C. H. Best

Diet and the insulin content of pancreas.

IT is well known that fasting, or feeding diets rich in fat and poor in carbohydrate, leads to a change in the metabolism of sugar as judged by (1) the glucosuria following glucose administration, (2) the diabetic type of sugar tolerance curve when glucose is given, and (3) the absence of the normal rise in respiratory quotient after glucose administration. There is evidence that the administration of insulin at least partially restores the normal metabolism of carbohydrate. An excellent review of the relevant literature has recently been published by W. H. Chambers [1938]. We have reported in a preliminary communication [Haist, Ridout & Best, 1939] that a very definite change in the insulin content of pancreas may be produced by alterations in diet. In order to investigate this subject it is necessary to have available (1) an experimental animal which will ingest the diets provided and one from which all the pancreatic tissue can be removed without undue difficulty; (2) an extraction procedure which consistently gives optimal yields of insulin from pancreatic tissue; and (3) a method of testing which gives accurate results when relatively small amounts of insulin are available. The first requirements are satisfied when the Wistar rat is used as the test animal. A suitable method for the extraction of insulin is that outlined by Jephcott [1932] and by Scott & Fisher [1938 a]. Satisfactory assays of the insulin content of the extracts are obtained by the mouse method of testing. The results of these investigations which will now be described demonstrate, among other points, that a very definite decrease in the insulin content of pancreas is brought about by fasting or by the ingestion of diets rich in fat.

Atheromatous Changes in Aorta, Carotid and Coronary Arteries of Choline-Deficient Rats.

Summary Lesions have been observed in the aortas, carotid and coronary arteries in 26 of 116 rats fed a low choline diet for periods up to 216 days. None of these animals exceeded 300 days of age at the time of sacrifice. The initial lesion consists microscopically of the deposition of stainable lipid in the endothelial cells of the intima. In later stages, a proliferation of intimal cells had taken place so that small plaques resulted. In the large vessels (aortas and carotids) the subjacent media had undergone necrosis and eventual calcification. The possible relation of these lesions to atheroma in man is discussed.

The Pathologic Effects of Large Amounts of Glucagon

The work of Ingle and Cavallero has shown that glucagon will produce a temporary or mild glucosuria in partially depancreatized animals and in animals pretreated with cortisone, but a significant diabetogenic action of glucagon administered alone to intact rats has not previously been shown. The results of early attempts made in our laboratory to show a diabetogenic action of the pancreatic hyperglycemic factor were disappointing. We reinvestigated the problem using much larger doses of glucagon attempting, at the same time, to prolong its activity by suspending it in corn oil and administering it subcutaneously at eight-hour intervals. Glucagon administered under these conditions appeared to exert a profound effect. • Figure i shows the average changes in weight and food intake of intact male controls injected with corn oil, and of normal male rats injected three times daily with 300 ug. of glucagon (Lilly, lot no. 258-234B-33) suspended in corn oil. Weight changes are shown for a second set of controls limited to the amount of food consumed by the glucagon-treated animals. It is apparent that the glucagon-treated rats consumed much less food than the controls and lost weight rapidly. The weight loss cannot be completely attributed to the reduction in food intake, as the pair-fed controls lost much less weight. The glucagon-treated animals were not glucosuric but animals similarly treated and encouraged to eat bread, frequently showed a transient but intense glucosuria. This stimulated us to investigate the effect of glucagon in force-fed rats. Male Wistar rats weighing 150 to 160 gm. were fed by stomach tube at 8:00 a.m., 4:00 p.m., and 12:00 m., the high carbohydrate diet described by Reinecke, Ball and Samuels. The volume administered was slowly increased

Dietary choline and the maintenance of the cardiovascular system in rats.

The pathological changes in the livers of a number of animal species and in the kidneys of rats resulting from an inadequate intake of lipotropic factors are well known. These effects have been extensively investigated during the two decades that have elapsed since the discovery of the lipotropic action of choline (Best and Huntsman, 1932). But the results of choline deficiency on organs other than liver and kidney have received relatively little attention. In 1945 Kesten, Salcedo, and Stetten reported cardiac necrosis in rats fed a high level of a synthetic ester of a medium-chain fatty acid (ethyl laurate, 35-40%) with a cholinedeficient diet. These investigators found that choline chloride (0.35%) added to this diet completely prevented cardiac necrosis. Several years ago we noted the development of gross necrotic lesions in the hearts of choline-deficient weanling rats fed a diet containing 12% beef fat. We were uncertain of the significance of this finding and did not publish it. The report from Stettens laboratory and our preliminary observations on cardiac necrosis in rats given natural fats prompted the present series of investigations. The results have indicated that the dietary condition essential for the production of cardiac necrosis is choline deficiency per se. Kesten, Salcedo, and Stetten (1945) found that lesions developed only when choline deficiency was produced in rats fed ethyl laurate; other synthetic lipids did not have this effect in their experiments. Our investigations have revealed that in susceptible animals, maintained under severely hypolipotropic conditions, cardiac lesions appear when a number of different fats are used. Ethyl laurate, lauric acid, other synthetic fatty acids, and the naturally occurring fats (lard, beef fat, corn oil, and coconut oil), at levels of 25-35%, are all effective in the production of cardiac lesions in varying proportions of rats on choline-deficient diets. Even in the complete absence of dietary fat, two of thirteen rats fed a hypolipotropic diet developed significant degrees of cardiac damage. Previous investigators (McCormick and Holman, 1949) have shown that high levels of dietary fat, in the presence of experimentally induced renal damage, are capable of producing cardiovascular lesions in dogs. Lehr and Churg (1952) produced aortic sclerosis and cardiac necrosis in rats that had developed renal injuries following the administration of sodium N-acetyl sulphathiazole. Our experiments indicate that in rats, even in the presence of apparently normal kidneys, dietary fats, inadequately balanced by lipotropic factors, may result in damage to heart and vessels. The higher the level of fat in the diet the more extensive and the more frequent are the lesions. But the production of cardiac lesions in even a small number of animals fed a fatfree choline-deficient diet emphasizes the possibility that lack of choline possesses basic aetiological significance in the production of this necrosis. High levels of dietary fat and the presence of renal damage are very important factors, but they do not appear to be the only ones essential for the production of this type of cardiac injury. There is no histological evidence suggesting any infective element in the production of these cardiovascular lesions, but the possibility cannot be completely dismissed that some such process might flourish in the absence of dietary choline. We are unaware of any convincing evidence to indicate that hypolipotropic diets are involved in the aetiology of any type of cardiovascular disease in man. Nevertheless, the consistent production of severe cardiac injury by the removal of a naturally occurring component from a diet containing natural fats, and the prevention of this lesion by restoring the missing factor, choline, may prove to be of fundamental importance in the understanding of factors responsible for the maintenance of the cardiovascular system.

THE LIPOTROPIC FACTORS

The nutritional role of choline was in 1932, so that this year it comes of age as a dietary factor. Choline was the first of the accessory food factors which was shown to protect the liver; i.e., in its absence from the diet, pathological changes appear promptly in the hepatic cells. The work on choline, as a survey of the literature clearly shows, stimulated great interest in liver fat, in the effect of large accumulations of fat on hepatic function, and in the part played by other dietary factors in the protection of the liver. The very general and broad title assigned to us here has made us somewhat uncertain about what is desired. The topics to be dealt with by our learned colleagues on this comprehensive program leave us but little in the way of definite responsibility, and we are therefore quite free to discuss certain general aspects of our subject. We are all familiar with the fact that choline, betaine, methionine and, under some conditions, vitamin B I ~ and inositol, protect rats in varying degrees against the development of fatty livers. I n the absence of choline from the diet, many species of animals have been found to develop this abnormality. The rat,l* *, 5 mouse,3’ rabbit,8 h a m ~ t e r , ~ calf,I0 pig,” and ducklingI2 exhibit fatty livers when the diet lacks sufficient of the lipotropic agents, choline or its precursors. Some of the classical lesions are illustrated in Professor Hartroft’s presentation (page 633). I n some species, such as the guinea pig, it is very difficult to produce an abnormal deposition of fat in the liver by feeding diets low in ch01ine.’~ We must always keep in mind that interference with the function of hepatic cells by means other than deficiency of the lipotropes may lead to fatty livers. Fatty livers of guinea pigs on scorbutic diets14 are probably not attributable to a lack of dietary choline. Many readers are doubtless familiar with the hemorrhagic kidney syndrome15 which may be consistently produced in young rats on diets low in the lipotropes. With the hypolipotropic diets currently used in our laboratory we are able consistently to produce kidney lesions in mature animals. Somewhat similar renal lesions have been reported in young p i g P and calves.1° The failure to produce this kidney lesion in young mice and puppies, both of which develop fatty livers readily enough, reveals a problem which may possibly require extended study. A comprehensive investigation and comparison of the relevant enzyme systems will undoubtedly help. An interesting sequel to the renal lesion in rats is the appearance in later life of hypertension.” This has not yet been reported in other species. You are also familiar with the failure of baby chicks,’* turkey poults19 and ducklings“o* to grow and their tendency to develop slipped tendon disease on choline-deficient diets. In our early studies on the effects of choline in preventing and curing the pathological changes in the diabetic dog, we mentioned “areas of beginning fibrosis” which did not appear to be as favorably affected by choline as the fatty changes,22

The lipotropic action of methionine

IN 1937 Tucker & Eckstein demonstrated that methionine exerts a lipotropic effect. This has been confirmed in our laboratory [Best & Ridout, 1938] and by Channon, Manifold & Platt [1938]. The latter workers have shown that methionine, under the conditions of their experiments, exerted very little effect upon the deposition of fat in the liver unless the basal diet was such that large amounts of fat were deposited in the livers of the control animals. We have been interested in the lipotropic effects of dand l-methionine and in the failure of large doses of the racemic mixture to produce greater effects than small doses under certain experimental conditions.

The Croonian Lecture: The Lipotropic Agents in the Protection of the Liver, Kidney, Heart and Other Organs of Experimental Animals

Our successors in medical research will read, with amused tolerance, our frequent agonized protests that the volume of literature which we should comprehend exceeds our capacity. Those of us who teach and write text-books for medical students and practitioners are forced to depend more and more upon the review journals. The number and breadth of the fields in which one can pose, even transiently, as an authority are diminishing rapidly. But if one looked only at this aspect of the picture an ageing outlook on the most fascinating of all games, medical research and teaching, might be suggested. The appearance of new authorities is most stimulating, and it is refreshingly obvious that the soaring rate of publication has not inhibited research. We must co-ordinate the abstracting services in our field, continue the struggle to eliminate unworthy papers, hope for even better colloquia and review journals, and confidently expect that the next fifty years will contribute more to medical science than has the past vigorous half-century. The award of the Croonian Lectureship has given me exceptional pleasure and a sense of great responsibility. In selecting a title I have considered the principal subjects in which I have endeavoured to keep abreast, and the choice has thus been narrowed to insulin and experimental diabetes, heparin and thrombosis, and the dietary factor choline and its precursors, which we have termed the lipotropic agents. Certain of the effects of these three substances might be discussed in a single lecture, since they all affect either the formation, distribution or the state of fat in the body. The action of a lipokinetic constituent of the anterior pituitary, first clearly demonstrated in our laboratory in 1936 (Best & Campbell), which increases the rate of mobilization of depot fat to the liver (Barrett, Best & Ridout 1938; Stetten & Salcedo 1944), might also have been included. The fat-mobilizing effect of anterior pituitary extracts may be due to Evans’s somatotropin, to the adrenocorticotropic hormone, to a more specific but as yet unidentified substance or, of course, to more than one of these. The four factors, insulin, choline, heparin and ‘adipokinin’ (Weil & Stetten 1947) have given us a measure of control over fat metabolism which our predecessors did not enjoy. There are, of course, other dietary and hormonal agents affecting these processes which one would have to discuss in a comprehensive treatment of the field. I shall not even list these and, indeed, after a very brief consideration of insulin and heparin, particularly in relation to fat metabolism, I shall limit my discussion to ‘the lipotropic agents’.

Insulin Extractable from the Pancreas and Islet Cell Histology: Comparative Studies in Spontaneous Diabetes in Dogs and Human Subjects

The characteristics of diabetes mellitus in man have long been compared, either directly or by inference, with those of diabetes produced in animals by experimental means. From the earliest studies in this field by von Mering and Minkowski1, in the work leading to the discovery of insulin2 (1921), and in many later investigations, the dog has occupied a prominent and productive place amongst experimental animals. In recent years, special aspects of the spontaneous diabetes of dogs have come under study in this laboratory and in that of Professor Henry Ricketts in Chicago. It is the purpose of this paper to examine the available data to determine the degree to which the spontaneous and the experimentally induced diabetes mellitus of dogs correspond with each other and with the spontaneous diabetes mellitus of man, in terms of the insulin extractable from the pancreas and the histological state of the islands of Langerhans. The relationship of these factors to the diabetic state with which they are associated will also be considered.

Intramuscular Cortisone Administration to Dogs

Summary A syndrome closely resembling diabetes insipidus was produced in 4 mongrel dogs by daily injections of cortisone in doses of 50 to 300 mg a day. The period of hormone injection varied from 4 days to 14 days in different dogs. The polydipsia reached values as high as 6200 cc per day and the polyuria reached values as high as 5790 cc per day. Frequent fasting blood sugar determinations and glucose tolerance tests did not reveal any changes in carbohydrate metabolism.

Pituitary growth hormone and the question of pancreatic secretion of glucagon.

Experiments are presented in which normal dogs were injected with a highly purified growth hormone preparation. One hour later blood samples were drawn simultaneously from the pancreatico-duodenal and jugular veins. The ‘central’ and ‘peripheral’ blood samples were injected into depancreatized dogs. While the injection of ‘peripheral’ blood produced no appreciable change in the blood sugar level of the diabetic recipient, the ‘central’ blood caused a definite but transient rise in blood sugar. The rise could be prevented by treating the recipient diabetic animal with the adrenergic blocking agent dihydroergotamine. It has been concluded that the hyperglycemic factor present in the blood of the pancreatico-duodenal vein after injection of growth hormone preparations is not identical with glucagon-Lilly, the effect of which is not influenced by dihydroergotamine. In two experiments in which completely depancreatized, instead of intact, dogs were used as donors the administration of growth hormone produced the characteristic hyperglycemic response when ‘central’ blood samples were injected into other depancreatized dogs. These findings cast doubt on the pancreatic origin of the hyperglycemic material and add interest to the search for its source and nature. Further experiments using depancreatized dogs are in progress.