Dissociation of 125I-insulin from adipocyte insulin receptors was studied as a function of receptor occupancy. When cells were equilibrated with a tracer amount of 125I-insulin and the insulin-receptor complexes allowed to dissociate in diluted 125I-insulin–free buffer, in the presence or absence of unlabeled insulin, the unlabeled insulin led to acceleration of 125I-insulin dissociation rates. Thus, the unlabeled insulin (dilution + insulin) led to higher fractional receptor occupancy and a clear-cut acceleration of 125I-insulin dissociation rates, a phenomenon consistent with negatively cooperative site-site interactions. This effect was dependent on the total insulin concentration employed and was also influenced by the temperature of the incubations. The insulin derivatives, desalanine desasparagine and desoctapeptide insulin, did not accelerate dissociation of 125I-insulin. Therefore, we find evidence that adipocyte insulin receptors exhibit similar properties to those originally described by De Meyts et al. (J. Biol. Chem. 251: 1877, 1976) for other insulin receptor systems. However, we also find that, under conditions where the accelerating effect of native insulin is either not appreciable or is maximal, 125I-insulin does not dissociate from its receptors as a first order process. This suggests that, even in the absence of negatively cooperative effects, adipocyte insulin receptors do not behave as a kinetically homogeneous population. Furthermore, when 125I-insulin was allowed to associate with cells at high initial levels of receptor occupancy, the subsequent dissociation of the 125I-insulin was faster than when only the negatively cooperative effect was determined. Therefore, something in addition to the cooperative effect led to further acceleration of insulin dissociation. If receptors exist with functionally distinct binding characteristics such that one group of receptors has a high affinity and a low capacity while another group has a lower affinity and a high capacity, then, when cells are associated with 125I-insulin plus high concentrations of unlabeled insulin, most of the 125I-insulin binds to the low affinity (fast-dissociating) sites. Therefore, dissociation rates are faster when high insulin concentrations are used in the association phase, since negative cooperativity as well as rapid dissociation from a functionally lower affinity receptor are being determined. When the material that dissociates from the cells was examined, it was found that the rapidly dissociating material was intact insulin while the more slowly dissociating material contained a significant proportion of degraded material. The rate of dissociation of intact insulin could be accelerated, whereas the rate of dissociation of degraded insulin could not be accelerated in the presence of unlabeled insulin in the buffer during the dissociation phase.
Thus, negatively cooperative site-site interactions do not account for all the kinetic characteristics that were observed, and the data are best explained by a model consisting of functionally heterogeneous binding sites, i.e. low affinity, high capacity sites, which are susceptible to the cooperative effect but don't degrade insulin, and high affinity, low capacity sites, which participate in the degradative process but are not susceptible to the cooperative effect.