On the basis of a monomeric insulin standard, ∼28% of total circulating immunoreactive insulin in insulin-dependent diabetes mellitus (IDDM) is a covalent aggregate of insulin. This aggregate probably originates in therapeutic insulin preparations. In this study, the activity of these aggregates was compared with that of monomeric insulin with regard to behavior in the radioimmunoassay, binding to insulin receptors, and biologic activity in isolated rat adipose cells. Molar activity of the aggregate in the insulin radioimmunoassay was approximately twice (240%) that of monomeric insulin, whereas the log-logit slope produced by the aggregate was indistinguishable from that of monomeric insulin. Insulin-receptor binding was determined by displacement of 125I-labeled A14-insulin by insulin or insulin aggregate (10−10–10−5 M). The free-insulin and aggregate concentrations required for half-maximal displacement of 125I-insulin were 4.0 × 10−10 and 2.25 × 10−9 M, respectively. [1-14C]glucose incorporation into 14CO2, glyceride-glycerol, and fatty acids was measured over a wide range of insulin monomer and aggregate concentrations (0–8 nM). In the bioassay, the maximal rates of glucose metabolism were equal (normal responsiveness). However, the concentration of insulin aggregates producing half-maximal stimulation of glucose metabolism was threefold greater than that of insulin (140 vs. 46 pM, respectively), indicating decreased sensitivity of the adipose cells to the aggregates. This was associated with a sixfold decrease in the Kd for binding of aggregates to adipose cell insulin receptors compared with binding of monomeric insulin. Taken together, these findings may partially explain the apparent discrepancies between measured concentrations of circulating immunoreactive insulin and bioactive insulin in IDDM. These data are consistent with the hypothesis that the aggregates are largely composed of insulin dimers.

This content is only available via PDF.