5
2
+ 3.5
5.0
f
2 17.0
+ 8.5
+ 15.5
2 2.5
2 2.5
2 6.0
+ 8.0

29
55
111
249
226
397
52
89
106
206

30
60

Time
(min)

30

Glucose
tmg/lOO ml)

(mg/lOO ml)

GS,

f
+
f
k
+
-t
2
k

1.5+
3.0+
7.5t
3.5
3.5+
3.5+
2.5*
4.5+
90 2 4.0
85 f 4.5

25
31
79
100
26
27
72
99

19 + 1.0+
22 + 1.0

-t
2
f
2
+
2
2
2
+
2

2.5***
1.5***
3.0**
1.5**
7.0*
3.0*
2.5**
2.0**
2.5***
9.0***
202 *
2.0***
175 + 13.0***

27
31
34
39
87
116
41
39
96
196

DGS,

2 2.0**
2 2.0**
f
3.0***
+ 20.0***

1.5**
1.0**
7.5*
4.5*

1.5***
0.5***

194 k 18.0***
140 -r- 14.5***

35
37
94
136

28 f
31 f
36 f
40 2
92 f
120 +

DGS,

Insulin release (pIU/lO islets) into the medium at the end of 30 and 60 min is given. Each incubation
was preceded
by a 30.min preincubation
with
60 mg glucose/l00
ml. The results report the mean of four observations
of each pair of rats for 5 pairs of animals in each group (mean + S.E.M.).

EFFECT
OF GLUCOSE
CONTROLLED-DIABETIC

TABLE

277

150mg

Glucose

150mq

150mg Glucose

teomg

Glucosetl60mg

GS4

GS,

30 60

30 60

30 60

30

60
Time

30 60 30 60

30

60

30

60

30

60

(min)

Fig. 5. Insulin released in 30 and 60 min when islets from 0, normal; 0, diabetic; and W, DGS,
groups were incubated with glucose at 160 mg/dl with and without GS, at 80 and 160 mg/dl.
Mean insulin ( f S.E.M.) released into the media by 10 islets are represented.

of 60 min, when compared to the hormone in the medium at 30 min, which

may be due to inactivation by GS, or reuptake into the islets.
Discussion

The data presented give evidence for increased insulin secretion and beta
cell number after the administration
of leaf extracts of G. sylvestre, suggesting possible regeneration
or repair of the islets of Langerhans in streptozototin-treated rats.
Experiments
of Loubatieres
(1944, 19461, first established that the probable mode of action of sulphonylureas
was by the release of pancreatic insulin,
which was confirmed by Houssay and Penhos (1956). Unlike the sulphonylureas, GS, does not enhance insulin release in normal rats under normoglycaemic (100 mg/dl) conditions, but enhances hormone release in diabetic
islets. It is here that G. sylvestre leaf extracts assume significance, because
of their capacity to regenerate
partially the damaged endocrine tissue, such
that the insulin content/islet number increases by therapy.
Hitherto, it was believed that the destruction
of beta cells in juvenile,
maturity-onset
and experimentally
induced diabetes was irreversible
and

278

300

mg Glucose

LOO

300 mg Glucose + 80 mg GS4

z
al
5

300

300 mg Glucose + 160mg GS4

0
<
3

TT

200

I%
4
a
.c3
;
H

100

30

60

30

60

30

60

30
Time

60

-30 60

30

60

30

60

30

60

30

60

(min)

Fig. 6. Insulin released in 30 and 60 min when isolated islets from q, normal; Ha, diabetic; and W.
DGS, groups were incubated with glucose at 300 mgldl with and without GS, at 80 and 160 mg/
dl. Mean insulin ( + S.E.M.) released into the medium by 10 islets are represented.

this report is the first giving evidence of a reversal of the damage to the
insulin producing cells by a drug. Mitotic studies are needed to confirm the
appearance of new cells.
The chemical composition of GS,, the active beta cell regenerative portion
of the leaves of G. sylvestre, remains unknown. The leaves have an antisaccharogenic property (suppression of sweet taste for several hours when the
leaves are chewed) that has been recorded repeatedly in the literature (Chopra et al., 1928). Hooper (1889) isolated a crude alcohol-soluble extract having
this antisaccharogenic property and named it as gymnemic acid. Gymnemic
acid (a glycoside) was purified and characterised by Stocklin (1967) as a D
glucuronide of a hexa-hydroxy A12-oleanene.However, this purified isolate
did not show any antidiabetic activity. Mhasker and Caius (1930) analysed
the leaves and reported the presence of 0.95% hentriacontane, in addition to
chlorophylls, phytin, resins and anthraquinone derivatives which give a pur-

279

gative effect to the leaves. None of the isolates were found to have antidiabetic activity, although the whole leaf powder was able to bring down blood
glucose levels not only in experimental diabetes, but also in man (Mhaskar
and Caius, 1930). However, in pancreatectomised animals, G. sylvestre leaves
did not show any hypoglycaemic effect, suggesting that residual pancreatic
function is essential for its antidiabetic activity. When these observations
are viewed with the data presented in this paper, it may be concluded that
the extract of G. sylvestre may bring down blood glucose levels by regeneration of the pancreatic islets and beta cells.
References
Bernfield, P. (1955) In: S.P. Colowich and N.P. Kaplan (Eds.1 Methods in Enzymology
1, 149- 165.
165.
Cherry, I.S. and Crandale, I.A. (19321 The specificity of pancreatic lipase: Its appearance in the
blood after pancreatic injury. American Journal of Physiology
100,266-273.
Chopra, R.N., Bose, J.P. and Chatterjee, N.R. (19281 Indigenous drugs of India. Indian Journal of
Medical

Research

16,115-

124.

Cole, V.V. and Harned, B.K. (19381 Diabetic traits in a strain of rats. Endocrinology
23, 318-326.
Cooper, G.R. and McDaniel, V. (19701 In: R.P. MacDonald (Ed.1 Standard Methods of Clinical
Chemistry,
Vol. 6, Academic Press, London, p. 159.
Gomori, G. (19411 Observations with differential stains on human islets of Langerhans. American
Journal

of Pathology

17, 395-406.

Herbert, V., Law, K.S. Gotilieb, C.N. and Bleicher, S.J. (19651 Coated charcoal immunoassay of
insulin. Journal of Clinical Endocrinology
25, 1375- 1384.
Hooper, D. (18891 Gymnema sylvestre,
R. Br., Chemical News 59, 159 and Chemical Zentrablatt
I, 632.
Houssay, B.A and Penhos, J.C. (19561 Action of the hypoglycaemic sulphonyl compounds in hypophysectomised, adrenalectomised
and pancreatectomised
animals. Metabolism 5, 727-738.
Jaffe, F.A. (19511 A quantitative study of the islets of Langerhans in the rabbit. Anatomic Recurie 111,109-121.
James, E.V. and Keith, D.B. (19681 Effect of glucose concentration on insulin and glucagon from
isolated islets of Langerhans of rats. Diabetes 17, 187- 193.
Lacy, P.E. and Kostianovsky. M. (19671 Method for the isolation of intact islets from the rat pancreas. Diabetes 16,35-40.
Loubatieres, A. (19441 Analyse de mechanisma delaction hypoglycaemisule due p-aminobenzene
sulphamidothiodiazol(2254
R.P.1, Comptes Rendus des Seances de la Societe de Biologie et de
ses Filiales

(Paris) 138,766

- 778.

Loubatieres, A. (19461 Etude physiologique et pharmacodynamique

de certains derives fulfamides hypoglycaemiant. Archives Internationales
de Physiologie
54, 174-184.
Mhaskar, K.S. and Caius J.F. (19301 A study of Indian medicinal plants. II: Gymnema sylvestre
R. Br. Indian Medical Research Memoirs 16, 2- 75.
Shanmugasundaram,
K.R. Panneerselvam,
C., Samudram, P. and Shanmugasundaram,
E.R.B.
(19811 Insulinotropic activity of Gymnema sylvestre
R.Br., an Indian medicinal herb used in
controlling diabetes mellitus. Pharmacological
Research
Communications
13, 475-485.
Stocklin, W., Weiss, E. and Raichstein, T. (19671 Gymnema Saure, das antizaccharin Prinzip von
G. sylvestre,
R. Br. Helvetica Chimica Acta 50, 474-490.
Weibel, E.R. and Elias, H. (19671 General Stereologic
Principles.
Quantitative
Methods in Morphometry,
Springer Verlag, Berlin.
Wexler, B.C. and Fischer, C.T. (19631 Abnormal glucose tolerance in repeatedly bred rats and
with atherosclerosis. Nature 200, 133- 136.