Nobel Lecture, December 12, 1947

The Role of the Hypophysis in Carbohydrate Metabolism and in Diabetes

The nutritive substances used in greatest quantities by mammals are carbohydrates. – These are ingested with the food or produced by the liver, and after being converted into glucose they are utilized by all the cells of the organism, specially by the muscles, which constitute nearly half the mass of the body.

Production and consumption of carbohydrates is so well regulated that there is a constant blood sugar level; any accidental increase or fall in blood sugar is rapidly compensated. The constancy of the blood sugar level is maintained by a complex physiological mechanism, a homeostatic mechanism of the same order as those which maintain the body temperature, the blood pressure or the heart rate at normal levels and control many other functions. The blood sugar level itself is an important factor in many chemical and physiological processes.

The liver is the principal organ in the regulation of the blood sugar level. If the liver is removed there is a progressive hypoglycemia which ends in death. After hepatectomy it is not possible to provoke hyperglycemia, whatever the agent or procedure employed, whether the animal were diabetic or not before the liver was removed. Nor is it possible to produce any form of diabetes in hepatectomized animals. On the other hand, when there is hyperglycemia the liver diminishes or ceases the production of sugar, and stores it.

The production and consumption of glucose and hence the blood sugar level, are controlled by a functional endocrine equilibrium. This mechanism acts on the liver – the organ which produces and stores glucose – and on the tissues which are the consumers of glucose, by means of hormones which play a part in the chemical processes of carbohydrate metabolism.

The secretion of each endocrine organ is controlled by a physiological mechanism. For instance, the pancreas secretes insulin in adequate quantities so as to maintain a normal blood sugar level, and the blood sugar level regulates the amount of insulin secreted. Thus hyperglycemia increases the secretion of insulin and hypoglycemia diminishes or completely inhibits it. This regulation of insulin secretion is maintained even when the pancreas is surgically reduced to one seventh of its normal mass, or when besides the intact pancreas in situ four others are grafted into the circulation. In both cases the blood sugar is kept constantly at a normal level. The extrinsic innervation of the pancreas is not necessary for the regulation of insulin secretion; the response of the denervated pancreas to changes in blood sugar differs only in minor details from the response of the normal pancreas.

Not only is the secretion of each gland regulated according to the organic needs of each moment, but there is also an equilibrium between the secretions of the different glands. The blood sugar and the production and consumption of glucose are kept within normal bounds, therefore there is an equilibrium between the glands of internal secretions which reduce the blood sugar (pancreas) and those which raise it (anterohypophysis, adrenals, thyroid, etc.).

In 1907, when I was a medical student, I was attracted to the study of the hypophysis because the microscopic picture showed glandular activity and its lesions were accompanied by serious organic disturbances, such as acromegaly, dwarfism, etc.

The frequency of glycosuria or diabetes in acromegaly has been reported many times; Atkinson found it in 32.8% of 817 cases of the disease published up to 1938. Moreover, extracts of the posterior lobe of the hypophysis produce rapid pharmacological effects, amongst which is a transitory hyperglycemia (Borchardt, 1908). Therefore it was commonly accepted that if the hypophysis played a part in carbohydrate metabolism, it would be due to the activity of its posterior lobe.

One year after the discovery of insulin a systematic study of the influence of endocrine glands on its activity was organized in my laboratory. With Magenta (1924-1927-1929) I found that hypophysectomized dogs were hypersensitive to the toxic and hypoglycemic action of insulin. In 1925, I found the same fact in the toad (with Mazzocco and Rietti); and in 1929, assisted by Miss Potick I saw that the administration of pars distalis prevented or corrected hypersensitiveness to insulin in hypophysectomized toads.

The next step consisted in the removal of the pancreas in hypophysectomized dogs and toads (with Biasotti in 1930). The removal of the hypophysis or of its pars distalis produced a considerable attenuation of the severity of pancreatic diabetes. The injection or implantation of anterior hypophyseal lobe reestablished or even increased the usual severity of diabetes.

This work, done in 1930, showed that: (a) the anterior lobe of the hypophysis has an important part in the physiological control of metabolism; (b) the hypophysis is a factor conditioning the severity of diabetes; (c) the injection of anterior lobe of the hypophysis has a diabetogenic effect.

Later the diabetogenic effect was also demonstrated in dogs with a surgically reduced pancreas (Houssay, Biasotti, Di Benedetto, and Rietti, 1932) or with an intact pancreas (Evans et al., 1932, etc.). A permanent diabetes was produced by prolonged treatment with the extract of anterior lobe of hypophysis in dogs with a reduced (Houssay, Biasotti, Di Benedetto, and Rietti, 1932) or an intact pancreas (Young, 1937).

After total hypophysectomy, or removal of the pars distalis, when the animals are kept fasting, a severe, and sometimes fatal, hypoglycemia easily occurs. These animals are also very sensitive to hypoglycemic agents, such as insulin and phloridzin. Hyperglycemia produced by adrenaline or by glucose administration is followed in some cases by intense hypoglycemia and convulsions.

The injection of pars distalis increases the resistance to insulin and phloridzin. This effect has been called glycotropic by Young (1936-1938).

Hepatic and muscular glycogen falls rapidly in fasting hypophysectomized animals. Hypophyseal injections inhibit this decrease specially in respect to muscle glycogen and this action is also produced in absence of the adrenals (Russell, Bennett); The mobilization of hepatic glycogen by insulin or adrenaline is in some species slower in hypophysectomized than in normal animals.

Extirpation of the hypophysis, or of the pars distalis, markedly attenuates the severity of pancreatic and phloridzin diabetes. Hypophysectomized pancreatectomized dogs show the following differences with respect to pancreatectomized dogs: (a) they survive a long time, the wounds heal well and infection is much less frequent; (b) there is a slow gradual loss of weight; (c) the loss of nitrogen is less considerable; (d) hyperglycemia and glycosuria are less marked and are sometimes absent; fasting produces a notable fall in blood sugar and decrease in glycosuria; (e) spontaneous hypoglycemic crises occur, usually due to underfeeding or fasting and there is hypersensitiveness to insulin; (f) hypoglycemic crises can be prevented by carbohydrate or protein-rich diets, but not by fats; injection of glucose saves the animal’s life when it has fallen into a hypoglycemic crisis; (g) there is little decrease in hepatic and muscle glycogen if the diabetes is not severe, and in some cases the same concentrations as in the controls are found; (h) when glucose is injected only part of it and sometimes none, is excreted in the urine; (i) the blood sugar curve following glucose administration is intermediate between those of pancreatectomized and normal dogs; (j) after glucose is given, there is a moderate rise in the respiratory quotient in some cases, but not in all; (k) the increase in ketonemia and ketonuria is remarkably small; (l) the increase in lipemia and cholesterolemia is not considerable; (m) the rise in the basal metabolic rate is less than in the pancreatectomized controls; (n) the D/N ratio in urine is also lower than in the controls.

Hypophysectomized pancreatectomized animals utilize more glucose than pancreatectomized, but less than normal animals. Protein consumption does not increase as much as in the pancreatectomized; the small amount of ketonuria points to a diminished consumption of fat.

The diabetogenic effect of the hypophysis was first observed in animals in which both the hypophysis and the pancreas had been removed. Later, in 1937, it was demonstrated in: (a) dogs with a surgically reduced pancreas (Houssay, Di Benedetto, and Rietti); (b)normal rabbits (Baumann and Marine); (c) normal puppies.

Daily injections of anterohypophyseal extract provoke after two or three days a rise in blood sugar and hypophyseal diabetes in cats and dogs. In the dog the blood sugar rises from normal to 150 to 300 mg per cent; there is glycosuria; ketonemia and ketonuria increase; and there is polyuria and polydipsia. Injection of glucose is followed by a typical diabetic blood sugar curve and the respiratory quotient does not rise or rises very little. The animal becomes insulin resistant.

The hypophysis of all vertebrates has this diabetogenic effect on animals of its own and other species. The pars distalis is responsible for the diabeto-genie effect, the posterior lobe has little activity in this respect. The diabeto-genie principle is rapidly destroyed at room temperature, therefore the extracts should be stored at a temperature near 0°C.

The active principle is a protein. Young has obtained a concentrated extract, but it has not been prepared in a pure form, so it is not yet possible to say if it is a separate hypophyseal hormone, different from those already known. I am mistakenly supposed to have described a diabetogenic hormone. I have described a diabetogenic property, or factor, or action of the extract, without having said this effect was due to a diabetogenic hormone. The regulation of carbohydrate metabolism is a normal function of the hypophysis, but certainly the production of diabetes is not one of its normal functions. Nevertheless an active hypophysis, even when it is functioning normally, increases the severity of diabetes.

The diabetogenic effect of the hypophysis is dependent on the liver. It can not be obtained in the hypophysectomized pancreatectomized toad if the liver has been removed. Hepatectomy causes a rapid fall in the blood sugar of dogs or toads with a hypophyseal diabetes. On the other hand, destruction of the diencephalon or removal of the whole brain, or extirpation of the kidneys, the adrenals, the thyroid, the lungs and all the digestive tract do not prevent the diabetogenic effect.

In the toad and in the dog, adrenalectomy diminishes but does not suppress the diabetogenic effect of the anterior lobe extract, which can be obtained in adrenalectomized dogs, in which the pancreas has been surgically reduced, and which are kept alive by treatment with desoxycorticosterone and salt or even with sodium chloride alone. The experiments of Long show that the adrenocorticotrophic hormone of the hypophysis by releasing corticoadrenal hormones, plays a part in the diabetogenic effect; but my own experiments prove there is a hypophyseal diabetogenic action which takes place when there are no adrenals.

There are pancreatic and extrapancreatic factors in the production of hypophyseal diabetes.

Extrapancreatic: factors are demonstrated by: (1) the diabetogenic effect obtained in pancreatectomized hypophysectomized animals; (2) the increase in the severity of diabetes in pancreatectomized dogs treated with anterohypophyseal extract; (3) the increase in resistance to insulin which precedes and accompanies hypophyseal diabetes; (4) the fact that hyperglycemia and diabetes occur before that there are visible lesions in the beta cells of the islets, and before insulin secretion diminishes.

After two to four days of hyperglycemia provoked by the injection of an terohypophyseal extracts the beta cells show signs of damage and insulin secretion diminishes. The beta cells are degranulated, they swell, become vacuolized and go into hydropic degeneration. In most cases there is no insulin secretion or it is considerably reduced. This can be demonstrated by grafting the pancreas from these animals into the circulation of a diabetic animal. A normal pancreas reestablishes the normal blood sugar level in three to five hours and maintains it there, while the pancreas taken from an animal in hypophyseal diabetes has little or no effect on the blood sugar.

If after a few days the anterohypophyseal treatment is discontinued the diabetic condition disappears, the blood sugar returns to a normal level, and later the beta cells regain their normal aspect. If daily injections of anterohypo physeal extracts are continued for several weeks, beta cells first show pycnotic nuclei and then gradually disappear in great numbers. At this stage even if the hypophyseal treatment is discontinued the animals remain permanently diabetic, degeneration of the beta cells continues and the pancreas ceases to secrete insulin.

The diabetic condition which appears during hypophyseal treatment and which disappears when this treatment is discontinued should be called hypophyseal diabetes. The lesions in the beta cells are reversible.

The diabetic conditions which persists when the hypophyseal treatment is discontinued, should be called metahypophyseal diabetes. It is a pancreatic diabetes due to irreversible lesions of the beta cells. It is produced by the injection of hypophyseal extract, but persists without this treatment.

In hypophyseal diabetes there is a marked increase in insulin resistance, while in metahypophyseal diabetes the resistance to the action of insulin is normal or only slightly increased. It is possible that in acromegaly, in some stages of the disease there may be a hypophyseal diabetes and at others a metahypophyseal diabetes.

There is a certain antagonism between the hypophysis and the pancreas. Thus as the mass of the pancreas is progressively reduced by extirpation of increasingly larger portions, the dose of hypophyseal extract needed to provoke diabetes diminishes. Reciprocally in hypophysectomized animals there is hypersensitiveness to insulin, and in fasting, their blood sugar and glycogen falls more rapidly and markedly than in normal animals.

The anterohypophysis is not necessary for the normal development and maintenance of the pancreas, therefore there is no physiologic pancreatotrophic effect of the hypophysis on the pancreas. Hypophysectomy does not cause a loss in the weight of the pancreas in dogs, nor of the insular mass in several species in which it has been examined; on the contrary, there is an increase in the islets. The pancreatic content and secretion of insulin are normal.

Anterohypophyseal extracts can produce two opposite effects on the islets: (a) stimulation and hyperplasia; (b) atrophy and hypofunction. One or other effect is observed according to the dose given, the amount of islet tissue and the animal species.

The lesions in the islets and the decrease in insulin secretion are due to two factors: (a) the effect of the anterohypophyseal extract; and (b) hyperglycemia. They are not due exclusively to hyperglycemia as proved by the fact that the same levels of blood sugar have been maintained during four days by continuous injection of glucose and by hypophyseal extracts injections. Only the hypophyseal treatment produced lesions of the islets and decrease or absence of insulin secretion.

The mechanism by which the anterohypophysis controls carbohydrate metabolism is still imperfectly known.

Hypophysectomized dogs utilize protein and carbohydrate ingested in the food, but during complete fasting, or protein fasting, the disintegration of body protein is slower than in normal dogs. Hypophysectomy also reduces the disintegration of body protein in pancreatic and phloridzin diabetes. It would seem therefore that hypophysectomy diminishes the capacity of the organism to catabolize body protein and to form glucose from this protein.

Ketonuria is also below normal in hypophysectomized animals in basal conditions and in the diabetic state, whatever its origin, therefore the consumption of fat also seems to be reduced. There is good utilization of carbohydrate in hypophysectomized animals. Experiments in rats (Fisher, Russel, and Cori, 1936) and rabbits (Greeley, 1935-1940) show there is a preferential consumption of glucose, above the normal. These workers believe this to be the fundamental metabolic disturbance produced by hypophysectomy. In the dog this supernormal, preferential consumption of glucose has not been demonstrated. Moreover, it could be the effect of the subnormal and insufficient catabolism of proteins and fats.

The pituitary is a very important regulator of the homeostatic function of the liver. In hypophyseal insufficiency glucose formation by the liver does not increase during hypoglycemia as it does in normal animals. During fasting or diabetes, the glucose formation from proteins is also lower. On the other hand an excess of anterior pituitary produces hyperglycemia due to diminished peripheral consumption of glucose and to lack of inhibition of its hepatic formation.

Recently, Price, Cori, and Colowick (1945-1947) have shown that the anterohypophysis inhibits hexokinase and that insulin counteracts this inhibition. It is the first case in which it has been shown that a hormone, or rather two hormones, act on an enzymatic process of importance in carbohydrate metabolism. Nevertheless, this important discovery does not explain all the facts, e.g. the hypersensitiveness to insulin of hypophysectomized animals.

One of the most important metabolic functions of the hypophysis is the part it plays in the formation of protein on which growth depends. The growth hormone increases the fixation of nitrogenous substances and body protein formation. Hypophysectomized rats with forced feeding form more fat and less protein than the controls. Ingested protein is catabolized in a greater proportion than in normal animals.

Young (1940-1944)supposes that the growth hormone of the hypophysis induces the fixation of nitrogen and the utilization of glucose when insulin is present and in this way causes growth. When there is not enough insulin, sugar is not utilized but excreted, i.e. there is a diabetogenic effect. His main arguments are the following: (1) a dose of hypophyseal extract which is rapidly diabetogenic in the adult diabetic animal causes only growth in puppies; these young animals do not become diabetic until after several months of treatment they cease to grow; (2) in the rat the extract increases the islet tissue, the animal grows and there is no diabetes; (3) extracts which are diabetogenic for the dog only cause growth in the rat; (4) pure growth hormone increases glycosuria in subtotally pancreatectomized rats; (5) the fixation of nitrogen caused by hypophyseal extract is conditioned by the amount of islet tissue or of insulin.

Long (1942) believes that the metabolic action of the hypophysis depends on the growth hormone, the adrenotrophic and the thyrotrophic hormones.

The work that has been done in my laboratory on this subject has been carried out with the collaboration of Drs. Lewis, Giusti, Mazzocco, Rietti, Potick, Biasotti, Braier, Di Benedetto, Gerschman, Aubrun, Campos, Lanari, Curutchet, Cicardo, Etcheverry, Foglia, Sammartino, Magdalena, Ferrer- Zanchi, Savino, Hug, Novelli, Orias, Parodi, Leloir, Artundo, Lascano-Gonzalez, Gofialons, Riet-Correa, Stoppani, Dambrosi, Dosne, Fernandez, A. Houssay, H. Houssay, Smyth, Pasqualini, Marenzi, De Robertis, and others.

Carbohydrate metabolism and other metabolic processesare regulated by the balance maintained between the secretion of several endocrine glands. Diabetes and other metabolic diseases are a disturbance in this endocrine equilibrium. There are still many problems to be solved, but undoubtedly the hypophysis is one of the most important organs in the regulation of metabolism and the center of the endocrine constellation.