Mutation of the two carboxyl-terminal tyrosines results in an insulin receptor with normal metabolic signaling but enhanced mitogenic signaling properties

Our previous studies have shown that the deletion of the insulin receptor carboxyl terminus impairs metabolic, but augments mitogenic, signaling (McClain, D. A., Maegawa, H., Levy, J., Huecksteadt, T., Dull, T. J., Lee, J., Ullrich, A., and Olefsky, J. M. (1988) J. Biol. Chem. 263, 8904-8911; Thies, R.S., Ulrich, A., and McClain, D. A. (1989) J. Biol. Chem. 264, 12820-12825). To explore further the regulatory role of the insulin receptor carboxyl terminus, a mutant insulin receptor was constructed in which the two tyrosines (Y1316 and Y1322) on the carboxyl terminus were replaced with phenylalanines. Rat 1 fibroblasts expressing high levels of this mutant receptor (Y/F2 cells) exhibited normal insulin binding and normal insulin internalization. The absence of the two tyrosines in the carboxyl terminus did not affect the phosphotransferase activity of the beta-subunit and insulin-stimulated glucose transport. However, the Y/F2 cells showed markedly enhanced sensitivity for insulin-stimulated DNA synthesis. Dose-response curves for both insulin-stimulated thymidine uptake and 5-bromo-2-deoxyuridine incorporation in the Y/F2 cell lines were shifted to the left (4-10-fold) compared with those observed in the cells expressing similar numbers of wild type receptors. Thus, the two tyrosines of the insulin receptor carboxyl terminus do not modulate the kinase function of the insulin receptor, although they are autophosphorylated in native receptors. Moreover, these tyrosines are not necessary for stimulation of glucose transport. On the other hand, these results suggest that the two carboxyl-terminal tyrosine residues exert an inhibitory effect on mitogenic signaling in native insulin receptors.     

Insulin receptors with defective tyrosine kinase inhibit normal receptor function at the level of substrate phosphorylation

Rat 1 fibroblasts have been transfected with the cDNA encoding a kinase-defective mutant human insulin receptor (A/K1018). Expression of this cDNA results in a receptor that is not only biologically inactive but also inhibits normal insulin action through the normal endogenous rat receptors in this fibroblast line (McClain, D. A., Maegawa, H., Lee, J., Dull, T. J., Ullrich, A., and Olefsky, J. M. (1987) J. Biol. Chem. 262, 14663-14671). We have investigated the mechanism of this inhibition and show that: 1) rat receptors are expressed at normal to increased levels in two cell lines which also express A/K1018 receptors at low (A/K1018-A, 5700 total receptors) or high (A/K1018-B, 2.2 x 10(5) total receptors) levels. 2) The rat receptors in the A/K1018 lines can be normally autophosphorylated under the control of insulin in vitro. 3) A/K1018 receptors do not inhibit the kinase activity of normal receptors when mixed together in vitro. 4) In intact A/K1018-B cells, the ability of insulin to stimulate autophosphorylation of the rat receptor is unimpaired; furthermore, the autophosphorylated rat receptor becomes normally activated as a tyrosine kinase. 5) The expression of receptors for insulin-like growth factor I and stimulation of hexose uptake mediated by this receptor are unaffected in cells expressing inhibitory A/K1018 receptors. 6) Expression of the A/K1018 receptor inhibits insulin-stimulated phosphorylation of two endogenous protein substrates (pp220 and pp170) by the native rat receptors. We conclude that the inhibition of insulin action seen in the A/K1018 cells is not mediated at the levels of native receptor expression or activation, nor is the effector (hexose uptake) mechanism affected by the A/K1018 receptors. The expression of this kinase-defective receptor does, however, inhibit the phosphorylation of substrate molecules by the normally activated endogenous rat receptors.     

Evidence for hybrid rodent and human insulin receptors in transfected cells.

Chinese hamster ovary cells and NIH 3T3 cells overexpressing mutant human insulin receptors were examined for the presence of hybrid receptors composed of human and rodent insulin receptors. In the present studies, most of the endogenous rodent receptors were found to be immunoprecipitated from the transfected cells but not the parental cells with a monoclonal antibody specific for human receptor. These data indicate that in these transfected cells, most of the endogenous rodent receptors are in a hybrid complex with the overexpressed human receptor. These results together with the in vitro studies of Treadway et al. (Treadway, J.L., Morrison, B.D., Soos, M.A., Siddle, K., Olefsky, J., Ullrich, A., McClain, D.A., and Pessin, J.E. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 214-218) showing that hybrid receptors exhibit transdominant inhibition explain the prior finding indicating that overexpression of defective insulin receptors interferes with the normal signaling of endogenous receptors.     

Properties of a human insulin receptor with a COOH-terminal truncation. I. Insulin binding, autophosphorylation, and endocytosis

In order to test the contribution of the insulin receptor COOH terminus to insulin action, a truncation of 43 COOH-terminal amino acids was engineered by cDNA-based deletion mutagenesis. This cDNA (HIR delta CT), as well as cDNA encoding the complete receptor (HIRc) was transfected into Rat 1 fibroblasts. Cells expressing 6.4 X 10(3) and 1.25 X 10(6) normal receptors and 2.5 X 10(5) HIR delta CT receptors, as well as control Rat 1 fibroblasts were selected for further analysis. All cell lines exhibited insulin binding of similar affinity. Partial tryptic digestion and immunoprecipitation by region-specific antibodies verified that the HIR delta CT receptors were truncated at the COOH terminus. Purified HIRc and HIR delta CT receptors underwent autophosphorylation with similar insulin and ATP sensitivity, although the HIR delta CT receptors were slightly more active in the absence of insulin. Transfected HIRc and HIR delta CT receptors undergo endocytosis in a normal fashion. Insulin internalization and degradation in both HIRc and HIR delta CT cells is increased in proportion to receptor number. Intracellular insulin processing, degradation, and release were qualitatively comparable among the transfected cell lines. Complete and truncated receptors internalize, recycle, and down-regulate normally. We conclude the following: 1) the COOH-terminal portion of the insulin receptor is not necessary for partial autophosphorylation or endocytosis; 2) following internalization the intracellular itinerary of the receptor and ligand appear normal with the truncated receptor; and 3) truncation of the COOH terminus does not impair recycling of the receptor or retroendocytosis of internalized ligand.     

Properties of a human insulin receptor with a COOH-terminal truncation. II. Truncated receptors have normal kinase activity but are defective in signaling metabolic effects

We have previously shown that a mutant human insulin receptor with a COOH-terminal 43-amino acid deletion (HIR delta CT), when expressed in Rat 1 fibroblasts, binds insulin normally, autophosphorylates, and undergoes endocytosis after insulin binding in a manner comparable to the normal human insulin receptor (HIRc). In this paper we have examined the biologic activity of the truncated and normal insulin receptors. In vitro, the HIR delta CT receptors caused a 1.8-fold greater phosphorylation of a Glu4/Tyr1 polypeptide than did the HIRc receptors, but the two receptor types were nearly equivalent in their ability to phosphorylate a src-derived peptide. Furthermore, insulin preactivation of HIRc and HIR delta CT receptors in intact cells led to equivalent stimulation of tyrosine kinase activity as subsequently determined for histone in vitro. Expression of HIRc receptors in cells led to enhanced sensitivity to insulin of 2-deoxy-D-glucose uptake and glycogen synthase activation. This increased sensitivity was proportional to receptor number at low (Ro = 6400) but not at high (Ro = 1.25 X 10(6] levels of receptor expression. However, expression of HIR delta CT receptors (Ro = 2.5 X 10(5] led to little, if any, increase in insulin sensitivity of either 2-deoxy-D-glucose uptake or glycogen synthase activation. Furthermore, compared with HIRc cells, HIR delta CT cells respond poorly to an agonistic monoclonal antibody specific for the human insulin receptor. In conclusion, the HIR delta CT receptor retains intact protein kinase activity in vitro. Despite this, however, the receptor displays low activity in mediating the metabolic effects of insulin.     

Intracellular signaling by a mutant human insulin receptor lacking the carboxyl-terminal tyrosine autophosphorylation sites.

We have recently characterized a mutant insulin receptor (Y/F2) in which the two tyrosines in the carboxyl terminus (Tyr1316, Tyr1322) were mutated to phenylalanine. Compared with wild type receptors, the Y/F2 receptor exhibited markedly enhanced sensitivity to insulin-stimulated DNA synthesis with normal insulin-stimulated glucose uptake (Takata, Y., Webster, N. J. G., and Olefsky, J. M. (1991) J. Biol. Chem. 266, 9135-9139). In this paper, we present further evidence for the divergence of the metabolic and mitogenic signaling pathways utilized by the insulin receptor. The mutant receptor showed normal sensitivity and responsiveness for insulin-stimulated glucose incorporation into glycogen. The insulin sensitivity for phosphorylation of two substrates (pp180 and pp220) was the same in both Y/F2 cells and HIRc cells. Phosphotyrosine content, however, was greater in Y/F2 cells than in HIRc cells, especially in the basal state. Insulin stimulated S6 kinase activity 2-6-fold, with an ED50 of -10 nM in Rat 1 cells and 0.5 nM in HIRc cells. The sensitivity to insulin was enhanced in Y/F2 cells with an ED50 of 0.1 nM. These effects were insulin-specific, since insulin-like growth factor (IGF)-I-stimulated mitogenesis was normal. In summary: 1) Y/F2 receptors exhibit normal metabolic and enhanced mitogenic signaling; 2) the enhanced mitogenic signaling is specific for the insulin receptor in the Y/F2 cells, since IGF-I-stimulated mitogenesis is normal; 3) Y/F2 cells display increased endogenous substrate phosphorylation and augmented insulin-stimulated S6 kinase activity placing these responses among insulin's mitogenic effects; and 4) these results are consistent with the concept that the COOH-terminal tyrosine residues of the insulin receptor are normally inhibitory to mitogenic signaling.     

The role of insulin receptor autophosphorylation in signal transduction.

We have examined the role of autophosphorylation in insulin signal transmission by oligonucleotide directed mutagenesis of seven potential tyrosine autophosphorylation sites in the human insulin receptor. Chinese hamster ovary cells transfected with these receptors were analyzed for insulin stimulated 2-deoxyglucose uptake, thymidine incorporation, endogenous substrate phosphorylation, and in vitro kinase activity. We found that phosphorylation on tyrosine residues 953, 1316, and 1322 were not necessary for receptor-mediated signal transduction. Mutation of tyrosine 960 reduced but did not abolish the signaling capabilities of the receptor. Finally, the simultaneous mutation of tyrosine residues 1146, 1150, and 1151 (the numbering system is that of Ullrich et al. (Ullrich, A., Bell, J. R., Chen, E. Y., Herrera, R., Petruzzelli, L. M., Dull, T. J., Gray, A., Coussens, L., Liao, Y. C., Tsubokawa, M., Mason, A., Seeburg, P.H., Grunfeld, C., Rosen, O. M., and Ramachandran, J. (1985) Nature 313, 756-761) resulted in a biologically inactive receptor, suggesting that the insulin receptor can be inactivated by removal of key autophosphorylation sites.     

Phorbol ester-mediated protein kinase C interaction with wild-type and COOH-terminal truncated insulin receptors.

The effects of 12-O-tetradecanoylphorbol-13-acetate (TPA) and insulin were compared in wild-type human insulin receptors (HIRc cells) and human insulin receptors lacking 43 COOH-terminal amino acid residues (HIR delta CT cells). TPA increased total phosphorylation of the wild-type insulin receptor and inhibited insulin-stimulated autophosphorylation by 32 +/- 10% in HIRc cells. TPA inhibited insulin-stimulated autophosphorylation by 46 +/- 14% in HIR delta CT cells and also caused a 65% decrease in basal phosphorylation. Insulin-stimulated tyrosine kinase activity for poly(Glu4/Tyr1) was inhibited by TPA in HIRc and HIR delta CT cells by 50 and 40%, respectively. TPA decreased insulin-stimulated glucose incorporation into glycogen by 50% in HIRc cells and to near basal levels in HIR delta CT cells; this inhibitory effect of TPA was reversed in both cell lines by staurosporine. In conclusion, 1) TPA-induced inhibition of insulin receptor tyrosine autophosphorylation was linked to concomitant inhibition of the biological effects of insulin in cells expressing either wild-type or COOH-terminal truncated insulin receptors; and 2) the inhibitory effects of TPA were not dependent upon phosphorylation of COOH-terminal residues and furthermore appeared to be independent of phosphorylation of any insulin receptor serine/threonine residues. These findings suggest a novel protein kinase C mechanism that results in altered insulin receptor function without increasing phosphorylation of the receptor.     

Receptor-mediated internalization of insulin requires a 12-amino acid sequence in the juxtamembrane region of the insulin receptor beta-subunit

The juxtamembrane region of the insulin receptor (IR) beta-subunit contains an unphosphorylated tyrosyl residue (Tyr960) that is essential for insulin-stimulated tyrosyl phosphorylation of some endogenous substrates and certain biological responses (White, M.F., Livingston, J.N., Backer, J.M., Lauris, V., Dull, T.J., Ullrich, A., and Kahn, C.R. (1988) Cell 54, 641-649). Tyrosyl residues in the juxtamembrane region of some plasma membrane receptors have been shown to be required for their internalization. In addition, a juxtamembrane tyrosine in the context of the sequence NPXY [corrected] is required for the coated pit-mediated internalization of the low density lipoprotein receptor. To examine the role of the juxtamembrane region of the insulin receptor during receptor-mediated endocytosis, we have studied the internalization of insulin by Chinese hamster ovary (CHO) cells expressing two mutant receptors: IRF960, in which Tyr960 has been substituted with phenylalanine, and IR delta 960, in which 12 amino acids (Ala954-Asp965), including the putative consensus sequence NPXY [corrected], were deleted. Although the in vivo autophosphorylation of IRF960 and IR delta 960 was similar to wild type, neither mutant could phosphorylate the endogenous substrate pp185. CHO/IRF960 cells internalized insulin normally whereas the intracellular accumulation of insulin by CHO/IR delta 960 cells was 20-30% of wild-type. However, insulin internalization in the CHO/IR delta 960 cells was consistently more rapid than that occurring in CHO cells expressing kinase-deficient receptors (CHO/IRA1018). The degradation of insulin was equally impaired in CHO/IR delta 960 and CHO/IRA1018 cells. These data show that the juxtamembrane region of the insulin receptor contains residues essential for insulin-stimulated internalization and suggest that the sequence NPXY [corrected] may play a general role in directing the internalization of cell surface receptors.     

The protein-tyrosine kinase activity of the insulin receptor is necessary for insulin-mediated receptor down-regulation

We have previously demonstrated that the human insulin receptor, mutated in the ATP-binding domain of the beta-subunit, is kinase-defective and fails to mediate multiple post-receptor actions of insulin in stably transfected Chinese hamster ovary cells (Chou, C.-K., Dull, T. J., Russell, D. S., Gherzi, R., Lebwohl, D., Ullrich, A., and Rosen, O. M. (1987) J. Biol. Chem. 262, 1842-1847). This study addresses the role of protein-tyrosine kinase activity in insulin-mediated receptor down-regulation. Although the mutant insulin proreceptor was properly processed and able to bind insulin like the wild-type human receptor, it differed from the latter in the following respects: 1) it failed to mediate internalization of surface-bound radiolabeled ligand; 2) it did not undergo short- or long-term down-regulation in response to 1 microM insulin; 3) it did not exhibit ligand-promoted receptor turnover; and 4) it was not phosphorylated on either tyrosine or serine residues in response to insulin. Although the cells transfected with the mutant receptor failed to respond to insulin-mediated insulin receptor down-regulation, they were able to down-regulate their insulin-like growth factor I receptors in response to insulin-like growth factor I or high concentrations of insulin and were sensitive to monoclonal antibody-induced down-regulation of their insulin receptors. Antibody-mediated receptor internalization alone, however, was unable to mimic at least one action of insulin, thymidine incorporation into DNA, and did not lead to any phosphorylation of the receptor. It is concluded that either the protein-tyrosine kinase activity of the insulin receptor or its phosphorylation state is essential for ligand-mediated receptor down-regulation.     

Enhanced insulin-induced mitogenesis and mitogen-activated protein kinase activities in mutant insulin receptors with substitution of two COOH-terminal tyrosine autophosphorylation sites by phenylalanine.

We have studied the function of a mutant human insulin receptor in which two COOH-terminal autophosphorylation sites (Tyr-1316 and -1322) were replaced by phenylalanine (F/Y COOH-terminal 2 tyrosines (CT2)). In addition, we have also constructed a mutant receptor in which Lys-1018 in the ATP-binding site was changed to arginine (R/K 1018). Both the wild type insulin receptor (HIR) and the mutant receptors were expressed in Chinese hamster ovary (CHO) cells by stable transfection. Autophosphorylation of solubilized and partially purified F/Y CT2 was decreased by approximately 30% compared with the HIR. Tyrosine kinase activities of F/Y CT2 and HIR toward exogenous substrates were almost equal. When CHO cells transfected with F/Y CT2 (CHO-F/Y CT2) were stimulated with insulin, autophosphorylation of the beta-subunit of the insulin receptor and the phosphorylation of an endogenous substrate (pp185) in the intact cell were normal compared with cells expressing HIR (CHO-HIR). CHO-F/Y CT2 exhibited the same insulin sensitivity as CHO-HIR with respect to 2-deoxyglucose uptake. However, the dose-response curve of insulin-stimulated thymidine incorporation in CHO-F/Y CT2 was shifted to the left (approximately 5-7-fold) compared with that in CHO-HIR. There was no significant difference in insulin-like growth factor 1-stimulated thymidine incorporation between CHO-F/Y CT2 and CHO-HIR. Furthermore, the dose-response curve of insulin-stimulated kinase activity toward myelin basic protein in CHO-F/Y CT2 was also shifted to the left (approximately 5-fold) compared with that in CHO-HIR. Kinase assays in myelin basic protein-containing gels revealed that both species of MAP kinases (M(r) 44,000, 42,000) were more sensitive to activation by insulin in CHO-F/Y CT2 than in CHO-HIR. This observation was confirmed in immune complex kinase assays toward microtubule-associated protein 2 (MAP2) using specific antibodies against mitogen-activated protein (MAP) kinase. R/K 1018 mutant insulin receptors showed an absence of insulin-stimulated kinase activity and CHO cells transfected with R/K 1018 (CHO-R/K 1018) failed to enhance 2-deoxyglucose uptake or thymidine incorporation in response to insulin. In addition, R/K 1018 kinase-defective insulin receptors were unable to mediate insulin-stimulated MAP kinase activation. These data suggest that: 1) tyrosine kinase activity of the insulin receptor is required for activation of insulin-stimulated MAP kinases and 2) phosphorylation of COOH-terminal tyrosine residues may play an inhibitory role in mitogenic signaling through regulation of MAP kinases.     

The metabolic and mitogenic effects of both insulin and insulin-like growth factor are enhanced by transfection of insulin receptors into NIH3T3 fibroblasts

Insulin and insulin-like growth factor (IGF-I) have 50% sequence homology and regulate similar cellular functions. Their membrane receptors also share 84% homology in a tyrosine kinase domain essential to transmembrane signaling and may thus share common postreceptor paths. To probe action mechanisms for these related hormones, we examined the receptor and postreceptor overlap of responses stimulated by insulin and IGF-I. NIH3T3 mouse fibroblasts have few endogenous insulin receptors and are insensitive to insulin; they have IGF-I receptors and are responsive to IGF-I. Stable transfection of these cells with cDNA for the human insulin receptor yielded a cell line (3T3/HIR) expressing greater than 6 x 10(6) receptors/cell that was highly sensitive and responsive to insulin for stimulation of deoxy[14C]glucose uptake and [3H]thymidine incorporation. The cells also showed increased responses to IGF-I, although the sensitivity was less than that for insulin. The receptor specificity of such responses was examined with a monoclonal antibody MA10 that bound to insulin receptors, but elicited no responses. When 3T3/HIR cells were preincubated with MA10, subsequent insulin- or IGF-I-stimulated deoxy[14C]glucose uptake was markedly inhibited. Likewise, the presence of MA10 caused a 10-fold increase in the concentration of insulin needed to stimulate half-maximal incorporation of [3H]thymidine and also led to diminished IGF-I-stimulated responses. These results showed that the transfected human insulin receptors coupled readily with existing effector pathways in the mouse fibroblasts and mediated metabolic and mitogenic responses to both insulin and IGF-I. Such findings indicate that insulin and IGF-I regulate common cellular functions using both overlapping receptor and postreceptor signaling pathways.     

Monoclonal antibodies mimic insulin activation of ribosomal protein S6 kinase without activation of insulin receptor tyrosine kinase. Studies in cells transfected with normal and mutant human insulin receptors

The effects of species-specific monoclonal antibodies to the human insulin receptor on ribosomal protein S6 phosphorylation were studied in rodent cell lines transfected with human insulin receptors. First, Swiss mouse 3T3 fibroblasts expressing normal human insulin receptors (3T3/HIR cells) were studied. Three monoclonal antibodies, MA-5, MA-20, and MA-51, activated S6 kinase in these cells but had no effects in untransfected 3T3 cells. Both insulin and MA-5, the most potent antibody, activated S6 kinase in a similar time- and dose-dependent manner. To measure S6 phosphorylation in vivo, 3T3/HIR cells were preincubated with [32P]Pi and treated with insulin and MA-5. Both agents increased S6 phosphorylation, and their tryptic phosphopeptide maps were similar. MA-5 and the other monoclonal antibodies, unlike insulin, failed to stimulate insulin receptor tyrosine kinase activity either in vitro or in vivo. Moreover, unlike insulin, they failed to increase the tyrosine phosphorylation of the endogenous cytoplasmic protein, pp 185. Next, HTC rat hepatoma cells, expressing a human insulin receptor mutant that had three key tyrosine autophosphorylation sites in the beta-subunit changed to phenylalanines (HTC-IR-F3 cells), were studied. In this cell line but not in untransfected HTC cells, monoclonal antibodies activated S6 kinase without stimulating either insulin receptor autophosphorylation or the tyrosine phosphorylation of pp 185. These data indicate, therefore, that monoclonal antibodies can activate S6 kinase and then increase S6 phosphorylation. Moreover, they suggest that activation of receptor tyrosine kinase and subsequent tyrosine phosphorylation of cellular proteins may not be crucial for activation of S6 kinase by the insulin receptor.     

Endocytosis of insulin receptors is not required for activation or deactivation of the hormone response

To examine the role of endocytosis in insulin action, hormone responsiveness was studied in transfected Rat 1 fibroblasts stably expressing a noninternalizing insulin receptor. The latter receptor (hIR delta ex16) was engineered by deleting the immediately submembranous 22 amino acids encoded by the 16th exon of the human insulin receptor and has previously been shown not to internalize despite having normal insulin-stimulated tyrosine kinase activity. It is shown in the present study that hIR delta ex16 receptors do mediate insulin action. Insulin dose-response curves both for activation of glycogen synthetase and for mitogenic stimulation demonstrate greater insulin sensitivity in hIR delta ex16 cells compared with untransfected Rat 1 cells. In addition, increases in the absolute levels of glycogen synthetase activity are seen in the hIR delta ex16 cells. Species-specific agonistic antibodies to the insulin receptor also stimulate hIR delta ex16 cells, confirming the activity of the mutant receptor. The non-internalizing receptors are rapidly dephosphorylated after removal of insulin, and the activation of glycogen synthetase decays no more slowly in hIR delta ex16 cells than in cells expressing wild-type receptors. The results demonstrate that receptor endocytosis is not necessary for activation or deactivation of the insulin response.     

Overexpression of insulin receptors in fibroblast and ovary cells induces a ligand-mediated transformed phenotype.

To investigate whether overexpression of the insulin receptor results in altered cell growth we used NIH 3T3 cells transfected with a bovine papilloma virus/insulin receptor cDNA construct (3T3/HIR). These cells expressed high numbers of insulin receptors (mean +/- sd, 631.0 +/- 16.7 ng receptors/10(6) cells). Insulin significantly stimulated the growth of 3T3/HIR cells maintained in serum-free medium. Moreover, in these cells, insulin induced marked phenotypic changes, including alterations in cell shape, loss of contact inhibition, and focal growth. In contrast to 3T3/HIR cells, insulin was without effect in either wild-type 3T3 cells (3T3/wt), 3T3 cells transfected with the neomycin resistance gene (3T3/NEO), or the bovine papilloma virus (3T3/BPV). To assess the presence of anchorage-independent growth, cells were seeded in soft agar and inspected for colony formation. 3T3/HIR cells showed absent or minimal colony growth in the absence of insulin. However, there was a dose-dependent insulin-stimulated increase in both colony size and number. Insulin-stimulated colony formation was specifically inhibited by an insulin antagonist, monoclonal antibody MA-10. In the presence of 100 nM insulin, about 3% of cells formed large colonies. Insulin neither stimulated growth nor induced colony formation in 3T3/wt cells or 3T3/NEO cells. Insulin also stimulated colony formation in CHO cells transfected with an insulin receptor cDNA construct. In conclusion, overexpression of normal insulin receptors induces a ligand-dependent transformed phenotype. This phenomenon may have clinical relevance by conferring a selective growth advantage to tumor cells with high numbers of insulin receptors.     

Cholesterol depletion disrupts caveolae and insulin receptor signaling for metabolic control via insulin receptor substrate-1, but not for mitogen-activated protein kinase control

Insulin exerts its cellular control through receptor binding in caveolae in plasmalemma of target cells (Gustavsson, J., Parpal, S., Karlsson, M., Ramsing, C., Thorn, H., Borg, M., Lindroth, M., Peterson, K. H., Magnusson, K.-E., and Str?lfors, P. (1999) FASEB. J. 13, 1961-1971). We now report that a progressive cholesterol depletion of 3T3-L1 adipocytes with beta-cyclodextrin gradually destroyed caveolae structures and concomitantly attenuated insulin stimulation of glucose transport, in effect making cells insulin-resistant. Insulin access to or affinity for the insulin receptor on rat adipocytes was not affected as determined by (125)I-insulin binding. By immunoblotting of plasma membranes, total amount of insulin receptor and of caveolin remained unchanged. Receptor autophosphorylation in response to insulin was not affected by cholesterol depletion. Insulin treatment of isolated caveolae preparations increased autophosphorylation of receptor before and following cholesterol depletion. Insulin-increased tyrosine phosphorylation of an immediate downstream signal transducer, insulin receptor substrate-1, and activation of the further downstream protein kinase B were inhibited. In contrast, insulin signaling to mitogenic control as determined by control of the extracellular signal-related kinases 1/2, mitogen-activated protein kinase pathway was not affected. Insulin did not control Shc phosphorylation, and Shc did not control extracellular signal-related kinases 1/2, whereas cholesterol depletion constitutively phosphorylated Shc. In conclusion, caveolae are critical for propagating the insulin receptor signal to downstream targets and have the potential for sorting signal transduction for metabolic and mitogenic effects.     

Insulin-Induced Circular Membrane Ruffling on Rat 1 Cells Expressing a High Number of Human Insulin Receptors: Circular Ruffles Caused by Rapid Actin Reorganization Exhibit High Density of Insulin Receptors and Phosphotyrosines

The early effects of insulin on morphological changes were examined using Rat 1 cells expressing one million human insulin receptors (Rat1HIR cells). Insulin transiently induced circular membrane ruffling on Rat1HIR cells. The circular ruffles were detected by phase microscopy as well as by fluorescence microscopy when actin was stained with rhodamine-conjugated phalloidin. The circular ruffles were also visualized by fluorescence microscopy when stained with either antiinsulin receptor monoclonal antibody, αIR-1, or antiphosphotyrosine antibody, followed by fluorescein isothiocyanate-labeled second antibodies. Control Rat1 cells or Rat1HIR A/K1018 cells expressing kinase-defective insulin receptors did not show any circular ruffles when treated with insulin. These results suggest that the circular ruffles are formed by actin reorganization induced by insulin and that this process requires the protein-tyrosine kinase activity of the receptor. Detection of insulin receptors and phosphotyrosines in the circular ruffles suggests the possibility that the insulin receptors in the ruffles may be highly active and phosphorylating either the receptors themselves or proteins associated with the ruffles. These results are consistent with our previous in vitro observations that a large aggregate form of the purified insulin receptor is a highly active protein-tyrosine kinase.     

Tyrosine kinase-defective insulin receptors undergo insulin-induced microaggregation but do not concentrate in coated pits

Biologically active colloid-gold complexes were used to compare ligand-induced microaggregation, redistribution, and internalization of insulin receptors on Rat 1 fibroblasts expressing wild type (HIRc) or tyrosine kinase-defective (HIR A/K1018) human insulin receptors. Insulin-like growth factor I (IGF I) and alpha 2-macroglobulin receptors also were compared. On both cell types, all four unoccupied receptor types occurred predominantly as single receptors. Ligand binding caused receptor microaggregation. Microaggregation of wild type or kinase-defective insulin receptors or IGF I receptors was not different. alpha 2-Macroglobulin receptors formed larger microaggregates. Compared to wild type insulin or IGF I receptors, accumulation of kinase-defective insulin receptor microaggregates in endocytic structures was decreased, and the size of microaggregates in coated pits was significantly smaller. As a result, receptor-mediated internalization of gold-insulin by HIR A/K1018 cells was less than 6% of the cell-associated particles compared to approximately 60% of the particles in HIRc cells. On HIR A/K1018 cells, alpha 2-macroglobulin and IGF I were internalized via coated pits demonstrating that those structures were functional. These results suggest that: 1) ATP binding, receptor autophosphorylation, and activation of receptor kinase activity are not required for receptor microaggregation; 2) receptor microaggregation per se is not sufficient to cause ligand-induced receptor-mediated internalization or the biological effects of insulin; and 3) autophosphorylation of the beta-subunit or activation of the receptor kinase activity is required for the insulin-induced concentration of occupied receptors in coated pits.