Trypsin exerts insulin-like effects in intact cells and on partially purified preparations of insulin receptors. To elucidate the mechanism of these insulinomimetic effects, we compared the structures of insulin- and trypsin-activated receptor species with their functions, including insulin binding, autophosphorylation, and tyrosine kinase activity. In vitro treatment of wheat germ agglutinin-purified receptor preparations with trypsin resulted in proteolysis of both alpha- and beta-subunits. The activated form of the receptor had an apparent molecular mass of 110 kDa under nonreducing conditions, compared to the 400-kDa intact receptor, and was separated following reduction into an 85-kDa beta-subunit related fragment and a 25-kDa alpha-subunit related fragment. Treatment of whole cells with trypsin prior to isolation of the insulin receptor resulted in proteolytic modification of the alpha-subunit only. In this case, the total molecular mass of the activated species was 116 kDa, comprised of an intact 92-kDa beta-subunit and again a 25-kDa alpha-subunit related fragment. Values of Km for peptide substrate phosphorylation and Ki for inhibition of receptor autophosphorylation, and sites of autophosphorylation within the beta-subunits were similar for receptors activated either by insulin or trypsin. Insulin had no additional effect on the rate of autophosphorylation of the truncated receptor, and no binding of insulin by the truncated receptor was detected either by direct assay or cross-linking with bifunctional reagents. Based on the deduced amino acid sequence of the insulin receptor and the structural studies presented here we concluded that this activated form of the receptor resulted from tryptic cleavage at the dibasic site Arg576-Arg577. This was accompanied by loss of the insulin binding site and separation of alpha-beta heterodimers. As truncation of the alpha-subunit results in beta-subunit activation, it appears that the beta-subunit is a constitutively activated kinase and that the function of the alpha-subunit in the intact receptor is to inhibit the beta-subunit.

To evaluate the synthesis and initial processing of the insulin receptor precursor, we compared cell-free translation of rat liver poly(A)+ RNA in a reticulocyte lysate system with metabolically labeled rat hepatoma (Fao) cells. In in vitro translation assays, the primary L-[35S]cysteine-labeled products of rat liver mRNA specifically immunoprecipitable with insulin receptor antiserum were two closely migrating polypeptides with a Mr range of 160,000-164,000 (n = 7). This is similar to the size predicted by the insulin receptor cDNA sequence. When heterologous microsomal membranes were included in the cell-free system to process newly synthesized proteins co-translationally, the receptor precursors migrated as larger species of 180 +/- 2 kDa (n = 3). For comparison, when Fao cells were treated with tunicamycin to block core N-glycosylation and pulse-labeled with L-[35S]methionine, two closely migrating precursors were labeled that co-migrated with the unprocessed in vitro translation products (approximately 160 kDa). Pulse labeling of Fao cells in the absence of tunicamycin revealed receptor precursor species of 188 and 198 kDa that rapidly disappeared (t1/2 = 54 min) as the receptor subunits were observed. Thus, the initial products of insulin receptor mRNA translation are two approximately 162-kDa polypeptides that are rapidly processed in intact cells and can only be observed by in vitro studies or by using inhibitors of core glycosylation. Insulin proreceptor species can also be partially glycosylated during cell-free translation by added microsomal membranes. This is the first description of cell-free translation of the insulin proreceptor in a system that will allow detailed characterization of the earliest steps in insulin receptor biogenesis.

Using the artificial beta-cell (Biostator), we determined the insulin requirements in five nonobese type I (insulin-dependent) diabetic subjects who received isocaloric 40 and 60% mixed-carbohydrate diets in a crossover randomized fashion for 4 days, each day consisting of four equal meals. This was followed on day 5 by a "Big Mac Attack" lunch consisting of a Big Mac, french fries, and milk shake. Insulin requirements to maintain normoglycemia were calculated for each 24-h period and for the 2 h after each meal. The mean 24-h insulin requirements to maintain normoglycemia was greater for the 60% carbohydrate diet than the 40% diet. Although the four meals were of equal size, in all patients the insulin required to cover breakfast greater than lunch greater than dinner greater than or equal to snack. Expressed as milliunits per kilocalorie, the amount of insulin to cover breakfast was greater for the 60% (P less than .05) than the 40% carbohydrate diet and greater for breakfast than the other meals (P less than .01). Insulin requirements for the Big Mac (43% carbohydrate) were 58% greater than for the 40% carbohydrate diet, even after correction for caloric differences. In summary, 1) increasing dietary carbohydrate from 40 to 60% results in an increased insulin requirement for meals only; 2) insulin requirements are greater in the morning than in the evening, even when meal size is constant; and 3) very large meals with high fat and carbohydrate content result in a major increase in insulin requirement. These data indicate that diet has an important impact on insulin requirements in diabetes.

Tyrosyl phosphorylation is implicated in the mechanism of insulin action. Mutation of the beta-subunit of the insulin receptor by substitution of tyrosyl residue 960 with phenylalanine had no effect on insulin-stimulated autophosphorylation or phosphotransferase activity of the purified receptor. However, unlike the normal receptor, this mutant was not biologically active in Chinese hamster ovary cells. Furthermore, insulin-stimulated tyrosyl phosphorylation of at least one endogenous substrate (pp185) was increased significantly in cells expressing the normal receptor but was barely detected in cells expressing the mutant. Therefore, beta-subunit autophosphorylation was not sufficient for the insulin response, and a region of the insulin receptor around Tyr-960 may facilitate phosphorylation of cellular substrates required for transmission of the insulin signal.

We have previously reported that despite an increase in receptor concentration, there is a decrease in autophosphorylation and tyrosine kinase activity of the insulin receptor in insulin-deficient diabetic rats. To determine if other tyrosine kinases might be altered, we have studied the epidermal growth factor (EGF) receptor kinase in wheat germ agglutinin-purified, Triton X-100-solubilized liver membranes from streptozotocin (STZ)-induced diabetic rats and the insulin-deficient BB rat. We find that autophosphorylation of EGF receptor is decreased in proportion to the severity of the diabetic state in STZ rats with a maximal decrease of 67% (P less than 0.01). A similar decrease in autophosphorylation was observed in diabetic BB rats that was partially normalized by insulin treatment. Separation of tryptic phosphopeptides by reverse-phase high-performance liquid chromatography revealed a decrease in labeling at all sites of autophosphorylation. A parallel decrease in EGF receptor phosphorylation was also found by immunoblotting with an anti-phosphotyrosine antibody. EGF receptor concentration, determined by Scatchard analysis of 125I-labeled EGF binding, was decreased by 39% in the STZ rat (P less than 0.05) and 27% in the diabetic BB rat (not significant). Thus autophosphorylation of EGF receptor, like that of the insulin receptor, is decreased in insulin-deficient rat liver. In the case of EGF receptor, this is due in part to a decrease in receptor number and in part to a decrease in the specific activity of the kinase.(ABSTRACT TRUNCATED AT 250 WORDS)

Dexamethasone-induced changes in insulin and epidermal growth factor (EGF) receptor number, autophosphorylation, and kinase activity were studied in intact rat hepatocytes. Hepatocytes were freshly isolated from Sprague-Dawley rats treated with dexamethasone (1 mg/kg) for 4 days and from untreated littermates. Dexamethasone had no effect on insulin receptor number, while EGF receptor binding was slightly increased (21.3% vs. 17.2% binding/10(6) cells) after dexamethasone treatment. In hepatocytes from both control and dexamethasone-treated animals labeled with 32P, insulin induced tyrosine phosphorylation of the beta-subunit of the insulin receptor as well as of a 175K protein believed to be its endogenous substrate. The degree of phosphorylation of the insulin receptor was decreased 34% by dexamethasone treatment compared to the control value when studied in fasted animals. In contrast, phosphorylation was increased to a similar extent by dexamethasone treatment in fed animals. In addition, the beta-subunit of the insulin receptor extracted from dexamethasone-treated animals migrated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis with a slightly increased mobility compared to normal (89 +/- 1.2K vs. 92.5 +/- 0.4K). EGF induced tyrosine phosphorylation of its own receptor and of a 120K protein in intact hepatocytes. Their degree of phosphorylation was decreased by 30% as a result of dexamethasone treatment in the fasted animal and was unchanged in the fed animals. Our data indicate that glucocorticoids modulate insulin and EGF receptor kinase activity, but the nature of their effect depends on other factors, including the dietary state of the animal. These studies also suggest that postreceptor changes account for a major component of glucocorticoid-induced insulin resistance.

The effect of 12-O-tetradecanoylphorbol-13-acetate (TPA) on the function of the insulin receptor was examined in intact hepatoma cells (Fao) and in solubilized extracts purified by wheat germ agglutinin chromatography. Incubation of ortho[32P]phosphate-labeled Fao cells with TPA increased the phosphorylation of the insulin receptor 2-fold after 30 min. Analysis of tryptic phosphopeptides from the beta-subunit of the receptor by reverse-phase high performance liquid chromatography and determination of their phosphoamino acid composition suggested that TPA predominantly stimulated phosphorylation of serine residues in a single tryptic peptide. Incubation of the Fao cells with insulin (100 nM) for 1 min stimulated 4-fold the phosphorylation of the beta-subunit of the insulin receptor. Prior treatment of the cells with TPA inhibited the insulin-stimulated tyrosine phosphorylation by 50%. The receptors extracted with Triton X-100 from TPA-treated Fao cells and purified on immobilized wheat germ agglutinin retained the alteration in kinase activity and exhibited a 50% decrease in insulin-stimulated tyrosine autophosphorylation and phosphotransferase activity toward exogenous substrates. This was due primarily to a decrease in the Vmax for these reactions. TPA treatment also decreased the Km of the insulin receptor for ATP. Incubation of the insulin receptor purified from TPA-treated cells with alkaline phosphatase decreased the phosphate content of the beta-subunit to the control level and reversed the inhibition, suggesting that the serine phosphorylation of the beta-subunit was responsible for the decreased tyrosine kinase activity. Our results support the notion that the insulin receptor is a substrate for protein kinase C in the Fao cell and that the increase in serine phosphorylation of the beta-subunit of the receptor produced by TPA treatment inhibited tyrosine kinase activity in vivo and in vitro. These data suggest that protein kinase C may regulate the function of the insulin receptor.

Insulin modifies cellular responsiveness to some hormones which operate via guanine nucleotide binding proteins (G-proteins); also, G-proteins have been implicated in some actions of insulin. Using pertussis toxin-catalyzed [32P]ADP-ribosylation of Gi as an index of G-protein conformation, we evaluated interaction of insulin receptors with G-proteins. In isolated rat liver plasma membranes, insulin treatment for 10 min inhibited [32P]ADP-ribosylation of Gi by 50%. This effect was half-maximal at 2 x 10(-8) M. A similar effect was observed with rat adipocyte plasma membranes with half-maximal effect at 1 x 10(-8) M. Pertussis toxin activity itself was uninfluenced by insulin, as ribosylation of tubulin or heat-treated bovine serum albumin was unaltered. Elevated Mg2+ diminished basal ADP-ribosylation, but insulin inhibition occurred at all Mg2+ levels between 0 and 1 mM. Insulin inhibition was independent of ATP (20 microM to 10 mM), and GTP (0-100 microM) concentrations. Because both protein kinase C and purified insulin receptor phosphorylate purified Gi in vitro, we examined Gi as a substrate for the insulin receptor tyrosine kinase in vivo. Triton-extracts of isolated rat hepatocytes which had been 32Pi labeled and treated with insulin were immunoprecipitated with a polyclonal anti-Gi antiserum. The dominant labeled phosphoprotein had a molecular weight of 42 kDa, consistent with the alpha-subunit of Gi, contained only phosphoserine, and was unaffected in its phosphorylation by insulin. These results indicate the existence of a novel pathway for physiological "cross-talk" between insulin and other hormones and further suggests that the insulin receptor may interact with regulatory G-proteins via biochemical mechanisms not directly involving the tyrosine kinase activity of the insulin receptor.