Symposium on Diabetes Mellitus
Pathophysiology of Diabetes Mellitus Robert Sherwin, M.D.,* and Philip Felig, M.D. **
The importance of insulin deficiency in the pathophysiology of diabetes mellitus has been firmly established since the studies of Minkowski in pancreatectomized animals and the epoch-making discovery of insulin by Banting and Best over 50 years ago. Later work by Houssay and by Long and Lukins indicated that pituitary and adrenal hormones interfere with the action of insulin and may contribute to hyperglycemia. The subsequent introauction of radioimmunoassay techniques for the measurement of polypeptide hormones led to the demonstration that insulin deficiency characterizes juvenile onset diabetes and most forms of adult-onset diabetes. 66 More recently techniques have been developed for examining the binding of insulin to its target cells 45 and for evaluating the in vivo action of insulin. I. 52 Such studies have rekindled interest in the possible role of insulin resistance in adult-onset diabetes. 4o In addition, a host of new data have accumulated regarding the contribution of glucagon to the diabetic syndrome. 53 • 62 The present discussion will thus review insulin secretion, binding, and action in normal man and diabetes, the metabolic consequences of insulin lack and the role of glucagon in diabetes.
Insulin Secretion in Diabetic Man Insulin is initially synthesized in the beta cells of the pancreas as a larger single-chain polypeptide precursor known as proinsulin. 57 During its storage in the beta cell, the single-chain proinsulin is cleaved, resulting in removal of a connecting strand CC-peptide) and the appearance of the smaller, double chain insulin molecule. Insulin and the C-peptide remnant are packaged in membrane-bound storage granules, which move toward the cell membrane and ultimately discharge their contents via a process Cemiocytosis, exocytosis) in which there is fusion of the granule membrane with the cell membrane. Stimulation of insulin secretion results in release of these granules and the discharge of equimolar amounts of insulin and C-peptide. 47 In normal subjects the concentration of glucose in the blood per*Assistant Professor of Medicine, Yale University School of Medicine, New Haven, Connecticut "":'C. N. H. Long Professor and Vice Chairman, Department of Medicine; Chief, Section of Endocrinology, Yale University School of Medicine, New Haven, Connecticut Medical Clinics of North America- Vo!. 62, No. 4, July 1978
ROBERT SHERWIN AND PHILIP FELIG
Figure 1. Glucose-induced insulin secretion by the beta cell is initiated by the interaction of glucose with a receptor on the cell surface and/or the intracellular metabolism of glucose. Release of insulin from membrane-bound storage granules involves a calcium-dependent contraction of intracellular microfilamentous structures.
fusing the pancreas is the most important determinant of insulin secretion. The response of the pancreatic beta cells to an increment in glucose is biphasic, consisting of an initial rapid secretory burst followed by a second more gradual output. The former results from the release of performed insulin from beta cell storage granules and the latter largely from stored insulin as well as synthesis of new insulin. In both phases the trigger mechanism initiating release of insulin appears to involve interaction with a glucose receptor at the cell membrane. 37 On the other hand, alterations in intracellular glucose metabolism may also be of importance in insulin release. 3o In either case, the "final common path" involves a calcium dependent contraction of intracellular microfilamentous structures 28 (Fig. 1). That the route of glucose entry as well as its concentration in blood is an important determinant of insulin secretion is indicated by the higher levels of circulating insulin observed following an oral glucose load as compared to an equivalent intravenous dose. This "alimentary augmentation" of insulin secretion is attributed to the release of upper gastrointestinal hormones upon contact of glucose with the gut mucosa. Current evidence suggests that gastric inhibitory polypeptide (GIP) is the primary intestinal hormonal factor augmenting insulin release.H In addition to glucose, amino acids are also capable of stimulating insulin secretion. Aminogenic stimulation largely accounts for the rise in insulin concentration observed after ingestion of pure protein meals. 64 As in the case of glucose, here too gastrointestinal factors play some role. Hypokalemia and potassium depletion are associated with impairment of insulin response to oral glucose or intravenous amino acids (arginine), and may be clinically important in this regard. 38a,38b While insulin secretion is generally considered in terms of its response to nutrients such as carbohydrate and protein, it should be emphasized that insulin secretion is ongoing and detectable in the blood in the interval between meals (basal or postabsorptive state). Basal pancreatic insulin secretion in normal man is approximately 0.5 to 1.0 unit per hour. 52 Furthermore, the basal level of insulin secretion is not fixed but varies with the fasting blood glucose concentration. In circumstances of hypoglycemia not associated with hyperinsulinism, an appropriate reduction in basal insulin secretion is observed.
Insulin Secretion in Diabetes
Current evidence favors the concept proposed over 50 years ago that a deficiency in the insulin secretory mechanism of the beta cell is the predominant or primary lesion in most forms of diabetes. This secretory abnormality may vary from virtually complete failure to a partial defect apparent only in circumstances of increased peripheral demands such as obesity, pregnancy and aging. In the juvenile-onset diabetic, insulin secretion is either totally defective or severely impaired. Endogenous insulin production in these patients generally varies inversely with the duration of the disease. In fact, endogenous insulin secretion as determined from circulating Cpeptide levels is generally not detectable in juvenile-onset diabetics who have received exogenous insulin therapy for more than 5 years. 5 Correlating with this functional disability is the frequent anatomic evidence of hyalinization and fibrosis of the islets in juvenile diabetics of long standing. Despite its severity, the secretory deficiency in the juvenile diabetic is often temporarily reversible in its initial stages as indicated by the tendency toward remission ("honeymoon phase" or Brush effect). This remission is characterized by a temporary restoration of insulin secretion as indicated by studies of C-peptide release. 6 The extent to which residual insulin secretion is maintained in the insulin-requiring diabetic is also an important factor in the stability of long-term diabetic control. The degree of diabetic instability, as determined by multiple fasting blood glucose measurements is inversely correlated with C-peptide secretory capacity (a measure of residual beta-cell secretion).51 These findings thus suggest that the severity of the insulin secretory defect is the major factor that distinguishes the more readily manageable patient from the "brittle" diabetic. In the maturity-onset diabetic the secretory failure is less severe. Basal insulin concentration is generally normal or increased, whereas glucose-stimulated insulin secretion is generally diminished. In its mildest forms, the beta-cell defect involves only the initial secretory phase; the slower insulin response remains intact. In such individuals the loss of responsiveness to glucose is demonstrable despite normal responsiveness to secretogogues such as beta aderenergic stimulation. 42 These observations suggest a specific abnormality of the glucose receptor in the beta cell in the earliest stages of diabetes. Other factors which have been implicated in this loss of glucose-stimulated insulin secretion are an increase in alpha-adrenergic activity43 and increased prostaglandin release. 44 Unopposed alpha adrenergic activity caused by chronic administration of propranolol has resulted in further reduction of insulin secretion in maturity onset diabetes mellitus and precipitation of hyperosmolar nonketotic coma. 3Se Of note is the fact that reduced insulin secretion in diabetes mellitus cannot be explained on the basis of diminished secretion of gastrointestinal hormones. In fact, GIP release after glucose ingestion is actually augmented in diabetes. 46 It is of interest that the initial reports by Berson and Yalow regarding plasma insulin levels in maturity-onset diabetes emphasized the presence of hyperinsulinemia. 66 The seeming hyperinsulinemia of the
ROBERT SHERWIN AND PHI LIP FELIG
maturity-onset diabetic was subsequently shown to be more apparent than real when body weight and ambient blood glucose levels are considered. Approximately 80 per cent of maturity-onset diabetics are obese, and the incidence of diabetes in obese individuals is substantially greater than in the general population. Obesity per se is accompanied by hyperinsulinemia and is associated with resistance on the part of target tissues (muscle, liver and adipose tissue) to the action of insulin (see below). In fact, when comparison is made between obese maturity-onset diabetics and an appropriate weight-matched nondiabetic control group, it becomes apparent that insulin levels in obese diabetics are clearly below those observed in obese subjects with normal glucose tolerance. 26 Obesity acts as a diabetogenic factor because the accompanying insulin resistance increases the demand for insulin. This increased demand thus unmasks a deficient secretory mechanism in those individuals with an underlying (presumably inherited) defect in beta cell function. In the non-obese maturity-onset diabetic, insulin secretion is somewhat variable. In most such patients the seeming hyperinsulinism is a consequence of the accompanying hyperglycemia. 26 In other words, when the glucose tolerance curve of the mild diabetic is simulated in the normal individual the insulin response is substantially greater than that of the diabetic. On the other hand, in some mild maturity-onset diabetics hyperinsulinemia may not be accounted for by either obesity or hyperglycemia. In such patient~ altered sensitivity to insulin (insulin resistance) may be an important pathogenetic factor. 4o Insulin Resistance in Diabetes Mellitus The biological action of insulin is not only related to the concentration of this hormone in the interstitial fluid of target tissues,52 but is also a function of the ability of insulin to activate cellular events. The first step in the cellular action of insulin is the binding of the hormone to a specific receptor on the cell surface. 22 Resistance to endogenous and exogenous insulin in obese man is associated with a reduction in insulin receptors on adipocytes and circulating mononuclear cells. 35 A decrease in the concentration of insulin receptors has also been reported in the plasma membranes of adipocytes,2l muscle cells,58 and hepatocytes 36 from obese animals. On the basis of these studies it has been suggested that decreased insulin binding to target tissues contributes to the insulin resistance observed in obesity.2 In fact, recent studies have demonstrated a remarkably close correlation between insulin binding to circulating monocytes and the in vivo sensitivity to exogenous insulin in nonobese and obese subjects when studied in the postabsorptive state. 13 Altered binding, however, is only one of many factors which may influence the action of insulin. For example, following starvation and refeeding the close relationship between insulin binding and insulin action is lost.1 3 The rate-limiting step in such circumstances involves post-receptor intracellular events and under these conditions insulin binding improves while sensitivity to insulin decreases.1 3
With respect to insulin resistance in diabetes, reduced insulin binding to circulating mononuclear cells has been observed in maturity-onset diabetics, particularly when basal hyperinsulinemia is present.3 4 Furthermore, in vivo resistance to the action of insulin has been reported in such patients using an infusion technique which employs exogenous insulin, glucose, epinephrine (to block endogenous insulin release), and propranolol (to block the contra-insulin effects of the infused epinephrine).40 In contrast, in juvenile-onset diabetics and in more severely hyperglycemic maturity-onset diabetics in whom there is absolute insulinopenia, insulin binding is not reduced and the response to the insulin-glucose-epinephrine-propranolol infusion is also normal.2 4. 38 It should be noted, however, that the epinephrine-propranolol infusion procedure may not be a definitive assessment of insulin action since variations (between diabetics and non-diabetics) in the response to the hyperglycemic effects of epinephrine have not been excluded. Overall, the available data suggest the presence of heterogeneity with respect to insulin sensitivity as well as insulin secretion in maturity-onset diabetes. In most circumstances maturity-onset diabetes is brought about by a failure of insulin secretion to keep pace with the augmented demands for insulin engendered by obesity. However, in some patients insulin resistance may be present even in the absence of obesity and may bring about diabetes despite hypersecretion of insulin. Metabolic Effects of Insulin Insulin is the primary factor which controls the storage and metabolism of ingested metabolic fuels. After a meal, secretion of insulin facilitates the uptake, utilization, and storage of glucose, fat, and amino acids. Conversely, a reduction in circulating insulin leads to mobilization of endogenous fuels and reduced uptake of ingested nutrients. The action of insulin involves three major metabolic fuels, carbohydrate, protein, and fat, and occurs in three principal tissues, liver, muscle, and adipose tissue. In each of these tissues there are anticatabolic as well as anabolic effects of insulin which act to reinforce each other (Table 1). Table 1.
Action of Insulin
Decreased glycogenolysis Decreased gluconeogenesis Decreased ketogenesis
Decreased protein catabolism Decreased amino acid output
Increased glycogen synthesis Increased fatty acid synthesis
Increased glycerol synthesis Increased fatty acid synthesis
Increased amino acid uptake Increased protein synthesis Increased glycogen synthesis
700 Table 2.
ROBERT SHERWIN AND PHILIP FELIG
Disposal of a 100 Gram Oral Glucose Load in Normal Postabsorptive Man';'
Hepatic Uptake Glycogen synthesis Triglyceride synthesis Glycolysis Hepatic Glycogen-Sparing Uptake of ingested glucose by noninsulin-dependent tissues (brain and blood cells) thereby sparing liver glycogen. Increased Peripheral Utilization of Glucose Insulin-dependent glucose uptake by fat and muscle tissue.
*Based on the data of Felig et al."
CARBOHYDRATE. The liver is quantitatively the most important site of insulin action in the disposal of an oral glucose load. Since glucose is absorbed via the portal system, the extent to which an oral glucose load is available for uptake by peripheral tissues depends upon its escape from the splanchnic bed. During the 3 hour period following the ingestion of 100 gm of glucose, approximately 60 gm are retained within the liver}7 The glucose retained within the liver is used for glycogen synthesis and triglyceride formation. In fact in human beings, the liver is quantitatively a far Ipore important site of conversion of dietary carbohydrate into fat than is adipose tissue. The total amount of glucose escaping hepatic uptake (40 gm) exceeds the basal rate of hepatic glucose output by only 15 gm. Thus, the total amount of glucose made available for peripheral (i.e., nonhepatic) insulin-dependent utilization is only 15 per cent of the ingested load (Table 2). A 100 gram oral glucose load is, however, not representative of either the magnitude of carbohydrate intake or the amplitude of blood glucose excursions observed in healthy individuals ingesting mixed meals. In circumstances of mixed meal intake, the blood glucose in normal man generally varies by no more than 30 to 40 mg per 100 ml over 24 hours.5o This "fine tuning" of blood glucose regulation is determined by the exquisite sensitivity of the liver to small changes in insulin secretion. When the blood glucose rises by only 10 to 15 mg per 100 ml, there is 60 to 100 per cent increase in peripheral insulin levels and a virtually complete inhibition of hepatic glucose output (i.e., sparing of hepatic glycogen) with no stimulation of peripheral glucose utilization. 16 Thus as compared to the liver, muscle and adipose tissue represent relatively minor sites of disposal of large glucose loads, and are less sensitive than the liver with respect to responses to small increases in plasma insulin. Nevertheless, with significant peripheral hyperinsulinemia, glucose uptake by fat and muscle tissue helps minimize the fluctuations in systemic blood glucose levels. The rise in insulin which accompanies carbohydrate feeding acts not only to promote glucose uptake and storage by the liver, but also
serves to inhibit gluconeogenesis. This action of insulin involves an inhibition of hepatic uptake of alanine, the key gluconeogenic precursor. Interestingly, the inhibition of gluconeogenesis requires greater concentrations of insulin than does inhibition of glycogenolysis.1 2 , 16 In addition to regulating carbohydrate disposal in the fed state, the basal concentration of insulin plays an important role in maintaining glucose homeostasis following an overnight fast. By its restraining effects on hepatic glycogenolysis and gluconeogenesis and its ability to minimize the stimulatory influence of glucagon on these processes, the basal insulin levels insure that glucose production is maintained within relatively narrow limits (2 to 3 mg per kg per min). In this manner glucose is made available to the brain while hyperglycemia is prevented. PROTEIN. Since muscle is in negative nitrogen balance in the fasting state, repletion of muscle nitrogen depends on a net uptake of amino acids in response to protein feeding (Fig. 2). The transfer of amino acids from the gut to muscle after protein ingestion is facilitated by the action of insulin. 19 A special role for the branched chain amino acids (leucine, isoleucine, and valine) has been demonstrated in muscle nitrogen metabolism after protein feeding. These amino acids are unique in their ability to escape hepatic uptake and/or metabolism after intestinal absorption. Valine, isoleucine, and leucine together account for more than 60 per cent of the total amino acids entering the systemic circulation after a protein meal. 64 Complementing the pattern of splanchnic
FASTE:D (POSTABSORPTIVE) STATE
Figure 2. Interorgan exchange of amino acids in normal man in the postabsorptive (overnight fasted) state and after protein feeding. In the postabsorptive state alanine and glutamine are released by muscle and alanine is released from the gut and kidney in small quantities. Alanine is taken up by liver where its carbon skeleton is used for glucose production. Glutamine is taken up by kidney where it contributes nitrogen for ammonia production, and by the gut where it is converted to alanine. In the protein-fed state alanine and glutamine continue to be released from muscle. Repletion of muscle nitrogen occurs via uptake of branched chain amino acids in ingested protein. The remainder of the amino acids in the protein meal are largely catabolized in the liver and are not delivered to muscle tissue.
AI.ni.. ... MUICII----..;::.::::::::.:..------
ROBERT SHERWIN AND PHILIP FELIG
amino acid output, the branched chain amino acids are responsible for most of the amino acid uptake after a protein meal. They account for more than half of the total peripheral amino acid uptake in the first hour after protein ingestion, and greater than 90 per cent at 2 to 3 hours.64 Interestingly, the high circulating and intracellular levels of branched chain amino acids may have importance beyond delivery of nitrogen. Recent data suggest that the branched chain amino acids have a specific regulatory role in enhancing protein synthesis. 10 The rise in plasma insulin which accompanies ingestion of a protein meal is the key factor in the uptake of ingested amino acids by muscle tissue. The stimulatory effect of insulin on muscle amino acid uptake is most marked for the branched chain amino acids. 39 Secretion of endogenous insulin or infusion of exogenous insulin results in an exaggerated decline in circulating branched chain amino acids 15 and a marked increase in the turnover of these amino acids in normal man. 55 Furthermore, administration of oral or intravenous glucose during protein feeding blunts the expected rise in branched chain amino acids. Thus by augmenting insulin secretion, glucose feeding facilitates the uptake of ingested branched chain amino acid by muscle. 55 Furthermore, the presence of insulin accelerates the rate of protein synthesis in addition to stimulating the intracellular transport of amino acids. 31 Besides its effects on muscle protein metabolism in the fed state, insulin is a key regulator of muscle nitrogen balance in the interval between feeding. Insulin is known to inhibit protein catabolism and the output of amino acids from muscle. 23 ,39 The restraining effects of insulin on muscle amino acid outflow thus serve to limit the rate of dissolution of body protein. Furthermore, this action of insulin in reducing the availability of circulating amino acids complements its direct inhibitory effect on hepatic gluconeogenesis. F AT. Fat homeostasis as it occurs in normal man is shown in Figure 3. Insulin accelerates the removal by adipose tissue of circulating triglycerides derived from exogenous or endogenous sources by its stimulatory effect on lipoprotein lipase. In addition, insulin is extremely effective in inhibiting a hormone-sensitive lipase within the fat cell, which catalyzes the hydrolysis of stored triglycerides and the liberation of free fatty acids. This antilipolytic action of insulin occurs at concentrations of insulin below those necessary to affect glucose transport. 67 Furthermore, a large proportion of insulin-mediated glucose uptake in the fat cell is utilized for the formation of alpha-glycerophosphate which is necessary for esterification of fatty acids to form triglycerides. Smaller amounts of insulin-stimulated glucose uptake in adipose tissue are used for in situ synthesis of fatty acids. However, as noted above the major site of insulin-mediated fat synthesis is the liver. The net effect of the antilipolytic, fat synthetic, and glycerogenic actions of insulin is to increase total fat storage. In addition to affecting fatty acid metabolism, insulin has a profound suppressive effect on circulating blood ketones. The formation and accumulation of ketone acids (beta-hydroxybutyrate and acetoace-
INTAKE and PRODUCTION
FFA CHYLOMICRON- TG
Figure 3. Intake, production (synthesis), storage, and utilization of fat in normal man. Normal fat homeostasis is dependent on the action of insulin. Within the liver insulin stimulates the synthesis of fatty acids (FF A) from glucose and their esterification to form triglycerides (TG). Both exogenously derived (dietary) triglycerides (chylomicron-TG) and endogenously synthesized triglycerides (lipoprotein-TG) are a source of fatty acid delivery to adipose tissue. Insulin accelerates the uptake of FF A by adipose tissue by its stimulatory effect on lipoprotein lipase. Fat storage within adipose tissue is also enhanced by insulin's glycerogenic effects and to a lesser extent by its stimulation of fat synthesis from glucose. The antilipolytic actions of insulin (inhibition of tissue lipase) also enhance fat storage while reducing the availability of circulating fatty acid substrates. Free fatty acids are released from adipose tissue when insulin levels fall (e.g., during fasting, exercise) and are taken up by muscle, heart, kidney and liver.
tate) is a consequence of three distinct metabolic events: (1) delivery of fatty acids from adipose tissue; (2) hepatic oxidation of free fatty acids to acetyl CoA which is converted to ketones (ketogenic capacity) and (3) a reduction in ketone utilization by peripheral tissues. As already noted, insulin is a powerful antilipolytic hormone. In addition, insulin decreases the liver's capacity to oxidize fatty acids irrespective of substrate availability.4 This antiketogenic action of insulin is intimately related to (a) its regulation of hepatic carnitine levels, and (b) its stimulation of fatty acid synthesis. The oxidation of fatty acids by the liver is augmented by increasing the level of carnitine, which is necessary for the transfer of fatty acids across the mitochondrial membrane (by the enzyme acyl-carnitine transferase) to the site of beta-oxidative enzymes. 32 The carnitine transferase enzyme is however, inhibited by malonyl CoA, the first intermediate product in the synthesis of fatty acids from acetyl CoA. The effect of insulin is to lower hepatic carnitine levels while increasing (by virtue of its stimulation of fat synthesis) the level of malonyl CoA.33 The net result is a reduction in the ketogenic capacity of the liver. These effects of insulin on fat mobilization and oxidation are accompanied by actions on ketone utilization. In the presence of insulin the uptake and oxidation of ketone acids by muscle tissue is accelerated. 54 Metabolic Dysfunction in Diabetes The metabolic alterations observed in diabetes primarily reflect the degree to which there is an absolute or relative deficiency of insulin. Viewed in the context of the role of insulin as the major storage hormone, a minimal deficiency results in a diminished ability to effec-
ROBERT SHERWIN AND PHILIP FELIG
tively increase the storage reservoir of body fuels because of inadequate disposal of ingested foodstuffs (e.g., glucose intolerance). With a major deficiency of insulin, not only is fuel accumulation hampered in the fed state, but excessive mobilization of endogenous metabolic fuels also occurs in the fasting condition (e.g., fasting hyperglycemia, hyperaminoacidemia, and elevated fatty acids). In its most severe form (diabetic ketoacidosis) there is overproduction of glucose and marked acceleration of catabolic processes (lipolysis, proteolysis) (Fig. 4). CARBOHYDRATE. In the patient with mild maturity-onset diabetes, insulin deficiency is apparent only with respect to the initial stimulative response to glucose. The fasting blood glucose level is consequently normal since basal insulin secretion is adequate to prevent glucose overproduction or underutilization in the postabsorptivestate. However, ingested glucose fails to elicit adequate early insulin release and consequently glucose is not taken up by the liver in normal amounts and is more slowly metabolized in the periphery. The quantity of glucose that escapes uptake by the liver and enters the systemic circulation after oral glucose ingestion is increased nearly twofold in the maturity-onset diabetic. 20 The net result of this defect in glucose uptake is postprandial hyperglycemia. Interestingly, the major factor responsible for 75 per cent of this accumulation of blood glucose is decreased hepatic uptake of glucose. 2o When absolute or relative insulin deficiency occurs in the basal state, an elevation in fasting blood glucose ensues. In this circumstance normal basal levels of insulin may be maintained, but only at the expense of fasting hyperglycemia. In such patients, glucose production (determined either by radioactive tracer techniques or splanchnic balance studies) is generally normal or only slightly increased, while fractional glucose turnover is reduced. 41 ,63 Since only mild hyperglycemia in a normal individual is sufficient to inhibit hepatic glucose output,16 the diabetic with fasting hyperglycemia is always in a state of relative or absolute glucose overproduction. Furthermore, the Figure 4. Alterations in carbohydrate, protein, and fat metabolism accompanying severe insulin deficiency. Hyperglycemia results from reduced hepatic and peripheral glucose uptake and accelerated glucose production by the liver. Augmented amino acid release from muscle and hepatic extraction of these amino acids and the availability of energy-yielding equivalents derived from fatty acid oxidation account for hepatic glucose overproduction. Hypertriglyceridemia largely results from diminished triglyceride uptake by adipose tissue. Hyperketonemia is a consequence of accelerated conversion of fatty acids to ketones and reduced ketone uptake by muscle. Branched chain amino acid (BCAA) uptake by muscle is reduced, leading to diminished repletion of muscle nitrogen and hyperaminoacidemia.
Correlation Between Changes in Insulin Secretion and Glucose Regulation in Diabetes GLUCOSE INSULIN SECRETION
Abnormal glucose tolerance Fasting hyperglycemia Ketoacidosis
relative contribution of gluconeogenesis is increased twofold. 63 This enhancement of gluconeogenesis with moderate deficiencies of insulin is in keeping with the relatively greater amounts of insulin necessary to inhibit gluconeogenesis as compared to glycogenolysis. In the extreme situation of total beta cell failure, an ever-increasing fasting blood glucose level fails to elicit a secretory response. In the absence of the restraining influence exerted by insulin, hepatic glucose production may increase threefold or more above normal (largely as a consequence of accelerated gluconeogenesis) and glucose turnover is further reduced. The clinical correlate of this sequence of events is severe hyperglycemia as is observed in diabetic ketoacidosis or nonketotic hyperosmolar coma. The various gradations of disordered carbohydrate metabolism are shown in Table 3. PROTEIN. Alterations in protein metabolism in the fasted as well as the protein-fed state are demonstrable in the diabetic even in the absence of severe insulin deficiency. In such patients plasma alanine is reduced and the hepatic uptake of this key glycogenic amino acid as well as other glucose precursors is increased twofold more. As a consequence of increased hepatic extraction and conversion of substrates to glucose, gluconeogenesis can account for over 30 to 40 per cent of hepatic glucose production in the diabetic as compared to 15 to 20 per cent in normal man. 63 In addition to the increased utilization of amino acids for gluconeogenesis in the fasted state, repletion of muscle nitrogen after protein feeding is reduced in the diabetic. In contrast to the ongoing uptake of branched chain amino acids observed in normal subjects after a protein meal, in diabetes a net uptake is observed only transiently. As a consequence, the total uptake of these amino acids by muscle is markedly decreased and the levels of branched chain amino acids are consistently elevated following protein ingestion. Diabetes thus may be viewed as a disorder of protein tolerance as well as glucose tolerance. 64 These observations are in keeping with the known stimulatory effect of insulin on branched chain amino acid transport (see above). In addition, the turnover of leucine in diabetic man is reduced and is. restored to normal by administration of insulin. 55 While overall uptake of branched chain amino acids is reduced in diabetes,55 those amino acids
ROBERT SHERWIN AND PHILIP FELIG
which do enter muscle tissue are preferentially oxidized rather than utilized for protein synthesis. 11 While the alterations in gluconeogenesis and the protein intolerance associated with mild to moderate insulin lack are not generally discernible in the clinical laboratory, the consequences of severe insulin deficiency on protein metabolism are readily apparent. The stunted growth of the juvenile diabetic observed in the preinsulin era and the negative nitrogen balance and protein wasting of the diabetic in ketoacidosis represent obvious clinical examples. The augmented protein catabolism and accelerated output of amino acids from muscle also accentuate glucose overproduction in diabetes, since availability of protein-derived gluconeogenic substrate is dramatically increased. FAT. Abnormalities in fat metabolism are observed in diabetic patients with relatively mild insulin deficiency. Under these circumstances, circulating triglycerides are frequently increased presumably as a consequence of reduced disposal of both exogenous and endogenously derived triglycerides. Accelerated mobilization of body fat stores and ketone accumulation are observed under conditions of severe insulin deficiency. The failure to observe ketosis in many diabetics with fasting hyperglycemia derives from the exquisite sensitivity of lipolytic processes within the fat cell, ketogenic processes in the liver, and ketone uptake by muscle to small quantities of insulin. When insulin secretion is inadequate to suppress lipolysis, fatty acids are mobilized in increased quantities for adipose tissue. Hepatic metabolism of these fatty acids supplies energy for gluconeogenesis and substrate for the generation of ketone bodies. In addition, lack of insulin results in increased activity of the acylcarnitine transferase step within the liver. This stimulation is due to increased levels of carnitine and diminished concentrations of malonyl CoA (see above).33 As a result, the fatty acids delivered to the liver are rapidly oxidized and converted to ketones. Alterations in muscle ketone metabolism also contribute to ketone accumulation in the diabetic state. Interestingly, decreased ketone turnover has been observed in some diabetics with normal ketone production, suggesting that the rate of ketone utilization may be a more sensitive index of insulin deficiency than ketone overproduction. 54 Hyperketonemia in diabetes is thus a consequence of changes in metabolic processes in adipose tissue, liver and muscle (Fig. 5). The Role of Glucagon It has been suggested that the metabolic abnormalities associated with diabetes result not from insulin lack by itself, but rather from a bihormonal disturbance of alpha cell and beta cell function. 62 The importance of glucagon in the development of the diabetic syndrome is suggested by the demonstration that suppression of glucagon by glucose is lost in diabetes and that protein-stimulated glucagon secretion is augmented. However, the presence of relative or absolute hyperglucagonemia does not of itself prove that this hormone is a necessary or primary factor in the abnormal fuel homeostasis of diabetes. In fact,
PATHOPHYSIOLOGY ADIPOSE TISSUE
! KETONE UTILIZATION /fr7/:/J lift 0<,
.. KETOGENESIS ......
Figure 5. The development of hyperketonemia in diabetes is a consequence of three distinct metabolic events: (1) accelerated delivery of free fatty acids (FFA) from adipose tissue: (2) augmented beta-oxidation of free fatty acids to ketones as a result of elevated cartinine levels and reduced concentrations of malonyl CoA; and (3) a reduction in ketone utilization by muscle. Each of these processes is reversed by the action of insulin.
recent studies indicate that glucagon contributes to the diabetic state only under circumstances of absolute insulin deficiency. 53 With respect to normal physiology, the major role of glucagon is to prevent hypoglycemia during non-glucose (e.g., protein)-stimulated insulin secretion. 60 The rise in glucagon observed after a protein meal permits the accompanying rise in insulin to facilitate the uptake of ingested amino acids without running the risk of hypo glycemia. In the postabsorptive state, glucagon also may act to antagonize the inhibitory action of even basal levels of insulin on hepatic glucose production. 29 With respect to the pathogenesis of diabetes, elevations in plasma glucagon (within the range observed in most hyperglucagonemic states) have no effect on glucose tolerance in normal man or in diabetic patients so long as endogenous or exogenous insulin is available. 53 On the other hand, glucagon contributes to endogenous hyperglycemia when insulin is deficient. 53 While the stimulatory effect of glucagon on hepatic glucose production is transient (less than 45 minutes)18 and of similar magnitude in normal and diabetic subjects,9 the glycemic response is excessive in the diabetic. 53 This is because insulin lack precludes rapid disposal of glucose transiently released by the liver. Thus, the exaggerated glycemic response to glucagon in diabetes is dependent on insulin deficiency.1 4 Similarly, hyperglucagonemia is insufficient to cause hyperketonemia in normal man (even if free fatty acid delivery is increased) or in diabetics in whom insulin is available. 49 ,53 Glucagon can, however, accelerate the development of ketosis in circumstances of absolute insulin lack. 48 The contribution of glucagon to hyperketonemia derives solely from this hormone's capacity to augment the hepatic conversion of fatty acids to ketones. Thus, glucagon secretion may exaggerate the metabolic alterations accompanying severe insulin deficiency. However, relative or absolute insulin lack is the essential factor necessary for the changes in fuel mobilization and utilization which characterize the diabetic state. 14 In addition to questions regarding the capacity of glucagon to bring about or worsen the diabetic state, the possibility of an "essential" role for glucagon in diabetes has been raised. 61 ,62 Such conclusions were based largely on observations involving the hypoglycemic response to relatively short term infusions of somatostatin, a tetradecapeptide iso-
ROBERT SHERWIN AND PHILIP FELIG
SRIF or SALINE INFUSION 0.4 TOTAL
BRANCHED CHAIN 0.2 AMINO ACIDS (fLM)
,B-HYDROXYBUTYRATE 400 (mM)
Figure 6. Effect of somatostatin (SRI F) administration on blood beta-hydroxybutyrate total plasma branched chain amino acids and plasma glucose in maturity onset diabetics. Somatostatin produced a prompt increase in betahydroxybutyrate and total branched chain amino acids and delayed hyperglycemia despite 50 per cent suppression of plasma glucagon concentration. There was an inverse correlation between the degree of hyperglycemia and the plasma insulin concentration (based on the data of Tamborlane et al."').
(mg/ 100 ml)
120 180 240 TIME (min)
lated from the hypothalamus, pancreatic islet cells and gastrointestinal tract, which is a potent inhibitor of glucagon as well as insulin secretion. Subsequent studies with prolonged infusions of somatostatin have shown that ongoing insulinopenia results in fasting hyperglycemia despite glucagon suppression. 56 Furthermore, in maturity-onset diabetics, prolonged infusions of somatostatin are accompanied by an intensification (rather than an amelioration) of hyperglycemia, hyperaminoacidemia, and hyperketonemia despite suppression of glucagon (Fig. 6).59 In like manner, in pancreatectomized human beings, insulin withdrawal results in hyperglycemia and ketosis despite the absence of detectable circulating glucagon. 3 Thus while glucagon may worsen the metabolic consequences of insulin deficiency its presence is not essential for clinical diabetes to be manifest. 14a, 56 Somatostatin In the above discussion some of the hormonal and metabolic effects of infusion of exogenous somatostatin have been described. In addition to its action on insulin and glucagon secretion, somatostatin interferes with carbohydrate absorption,65 reduces gastric motility,27
and inhibits the secretion of gastrin and secretion. 7 ,8 Recent in vitro studies have revealed that glucose stimulates the release of somatostatin from perfused islet cells. 25 Whether such changes occur in vivo and whether the magnitude of the response in circulating levels of somatostatin is sufficient to alter gastrointestinal or pancreatic alpha and beta cell function remains to be established. Clarification of the physiologic role of somatostatin in body fuel metabolism awaits the determination by reliable assay of the circulating concentration of this peptide and subsequent examination of the metabolic response to physiologic doses.
REFERENCES 1. Andres, R., Cader, G., and Zierler, K. L.: The quantitatively minor role of carbohydrate in oxidative metabolism by skeletal muscle in intact man in the basal state. Measure· ments of oxygen and glucose uptake and carbon dioxide and lactate production in the forearm. J. Clin. Invest., 35:671-682, 1956. 2. Bar, R S., Gorden, P., Roth, J., et a!.: Fluctuations in the affinity and concentrations of insulin receptors on circulating monocytes of obese patients. J. Clin. Invest., 58:11231135, 1976. 3. Barnes, A. ]., Bloom, S. R, Alberti, K. G. M. M., et a!.: Ketoacidosis in pancreatec· tomized man. New EngI. J. Med., 296:1250-1253,1977. 4. Bieberdorf, F. A., Chernick, S. S., and Scow, R 0.: Effect of insulin and acute diabetes on plasma FFA and ketone bodies in the fasting rat. J. Clin. Invest., 49:1685-1693, 1970. 5. Block, M. B., Mako, M. E., Steiner, D. F., et al.: Circulating C-peptide immunoreactivity. Studies in normals and diabetic patients. Diabetes, 21 :1013-1018,1972. 6. Block, M. B., Mako, M. E., Steiner, D. F., et al.: Diabetic ketoacidosis: Evidence for C·peptide and proinsulin secretion following recovery. J. Clin. EndocrinoI. Metab., 35:402-406,1972. 7. Bloom, S. R, Montimer, C. H., Thorner, M. 0., et al.: Inhibition of gastrin and gastric-acid secretion by growth hormone-release-inhibiting hormone. Lancet, 2:11061109,1974. 8. Boden, G., Switz, M. C., Owne, O. E., et al.: Somatostatin suppresses secretion and pancreatic exocrine secretion. Science, 190:163-164, 1975. 9. Bomboy, J. D., Jr., Lewis, S. B., Lacy, W. W., et a!.: Transient stimulatory effect of sustained hyperglucagonemia on splanchnic glucose production in normal and diabetic man. Diabetcs, 26:1 77-184, 1977. 10. Buse, M. G., and Reid, S. S.: Leucine: A possible regulator of protein turnover in muscle. J. Clin. Invest., 56:1250-1261, 1975. 11. Buse, M. G., Herlong, H. F., and Weigand, D. A.: The effects of diabetes, insulin, and the redox potential on leucin metabolism in isolated rat hemidiaphragm. Endocrinology, 98:1166-1175,1976. 12. Chiasson, J. L., Liljenquist, J. E., Finger, F. E., et al.: Differential sensitivity of glycogenolysis and glucogenogenesis to insulin infusions in dogs. Diabetes, 25 :283291, 1976. 13. DeFronzo, R, Sherwin, R, Soman, V., et al.: Dissociation of insulin binding to monocytes and insulin action in intact man during starvation and refeeding. Clin. Res., 25:494A, 1977. 14. Dupre, J., Ross, S. A., Watson, D., et al.: Stimulation of insulin secretion by gastric inhibitory polypeptide in man. J. Clin. EndocrinoI. Metab., 37:826-828, 1973. 14a. Felig, P., Wahren, J., Sherwin, R, et a!.: Insulin, glucagon, and somatostatin in normal physiology and diabetes mellitus. Diabetes, 25:1091-1099,1976. 15. Felig, P., Marliss, E., and Cahill, G. F., Jr.: Plasma amino acid levels and insulin secretion in obesity. New EngI. J. Med., 281 :811-815, 1969. 16. Felig, P., and Wahren, J.: Infleunce of endogenous insulin secretion on splanchnic glucose and amino acid metabolism in man. J. Clin. Invest., 50:1702-1711, 1971. 17. Felig, P., Wahren, J., and Hendler, R: Influence of oral glucose ingestion on splanchnic glucose and gluconeogenic substrate metabolism. Diabetes, 24:468-475,1975. 18. Felig, P., Wahren, J., and Hendler, R.: Influenceof physiologic hyperglucagonemia on basal and insulin inhibited splanchnic glucose output in normal man. J. Clin. Invest., 58:761-765,1976.
ROBERT SHERWIN AND PHILIP FELIG
19. Felig, P., Wahren, J., Sherwin, R., et al.: Amino acid and protein metabolism in diabetes mellitus. Arch. Intern. Med., 137:507-513, 1977. 20. Felig, P., Wahren, J., and Hendler, R.: Influence of maturity-onset diabetes on splanchnic glucose balance after oral glucose ingestion. Diabetes, in press. 21. Freychet, P., Laudat, M. H., Laudat, P., et a!.: Impairment of insulin binding to the fat cell plasma membrane in the obese hyperglycemic mouse. Fed. Europ. Biochem. Soc. Lett., 25:339-342,1972. 22. Freychet, P.: Interactions of polypeptide hormones with cell membrane specific receptors: Studies with insulin and glucagon. Diabetologia, 12:83-100, 1976. 23. Fulks, R. M., Li, J. B., and Goldberg, A. L.: Effects of insulin, glucose, and amino acids on protein turnover in rat diaphragm. J. Bio!. Chem., 250:290-298, 1975. 24. Ginsberg, H. N.: Investigation of insulin sensitivity in treated subjects with ketosisprone diabetes mellitus. Diabetes, 26:278-283, 1977. 25. Ipp, E., Dobbs, R. E., Arimura, A., et a!.: Release of immunoreactive somatostatin from the pancreas in response to glucose, amino acids, pancreozymin-cholecystokinin, and tolbutamide. J. Clin. Invest., 60:760-765, 1977. 26. Kipnis, D. M.: Insulin secretion in normal and diabetic individuals. Adv. Intern. Med., 16:103-134,1970. 27. Konturek, S.: Somatostatin and the gastrointestinal secretions. Scand. J. Gastroentero!., 11: 1-4, 1976. 28. Lacy, P. E., Howell, S. L., Young, D. A., et al.: New hypothesis of insulin secretion. Nature, 219:1177-1179, 1968. 29. Liljenquist, J. E., Mueller, G. L., Cherrington, A. D., et al.: Evidence from an important role for glucagon in the regulation of hepatic glucose production in normal man. J. Clin. Invest., 59:369-374, 1977. 30. Malaisse, W. J., Sener, A., Koser, M., et al.: Stimulus-secretion coupling of glucose-induced insulin release. Metabolism of a and ,B-D-glucose in isolated islets. J. Bio!. Chem., 251 :5936-5943, 1976. 31. Manchester, K. L., and Young, F. G.: The effect of insulin on incorporation of amino acids into protein of normal rat diaphragm. Biochem. J., 70:353-358, 1958. 32. McGarry, J. D., and Foster, R. P.: Hormonal control of ketogenesis: Biochemical considerations. Arch. Intern. Med., 137:495-501, 1977. 33. McGarry, J. D., Mannaerts, G. P., and Foster, D. W.: A possible role for malonyl-COA in the regulation of hepatic fatty acid oxidation and ketogenesis. J. Clin. Invest., 60:265270, 1977. 34. Olefsky, J. M., and Reaven, G. M.: Decreased insulin binding to lymphocytes from diabetic patients. J. Clin. Invest., 54:1323-1328, 1974. 35. Olefsky, J. M.: Decreased insulin binding to adipocytes and circulating monocytes in obesity. J. Clin. Invest., 57:1165-1172,1976. 36. Olefsky, J. M., Bacon, V. C., and Baur, S.: Insulin receptors of skeletal muscle: specific insulin binding sites and demonstrations of decreased number of sites in obese rats. Metabolism, 25:179-191,1976. 37. Pagliara, A. S., Stillings, S. N., Hover, B., et al.: Glucose modulation of amino acid-induced glucagon and insulin release in the isolated perfused rat pancreas. J. Clin. Invest., 54:819-832,1974. 38. Pedersen, 0., Beck-Nielsen, H., Heding, L., et al.: Insulin receptors in juvenile diabetes. Diabetologia, 13:424, 1977. (Abstract.) 38a. Podolsky, S., Zimmerman, H. J., Burrows, B. A., Cardarelli, J. A., and Pattavina, C. G.: Potassium depletion in hepatic cirrhosis: A reversible cause of impaired growth hormone and insulin response to stimulation. New Eng!. J. Med. 288:644-648, 1973. 38b. Podolsky, S., and Melby, J.: Improvement of impaired growth hormone response to stimulation in primary aldosteronism with correction of potassium depletion. Metabolism, 25:1027-1032,1976. 38c. Podolsky, S., and Pattavina, C. G.: Hyperosmolar nonketotic diabetic coma: A complication of propranolol therapy. Metabolism, 22:685-693, 1973. 39. Pozefsky, T., Felig, P., Tobin, J. D., et a!.: Amino acid balance across the tissues of the forearm in postabsorptive man: Effects of insulin at two dose levels. J. CUn. Invest., 48:2273-2282, 1969. 40. Reaven, G. M., Bernstein, R., Davis, B., et al.: Nonketotic diabetes mellitus: Insulin deficiency or insulin resistance? Amer. J. Med., 60:80-88, 1976. 41. Reichard, G. A., Jr., Moury, N. F., Hochella, N. J., et a!.: Quantitative estimation of the Cori cycle in the human. J. Bio!. Chem., 238:495-501,1963. 42. Robertson, R. P., and Porte, D., Jr.: The glucose receptor. A defective mechanism in diabetes mellitus distinct from the beta adrenergic receptor. J. Clin. Invest., 52:870876,1973. 43. Robertson, R. P., Halter, J. B., and Porte, D., Jr.: A role for alpha-adrenergic receptors in abnormal insulin secretion in diabetes mellitus. J. Clin. Invest., 57:791-795,1976.
44. Robertson, R. P., and Chen, M.: A role for prostaglandin E in defective insulin secretion and carbohydrate intolerance in diabetes mellitus. J. Clin. Invest., 60:747-753,1977. 45. Roth, J.: Peptide hormone binding to receptors. A review of direct studies in vitro. Metabolism, 22: 1059-1073, 1973. 46. Ross, S. A., Brown, J. C., and Dupre, J.: Hypersecretion of gastric inhibitory polypeptide following oral glucose in diabetes mellitus. Diabetes, 26:525-529, 1977. 47. Rubinstein, A. H., Clark, J. L., Melani, F., et al.: Secretion of proinsulin C-peptide by pancreatic f3 cells and its circulation in blood. Nature, 224:697-699,1969. 48. Schade, D. S., and Eaton, R. P.: Glucagon regulation of plasma ketone body concentration in human diabetes. J. Clin. Invest., 56:1340-1344, 1975. 49. Schade, D. S., and Eaton, R. P.: Modulation of fatty acid metabolism by glucagon in man. IV. Effects of a physiologic hormone infusion in normal man. Diabetes, 25:978983,1976. 50. Service, F. J., Molnar, G. D., Rosevear, J. M., et al.: Mean amplitude of glycemic excursions, a measure of diabetic instability. Diabetes, 19:644-655, 1970. 51. Shima, K., Tanaka, R., Morishita, S., et al.: Studies on the etiology of "brittle" diabetes: Relationship between diabetic instability and insulinogenic reserve. Diabetes, 26:717725,1977. 52. Sherwin, R. S., Kramer, K. J., Tobin, J. D., et a!.: A model of the kinetics of insulin in man. J. Clin. Invest., 53:1481-1492,1974. 53. Sherwin, R. S., Fisher, M., Hendler, R., et a!.: Hyperglucagonemia and blood glucose regulation in normal, obese and diabetic subjects. New Engl. J. Med., 294:455-461, 1976. 54. Sherwin, R. S., Hendler, R., and Felig, P.: Effect of diabetes mellitus and insulin on the turnover and metabolic response to ketones in man. Diabetes, 25: 776-784, 1976. 55. Sherwin, R. S., Rosenzweig, J., Soman, V., et al.: Effect of insulin and diabetes on protein and branched chain amino acid utilization. Diabetes, 25 (Suppl. 1):332, 1976. 56. Sherwin, R. S., Hendler, R., DeFronzo, R. A., et a!.: Glucose homeostasis during prolonged suppression of glucagon and insulin secretion by somatostatin. Proc. Natl. Acad. Sci., U.S.A., 74:348-352, 1977. 57. Stein er, D.: Proinsulin and the biosynthesis of insulin. New Engl. J. Med., 280:11061112, 1969. 58. Soli, A. H., Kahn, C. R., and Neville, D. M.: Insulin binding to liver plasma membranes in the obese hyperglycemic (ob/ob) mouse. Demonstration of a decreased number of functionally normal receptors. J. BioI. Chem., 250:4702-4707,1975. 59. Tamborlane, W. V., Sherwin, R. S., Hendler, R., et al.: Metabolic effects of somatostatin in maturity-onset diabetes. New Engl. J. Med., 297:181-183,1977. 60. Unger, R. H., Ohneda, A., Aguillar-Parada, E., et al.: The role of amino genic glucagon secretion in blood glucose homeostatis. J. Clin. Invest., 48:810-822, 1969. 61. Unger, R. H., and Orci, L.: The essential role of glucagon in the pathogenesis of the endogenous hyperglycemia of diabetes. Lancet, 1: 14-16, 1975. 62. Unger, R. H.: Diabetes and the alpha cell. Diabetes, 25:136-151,1976. 63. Wahren, J., Felig, P., Cerasi, E., et al.: Splanchnic and peripheral glucose and amino acid metabolism in diabetes mellitus. J. Clin. Invest., 51: 1870-1878, 1972. 64. Wahren, J., Felig, P., and Hagenfeldt, J.: Effect of protein ingestion on splanchnic and leg metabolism in normal man and in patients with diabetes mellitus. J. Clin. Invest., 57:987-999, 1976. '65. Wahren, J., and Felig, P.: Influence of somatostatin on carbohydrate disposal and absorption in diabetes mellitus. Lancet, 2:1213-1216,1976. 66. Yalow, R. S., and Berson, S. A.: Immunoassay of endogenous plasma insulin in man. J. Clin. Invest., 39: 1157-1175, 1960. 67. Zierler, K. L., and Rabinowitz, D.: Effect of very small concentrations of insulin on forearm metabolism: Persistence of its action on potassium and free fatty acids without its effect on glucose. J. Clin. Invest., 43:950-962,1964. Department of Internal Medicine Yale University School of Medicine 333 Cedar Street New Haven, Connecticut 06510