Effects of long-term restricted insulin production in obese-hyperglycemic (genotype ob/ob) mice
C.N. Boozer 0
J. Mayer 0
0 Dept. of Nutrition, Harvard School of Public Health , Boston, Massachusetts , U. S.A
Summary. Primary hypersecretion of insulin has been suggested as one possibility for the genetic fault of ob/ob mice. To test this hypothesis, streptozotocin (SZO) was used to reduce permanently insulin secretion in young lean and obese mice. After establishment of hyperglycaemia and weight reduction in treated obese mice (obese-SZO), daily insulin replacement was begun in some (obese-SZO-Ins). Obese-SZO mice maintained insulin levels and body weights similar to lean controls, though they were shorter and fatter, while food intake and blood sugar levels exceeded lean values. Obese-SZO-Ins mice with reduced islet hyperplasia, but great insulin resistance, gained more weight than obese-SZO mice; had high serum insulin and controlled blood glucose; and exhibited hyperphagia. These results suggest that primary hypersecretion of insulin cannot be the genetic defect, as ob/ob mice are hyperphagic, hyperglycaemic, insulin resistant, and "obese" even when insulin levels are restricted.
Mutation ob/ob; obese mouse; streptozotocin; insulin resistance; hyperphagia; spontaneous diabetes; diabetes in mice; obesity in mice; hereditary obesity; obese-hyperglycaemic mice; B-cells
The obese-hyperglycaemic mutation in mice (ob/ob)
has been the subject of a great number of studies
because of its intriguing inherited syndrome
characterized by hyperglycaemia, hyperinsulinaemia,
hyperphagia, obesity and insulin resistance. Yet the
fundamental question of the nature of the primary genetic
fault is still unanswered.
From the earliest descriptions of this mutation,
insulin and the pancreas have been of great interest.
Resistance to massive injections of insulin was shown
in these mice in 1951 [
] and greatly enlarged islets of
Langerhans were described in 1952 [
]. In 1955,
increased extractable insulin from the pancreas was
] and elevated circulating levels of insulin
have been shown many times by biossay or
radioimmunoassay procedures [
4, 5, 6, 7
]. In more recent
years, several authors have suggested that
hypersecretion of insulin could be the primary pathogenetic
mechanism leading to compensatory responses which
result in the abnormalities characteristic of these
8, 9, 10
The experiment reported here was designed to test
this hypothesis and to study the long-term effects of
reduced endogenous insulin secretion in ob/ob mice.
To do this, a dose of streptozotocin, a drug specifically
cytotoxic to pancreatic B-Cells, was found that would
destroy sufficient B-cells in young ob/0b mice to result
in insulin levels equivalent to those of lean animals.
The consequence of controlled insulin levels was
studied over a sixteen week period for effect on: blood
glucose, glucosuria, obesity, resistance to exogenous
insulin, food intake, body weight and body length. The
permanence of streptozotocin damage to B-cells was
determined by microscopic examination of the
Materials and Methods
Male, weanling, ob/ob mice (strain C57 BL/6J-ob)
and their lean littermates (ob/+ or + / + ) were
purchased from the Jackson Laboratory, Bar Harbor,
Maine. They were housed individually in suspended
stainless steel cages and fed ad libitum Purina Mouse
Chow (Ralston Purina Co., St Louis, Missouri).
Streptozotocin, provided by the U p j o h n Company
(Kalamazoo, Michigan), was dissolved in saline (10
mg/ml) and adjusted to pH 4.0 with HCI immediately
before injection. Injections were given
intraperitoneally at 9 : 0 0 A.M. in varying doses (125, 150, 175, or
200 mg/kg body weight) to mice that had been fasted
overnight. Control mice were injected with equivalent
volumes of saline. All mice were between 4 and 5
weeks of age at the time of injection.
Insulin replacement was initiated in some obese
mice two weeks after streptozotocin treatment. N P H
insulin, diluted in sterile saline to an appropriate
concentration, was injected subcutaneously once per day
at 4:00 P.M. Volumes injected were about 0.5 ml.
Daily injections were continued in these mice in
increasing doses for the remaining 14 weeks of the
experiment. Initial doses were 2 U / 1 0 0 g body weight.
This was increased to a maximum of 12 U/100 g in one
group and 30 U/100 g in another.
Daily subcutaneous injections of insulin were also
given to 14 lean mice which had not been previously
injected with streptozotocin. These mice were either 5
or 7 weeks of age at first injection.
Blood Glucose and Serum Insulin
Blood samples were taken from the sub-orbital sinus
on the morning prior to streptozotocin injection and at
biweekly intervals thereafter until the conclusion of
the experiment 16 weeks later. These samples (about
150 2 of blood) were collected in small
nonheparinized capillary tubes which were spun in a
hematocrit centrifuge to separate serum, which was
subsequently frozen in small plastic tubes. Serum
samples were later analyzed for glucose by a
micromodification of the glucose oxidase technique [
and for insulin by radioimmunoassay (Phadebas
Insulin Test), using porcine insulin as a standard.
Body Weights, Urinary Glucose and Food Intake
Body weights and tests for urinary glucose (Clinistix)
were recorded weekly. Quantitative determinations
of urinary glucose were made at about 14 weeks past
injection by using metabolism cages which allowed
separation and collection of 24 hours urine samples.
Urine was collected into a solution of 0.5% sodium
benzoate and 0.5% sodium fluoride, with 2 drops of
toluene, and was analyzed for glucose by the glucose
Food intake was measured at 5 different intervals
of 3 to 6 days each. Amount of food consumed was
ascertained by difference between the weights of the
; 3 o
SZO // i"
20 INJ. /"
/ / /
/ / "
/ / /
/ /" i T r OBESE(5] /
TE OBESESZO-INS~.. t
. / ~
. ~i/ / . . ."r L.EAN.(5)
1N'-0BESE- SZO-INS (HIGH) (.3/~/'
~ E C T I O N S
WEEKS PAST STREPTOZOTOCIN INJECTION
ly lighter than their lean controls, with a final average
weight of 21 g.
Values for serum insulin are plotted in Figure 2.
Typical values for lean mice were about 20 ~U/ml, and
remained fairly constant during the experiment.
Values for untreated obese mice rose from 51 to 289 FtU,
but fell again as the mice got older. Insulin levels of the
obese-SZO mice were significantly reduced below
obese controls and were similar to untreated lean mice
throughout the experiment. Insulin values of
obeseSZO mice given maximum insulin injections increased
ground chow plus food cup on different days at the
same hour, corrected for food spilled. All weights
were taken to the nearest tenth of a gram.
Sixteen weeks after streptozotocin injection, the
experiment was terminated. The mice were
anesthetized with a CO2:O 2mixture (50:50) and measured
for length before sacrifice by decapitation. Pancreata
were removed and fixed in Bouin's solution.
Paraffinembedded sections were stained with H & E and
Statistical analyses were carried out using a
onesided Student's t-test.
The doses of streptozotocin found to have a diabetic
effect in obese mice, as tested by glucose spillage in the
urine, were 175 and 200 mg/kg body weight, with the
higher doses giving consistent and severe effects. Of
the 13 obese mice injected at these doses, in only one
did the drug seem ineffective. Few of the lean mice
were injected with streptozotocin, but loss of weight
and glucosuria did result from injections of 150 or
175 mg/kg body weight, with poor survival at both
Body weights during the course of the experiment are
shown in Figure 1. All mice showed an initial loss of
weight due to the overnight fast prior to
streptozotocin or saline injection. Thereafter, the lean and obese
controls displayed typical growth patterns with the
obese gaining rapidly over the lean and reaching
average weights of 47 and 27 g, respectively, at 16 weeks
past injection (20 weeks of age).
It can be seen that streptozotocin had an
immediate effect in restricting weight gain of obese mice
to weights below even those of lean controls. After 7
weeks, the obese-streptozotocin (obese-SZO) mice
had reached the weight of the lean controls and
thereafter weights for the two groups were similar. Even at
16 weeks past injection these groups weighed 30 and
27 g, respectively, and were not significantly different.
At two weeks past injection, insulin replacement
by daily subcutaneous injection was begun in some of
the obese-SZO mice. Initially 2 U / 1 0 0 g body weight
were injected, and this dosage was gradually increased
to 12 U/100 g body weight and in 3 of the mice to 30
U/100 g body weight. Insulin injections did result in
greater weight gain in SZO-mice, but, at the end of the
experiment, they still weighed significantly less than
Two lean-SZO mice (not shown) were
consistent[4-12LU I 0
6 7 8 910 12 15 20 30
DOSE OF INSULIN INJECTED (U/IO0 groins body weight}
Pig. 3. Comparison of survival rates of lean and
obese-streptozotocin mice at various doses of exogenous insulin. Note that the amount
of insulin injected is presented on a logarithmic scale
) " ~ t'~,,,~
/" .,-~'~c ~
- lrr OBESE-SZO(51
,'-z 3'-4 5'-6 T'8 9'-,o ,'-12 ,3'-,4 ,;
WEEKS PAST STREPTOZOTOCIN INJECTION
Fig. 4. Changes in serum glucose values of lean and ob/ob control
mice and obese mice injected with streptozotocin, followed by
insulin replacement (obese-SZO-Ins), or no insulin replacement
(obese-SZO). Values are means _+ standard error of the mean
steeply and reached the remarkable level of 1777
~tU/ml serum by the end of the experiment.
ObeseSZO mice injected with lower doses of insulin reached
an average of 516 ~tU insutin/ml serum, while values
for lean-SZO mice (not shown) average about 10
~tU/ml, only half the average for lean controls.
All of the 14 lean mice given daily injections of
insulin died, most within a few days of the first
injection (Fig. 3). Maximum dosage of insulin that could be
given before resulting in death was 5 U / 1 0 0 g body
weight, with many of these lean mice unable to
tolerate even 2 U/100 g body weight.
In Fig. 4, it can be seen that serum glucose values were
higher for obese than lean controls, with the
difference increasing during the experiment to 327 and 215
mg/100 ml serum respectively at 16 weeks. As
expected, but in contrast to results reported by
Batchelor et al. [
], streptozotocin injection of obese
animals resulted in greatly elevated glucose values
(above 500 mg/100 ml serum). These values
remained high in the group not receiving insulin
replacement (453 mg/100 ml at 16 weeks). The very high
levels of insulin which were present in the
obese-SZOI n s group were effective in lowering the glucose levels
of this group to 260 mg/100 ml at 16 weeks past
streptozotocin injection. Lean-SZO mice had an
average of 495 mg/100 rnl serum at 16 weeks past
injection of streptozotocin (not shown).
Average food intake data from five separate periods
of measurement are shown in Fig. 5. Lean, control
mice ate significantly less food than lean-SZO mice
and all groups of obese mice (p _z 0.005). There were
no significant differences between the groups of obese
mice, or between the lean-SZO mice and any obese
Table 1 shows the results obtained from glucose
analysis of 24 hour urine samples. As expected, there
was almost no glucose in the urine of untreated lean
mice. Obese mice, however, spilled a significant
amount of glucose even with no treatment.
Streptozotocin injection resulted in greater urinary toss in
both lean and obese mice, while insulin-injected-SZO
mice spilled less than obese controls. At this time (14
weeks past streptozotocin injection) serum insulin
levels were quite high in these mice.
Body lengths at the end of the experiment are shown
in Table 2. Setting the lean control mice at 100%, it
can be seen that obese controls are the same length
whereas all mice treated with streptozotocin were
stunted. Insulin replacement improved body length
I II 171" TV" "V"
n = L(E5A)N LS(EZ2A)ON OB(5E)SE OSB(Z3EO)SE OSBINZEOSSE
from 91% of lean in obese-SZO, to 96%. It is of
interest also that in the few lean-SZO mice surviving,
body length was not as severely reduced as in the
Weights of epididymal fat pads are presented in Fig. 6.
The weights of fat pads of all obese mice were greater
than those of lean mice, even though the body weights
of obese-SZO mice were not significantly greater than
lean. Streptozotocin injected mice had smaller fat
pads than their controls, whether obese or lean, and
insulin injections in SZO-mice did not increase these
Because obese-SZO mice weighed less than obese
controls and obese-SZO-Ins mice, their fat pads
represented a greater proportion of total body weight
(Table 3), 6.6%, compared to 5.9% for obese, 5.8%
for obese-SZO-Ins, and 2.1% for lean. The lean-SZO
mice, however, did not preserve their relative
adiposity which fell to less than 0.5%.
Pancreatic Islet Morphology
Islets of untreated, non-obese mice were small,
welldefined and had well granulated B-cells when stained
with aldehyde fuchsin. The islets varied in the number
of apparent A-cells, but these were always located at
the periphery of the islet, with B-cells in the interior.
Islets of lean streptozotocin-diabetic mice were, in
general, smaller in size than their controls, and
contained markedly degranulated B-cells. Many islets
also showed an admixture of A- and B-cells in the islet
Obese untreated mice had islets that were
characteristically very large, with degranulated B-cells. Islets
from obese mice treated with streptozotocin were still
larger than lean controls, but smaller than those from
untreated obese mice, due to the destruction o f
B-cells. Like the islets from lean, treated mice, these
were characterized by the presence of A-cells in the
islet interior. Some islets from treated, obese mice had
less defined configurations with acinar tissue intruding
into the islets.
Some of these effects of streptozotocin on the islet
are similar to those reported by others [
adult obese mice treated with this drug. The presence
of A-cells scattered within the islet interior has also
been reported by Like and Chick  for diabetic
mutant mice in the terminal stage of the syndrome.
In this experiment, a dose of streptozotocin was found
which was effective in destroying sufficient B-cells in
weanling ob/ob mice to result in insulin levels
equivalent to those of lean mice. This effect was permanent,
due to insufficient regeneration of B-cells following
destruction by this drug.
Other investigators, using streptozotocin in older
animals, have reported islet regeneration after initial
B-cell necrosis [
]. The Birmingham ob/ob mice
used in their study were 3 to 4 months old at
treatment. Coleman [
] also found only transitory effects
with streptozotocin in older ob/ob mice. The
difference in permanence of effect may depend on the age at
treatment and consequently the number of B-cells
initially present. The background strain of the mice
could also influence regeneration capability.
In this experiment, lean treated mice became
diabetic, as their insulin levels averaged less than half
normal values, and persistent hyperglycaemia and
glucosuria resulted. The hyperphagia s e e n in these
mice was undoubtedly a secondary result of the
induced insulin insufficiency.
Obese mice treated with streptozotocin had
insulin values that remained much lower than those of
obese controls, yet were not lower than those of lean
controls. Therefore, the hyperglycaemia and
glucosuria seen in these mice is not a result of low
levels of insulin, but rather of hyperphagia. That obese
mice continue to overeat whether their insulin levels
are high or low (comparable to lean mice), makes it
apparent that their hyperphagia cannot be the result
of hyperinsulinaemia. Rather, the hypersecretion of
insulin normally seen in these obese mice must be an
effect that is secondary to the increased food
consumption and/or insulin resistance.
Marked insulin resistance is seen in these
obesestreptozotocin mice without prior hyperinsulinaemia
and without increased body weight. While lean,
untreated mice could withstand no more than 5 U of
exogenous insulin, obese-SZO mice were able to
tolerate repeated injections of 30 U. It is not known what
would be the maximum level these mice could survive,
but clearly it is at least a factor of six times the lethal
dose in normal lean mice.
Mahler and Szabo [
] have reported that
insulin sensitivity in the obese mouse can be restored
by suppression of pancreatic islet cell hyperplasia.
Their studies involved injection of alloxan in adult
ob/ob mice after which they reported a reduction of
immunoreactive insulin levels, but no increase in
blood sugar, and no reduction in body weight. Perhaps
this discrepancy between their results and those
reported here is again due to the age of the mice treated,
and to the fact that the older mice had already become
obese and hyperinsulinaemic by the time of their
study. Certainly in these younger treated mice the
ability to tolerate huge amounts of injected insulin is
evidence for persisting abnormal resistance to the
hormone, even after islet hyperplasia has been
While it is possible that this extreme resistance to
exogenous insulin could be explained by greater
development of insulin-binding antibodies in the ob/ob
mice than in o b / + or + / + mice, this seems unlikely.
More probable would be the explanation of insulin
resistance due to the relative obesity of these mice.
It is impressive that the epididymal fat pads of
these mice weighed almost four times as much as did
those of lean mice, though body weights and insulin
levels were not significantly different. That these mice
were shorter, yet fatter, implies that adipose tissue is
C. N. Boozer and J. Mayer: Effects of Long-Term Restricted InsulinProduction
more sensitive to insulin than is lean body mass.
Stauffacher and Renold [
] have shown a relative
resistance to insulin in ob/ob mice of muscle over adipose
tissue when they compared glucose incorporation into
adipose tissue lipids, following intraperitoneal
injection of insulin and glucose 14C.
The results of this long-term study clearly establish
that hyperinsulinaemia is not necessary for insulin
resistance or "obesity" in ob/ob mice. These
conditions are most probably the result of early and
persisting hyperphagia. The drive to overeat could lead to
increased nursing and perhaps earlier and greater
consumption of solid food in pre-weanling ob/ob mice.
Overeating would result in greater deposition of fat
and consequently resistance to endogenous insulin
and hyperglycaemia and hyperinsulinaemia. Such a
sequence is consistent with experiments suggesting
that fat stores are increased in obese mice before
circulating insulin levels are elevated [
1. Mayer , J. , Bates , M.W. , Dickie , M.M.: Hereditary diabetes in genetically obese mice . Science 113 , 746 - 747 ( 1951 )
2. Bleisch , V.R. , Mayer , J. , Dickie , M. M.: Familial diabetes mellitus and insulin resistance associated with hyperplasia of the islands of Langerhans in mice . Amer. J. Path . 28 , 369 - 385 ( 1952 )
3. Wrenshall , G.A. , Andrus , S.B. , Mayer , J. : High levels of pancreatic insulin coexistent with hyperplasia and degranulation of beta-cells in mice with the hereditary obese-hyperglycemicsyndrome . Endocrinology 56 , 335 - 340 ( 1955 )
4. Christophe , J. , Dagenais , Y. , Mayer , J.: Increased circulating insulin-like activity in obese-hyperglycemic mice . Nature (Lond.) 184 , 61 - 62 ( 1959 )
5. Stauffacher , W. , Lambert , E. , Vecchio , D. , Renold , A.E. : Measurements of insulin activities in pancreas and serum of mice with spontaneous ("obese" and "New Zealand obese") and induced (goldthioglucose) obesity and hyperglycemia, with consideration on the pathogenesis of the spontaneous syndrome . Diabetologia 3 , 230 - 237 ( 1967 )
6. Genuth , S.: Hyperinsulinism in mice with genetically determined obesity . Endocrinology 84 , 386 - 391 ( 1969 )
7. Abraham , R.R. , Beloff-Chain , A. : Hormonal control of intermediary metabolism in obese hyperglycemic mice. I. The sensitivity and response to insulin in adipose tissue and muscle in vitro . Diabetes 20 , 522 - 534 ( 1971 )
8. Bray , G.A. , York, D.A.: Genetically transmitted obesity in rodents , Physiol. Rev . 51 , 598 - 646 ( 1971 )
9. Stauffacher , W. , Orci , L. , Cameron , D.P. , Burr , I. M. , Renold , A.E. : Spontaneous hyperglycemia and/or obesity in laboratory rodents: An example of the possible usefulness of animal disease models with both genetic and environmental components . Recent Progr. Hormone Res . 27 , 41 - 95 ( 1971 )
10. Stern , J.S. , Hirsch , J.: Obesity and pancreatic function . In: Handbook of Physiology, Section 7 (eds. R. O. Greep , E. B. Aspwood), pp. 644 - 651 . Washington: American Physiological Society 1972
11. Fales , F.W. : Standard methods of clinicalchemistry , Vol. 4 (ed. D. Seligson), p. 101 . New York: Academic Press 1963
12. Batchelor , B.R. , Stern , J.S. , Johnson, P. R. , Mahler , R.J.: Elfects of streptozotocin on glucose metabolism, insulin response and adiposity in ob/ob mice . Metabolism 24 , 77 - 91 ( 1975 )
13. Brosky , G. , Logothetopoulos , J.: Streptozotocin diabetes in the mouse and guinea pig . Diabetes 18 , 606 - 611 ( 1969 )
14. Findlay , J. A. , Rookledge , K. A. , Beloff-Chain , A. , Lever , J. D.: A combined biochemical and histological study on the islets of Langerhans in the genetically obese hyperglycemic mouse and in the lean mouse, including observations on the effects of streptozotocin treatment . J. Endocr . 56 , 571 - 583 ( 1973 )
15. Like , A. A. , Chick , W. L. : Studies in the diabetic mutant mouse: 2. Light microscopy and radioautography of pancreatic islets . Diabetologia 6 , 207 - 215 ( 1970 )
16. Coleman , D.L. : Personal communication ( 1974 )
17. Mahler , R.J. , Szabo , O. : Restoration of insulin sensitivity in the obese mouse following suppression of pancreatic islet cell hyperplasia . Israel J. med. Sci. , 8 , 810 - 811 ( 1972 )
18. Mahler , R.J. , Szabo , O. : Amelioration of insulin resistance in obese mice . Amer. J. Physiol . 221 , 980 - 983 ( 1971 )
19. Stauffacher , W. , Renold , A.E. : Effect of insulin in vivo on diaphragm and adipose tissue of obese mice . Amer. J. Physiol 216 , 98 - 105 ( 1969 )
20. Chlouverakis , E. , Dade , F. , Batt , R.A.L. : Glucose tolerance and time sequence of adiposity, hyperinsulinemia and hyperglycaemia in obese-hyperglycemic mice (ob/ob) . Metabolism 19 , 687 - 693 ( 1970 )
Received: October 1O , 1975 , and in revised form: February 13 , 1976 C.N. Boozer , D. Sc . 5N Hibben Apartments Faculty Road
Princeton , N.J. 08540 USA