Inhibitors, Antagonists, and Inactivators in the Etiology of Diabetes Mellitus in Man
I. Arthur Mirsky
I . Introduction 383 I I . Insulinogenesis 384 I I I . Release of Insulin 384
I V . Impairment of Insulinogenesis 384 V . Impairment of Insulin Release 386 V I . Impairment in Transport of Insulin 387
V I I . Inhibitors of Insulin 389 A . Insulin-binding Antibodies 389
B. Inhibitors Associated with Albumin Fraction of Serum Proteins. 390 C. Inhibitors Associated with ai-Globulin Fraction of Blood 391
D . Mechanism of Action of Insulin Inhibitors 392 V I I I . Impairment of Insulin Receptor Sites 393
I X . Antagonists of Action of Insulin 395
X . Excretion of Insulin 396 X I . Degradation of Insulin 397 X I I . Inhibitors of Insulin Degradation 398
X I I I . Summary 401 References 402
I. INTRODUCTION
The metabolic derangement of diabetes mellitus in man is due to an insufficiency of insulin relative to the requirements of the insulin-de
pendent tissues. Such an insufficiency can be attributed to any one or combination of the following: (a) an impairment in the mechanism re
sponsible for the synthesis of insulin by the beta cells of the islets of Langer- hans; (b) an impairment in the mechanism responsible for the release of
383
384 I. A. M I R S K Y
insulin into the circulation; (c) the synthesis of an abnormal insulin mole
cule with decreased hormonal activity; (d) an impairment in the transport of insulin to tissue receptors by insulin-binding components of the plasma;
(e) an inactivation of insulin by inhibitory agents in the circulation; (/) an increase in the production of agents which antagonize the physiological effects of insulin; (g) a block of the insulin receptors of insulin-dependent tissues; and (h) an increase in the rate at which insulin is degraded.
II. INSULINOGENESIS
Insulin is synthesized by the beta cells of the islets of Langerhans of the pancreas. It is comprised of two polypeptide chains linked together by two disulfide bridges (Fig. 1): the A chain consisting of 21 amino acids and an intrachain disulfide link between the half-cystine residues in posi
tion 6 and 11, and the Β chain consisting of 30 amino acids (1). The amino acid sequences of insulins obtained from the pancreas of six species are identical except for the sequence of the 3 amino at positions 8, 9, and 10 of the A chain (Table I ) . Human insulin, however, differs from that of other species in that the C terminal amino acid of the Β chain is threonine instead of alanine (2).
III. RELEASE OF INSULIN
The insulin monomer isolated from the pancreas has a molecular weight of 6000 and is soluble at a neutral pH. In the beta cell, however, insulin is stored as a polymer with a molecular weight of 24,000-48,000. This polymer is insoluble at the pH of the cell and comprises the beta granules.
These granules are enclosed in smooth, membranous capsules in the cytoplasmic matrix of the cell.
Little is known about the mechanism whereby the aggregated insulin stored as beta granules is stimulated to traverse the many structures between the capsule of the granule and the circulation. Although it is probable that the beta cells continuously secrete a small amount of insulin into the circulation, the major stimulus to the release of insulin appears to be an increase in the concentration of circulating glucose. Such an increase is followed by degranulation which is assumed to be due to dis
aggregation and solubilization of the insulin.
IV. IMPAIRMENT OF INSULINOGENESIS
The hypothesis that a primary impairment in the production and/or release of insulin by the beta cells is responsible for the insulin insufficiency
N H2| S S j N H2 N H2 N H2
A Gly.Heu. Vol .Glu . Glu. Cy.Cy. Alo. Ser. Vol. Cy. Ser. Leu. Tyr. Glu. Leu. Glu. Asp. Tyr. Cy. Asp I 2 3 4 5 6 Ι β 9 10 II 12 13 14 15 16 17 18 19 I 21
f f
N H , N H2 S
s
T U ,w
| | I j — — — — — — — — Thi» (humon) B Phe Vol AsDGIuHisLeu.Cy.Gly.Ser.His.Leu.Vol. Glu. Alo. Leu. Tyr. Leu. Vol. Cy. Gly. Glu. Arg. Gly. Phe. Phe. Tyr. Thr. Pro. Lys. Ala (other species)
1 * 2 3 4 * 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Fig. 1. The amino acid sequences of the A and Β chains of insulin ( i ,
S MELLITUS IN MAN 385
386 I. A . M I R S K Y
of human diabetes is held by many. Although some degree of islet damage may be found in most patients who die with diabetes, only rarely is it of sufficient severity to account for a significant impairment in insulino
genesis (8, 4 ) . The major atrophic lesions of the islets are found in the less than 10% of patients who comprise the "growth-onset" type of dia
betes (onset before 20 years of age) rather than in the more than 90% of patients who comprise the "maturity-onset" type of diabetes (onset after 20 years of age) (5, 6). The probability that even the growth-onset type of diabetes does not start with a decrease in the activity of the beta cells is suggested by the normal or hyperplastic islets observed in young pa
tients who die within eight weeks after the onset of the syndrome (7).
Thus, instead of indicating an impaired insulinogenesis, the morphological abnormalities and the low concentrations of insulin that can be extracted from the pancreases of some patients (6, 6) may be the result of a chronic increase in the rates of synthesis and release of insulin. In accord is the frequency with which insignificant or no lesions of the beta cells are found among patients with diabetes (8, 4> 8), the frequency with which pancre
atic lesions occur in the absence of diabetes (8, 4)> the demonstration that some of the lesions are due to the metabolic derangement (9), and the beta cell damage that may follow the hyperglycemia induced by glucose (10) and other agents (11, 12).
V. IMPAIRMENT OF INSULIN RELEASE
The majority of patients with the maturity-onset type of diabetes and children during the first few months after the onset of diabetes develop hypoglycemia in response to tolbutamide (18, H). Since this agent is effective only in the presence of more than 30% of the normal amount of functional islet tissue (16), such hypoglycemic responses suggest that the insulin insufficiency responsible for the existing metabolic syndrome is not due to a primary deficiency in the insulinogenic and insulin-releasing mechanisms. In agreement is the demonstration that the insulin concen
tration and the insulin-like activity of the plasmas of maturity-onset, tolbutamide-responsive patients is essentially normal (16-25) and, as in healthy subjects, increases promptly after the ingestion of a single dose of tolbutamide (20, 28, 26) or of glucose (20, 24-27). In contrast, the plasmas of patients with relatively severe diabetes of the growth-onset type show no insulin or insulin-like activity after tolbutamide (22, 28) or after glucose (27). Yet, during the early phases of the growth-onset diabetes, even larger than normal concentrations of insulin-like activity can be demon
strated (27a). Thus, although insulin may be detected in the circulation of
patients with diabetes during the early phases of the syndrome, it may not be present after many years of the disease.
Nondiabetic subjects and patients with early maturity-onset type of diabetes exhibit essentially the same increase in the concentration of insulin in the plasma the first hour after the ingestion of glucose (25).
During the second hour after the glucose the nondiabetic subjects develop a decrease in both the blood sugar and plasma insulin concentrations. The diabetic subjects, however, show a further increase in the plasma insulin concentration, while the blood sugar remains elevated (25). The per
sistence of hyperglycemia in the presence of relatively high concentrations of insulin suggests that some extrapancreatic factor rather than a pancre
atic deficiency must be responsible for the sustained hyperglycemia.
The possibility that an abnormal insulin with relatively poor hormonal activity is secreted by the patient with diabetes would appear to be obvi
ated by the essentially similar qualitative and quantitative in vivo and in vitro insulin-like effects of the plasmas from normal subjects and from patients with the tolbutamide-responsive maturity-onset type of diabetes.
Likewise, the insulin extracted from the pancreases of diabetic subjects by the commonly employed acid-alcohol procedure does not appear to act differently from that of nondiabetic subjects (5). Since some of the insulin-like effects on isolated tissues can be produced by agents other than the intact molecule of insulin (28-31), it may be erroneous to con
clude that there is no difference in the molecular structure of the insulins secreted by nondiabetic and diabetic subjects. Comparison of the amino acid sequence of insulins prepared from the pancreas of both groups of subjects may provide a more definitive answer.
VI. IMPAIRMENT IN TRANSPORT OF INSULIN
A sustained hyperglycemia in the presence of a persistent and significant increase in the plasma insulin concentration after a standard dose of glu
cose may be due to some aberration in the mechanism whereby insulin is transported in the circulation. That insulin may be transported in some complex linkage is suggested by the insulin-like activity of plasma pro
teins separated by various techniques (32-44)- Since labeled or unlabeled insulin added to plasma migrates electrophoretically between albumin and the α-globulins (33, 39, 44) j it has been proposed that the insulin-like activity associated with the β- and γ-globulin fractions is due to a "bound"
form of insulin (88, 89, 40). In accord is the demonstration that cationic resins can remove the insulin-like activity of normal plasma but do not affect that due to exogenous "free" insulin (82, 84-86). Insulin extracted
388 I. A. MIRSKY
in a similar manner from human and bovine pancreases is bound to a basic protein (46). Accordingly, it has been proposed that such an insulin-basic protein complex may represent the bound insulin of the plasma (84, 85, 45, 46).
Whereas the insulin bound to the β- and γ-globulins of the blood ap
pears to exert its usual effect, the insulin-basic protein complex extracted from the pancreas by cationic resins exerts no insulin-like action on the isolated diaphragm of the rat (86, 45). Dissociation of the complex at pH
10, however, releases the free insulin and permits the usual stimulation of glucose uptake by the diaphragm (46). Likewise, a complex prepared from human blood in a similar manner exhibits no insulin-like action on the isolated rat diaphragm but does do so after apparent dissociation at pH 9.8 (46). The presence of such a bound, biologically inactive form of insulin in the plasma may explain the observation that the insulin-like action of normal plasma on the uptake of glucose by the rat diaphragm is markedly increased after dilution of the plasma (19, 48)- Since the insulin concentration of plasma determined by the immunological procedure is decreased in proportion to the dilution of the plasma (25), it is quite possible that the apparent increase in the biological activity induced by dilution reflects the release of free insulin from its bound, inactive form.
It is pertinent to note, however, that whereas the bound insulin extracted by cationic resins from blood and pancreas exerts no significant action on the rat diaphragm in vitro, it does exert an insulin-like effect on the oxida
tion of glucose by epididymal adipose tissue (45~47). This effect is at
tributed to the presence of a factor in adipose tissue which dissociates the insulin-protein complex (47).
These various observations suggest that insulin may be released by the pancreas and transported in the circulation principally in a bound form which can enhance the utilization of glucose and lipogenesis by the adipose tissues but cannot influence the metabolism of muscles until it is dis
sociated. Thus, the physiological activity exerted by the insulin secreted by the pancreas may be determined by the character of the binding and by the ease with which the insulin complex is dissociated. In accord is the observation that whereas the administration of glucose results in a rapid dissociation of the bound insulin in the circulation of nondiabetic sub
jects, it does not induce the same degree of dissociation of the bound insulin in the blood of patients with diabetes (47a). Accordingly, it is con
ceivable that the formation of some undissociable complex or an impair
ment in the mechanism responsible for the dissociation of the insulin-basic protein complex in vivo may play a role in the pathogenesis of diabetes.
Validation of this and other related hypotheses is dependent upon the ac
cumulation of much more definitive data on the nature of the physical state of insulin in the circulation.
VII. INHIBITORS OF INSULIN
In spite of a normal or even excessive rate of insulinogenesis, hyper
glycemia may result from an inhibition of the action of circulating insulin by some humoral agent. Such inhibition may be due to (a) binding of the insulin with the humoral agent and the formation of a complex which is not accessible to the insulin-dependent tissues, (b) a reaction with the humoral agent which induces a change in molecular structure and thereby renders the insulin physiologically inactive, or (c) an action of the humoral agent on the insulin-dependent tissues which renders them resistant to the action of insulin.
A. Insulin-Binding Antibodies
It has been known for some time that the sera of patients with insulin resistance, i.e., patients who require more than 200 units/day for the treatment of their metabolic derangement, may contain agents which protect animals from insulin-induced hypoglycemia (49-60) and which inhibit the effect of insulin on the utilization of glucose in vitro by the rat diaphragm (58, 61). A variety of studies suggested that the insulin-neutral
izing agent was associated with the γ-globulins of the plasma (54, 55, 58, 59, 62, 68) and had the characteristics of an antibody (50, 51, 56). More recently, it became evident that almost all human subjects who are treated with insulin for longer than three months develop a similar antibody (or antibodies) which inhibits the action of insulin both in vivo and in vitro (88, 64~71). The immunological response is attributed to the relatively minor differences between the amino acid sequence of human insulin and those of the beef, sheep, or pig insulins that are usually administered
(Table I ) . Electrophoretic and ultracentrifugal studies revealed that the
T A B L E I
T H E AMINO A C I D SEQUENCE AT POSITIONS 8, 9, 10 OF THE A CHAIN OF D I F F E R E N T SPECIES OF INSULIN
Position
Species 8 9 10
Cattle Ala Ser Val Pig, sperm whale, man Thr Ser lieu Sei whale Ala Ser Thr Sheep Ala Gly Val Horse Thr Gly lieu
390 I. A. M I R S K Y
apparent inhibition is achieved through a binding of the insulin with one or more fractions of serum proteins in the 7- or inter-/3-gamma globulin zone, irrespective of whether the insulin is injected in vivo or added in vitro to the plasma (38, 66-71). The insulin-antibody complex dissociates slowly and, consequently, intravenously injected insulin is removed more slowly from the circulation of insulin-treated patients than from untreated subjects (33, 64, 69). The fact that the complex does dissociate, however, is responsible for the persistent hypoglycemia that may occur in patients with diabetic acidosis who have been given large quantities of insulin.
Although exogenous I 131
-labeled insulin can be bound to various protein fractions of serum from nondiabetic subjects (38, 44, 71a, 71b), it has been attributed to either radiation damage (83, 71a) or degradation (70) of the insulin. Yet, prolonged incubation of normal serum at 35°C with undam
aged insulin results in the binding of the latter to a fraction which has the same mobility as a2-globulin (71b). Incubation of normal serum alone at 35°C increases the binding of insulin to the a2-globulin fraction on subse
quent incubation with insulin (71b). Accordingly, the binding is due in part at least to a factor which is released in vitro.
B. Inhibitors Associated with Albumin Fraction of Serum Proteins
The existence of humoral agents which can inhibit the action of insulin by some mechanism other than the binding of insulin is suggested by the observation that whereas the plasma of untreated nonketotic patients with a severe diabetes of the growth-onset type may exhibit no insulin-like activity on the uptake of glucose by the rat diaphragm in vitro, it may diminish the action of added insulin (20, 72, 78). After a fourfold dilution, however, such plasma loses its apparent inhibitory action and exerts an insulin-like activity (19, 74)· In contrast, the undiluted plasmas of healthy subjects and of patients with the maturity-onset type of diabetes exhibit an insulin-like activity and exert no inhibitory activity. Likewise, after effective therapy with insulin, the plasma of the patient with the severe growth-onset type of diabetes may exhibit a normal insulin-like action and no inhibitory effect on added insulin. Accordingly, it has been pro
posed that insulin may be present in the plasma of the untreated diabetic patient but is masked by the presence of an inhibitor. Dilution of the plasma is postulated to inactivate the inhibitor and thereby release the insulin (74)-
Since a fourfold dilution of the plasma from normal subjects and from patients with the maturity-onset type of diabetes results in an increase in insulin-like activity, the inhibitor is assumed to be present in normal plasma as well as in that from patients with diabetes but is masked by the presence of relatively large quantities of insulin (74, 75). A similar
conclusion was derived from studies in which insulin was extracted from plasma with a mixture of organic solvents and the native plasma and the extract and the residue were subsequently assayed (60, 76). Irrespective of whether or not the native plasma exhibited any insulin-like activity, the extract stimulated the uptake of glucose by the isolated diaphragm, while the residue inhibited this action.
Fractionation of the plasma from untreated insulin-requiring patients by a rather drastic acid-alcohol procedure which destroys the globulins
(77) revealed the inhibitory activity to reside in the albumin fraction (75).
Although relatively high levels of the insulin inhibitor are found in the plasma of both obese diabetics and prediabetics (78a), similar preparations of albumin from the plasma of normal subjects also exert an inhibitory effect on the action of insulin in vitro (75, 78, 78b). The inhibitory activity of albumin preparations appears to be due to an associated compound of low molecular weight which is resistant to the action of various proteolytic enzymes (78a, 79). Since the albumin fractions of plasma taken from several nondiabetic hypophysectomized patients did not exert any insulin- inhibitory activity in vitro, it has been proposed that the inhibitor either originates in or is dependent upon the integrity of the pituitary gland (78). Likewise, its absence from the blood of adrenalectomized patients suggests that the inhibitor is also dependent upon the presence of adrenal corticosteroids (78a). The relation of this inhibitor to another pituitary- dependent activity found in a variety of subfractions associated with the a2-globulins of normal blood is unknown (80, 81).
C. Inhibitors Associated with ^-Globulin Fraction of Blood
An inhibitor with somewhat different characteristics is found in the serum of some patients with acute severe acidosis who have not received insulin therapy, as well as in some new severe diabetics without acidosis
(73, 82-84)- The inhibitory activity is evident in vitro not only when insulin is added to the serum but also if the diaphragm is pretreated with serum before the addition of insulin. Yet, studies with I
131
-labeled insulin suggest that the inhibitor does not compete with the insulin for binding sites on the muscle (80). The inhibitory activity is not found in the sera of nondiabetic subjects or in patients with diabetes who do not require insulin. In some instances the inhibitory activity of the sera disappears in from 6 to 9 hours after institution of insulin therapy. The activity ap
pears to be dependent upon a compound that migrates electrophoretically with the ai-globulin fraction. It is destroyed by chymotrypsin but not by trypsin; it is stable to repeated freezing and thawing, but is destroyed by heating at 100°C. It appears to be independent of glucagon, somatotropin, or adrenal corticosteroids (82).
392 I. A . M I R S K Y
It is apparent that at least three types of inhibitors of insulin can be detected in the plasma of patients with diabetes. One, associated with the inter-/3-7-globulin fractions of the plasma proteins, is an antibody (or antibodies) which appears after the initiation of insulin therapy and therefore cannot be involved in the pathogenesis of diabetes. Another, consisting of a small molecular weight compound associated with the albumin fraction is found in untreated, insulin-requiring, nonketotic subjects with the growth-onset type of diabetes. The third inhibitor, found in association with the «i-globulin fraction, is detected in the plasma of patients with the ketotic acidosis that may initiate a severe diabetic state. Both the second and third types of inhibitors appear during un
controlled phases of the metabolic derangement of some insulin-requiring patients and disappear after effective insulin therapy. Neither is found in the native plasma of patients with the mild diabetes of the maturity-onset type in whom hyperglycemia persists in spite of ample concentrations of insulin in the plasma. Consequently, it is quite possible that these in
hibitors may be products of the abnormal metabolic state and play no role in the development of the insulin insufficiency responsible for the metabolic derangement. In accord is the appearance of an inhibitor in the plasma of some diabetic patients during an infection (73), as well as the appearance of various types of inhibitors in the circulation of some animals after the production of experimental diabetes. Thus, pancreatectomy re
sults in the appearance of an insulin-inhibitor in the α-globulin fraction of the plasma of the cat (85). Likewise, an inhibitor of the in vitro uptake of glucose by the diaphragm appears in the ^-lipoprotein fraction of serum after the production of alloxan diabetes in the rat (86-88).
D. Mechanism of Action of Insulin Inhibitors
The mechanism whereby circulating inhibitors other than antibodies impair the action of insulin is not known. There is no evidence that in
hibitors other than antibodies combine with insulin and thereby render it biologically inactive; nor is there any evidence that the molecular struc
ture of endogenous insulin is affected by the inhibitor; neither do they act by degrading the insulin (83). It is quite possible that some inhibitors may act by blocking the insulin receptor sites. Irrespective of whether these agents act by binding, neutralization, or blocking of receptor sites, their presence in the circulation should be reflected by a decrease in the responsiveness of the patient to a standard dose of insulin. Yet, with the exception of the relatively rare instances of insulin resistance with large quantities of binding antibodies in the circulation, patients with the insulin-requiring growth-onset type of diabetes are asserted to be very
sensitive to insulin, while those with the maturity-onset type are insulin- insensitive (48). The inference that insulin-sensitive patients have in
hibitors of insulin in the circulation while insulin-insensitive patients do not suggests that either the inhibitors have no physiological significance or that the method for evaluating insulin sensitivity in man is unreliable
(89). Some derangement in the insulin-dependent tissues, however, may be more effective in neutralizing the physiological action of insulin than the presence of inhibitors in the circulation. More definitive data must await the development of methods for determining directly the responsiveness of the insulin-dependent tissues in man. In vitro studies on adipose tissue taken from diabetic and nondiabetic subjects reveal no difference in re
sponsiveness to insulin (89a).
VIII. IMPAIRMENT OF INSULIN RECEPTOR SITES
The major impediment to an evaluation of the role of agents that may block the insulin receptors in the etiology of diabetes is the lack of in
formation about the nature of such receptors. Some inferences about them may be drawn from a variety of observations on the mechanism of insulin action. A mass of data supports the prevailing consensus that insulin acts by producing some change in the organization of the cellular membrane with a resultant increase in the rate of entry of glucose into the cell (90-92).
Direct evidence for this hypothesis is provided by electron microscopic studies, which reveal that, concomitant with its stimulation of glucose uptake by epididymal adipose tissue, insulin induces an invagination of the cellular membrane, the formation of membrane-bound vesicles, and all the other phenomena that characterize the process of pinocytosis (93). Yet, the transport of extracellular glucose into the intracellular compartment by means of the vesicles formed from the cellular membrane cannot ac
count for the increment in glucose uptake induced by insulin (93). It is possible that the initiation of pinocytosis produces changes in the perme
ability of the adjacent membrane and thereby permits an increase in the diffusion of glucose into the cell. It is equally possible that pinocytosis serves principally to transport insulin into the cell, where it activates some other mechanism to facilitate the transport of glucose.
The morphological studies suggest that the first step in the action of insulin must be its adsorption to the cellular membrane. Although the mechanism for such adsorption is unknown, an irreversible "binding" of insulin occurs after an extremely brief exposure of some tissues to a solu
tion of insulin in vitro (94)- It is quite possible that the binding of insulin involves some thiol-disulfide interchange between the S—S of insulin and
394 I. A . M I R S K Y
the —SH of the proteins that comprise the receptor sites on the surface of the membrane. Evidence for such a mechanism has been reported to account for both the fixation of vasopressin to its receptor sites in the tubules of the kidney and for its physiological activity (95). The forma
tion of a hormone-receptor disulfide and a subsequent series of sulfhydryl- disulfide reactions has been postulated to induce alterations of the tertiary structure of the membrane proteins and thereby open channels for the passage of water and specific solutes (Fig. 2A), to modify an ordered water lattice around the structural proteins or to separate disulfide-linked fibrillar elements (Fig. 2B). Such sulfhydryl-disulfide interchanges have
HORMONE R E C E P T O R H O R M O N E - R E C E P T O R C O M P L E X
FIG. 2. Hypothetical mechanisms whereby intrachain disulfide of hormone inter
changes with — S H of cellular membrane. Adapted from (95).
been proposed in explanation of some aspects of protein denaturation, blood coagulation, mitosis, and other physiological phenomena (96). That some similar mechanism may be involved in the action of insulin is sug
gested by the observation that pretreatment of the isolated rat diaphragm (97) or epididymal adipose tissue (97a) with concentrations of iodoacetate, p-chloromercuribenzoate, or iV-ethylmaleimide that produce no significant changes in glucose uptake, nevertheless inhibits the action of insulin.
Likewise, the action of insulin on the utilization of glucose by the perfused rat heart is inhibited by iV-ethylmaleimide (97b). Furthermore, the total number of —SH groups of rat diaphragm is decreased about 22% after incubation with insulin (98).
These observations support but do not establish the hypothesis that the action of insulin is dependent upon the binding of insulin to the insulin receptor sites of the tissues by means of a disulfide-thiol interaction. Thus, whereas iV-ethylmaleimide may inhibit the binding of insulin to the per
fused rat heart (97b), it does not do so to the rat diaphragm or epididymal adipose tissue (97a). Furthermore, whereas oxytocin and related disulfide peptides exert an insulin-like action on rat epididymal adipose tissue, they do not influence the utilization of glucose by muscle nor the action of insulin thereon.
Pertinent to the above is the demonstration that synthetic oxytocin and related disulfide peptides exert an insulin-like action in vitro on rat epididy
mal adipose tissue (98a, 98b). The administration of these agents to healthy and alloxan-diabetic dogs results in a prompt reduction in the free fatty acid concentration of the plasma (98c). These agents do not influence the utilization of glucose by the rat diaphragm, nor do they affect the action of insulin thereon.
Since —SH-binding agents inhibit the action of insulin in vitro while metabolic inhibitors like 2,4-dinitrophenol, cyanide, and salicylate do not (99), the insulin insufficiency of diabetes mellitus may be due to the forma
tion of agents which bind the sulfhydryl groups of the insulin receptors or other essential groups and thereby make them unavailable to the circu
lating insulin. The production of such agents may account for the observa
tions that patients with severe burns may develop a temporary insulin- resistant form of diabetes which is ameliorated by the administration of 2,3-dimercapto-l-propanol ( B A L ) (100). Likewise, B A L has been re
ported to be effective in improving the diabetic status of other types of insulin-resistant diabetes (101, 102).
IX. ANTAGONISTS OF ACTION OF INSULIN
A distinction must be made between humoral agents which inhibit the action of insulin by preventing its access to the receptor sites and those which antagonize the physiological effects of the action of insulin. Whereas insulin induces an increase in the uptake of glucose by various tissues and directly or indirectly reduces hepatic gluconeogenesis and increases hepatic glycogenesis, antagonistic agents exert their action primarily through an increase in the rates of hepatic glycogenolysis and gluconeogenesis as well as through a decrease in the utilization of glucose. The contrainsulin effects of such agents are evident in the exacerbation of the metabolic de
rangements that follows their administration to the completely depancre- atized or alloxanized animal (11). Accordingly, it may be postulated that the excessive production of contrainsulin agents could result in an increase in the requirements for insulin and thereby induce an insufficiency.
396 I. A . M I R S K Y
The endogenous agents which can exert an antagonistic action to that of insulin may be classified as (a) catabolic products of infection or of local tissue damage and (b) endocrine secretions. The former produce a temporary impairment in carbohydrate metabolism and consequently may be significant only in the precipitation of the diabetic syndrome in the predisposed person (108). A much more potent role in the pathogenesis of diabetes, however, has been attributed to the action of hormones se
creted by the adenohypophysis, the adrenal medulla, the adrenal cortex, the gonads, the thyroid gland, and the alpha cells of the pancreas. It is impossible to review here the vast quantity of experimental and clinical data dealing with the action of the various hormones in the regulation of carbohydrate metabolism. The consensus, however, is that prolactin, thyroxine, and estrogens exert definite but quantitatively insignificant contrainsulin effects. Epinephrine and glucagon are somewhat more potent insulin antagonists in that both accelerate hepatic glycogenolysis; the pro
longed administration of glucagon may also induce an increased gluconeo- genesis and a temporary diabetes (104). Yet, there is no evidence that an excessive secretion of epinephrine, such as occurs in patients with pheo- chromocytoma, or an excessive secretion of glucagon is involved in the idiopathic syndrome of diabetes.
The most potent antagonistic agents of insulin are the growth hormone of the adenohypophysis and the 11,17-hydroxycorticosteroids of the adrenal gland (105). Consequently, many have attempted to relate an increased secretion of either or both hormones with the precipitation and perpetuation of the diabetic syndrome in man. There is no convincing evidence, however, of an increase in the activity of either gland in the patient with diabetes mellitus (98). Although many patients with the hyperpituitarism of acromegaly, and the hypercorticalism of Cushing's syndrome, may develop a diabetic syndrome, the stigmata of the endocrine disease are very evident, while the diabetic syndrome differs markedly from that which characterizes either the growth-onset or maturity-onset type of diabetes (106).
X. EXCRETION OF INSULIN
A fine balance between the rate of insulin production by the pancreas and the rate of excretion in the urine could be postulated to account for the insulin insufficiency of patients with diabetes mellitus who have normal amounts of insulin in the plasma. In spite of the relatively small size of the molecule, however, insulin is not excreted in significant quantities in the urine of man; only about 1% of intravenously injected insulin is ex-
creted in the urine (107). Patients with the mild diabetes of the maturity- onset type with ample concentrations of insulin in the circulation excrete even less insulin than nondiabetic subjects. Accordingly, an increased rate of excretion of insulin cannot account for the metabolic derangement of diabetes.
XI. DEGRADATION OF INSULIN
Extensive studies with extracts and slices of various tissues from animal and man (108, 109), perfused organs (110-112), eviscerated rabbits (118), intact animals (114-116), and man (83, 64, 69) have established that insulin is inactivated and degraded by an enzymic process. Although the liver and kidney exhibit the greatest insulin-inactivating and insulin- degrading activities, nearly all other tissues exhibit some activity (108, 109). Fresh animal or human plasma does not inactivate or degrade insulin (117, 118). Both the inactivation and the degradation of insulin induced by a liver homogenate appear to be due to the action of a heat- labile proteolytic system of the supernatant fraction (119-122). This system accounts for more than 70% of the concomitant insulin-inactivating and insulin-degrading activities of liver, kidney, and muscle homogenates (108, 109, 117, 120-122). The maximal activity of the heat-labile system of liver extracts occurs at a slight alkaline pH with a maximum velocity
(Vm) of 20 μg insulin/gm liver and a Michaelis constant (Km) of 8 X 10~
8 moles (120,124)· The system is inhibited by Cu+ +, Zn+ +, p-chloromercuri- benzoate, and iodoacetate, which suggests the participation of essential sulfhydryl groups. Similar properties are exhibited by a highly purified, apparently homogenous insulin-degrading system isolated from liver ex
tracts (125). The relative specificity exhibited by the heat-labile insulin- degrading system led to its tentative designation as "insulinase" (108).
It has been proposed that the hydrolysis of insulin is preceded by a cleavage of its disulfide bridges as the result of the action of an "insulin reductase" in the presence of reduced glutathione made available by the action of glutathione reductase (128). In accord is the demonstration that a highly purified preparation of liver catalyzes the reductive cleavage of disulfide bonds of insulin by reduced glutathione (128a, 125).
Irrespective of whether the destruction of insulin is due to a primary cleavage of the disulfide bridges or to a specific proteolytic enzyme or to a combination of enzymes, it is a physiological phenomenon and is in
fluenced by a variety of factors. Thus, prolonged fasting decreases the rate of destruction of insulin by intact mice and rats (108, 109) and con
comitantly increases the hypoglycemic response to a standard dose of
398 I . A . M I R S K Y
insulin (126). Likewise, hepatic insulinase activity is depressed in rats maintained on diets deficient in protein, riboflavin, and pantothenic acid (127). Refeeding fasted animals with a high carbohydrate diet results in a rapid restitution of the rate of insulin degradation, while only a gradual restitution occurs with protein (108). The production of obesity with gold thioglucose results in an increase in the insulinase activity of mice (128).
Hypophysectomy and thyroidectomy produce a decrease in the insulinase activity (109), but the chronic administration of somatotropin or adreno- corticotropin to normal animals produces no change.
The low Km and the high Vm of insulinase are consistent with the rapid rate and the very high capacity of the insulin-degrading system observed in intact mice (115), rabbits (116), and man (33, 64). When I
131 -labeled insulin is injected directly into the portal vein of animal or man, about 50% is retained by the liver during a single passage (129). Likewise, about 40% of the I
131
-labeled insulin entering the perfused rat liver at any mo
ment is removed during a single passage through the liver. The insulin that is retained is subsequently degraded (112). Accordingly, the concen
tration of insulin that reaches the systemic circulation after the ingestion of a carbohydrate meal represents only about half of the quantity re
leased by the pancreas into the portal vein. Thereafter, the insulin is re
moved from the circulation at a relatively constant rate of about 2%/hour (83). The rate of removal of unlabeled insulin parallels that of I
131 -labeled insulin (180). Since the disappearance of the I
131
-labeled insulin can be accounted for by the appearance of I
1 31
in the nonprotein fraction of the plasma, it is probable that the rate of removal of insulin is related to the rate at which it is degraded (33, 64, 180).
Very rapidly after the intravenous injection of labeled insulin into animal and man it is distributed in a space which exceeds the volume of the extracellular space (83, 114)- That this distribution is due to entry of insulin into the cells of the liver, kidney, muscle, and other tissues is re
vealed by autoradiographic as well as by subcellular fractionation of the tissues of rats injected with insulin (114, 131)- The rapid intracellular distribution may be due to a phenomenon similar to the pinocytosis in
duced in adipose tissue by insulin; the vesicles formed from the invaginated cellular membrane may carry the insulin into the cell and thereby make it available for eventual degradation by the insulinase present in the soluble portion of the cell.
XII. INHIBITORS OF INSULIN DEGRADATION
The destruction of insulin can be competitively inhibited in vitro and in vivo by a heat-stable, dialyzable, nonprotein fraction of extracts of liver and other tissues (108). The pep tide-like nature of the liver inhibitor
led to studies with other peptides and the demonstration that partial hydrolyzates of insulin, glucagon, adrenocorticotropin, somatotropin, casein, and a variety of other proteins exert a competitive inhibition of insulinase; complete hydrolysis destroys the inhibitory action (108).
Subsequent studies with amino acids and synthetic peptides revealed that L-tryptophan and some L-tryptophyl peptides exert an insulinase- inhibitory activity in vitro and in vivo and induce a concomitant decrease in the blood sugar concentration when administered by stomach tube to rats (182). A similar hypoglycemic response and a concomitant inhibition of insulinase activity in vitro and in vivo was obtained with some of the metabolic products of L-tryptophan (e.g., indole-3-acetic acid, nicotinuric acid, anthranilic acid) (183). Some of these compounds (e.g., indole-3- acetic acid, nicotinic acid) are effective hypoglycemic agents when ap
propriate amounts are administered by mouth to man (108).
The possibility that the hypoglycemic and the insulinase-inhibitory actions of the derivatives of tryptophan were related to the capacity of these compounds to function as plant growth regulators (auxins) led to studies on the influence of a variety of natural and synthetic plant-growth regulators on the blood sugar of rats and on the in vitro and in vivo in
hibition of insulinase (Fig. 3). Thus, like indole-3-acetic acid, the synthetic compounds indole-3-butyric, indo!e-3-propionic, o-chlorophenoxyacetic, p-chlorophenoxyacetic, 2,4-dichlorobenzoic, 0-naphthaleneacetic and a- naphthaleneacetic acids exert a hypoglycemic response and a concomitant inhibition of insulinase in rats (108). Studies with more than 100 related compounds revealed many similarities between the structural requirements which determine plant-growth regulatory action and those which deter
mine the hypoglycemic and insulinase-inhibitory activity in rats (108).
Nearly all of the compounds which inhibit the destruction of insulin do so by a competitive inhibition of insulinase.
Although indole-3-acetic acid can reduce the blood sugar of patients with mild diabetes of the maturity-onset type (184), the unknown toxicity of the synthetic plant regulators precluded their study in man. The possi
bility that some insulinase inhibitor may be effective in the treatment of patients with diabetes led others to screen the above and numerous other compounds for their inhibitory activity (135). Many of the compounds that were found to inhibit the destruction of insulin in vitro proved to be noncompetitive inhibitors in vitro and produced an increase in the blood sugar concentration when given by mouth to rats. In contrast, the sulfonyl
ureas, carbutamide and tolbutamide, which exert a noncompetitive in
hibition of insulinase in vitro and in vivo (116, 136, 137) produce a hypo
glycemic response in vivo. It is probable, however, that the hypoglycemic response to the sulfonylureas is due primarily to the release of insulin by the pancreas and only partially to the inhibition of insulinase (108).
400 I . A . M I R S K Y
Plasma from nondiabetic and diabetic subjects has been reported to protect insulin from degradation by slices and extracts of liver and other tissues (38, 188). Whereas some claim that such protection occurs only with plasma from insulin-treated subjects and is due to binding of the insulin with an insulin antibody associated with the globulins of the plasma proteins (124), others report that similar protein fractions of plasma from nondiabetic subjects devoid of insulin-binding components exert an insulinase-inhibitory action in vitro (138). The noncompetitive nature of the inhibition exerted by the plasma protein fractions favors the hypothesis that the protection is accomplished through some binding mechanism rather than by a direct inhibition of the insulin-degrading system.
The demonstration that the intracellular compartment contains an insulin-degrading system as well as an inhibitor of this system suggests that the balance between them determines the rate at which insulin is destroyed. An increase in the activity of the enzyme or a decrease in the concentration of the inhibitor will result in an increase in the rate at which insulin is degraded after it is brought into the cell. If the physio
logical effects of insulin are dependent to some degree upon its entrance into the intracellular compartment by pinocytosis, a decrease in its meta
bolic effectiveness will ensue from its accelerated degradation. Thus, an insulin insufficiency may occur even in the presence of ample concentra
tions of insulin in the extracellular compartment. Studies with I 1 3 l
-labeled insulin reveal no difference between the mean rate of removal of insulin from the circulation of nondiabetic and previously untreated diabetic subjects (83, 64). The marked variability in the rates of turnover that have been reported, however, make it apparent that even a significant increase in the rate of insulin degradation may escape detection by the methods employed. Until much more precise or more direct techniques are developed, it will be impossible to establish whether or not an increase in the rate of insulin degradation plays a role in the pathogenesis of di
abetes.
XIII. SUMMARY
It is apparent from the preceding that the mechanism responsible for the insulin insufficiency of diabetes mellitus in man is unknown. Although a primary defect in beta cell function may be responsible for the insufficiency of some patients, it cannot account for that of the majority of patients.
The morphological derangements found in the islets of the pancreases from the majority of patients with diabetes may be the result rather than the cause of the insufficiency. Likewise, the diminished reserves of insulin
402 I . A . M I R S K Y
in the islets of many patients with diabetes may be the result of a chronic drain induced by extrapancreatic factors such as inhibitors, antagonists, and/or inactivators of insulin. The lack of definitive evidence that any of these factors are involved in the pathogenesis of diabetes mellitus in man emphasizes the need for the development of new technical approaches.
R E F E R E N C E S
1. F. Sanger, Science 129, 1340 (1959).
2. D . S. H . W . Nicol, and L . F. Smith, Nature 187, 483 (1960).
3. W . S. Hartroft, Diabetes 5, 98 (1956).
4. Ε. T . Bell, "Diabetes Mellitus: A Clinical and Pathological Study of 2529 Cases."
C. C Thomas, Springfield, Illinois, 1960.
5. G. A . Wrenshall, A . Bogoch, and R . C. Ritchie, Diabetes 1, 87 (1952).
6. W . S. Hartroft and G. A . Wrenshall, Diabetes 4, 1 (1955).
7. N . Maclean and R . F. Ogilvie, Diabetes 8, 83 (1959).
8. E. Moschcowitz, A.M.A. Arch. Pathol. 61, 136 (1956).
9. W . E. Toreson, Am. J. Pathol. 27, 327 (1951).
10. Ε. M . Brown, Jr., F. C. Dohan, L . R . Freedman, P . de Moor, and F. D . W . Lukens, Endocrinology 50, 644 (1952).
11. J. Campbell, L . Chaikoff, G. A . Wrenshall, and R . Zemel, Can. J. Biochem. and Physiol. 37, 1313 (1959).
12. H . Buse, K . Gundersen, and F. D . W . Lukens, Diabetes 6, 428 (1957).
13. I . A . Mirsky, D . Diengott, and H . Dolger, Metabolism, Clin, and Exptl. 5, 875 (1956).
14. R . Klein, J. Marks, and I . A . Mirsky, Pediatrics 22, 289 (1958).
15. A . Loubatières , Diabetes 6 , 40 8 (1957) .
16. J . Bornstei n an d P . Trewhella , Med. J. Australia 1 , 11 9 (1951) . 17. J . Bornstei n an d R . D . Lawrence , Brit. Med. J. I , 73 2 (1951) . 18. J . Bornstei n an d R . D . Lawrence , Brit. Med. J. I I , 154 1 (1951) .
19. J . Groen , C . E . Kamminga , A . F . Willebrands , an d J . R . Blickman , J. Clin. Invest.
31, 9 7 (1952) .
20. J . Vallance-Owen , B . Hurlock , an d N . W . Please , Lancet ii , 58 3 (1955) . 21. A . F . Willebrand s an d J . Groen , Diabetes 5 , 37 8 (1956) .
22. J . Vallance-Owen , G . F . Joplin , an d R . Fraser , Lancet ii , 58 4 (1959) .
23. E . F . Pfeiffer , M . Pfeiffer , H . Ditschuneit , an d C . S . Ahn , Ann. Ν. Y. Acad. Sci.
82, 479 (1959).
24. R . S. Yalow and S. A . Berson, Diabetes 9, 254 (1960).
25. R . S. Yalow and S. A . Berson, J. Clin. Invest. 39, 1157 (1960).
26. R . S. Yalow, H . Black, M . Villazon, and S. A . Berson, Diabetes 9, 356 (1960).
27. H . S. Seltzer and W . L . Smith, Diabetes 8, 417 (1959).
27a. J. Steinke, K . W . Taylor, and A . E. Renold, Lancet i, 30 (1961).
28. P. J. Randle, Ciba Foundation Colloq. Endocrinol. 11, 115 (1957).
29. D . S. H . W . Nicol, in "The Mechanism of Action of Insulin" ( W . A . Broom, F.
W . Wolff, and F. G. Young, eds.), p. 3. Blackwell, Oxford, 1960.
30. R . G. Langdon, J. Biol. Chem. 235, 4 (1960).
31. A . E. Renold, J. A . Owen, Jr., and P. H . Forsham, Diabetes 9, 324 (1960).
32. P . M . Beigelman, H . N . Antoniades, F. C. Goetz, A . E. Renold, J. L . Oncley, and G. W . Thorn, Metabolism, Clin, and Exptl. 5, 44 (1956).
33. S. A . Berson, R . S. Yalow, A . Bauman, M . A . Rothschild, and K . Newerly, / . Clin.
Invest. 35, 170 (1956).
34. H . N . Antoniades, Science 127, 593 (1958).
35. H . N . Antoniades, P. M . Beigelman, R . B. Pennell, G. W . Thorn, and J. L . Oncley, Metabolism, Clin, and Exptl. 7, 266 (1958).
36. H . N . Antoniades, Endocrinology 68, 7 (1960).
37. P. M . Beigelman, Diabetes 7, 365 (1958).
38. P. J. Randle and K . W . Taylor, Lancet ii, 996 (1958).
39. P. J. Randle and K . W . Taylor, Endocrinol. 17, 387 (1958).
40. K . W . Taylor and P. J. Randle, J. Endocrinol. 19, 221 (1959).
41. R . E. Bolinger, H . Van der Geld, and A . F. Willebrands, Metabolism, Clin, and Exptl. 8, 39 (1959).
42. K . W . Taylor, L . Vargas, and P. J. Randle, Lancet i, 1313 (1960).
43. L . Vargas, K . W . Taylor, and P. J. Randle, Biochem. J. 77, 43 (1960).
44. M . L . Mitchell, J. Clin. Endocrinol. and Metabolism 20, 1319 (1960).
45. H . N . Antoniades, A . E. Renold, Y . M . Dagenais, and J. Steinke, Proc. Soc. Exptl.
Biol. Med. 103,677 (1960).
46. K . Gundersen and H . N . Antoniades, Proc. Soc. Exptl. Biol. Med. 104, 411 (1960).
47. H . N . Antoniades and K . Gundersen, Endocrinology 68, 36 (1960).
47a. H . N . Antoniades, K . Gundersen, and H . M . Pyle, Endocrinology 69, 163 (1961).
48. J. Vallance-Owen, Physiol. Revs. 40, 219 (1960).
49. A . Glen and J. C. Eaton, Quart. J. Med. 7, 271 (1938).
50. F. C. Lowell, Proc. Soc. Exptl. Biol. Med. 50, 167 (1942).
51. F. C. Lowell, Clin. Invest. 23, 233 (1944).
52. F. C. Lowell, J. Clin. Invest. 26, 57 (1947).
53. J. Bornstein and R . D . Lawrence, Brit. Med. J. II, 1541 (1951).
54. V . De Filippis and A . Iannaccone, Lancet i, 1191 (1952).
55. M . Kaye, E. McGarry, and I . Rosenfeld, Diabetes 4,133 (1955).
56. M . H . Loveless and J. R . Cann, J. Immunol. 74, 329 (1955).
57. J. Kleeberg, D . Diengott, and J. Gottfried, J. Clin. Endocrinol. and Metabolism 16, 680 (1956).
58. T . Yankelowitch, S. Massry, and S. Gitter, Diabetes 5, 457 (1956).
59. A . R . Colwell and R . Weiger, J. Lab. Clin. Med. 47, 844 (1956).
60. C. W . Baird and J. Bornstein, Lancet i, 1111 (1957).
61. J. B . Marsh and N . Haugaard, / . Clin. Invest. 31, 107 (1952).
62. B . Dituri, New Engl. J. Med. 251, IS (1954).
63. A . H . Sehon, M . Kaye, E. McGarry, and B. Rose, J. Lab. Clin. Med. 45,765 (1955).
64. G. W . Welsh, E. H . Henley, R . H . Williams, and R . W . Cox, Am. J. Med. 21, 325 (1956).
65. R . W . Weiger and A . R . Colwell, Clin. Research Proc. 4, 123 (1956).
66. S. A . Berson and R . S. Yalow, J. Clin. Invest. 36, 873 (1957).
67. S. A . Berson and R . S. Yalow, Diabetes 6, 402 (1957).
68. S. A . Berson and R . S. Yalow, Am. J. Med. 25, 155 (1958).
69. N . Kalant, C. Gomberg, and R . Schucher, Lancet ii, 614 (1958).
70. S. A . Berson and R . S. Yalow, Clin. Invest. 38, 1996 (1959).
71. J. B . Field and M . L . Woodson, J. Clin. Invest. 38, 551 (1959).
71a. R . S. Yalow and S. A . Berson, Radiology 66, 106 (1956).
71b. W . G. Duncombe and C. B . Mann, Biochim. et Biophys. Acta 56, 193 (1962).
72. J. B . Field, Diabetes 9, 245 (1960).
73. J. B. Field and B . Rigby, / . Endocrinol. 19, 174 (1959).
404 I . A . M I R S K Y 74. J. Vallance-Owen, / . Postgrad. Med. 35, 266 (1959).
75. J. Vallance-Owen, E. Dennes, and P. N . Campbell, Lancet ii, 336 (1958).
76. C. W . Baird and J. Bornstein, J. Endocrinol. 19, 74 (1959).
77. J. R . Debro, H . Tarver, and A . Korner, Lab. Clin. Med. 50, 728 (1957).
78. J. Vallance-Owen, E. Dennes, and P. N . Campbell, Lancet ii, 696 (1958).
78a. J. Vallance-Owen and M . D . Lilley, Lancet i, 806 (1961).
78b. J. Vallance-Owen and M . D . Lilly, Lancet i, 804 (1961).
79. J. Vallance-Owen and M . D . Lilley, Biochem. J. 74, 18P (1960).
80. P. J. Randle, K . W . Taylor, and L . Vargas, Endocrinol. 20, 20 (1960).
81. K . W . Taylor, L . Vargas, and P. J. Randle, Lancet i, 1313 (1960).
82. J. B. Field and D . Stetten, Jr., Am. J. Med. 21, 339 (1956).
83. J. B. Field and D . Stetten, Jr., Diabetes 5, 391 (1956).
84. J. B. Field, F. Tietze, and D . Stetten, Jr., J. Clin. Invest. 36, 1588 (1957).
85. J. Vallance-Owen and F. D . W . Lukens, Endocrinology 60, 625 (1937).
86. J. Bornstein and C. R . Park, Biol. Chem. 205, 503 (1953).
87. J. Bornstein, J.'Biol. Chem. 205, 513 (1953).
88. E. D . Hendley, E. Bregman, and M . E. Krahl, J.Biol. Chem. 226, 459 (1957).
89. T . S. Danowski, "Diabetes Mellitus," Williams & Wilkins, Baltimore, Maryland, 1957.
89a. I . A . Mirsky and S. S. Kaufman, unpublished data, 1962.
90. R. Levine and M . S. Goldstein, Recent Progr. in Hormone Research 11, 343 (1945).
91. E. J. Ross, Medicine 35, 355 (1956).
92. C. R. Park, O. Reinwein, M . J. Henderson, E. Cadenas, and H . E. Morgan, Am.
J. Med. 26, 674 (1959).
93. R . J. Barnett and E. G. Ball, Biophys. Biochem. Cytol. 8, 83 (1960).
94. W . C. Stadie, N . Haugaard, and J. B. Marser, J. Biol. Chem. 188, 167 (1951).
95. I . L . Schwartz, H . Rasmussen, M . A . Schoessler, L . Silver, and C. T . O. Fong, Proc. Natl. Acad. Sci. U. S. 46, 1288 (1960).
96. Ε. V . Jensen, Science 130, 1319 (1959).
97. D . E. Haft and I . A . Mirsky, J. Pharmacol. Exptl. Therap. 104, 340 (1952).
97a. I . A . Mirsky and G. Perisutti, Biochim. et Biophys. Acta 62, 490 (1962).
97b. E. Cadenas, H . Kaji, C. R . Park, and H . Rasmussen, J. Biol. Chem. 236, 63P (1961).
98. G. Ungar and S. Kadis, Nature 183, 49 (1959).
98a. I . A . Mirsky and G. Perisutti, Biochim. et Biophys. Acta 50, 603 (1961).
98b. I . A . Mirsky and G. Perisutti, Endocrinology 71, 158 (1962).
98c. I . A . Mirsky, Proc. Soc. Exptl. Biol. Med. 110, 42 (1962).
99. P. J. Randle and G. H . Smith, in "The Mechanism of Action of Insulin" ( W A . Broom, F. W . Wolff, and F. G. Young, eds.), p. 65. Blackwell, Oxford, 1960.
100. W . J. H . Butterfield, Lancet i, 489 (1955).
101. W . J. H . Butterfield and R . H . S. Thompson, Clin. Sci. 16, 679 (1957).
102. T . E. Prout, J. A . Weaver, G. W . Scott, and S. P . Asper, Jr., Metabolism, Clin, and Exptl. 7, 240 (1958).
103. J. A . Greene, A . David, and G. Johnston, Am. J. Physiol. 136, 595 (1942).
104. J. Logothetopoulos, Β . B . Sharma, J. M . Salter, and C. H . Best, Diabetes 9, 278 (1960).
105. R . C. DeBodo and N . Altszuler, Physiol. Revs. 38, 389 (1958).
106. J. B . Field, Am. J. Med. 26, 662 (1959).
107. I . A . Mirsky, C. J. Podore, J. Wachman, and R . H . Broh-Kahn, J. Clin. Invest. 27, 515 (1948).
108. I. A . Mirsky, Recent Progr. in Hormone Research 13, 429 (1957).
109. R . H . Williams, J. S. Hay, and M . B . Tjaden, Ann. N. Y. Acad. Sci. 74, 513 (1959).
110. S. Weisenfeld, R . H . Jauregui, and M . G. Goldner, Am. J. Physiol. 188, 45 (1957).
111. G. E. Mortimore, F. Tietze, and D . Stetten, Jr., Diabetes 8, 307 (1959).
112. G. E. Mortimore and F. Tietze, Ann. Ν. Y. Acad. Sci. 82, 329 (1959).
113. A . N . Wick and D . Drury, Proc. Soc. Exptl. Biol. Med. 97, 514 (1958).
114. N . J. Elgee, R . H . Williams, and N . Lee, Clin. Invest. 33, 1252 (1954).
115. I . A . Mirsky, G. Perisutti, and R . Jinks, Endocrinology 56, 484 (1955).
116. S. A . Berson, R . S. Yalow, and B. W . Volk. Lab. Clin. Med. 49, 331 (1957).
117. I. A . Mirsky and R . H . Broh-Kahn, Arch. Biochem. 20,1 (1949).
118. I. A . Mirsky, G. Perisutti, and N . C. Davis, J. Clin. Invest. 38, 14 (1959).
119. I. A . Mirsky, G. Perisutti, and F. J. Dixon, Proc. Soc. Exptl. Biol. Med. 86, 228 (1954).
120. I. A . Mirsky, G. Perisutti, and F. J. Dixon, J. Biol. Chem. 214, 397 (1955).
121. H . H . Tomizawa, M . L . Nutley, H . T . Narahara, and R . H . Williams, Biol.
Chem. 214, 285 (1955).
122. M . Vaughan, Biochim. et Biophys. Acta 15, 432 (1954).
123. H . T . Narahara and R . H . Williams, J. Biol. Chem. 234, 71 (1959).
123a. H . H . Tomizawa, J. Biol. Chem. 237, 428 (1962).
124. R . S. Yalow and S. A . Berson, J. Clin. Invest. 36, 648 (1957).
125. H . H . Tomizawa and Y . D . Halsey, / . Biol. Chem. 234, 307 (1959).
126. I. A . Mirsky, G. Perisutti, and D . Diengott, Endocrinology 60, 303 (1957).
127. D . Diengott, S. Halevy, and K . Guggenheim, Endocrinology 65, 602 (1959).
128. S. Larsson, Acta Physiol. Scand. 46, 159 (1959).
129. L . L . Madison, B. Combes, R . H . Unger, and N . Kaplan, Ann. Ν. Y. Acad. Sci.
74, 548 (1959).
130. S. A . Berson and R . S. Yalow, Advances in Biol. and Med. Phys. 4, 349 (1958).
131. O. Stein and J. Gross, Endocrinology 65, 707 (1959).
132. I. A . Mirsky, G. Perisutti, and D . Diengott, Endocrinology 59, 369 (1956).
133. I . A . Mirsky, G. Perisutti, and R . Jinks, Endocrinology 60, 318 (1957).
134. I . A . Mirsky and D . Diengott, Proc. Soc. Exptl. Biol. Med. 93, 109 (1956).
135. R . H . Williams, F. B. Martin, E. D . Henley, and H . E. Swanson, Metabolism, Clin.
and Exptl. 8, 99 (1959).
136. I . A . Mirsky, G. Perisutti, and S. Gitelson, Ann. Ν. Y. Acad. Sci. 71, 103 (1957).
137. R . H . Williams and B. W . Tucker, Metabolism Clin, and Exptl. 5, 801 (1956).
138. N . Freinkel and C. J. Goodner, Personal communication, 1961.
