BRITISH MEDICAL JOURNAL
LONDON SATURDAY AUGUST 15 1953
INSULIN AS A GROWTH HORMONE
JAMES SALTER, M.A.
Research Assistant, Department of Physiology, University of Toronto, Toronto, Canada
CHARLES H. BEST, M.D., F.R.S.
Director, Department of Physiology and Banting and Best Department of Medical Research,
University of Toronto, Toronto, Canada
[WITH SPECIAL PLATE]
The effect of insulin on the growth of animals deprived
of the pituitary gland has not previously been reported.*
The well-known obstacle has been the great sensitivity
of the hypophysectomized animal to the antidiabetic
attractive to the animal and constantly available, we
have been able to stimulate rapid growth in completely
increase in all the main body constituents and a very
definite augmentation of skeletal growth.
used, the details of the results, and their potential signifi-
cance in physiology and medicine will be presented and
The hypophysectomized rats were secured from the Hor-
mone Assay Company.
Completeness of the operation was
confirmed at necropsy and no anterior pituitary tissue has
thus far been found.
One daily injection of protamine zinc
insulin (Toronto), 40 units per ml., has been administered
The first lot may have contained a little
detected in subsequent ones.
was as follows:
commercial dog food containing horse
meat 892 g., white bread crumbs 345 g., whole wheat flour
396 g., glucose 80 g., and
Animals in all groups ate the same diet ad libitum.
food intake and the weight of each
The composition of the diet
1 quart of whole milk.
rat were measured
*Preliminary notes on this work have appeared (Salter and
Best, 1952, 1953).
returned to the carcass, which, minus the contents of the
gastro-intestinal tract, was finely minced and then homo-
genized in a blender.Representative aliquots were thus
available for the determinations of protein, fat, and water,
which were carried out by standard and regularly checked
methods. The tibia test of Greenspan et al. (1949) was per-
formed to evaluate skeletal growth.
ment two tibiae were removed from each
was placed in neutral formalin for the above-mentioned
test, and the second was fixed in formol calcium.
sections were prepared and stained with H. and E. and
Mallory's connective-tissue stains.
Organs were removed at necropsy, weighed, and
In our second experi-
Experimental Procedures and Results
hours, a " pilot " experiment using two rats was performed.
It was found that small amounts of protamine zinc insulin
were well tolerated by the hypophysectomized animals, and
that when a suitable diet was made available at all times
constant attention became unnecessary.
reached in this experiment one animal succumbed to hypo-
glycaemia, but the other grew rapidly, nearly doubling its
weight in 20 days, and it served as a model for the many
larger series of animals which have subsequently been
In this rat it was found that the tail had lengthened
the thymus, kidney, and heart weights were within the
The liver was heavier than normal.
testes, adrenals, and thyroid were atrophic.
Experiment A.-In this experiment 40 male rats, 14 days
after hypophysectomy, were divided into three groups of
at many points in the 24
Before this goal was
I cth. during the 20 days of insulin therapy. At necropsy
TABLE I.-Organ Weights of Hypophysectomized Rats
Hypophysectomized controls killed
at start of experiment
Hypophysectomized controls; no
treatment 16 days
Hypophysectomized rats receiving
protamine zinc insulin 15 days
12 60 011
3 79 3-29
INSULIN AS A GROWTH HORMONE
TABLE II.-Carcass Composition of Hypophysectomized Rats
Hypophysectomized control skilled at
start of experiment..
treatment 16 days
protamine zinc insulin 1Sdays
10, 18, and 12. The 10 animals were sacrificed on the initial
day to secure basal values.
group received daily for 15 days a dose of protamine zinc
insulin, which was increased whenever the previous dose
failed to produce a definite increase in weight.
animals in group
3 received no insulin and were sacri-
ficed at the same time as those in group 2.
secured on the surviving animals
Tables I and II, and the growth curves are plotted in Fig. 1.
Each member of the second
Experiment B.-In this experiment 40 male rats, 14 days
after hypophysectomy, were divided into groups of 20, 10,
Each rat in group
one daily dose of protamine zinc insulin.
group 2 received a daily subcutaneous dose of growth hor-
mone, which was regulated to produce approximately the
same rate of growth as exhibited by the insulin-treated ani-
Group 3 served as a control and received the same
diet ad libitum as the other two groups.
were sacrificed on the 15th day of the experiment.
analytical findings in the surviving animals are presented in
Tables III and IV and the growth curves in Fig. 2.
FIG. 1.-Weight changes and food intake of hypophysectomized
,controls. ------,receiving protamine zinc insulin.
In Table I the figures demonstrate that insulin has pro-
duced an absolute increase in the weights of thymus, kidney,
heart, and liver.
gain in weight of the insulin-treated animals was due to
an absolute increase in the main body constituents.
results of the tibia test showed that very definite skeletal
greater extent than protein and water. There was, however,
a definite increase in protein.
reduction in total protein in the control animals, but this
may not be significant.
The curves in Fig.
II the figures show that the
There appears to be a*slight
1 received, as in Experiment A,
Each rat in
All the animals
The findings confirm those of Experiment A and indicate
that under the conditions of this study the insulin-treated
hypophysectomized rats stored more fat and less water than
similar ones treated with G.H. The animals receiving P.Z.I.
consumed more food (including protein), but their total
body protein was slightly less than that of the animals
treated with G.H. The insulin-treated rats retained a smaller
proportion of the ingested nitrogen.
gain and final weight were the same in both experimental
Insulin appeared to produce
weight of kidneys, heart, liver, and thymus.
mone exerted any definite effect on the adrenal, thyroid, or
Experiment C.-In this experiment 53 hypophysectomized
rats were divided into groups 1, 2, and 3, consisting respec-
tively of 14 controls, 14 treated with growth hormone, and
25 treated with protamine zinc insulin.
lowed was identical with that described in Experiment B.
During the experimental period 8 of the insulin-treated
All rats in groups
increase in weight of both insulin- and growth-hormone-
treated animals was 42 g. in the 15 days.
both materials was approximately the same as in Experiment
B (see Fig. 2).
The gain in protein was 2.6 g. per animal
in the insulin-treated and 4.0 in the G.H.-treated group. The
increase, in width of the epiphysial disk was 75
insulin-treated group and 154 ,u in the G.H.-treated.
normal rat weighing the same as the insulin-treated animal
the epiphysial disk
was approximately the same width.
The growth hormone increased the width of the disk to
almost twice that of a normal intact control rat of the
Fig. 3 (Special Plate) shows sections of the
epiphysialdisks of treated and untreatedhypophysectomized
The average weight
a greater gain in
The procedure fol-
1 and 2 survived.
The dosage of
/A in the
JAMES SALTER AND CHARLES
LEGENDS TO PLATE
zones of the epiphysial disk
trabecular bone is very coarse and the bone marrow
is very fatty in both epiphysis and metaphysis.
section illustrates the characteristic effects of hypophys-
ectomy on the tibia of a young rat.
Professor W. S. Hartroft chose
The histological picture approximates that of a
normal intact rat of the same weight (see Fig. C). The
the epiphysial and metaphysial
bones is coarse and irregular, with pronounced growth
indicated by irregular
marrow of the epiphysis and metaphysis
than that of the control animals.
through the epiphysial disk of a normal intact male
rat weighing the same
as the average weight of the
The organization and arrangement within the disk are
very orderly and no irregularities are apparent.
is less fatty
*Maximal amounts of growth hormone were not administered.
Somewhat larger doses of insulin might also have been given.
In this experiment the weight gains of the two groups were
matched by design.
AUG. 15, 1953
AUG. 15, 1953
J. M. SALTER AND C. H. BEST: INSULIN AS A GROWTH HORMONE
INSULIN AS A GROWTH HORMONE
Organ Weights of Hypophysectomized Rats
Hypophysectomized controls; no
treatment 15 days
Hypophysectomized rats; growth
hormone injected 14 days
amine zinc insulininjected14
10 13 7
10 15 7
0 33 0-23
TABLE IV.-Carcass Composition of Hypophysectoinized Rats
treatment 15 days .10
Hypophysectomized rats; growth hor-
mone injected 15 days
Hypophysectomized rats; protamine
zincinsulininjected 15 days
FIG. 2.-Weight changes and food intake of hypophys-
-, receiving protamine zinc insulin.
- - - - -,
It is well established that insulin plays an important part
processes, and that it reverses the series of
catabolic reactions which are collectively known as diabetes
Insulin causes resumption of growth in the de-
It increases the rate of formation of
glycogen and fat and of the incorporation of amino-acids
The only catabolic effect of physiological
amounts is to increase the combustion of sugar.
physiological conditions insulin works synergistically with
the somatotropic factor of the anterior pituitary and with
other hormones which in the complete absence of insulin
exert predominantly catabolic effects.
When the pituitary gland
Evans and Long,
at approximately the same weight
The growth rate is zero.
But in many species when the
source of insulin
loss in weight.
The growth rate is negative.
the pituitary from these depancreatized animals may restore
the growth rate to zero.
The hypophysectomized animal may develop fatal hypo-
glycaemia during even a short period of fasting, without
receiving any exogenous insulin, and this is probably the
is removed growth may be
is completely removed there is a rapid
principal reason why experiments such as we
are reporting have not previously been car-
We have used only one daily
dose of P.Z.I.
It is quite probable that a
reveal even more dramatic effects of insulin
The well-established anabolic effects of
somatotropin in the normal or hypophysec-
tomized aninmal do not require comment
The protein retention and skeletal
growth are both much greater than that
which we have produced with insulin alone.
insulin and the pituitary growth hormone
insulin produces no nitrogen retention or
skeletal growth (Milman et al., 1951).
experiments provide no further evidence on
this latter point or on the debatable question
of the liberation of insulin from the islets by
the indirect or direct action of somatotropin.
which produces a large increase in the volume of the pan-
creatic islets, causes a more effective liberation of insulin
than we have accomplished with one daily dose of P.Z.I.,
but there is no direct evidence as yet that somatotropin
causes any liberation of insulin.
It has been reported that somatotropin may cause libera-
the pancreatic hyperglycaemic
(Bornstein et al., 1951), and, recently, that glucagon may
stimulate growth, as evidenced by widening of the epiphy-
sial disks of hypophysectomized rats (Elrick, 1953).
latter experiments the most potent preparation of glucagon
did not produce a convincing widening of the epiphysial
Convulsions, relieved by sugar, were frequently a
late result of the injection of other preparations.
could have been produced by some residual insulin or per-
haps, as suggested by the author, by the reaction of the
beta cells to the hyperglycaemia.
ings by earlier investigators, Dr. 0. Sirek, in our laboratory,
has demonstrated that a purified growth preparation may
produce no hyperglycaemia in a normal dog but a large
in blood sugar in the same dog after
The blood-sugar changes that occur under
these conditions suggest that during a period of growth-
hormone administration the hypoglycaemic factor may be
G GOWTH HORMONE
is remotely possible that somatotropin,
In an extension of find-
AUG. 15, 1953
INSULIN AS A GROWTH HORMONE
liberated from the pancreas in more effective amounts than
is the hyperglycaemic principle.
The clinical significance of this experimental demonstra-
" pituitary "-dwarf animal by giving insulin-merits
Is a failure of insulin liberation one of the
pituitary dwarf ?Removal of the pituitary in animals does
not cause atrophy of the islands of Langerhans, but it pre-
vents their growth and presumably the augmented liberation
of insulin which this normal growth would require and
produce general body growth will cause growth of islet
true under most circumstances
study, since the excess of insulin would lessen the need for
islet activity, reduce the granule count of the beta cells,
and lower the insulin content of the pancreas.
Some human "pituitary" dwarfs are very sensitive to
pituitary at all is frequently most unconvincing.
ings show that in one species the presence of low blood-
sugar values and great sensitivity to insulin do not eliminate
of insulin as one means of producing a
dramatic growth effect.
To what extent the hypoglycaemic
effect of insulin may be modified by other hormones with-
out interfering with its growth effect is being investigated
both experimentally and clinically.*
clinical question is whether or not a failure of growth ever
occurs, apart from the frank diabetic state, in which the
limiting factor is the ability of the beta cells to produce
and liberate insulin.
Our findings do not show that exo-
genous insulin is necessary for growth of hypophysecto-
Somatotropin, when endogenous
available, is more effective.
But a clinical condition might
exist in which the pancreatic beta cells, while still capable
of preventing the diabetic
rate of growth and demand for insulin, are one of the
factors limiting normal development.
insulin together might produce results where one alone
would be inadequate.
The clinician watches for evidence
of decreased function of thyroid, adrenal, and gonads in
cases of suspected pituitary dwarfism.
stimulate these glands or to provide their products.
are suggesting that, in addition, a condition may exist in
which the stimulus exerted by somatotropin or by an in-
creased food intake may prove inadequate to provide an
optimum liberation of insulin from the pancreas.
this situation always be revealed by the appearance of dia-
Are there other procedures which may stimulate the
beta cells ?
Will insulin administration further decrease
the response of the beta cells ?
Some growth studies have been made in animals with
small amounts of insulin and large amounts of somato-
tropin available, but many other variations of dosage of
these two hormones and of insulin with other hormones will
certainly yield new and interesting findings.
gators have confused the issue by assuming, without justifica-
tion, that the alloxan-diabetic animal has nio insulin avail-
With our high-carbohydrate diets the effect of insulin on
fat synthesis has been conspicuous.
diets a different picture might be obtained, but the tolerance
for insulin would probably be reduced.
Querido's laboratory in Leyden (personal communication)
indicate that the duration of feeding periods and their
spacing throughout the day may affect the composition of
The time of administration and the duration
of action of the various hormones might affect appetite, the
pattern of food intake, and perhaps the composition of body
It may be argued that any factor or factors which
It would not be expected in this present
A more fundamental
at the existing reduced
He may attempt to
With very high protein
The findings in
*Mr. Robin Lawrence and one of us (J. S.) have found that
cortisone may prevent the effect of insulin on the widening of
the epiphysial disk, but the hypophysectomized animals receiv-
ing both hormones gained appreciably in weight.
As illustrated in Figs.
physectomized rats consumed more food than the controls.
There is evidence that insulin-treated normal rats (Maassen,
1951) grow faster than paired-fed controls, and this is cer-
tainly clearly seen when somatotropin is the growth stimu-
lant (Li and Evans, 1948; Young, 1952).
forced feeding on the growth of hypophysectomized rats is
Samuels, Reinecke, and Bauman (1943) state
that force-fed hypophysectomized rats exhibit an increase
in femur and body weight.
Levin (1944), in repeating these
experiments, reported that the increase in body weight could
be totally attributed to fat deposition.
to give much insulin to hypophysectomized rats which are
not permitted to eat more than the controls.
comparison of force-fed and insulin-treated hypophysec-
tomized animals with identical food intakes might provide
Great care in the interpretation of results
would be necessary, since the effect of insulin, liberated
physiologically by the demands of a forced food intake,
might not affect growth in exactly the same way as would
an identical food intake stimulated by an excess of exo-
genous insulin given in one or more daily doses.
1 and 2, the insulin-treated hypo-
The effect of
It will be difficult
Summary and Conclusions
The sensitivity to insulin of hypophysectomized ani-
mals has prevented investigation of the effects of the
antidiabetic hormone on their growth.
acting insulin in gradually increasing doses an average
increase in weight of 38 g. in 15 days has been secured in
completely hypophysectomized rats.
epiphysial disk of the tibia in these three experiments
showed, on the average, an increase of 65 %.
age increase in body protein was 2.6 g. per rat.
control hypophysectomized animals which ate smaller
amounts of the same diet remained essentially stationary
in weight, protein content, and skeletal size.
ings therefore demonstrate that insulin may stimulate
growth under these conditions and strengthen the possi-
bility that decreased liberation of insulin may be one of
the factors limiting growth in the hypophysectomized
The physiological and clinical significance of
these findings has been briefly discussed.
We are indebted to Professor W. Stanley Hartroft for his
interest and help.
This work has been supported by a grant from
the Canadian Life Insurance Officers' Association.
By using a slow-
The width of the
Allen, B. M. (1916).
Bornstein, J., Reid, E., and Young, F. G. (1951).
Elrick, H. (1953).. Proc. Soc. exp. Biol., N.Y., 82, 76.
Evans, H. M., and Long, J. A. (1921).
Greenspan, F. S., Li, C. H., Simpson, M. E., and Evans, H. M. (1949).
Endocrinology, 45, 455.
Haist, R. E. (1953).
Levin, L. (1944).
Amer. J. Physiol., 141, 143.
Li, C. H., and Evans, H. M. (1948).
Maassen, A. P. (1951).
Proc. kon. ned. Akad. Wet., 54, 160.
Milman, A. E.. De Moor, P., and Lukens, F. D. W. (1951).
Physiol., 166, 354.
Salter, J. M., and Best, C. H. (1952). Proc. Canad. Physiol. Soc.
Fed. Proc., 12, 122.
Samuels, L. T., Reinecke, R. M., and Bauman, K. L. (1943).
ology, 33, 87.
Smith, P. E. (1916).
Science, 44, 280.
Young. F. G. (1952).
posium sur les Hormones Protdiques et Derivees des Proteines.
Anat. Rec., 11, 486.
Nature, Lond., 168, 903.
Anat. Rec., 21, 62.
Recent Progr. Hormone Res., 3, 3.
lle Congres International de Biochimie.
The Council of the Yugoslav Academies has commenced
of a quarterly Scientific Bulletin
French, and German, the chief purpose of which is to give
information on scientific activities in that country.
scientific work in Yugoslavia appears only in periodicals
makes it difficult for those abroad to become acquainted
with current research work.
Bulletin should prove of considerable interest.
by Dr. A. Peterlin, Institut de Physique J. Stephan, Ljubljana,
For this reason alone the
It is edited