Insulin Binding to Liver Plasma Membranes in the Obese Hyperglycemic (o b/o b) Mouse DEMONSTRATION OF A DECREASED NUMBER OF FUNCTIONALLY NORMAL RECEPTORS

In previous studies, the insulin resistance of the obese hyperglycemic mouse (oblob) was found to be associated with decreased insulin binding to liver, fat, and lymphocytes. The present study demonstrates that the insulin receptors in the liver membranes of the oblob mouse are decreased in number, but are indistinguishable from normal by other criteria including affinity, kinetics of association and dissociation, temperature dependence of binding, and biological specificity of the binding reaction. The receptor in liver membranes of the ob/ob mouse is also indistinguishable with respect to insulin receptor site-site interactions. Degradation of both insulin and of receptor sites was studied and did not account for differences observed in binding. We conclude that the insulin receptor from the oblob mouse is a functionally normal receptor and that its presence in diminished number accounts for the observed decrease in insulin binding to liver plasma membranes. is a correlation between

that the insulin receptors in the liver membranes of the oblob mouse are decreased in number, but are indistinguishable from normal by other criteria including affinity, kinetics of association and dissociation, temperature dependence of binding, and biological specificity of the binding reaction. The receptor in liver membranes of the ob/ob mouse is also indistinguishable with respect to insulin receptor site-site interactions.
Degradation of both insulin and of receptor sites was studied and did not account for differences observed in binding.
We conclude that the insulin receptor from the oblob mouse is a functionally normal receptor and that its presence in diminished number accounts for the observed decrease in insulin binding to liver plasma membranes.
With obesity, both in rodents (l-5) and in man (6,7), there is a correlation between insulin resistance and decreased insulin binding to its specific receptor sites on the cell surface (8). In the obese hyperglycemic mouse (oblob), a model of extreme insulin resistance, insulin binding to the purified plasma membrane fractions of liver and fat and to isolated hepatocytes and thymocytes has been found to be decreased (l- 3,5). In previous studies this decrease has been shown to be due to a specific alteration in the insulin receptor population with no change in other hormone receptors or in membrane structure or function, as evidenced by changes in gross morphology, enzyme activity, or protein subunit composition of the membrane (1,2). In the present study, the detailed physicochemical properties of the insulin receptor in liver membranes of the oblob mouse were compared to those from its thin littermates. These studies demonstrate that, by all criteria, the insulin receptor of the ob/ob mouse is a normal receptor that is present in decreased concentration. MATERIALS AND temperature for the incubation is indicated in the figure legends. At the end of the incubation, without further dilution or transfer, the tubes were centrifuged for 1 min in a Beckman microfuge, the supernatants were aspirated and discarded, and the tips were excised and counted in a Nuclear-Chicago gamma counter. Except where noted, the nonspecific binding, defined as the radioactivity associated with the pellet in the presence of 50 fig/ml of unlabeled insulin, was subtracted from the total '%insulin binding to yield specific binding (14).
An attempt was made to use a membrane protein concentration of 0.2 mg/ml in all experiments, but some variation in membrane concentration between experiments occurred, necessitating normalization. This correction is possible since with '*61-insulin at tracer concentrations the bound to free ratio of labeled hormone is proportional to the total receptor concentration and, therefore, to the membrane protein concentration.' In order to validate this correction for membrane protein concentration, the specific binding of "'I-insulin at 0.1 rig/ml was determined at membrane concentrations from 0.05 to 0.3 mg/ml for both the oblob and normal liver membranes. A linear correlation was evident for both the oblob and normal membranes, with regression coefficients of 0.96 and 0.98, respectively. The normalized bound to free ratio of labeled hormone was then simply the product of the observed bound to free ratio and the ratio of the normalized to actual membrane concentration. Since small errors in membrane concentration are amplified by this correction, precision in the Lowry determination is quite important. In addition, precautions must be exercised in working with small volumes of a membrane suspension since with repeated agitation the membranes adhere to the wall of a test tube, decreasing the concentration of membrane that is dispensed.

Comparison
of Association and Dissociation-To assess kinetic properties of the insulin receptors of the oblob and normal liver membranes, association and dissociation of the '2SI-insulin with both types of membranes were compared at 20". The time courses of association of 'Y-insulin to the oblob and normal liver membranes when normalized to the same maximal binding were indistinguishable ( Fig. 1). In both, the half-time of complex formation was 2?/2 to 3 hours, and the maximal binding occurred at 5 to 6 hours. After 6 hours the specific binding of Y-insulin began to decrease, suggesting degradation of the hormone, the receptor, or the hormonereceptor complex (uide infra). When the initial rates of binding to liver membranes of the oblob and thin mice were compared at equal receptor concentrations (a membrane concentration of 0.29 mg/ml for the oblob and 0.11 mg/ml for the thin), no differences were observed (data not shown).
Dissociation of "'I-insulin was initiated after 1 hour of incubation by IOO-fold dilution of the incubation mixture (Fig.  2, A and B). The dissociation rates from both the normal and oblob liver membranes were indistinguishable, with a t, of 60 min. Since dissociation did not follow first order kinetics, a single rate constant could not be determined. rate of dissociation of '251-insulin produced by "infinite" dilution was further increased in the presence of unlabeled insulin in the dilution medium (15). Since decreased insulin binding in the oblob mice could be due to increased cooperative interactions, the extent of cooperativity was compared in the oblob and normal liver membranes by examining dissociation of 'Y-insulin-receptor complexes by dilution alone and by dilution in the presence of unlabeled insulin at 1 pg/ml (Fig. 2, A and B). With the receptors from oblob and normal mice at equal concentration and fractional saturation, the dissociation rate was increased in the presence of unlabeled insulin to a similar extent, with the half-time of dissociation reduced to about 25 min in both groups.
The negative cooperativity exerted by insulin on its own receptor is dependent on the insulin concentration (15); this dependence was indistinguishable in the oblob and normal liver membranes, with maximum effect found at 1000 rig/ml (Fig. 2C). The decrease in the effect observed at higher insulin concentrations is thought to be due to dimerization of insulin, which masks the cooperative site (16). Under the conditions of these experiments, the dissociation rate and the cooperative effect were independent of membrane concentrations in the range betwee 0.10 and 0.40 mg/ml for the thin and 0.20 and 0.70 mg/ml for the oblob membranes (data not shown).
Comparison of Affinity-As previously reported (2), when the percentage of 'Y-insulin that is bound is plotted as a function of insulin concentration, a marked decrease in the insulin binding to the liver membranes of the oblob mouse as compared to the thin is observed over the range of 0.1 to 1000 rig/ml (Fig. 3A). Changes in either affinity or capacity of the insulin receptor population could account for this decrease. When the bound to free ratio of the labeled hormone is plotted as a function of the bound hormone for the oblob and normal, as described by Scatchard (17), identically shaped J-In the simple mass action expression, the affinity constant K = curves are obtained (Fig. 3B). and is therefore proportional to -1.. ..-"--A :--~._L i--~~ ~~. .I:~ ~~ .itivity (I5), but may also reflect, in part, two or more discrete membrane concentration. Some error will result if this correction is populations of receptors, with different affinities (14). In a Xors, with dinerent attmities (14). In a used for an insulin concentration greater than 25 rig/ml. model with two classes of receptor sites but without cooperativ- ] is plotted as a function of the log-free hormone, where B is the insulin bound at a given free hormone concentration. B,., was considered to be the specific binding at 250 rig/ml since it is difficult to determine the exact contribution of nonspecific binding to total binding of Y-insulin above this concentration. A single line fit the data for both the oblob mice (0) and the thin mice (0).
In the presence of cooperativity, the Hill plot (20, 21) is useful for assessing the average affinity and the extent of cooperativity.
The data for the oblob and normal membranes superimposed on a single line (Fig. 3C) indicating that over this range both the average affinity and the extent of cooperativity were identical. The Hill coefficient (the slope of the Hill plot) was 0.75; coefficients less than 1 are consistent with negative cooperativity.
It should be noted that this Hill coefficient is likely to be an overestimate since the range of concentrations considered was somewhat restricted. The average affinity constant, which is the reciprocal of the product of the Hill coefficient and the free hormone concentration that produces half-maximal saturation of the receptor, was 1.5 x 10' M-' for both the oblob and thin membranes.
A reasonable approximation of the binding capacity of the specific receptor sites for insulin can be obtained by linear extrapolations of the Scatchard plot to the abscissa. By this technique the total receptor concentration in the oblob mouse was found to be 33% of that in the thin (4.3 * 0.2 versus 13.2 f 1.2 pmol/mg of membrane protein). At tracer concentration of hormone, the bound to free ratio of labeled hormone for the oblob membranes is 35% of that observed in the thin mice (Fig.  3B). When the free hormone concentration is very small, the bound to free ratio of the labeled hormone approaches the product of the affinity constant of the unfilled receptors and the total receptor concentration, ' and thus the decrease in bound to free ratio in liver membranes of the oblob mouse is fully accounted for by the decrease in receptor concentration with no change in the affinity of the receptor.
Biological Specificity of Insulin Receptors-In previous studies with rat liver the biological specificity of the insulin receptor has been validated by comparing the ability of several insulin analogues to inhibit Y-insulin binding (13). Consistent with these earlier studies, we have found that both the oblob (Fig. 4A) and normal (Fig. 4B) liver membranes bound various insulin analogues of different biological potency in a rank order proportionate to the ability of these analogues to stimulate glucose oxidation in adipose tissue (13). The finding that mouse insulin is only 20 to 30% as potent compared to the pork standard by this receptor assay could be due to higher affinity of the mouse receptor for pork insulin, or, more likely, to a decreased potency of our lot of mouse insulin due to damage in preparation or storage. Effects of Temperature-As shown previously (14), the steady state level of binding of insulin is inversely related to temperature (Fig. 5). For both the oblob and thin mouse, the "'I-insulin binding was higher at lower temperatures (Fig. 6); at all temperatures the relative decrease in insulin binding to the oblob membranes was constant. Multiple factors are responsible for the lower level of binding observed at higher temperatures, including decreased affinity,' decreased number of receptors (14), and enhanced degradation of insulin and receptors.
Degradation of Insulin-It has been shown with liver membranes (10) that binding to specific receptors and degradation of insulin are separate processes. Degradation of insulin is a function of time, temperature, and membrane concentration. As has been previously reported from studies at 30" (2), degradation of hormone at 20" was slower in the oblob membrane than in the thin membranes when similar membrane concentrations were used (Fig. 7) Competitive binding of insulin analogues to liver membranes. Purified liver membranes were incubated with '2sI-porcine insulin at 0.1 rig/ml for 6 hours at 20" in the presence of unlabeled porcine, mouse, guinea pig, or desalanine-desasparagine (DAA) insulins. The per cent of specifically bound ""I-insulin (mean of triplicates) is plotted as a function of total hormone concentration for the thin mice in A and for the oblob mice in B. Membrane concentrations were 0.10 mg/ml for the thin mice and 0.29 mg/ml for the obese mice; no corrections for membrane concentration were made.   6. Effect of temperature on insulin binding by liver membranes of the oblob and thin mouse. The binding of 0.1 rig/ml of '251-insulin to liver membranes of the thin (0) and oblob (0) mice was studied at 15, 20, 30, and 37". Duration of incubation was 9 hours at 15", 6 hours at 20", 60 min at 30", and 45 min at 37". Except for 15", these times of incubation were chosen to coincide with the point of maximum binding in Fig. 5. The data are corrected to a membrane concentration of 0.2 mg/ml and specific binding is plotted.
slower rate of hormone degradation in the oblob is unclear. The difference in rates of hormone degradation were much less pronounced than the differences in binding and the decreased rate of degradation would, if anything, have made the decrease in insulin binding in the oblob mouse less apparent (vide infra).
Degradation of Receptor-The degradation of the insulin receptor in the oblob and normal membranes was similar when compared at equal receptor concentrations.
Receptor degradation in membranes of the thin mouse was dependent on time, The per cent of hormone degraded was calculated by comparing the per cent of counts of '%insulin that bound to cultured human lymphocytes before and after incubation with liver membranes.
Each point is the mean of quadruplicate determinations. temperature, and membrane concentration.
Receptor degradation was rapid at higher temperatures, with 15% of the initial receptor population degraded per hour at 3'7" (membrane concentration of 0.28 mg/ml) (Fig. 8). At 20", receptor degradation was much slower; however, significant degradation did occur during the B-hour incubation period. When compared at equal membrane concentrations, receptor degradation in the oblob mouse was somewhat less than that in the thin mouse (Fig. 9A). However, when receptor degradation was related to receptor concentration, no difference between the ob/ob and normal membranes was evident (Fig. 9B). Determination of the rates of hormone and receptor degradation as a function of liver membrane concentration allowed correction of the binding data for these processes. These corrections (Fig. 3A) resulted in only minor quantitative differences which, if anything, increase the apparent deficiency of insulin binding to the ob/ob membranes. DISCUSSION The obese hyperglycemic syndrome in mice (oblob) is characterized by hyperphagia, marked obesity, and insulin resistance, manifest by hyperglycemia in the face of abnormally high circulating levels of biologically active insulin (23). In these oblob mice decreased insulin binding to specific receptors on the surface membranes of liver, fat, and thymic lymphocytes (l-3, 5) has been found as a biochemical correlate to their insulin resistance. In previous studies using purified liver membranes, this decreased insulin binding was shown to be a specific alteration in the membrane receptors for insulin; other hormone receptors, protein subunit composition, membrane marker enzymes, and gross membrane morphology were unchanged in the oblob mouse compared to normal (1,2). Although these studies were strong evidence for a specific defect in insulin binding to liver membranes, they did not answer the question of whether the insulin receptor in the oblob mouse was a defective receptor (product of an abnormal structural gene) or a normal receptor present in decreased concentration as a result of deranged regulation of receptor concentration, decreased stability of the receptor, or in response to the metabolic alterations of obesity. stability leading to more rapid degradation (24), and disordered regulation of synthesis (25). In the present study, by means of defining the functional characteristics of the insulin receptor, we sought to determine whether the impaired insulin binding in the oblob mouse was itself a direct consequence of the genetic defect. Studies of the kinetic properties of the insulin receptor of the oblob and normal mice revealed that the initial rates of association and the half-time of dissociation by dilution were indistinguishable. Insulin receptor site-site interactions as evidenced by the insulin-induced acceleration of dissociation rate were also comparable.
In the Hill analysis, the average affinity constants were equal. Comparison of the Scatchard curves at the same degree of fractional saturation indicated that the apparent affinities were the same, with the major alteration being the decrease in receptor concentration. The insulin receptor from the oblob mouse had an unaltered biological specificity for insulin. In addition the shift in insulin binding with temperature was parallel in the membranes of both groups of mice. Therefore the decreased insulin binding observed in the liver membranes of the oblob mouse is fully accounted for by the decreased concentration of an insulin receptor which by the above functional criteria is otherwise normal.
Factors involved in the regulation of the insulin receptor are in large part undefined, and a genetic mutation in regulation of the insulin receptor concentration also could lead to a decrease  9. Receptor degradation as a function of membrane concentra-ml) is the product of the membrane concentration and the receptor tion. A, receptor degradation in the liver membranes of the oblob and content per mg of membrane protein as determined by the Scatchard thin mice during a 6.hour, 20" preincubation was studied as described analysis (Fig. 3B). The amount of receptor degraded then is obtained in the legend of Fig. 8. The per cent of receptor degraded is plotted as a from the product of the receptor concentration and the percentage of function of the membrane concentration for the thin (0) and the oblob receptor degraded in the 6-hour experiment. Data for the oblob (A) (0) mice. B, the data from A is plotted as the receptor degraded versus and thin mice (0) are shown. the receptor concentration. The receptor concentration (picomoles per Since the insulin receptor is defined only by its functional characteristics, we cannot exclude an alteration in the receptor molecule at sites other than those responsible for the binding of hormone and for cooperative interactions.
A decreased insulin receptor concentration might occur with such an alteration if receptor stability was greatly impaired.
The in vitro studies with liver membranes reported here, however, show no appreciable difference in receptor degradation.
The pathophysiological mechanism by which the ob mutation results in the obese hyperglycemic syndrome remains unclear; based on the present results it seems unlikely that the insulin receptor deficiency is the primary defect. Rather, the insulin receptor in the oblob mouse is functionally normal. As in other forms of obesity there are fewer receptors per cell accounting for the decreased insulin binding. The insulin resistance displayed by these mice appears to be a result of this decreased insulin receptor concentration.
Since the filling of