Antibodies to insulin receptor tyrosine kinase stimulate its activity towards exogenous substrates without inducing receptor autophosphorylation

Insulin receptor kinase-independent signaling via tyrosine phosphorylation of phosphatase PHLPP1

Most insulin responses correlate well with insulin receptor (IR) Tyr kinase activation; however, critical exceptions to this concept have been presented. Specific IR mutants and stimulatory IR antibodies demonstrate a lack of correlation between IR kinase activity and specific insulin responses in numerous independent studies. IR conformation changes in response to insulin observed with various IR antibodies define an IR kinase-independent signal that alters the C-terminus. IR-related receptors in lower eukaryotes that lack a Tyr kinase point to an alternative mechanism of IR signaling earlier in evolution. However, the implied IR kinase-independent signaling mechanism remained obscure at the molecular level. Here we begin to define the molecular basis of an IR-dependent but IR kinase-independent insulin signal that is equally transmitted by a kinase-inactive mutant IR. This insulin signal results in Tyr phosphorylation and catalytic activation of phosphatase PHLPP1 via a PI 3-kinase-independent, wortmannin-insensitive signaling pathway. Dimerized SH2B1/PSM is a critical activator of the IR kinase and the resulting established insulin signal. In contrast it is an inhibitor of the IR kinase-independent insulin signal and disruption of SH2B1/PSM dimer binding to IR potentiates this signal. Dephosphorylation of Akt2 by PHLPP1 provides an alternative, SH2B1/PSM-regulated insulin-signaling pathway from IR to Akt2 of opposite polarity and distinct from the established PI 3-kinase-dependent signaling pathway via IRS proteins. In combination, both pathways should allow the opposing regulation of Akt2 activity at two phosphorylation sites to specifically define the insulin signal in the background of interfering Akt-regulating signals, such as those controlling cell proliferation and survival.

Interaction of the C-terminal acidic domain of the insulin receptor with histone modulates the receptor kinase activity

In this study, we investigated the role of the insulin receptor domain 1270-1280, an acid-rich sequence located in the receptor C-terminus. Antipeptide IgG raised against this sequence were obtained and used to analyze their effect on receptor function. Antipeptide IgG inhibited receptor autophosphorylation at Tyr1146, Tyr1150 and Tyr1151. These sites are known to be key modulators of the receptor activity. Autophosphorylation at other sites may also have been inhibited. The antipeptide antibody decreased the receptor kinase activity measured with poly(Glu80Tyr20) and a synthetic peptide corresponding to the proreceptor sequence 1142-1158. We provide evidence that the effect of the antibody on substrate phosphorylation may result from the control of the phosphorylation level of the receptor. Concerning the action of the antipeptide IgG on the receptor kinase activity, histone did not behave similarly to poly(Glu80Tyr20). The antibody recognizing sequence 1270-1280 competed with histone for an overlapping binding site. Histone also modulated insulin receptor autophosphorylation, supporting the idea that interference with domain 1270-1280 alters the receptor kinase. Our data suggest that the acidic region including residues 1270-1280 of the insulin receptor C-terminus is involved in the following events: (a) receptor binding with histone, an exogenous substrate of the receptor kinase, and (b) the regulation of receptor autophosphorylation and kinase activity. Based on these observations, we would like to propose that this insulin receptor domain could interact with cellular proteins modulating the receptor kinase.

A monoclonal anti-peptide antibody reacting with the insulin receptor beta-subunit. Characterization of the antibody and its epitope and use in immunoaffinity purification of intact receptors

A mouse monoclonal antibody (CT-1) was prepared against the C-terminal peptide sequence of the human insulin receptor beta-subunit (KKNGRILTLPRSNPS). The antibody reacted with native human and rat insulin receptors in solution, whether or not insulin was bound and whether or not the receptor had undergone prior tyrosine autophosphorylation. The antibody also reacted specifically with the receptor beta-subunit on blots of SDS/polyacrylamide gels. Preincubation of soluble receptors with antibody increased the binding of 125I-insulin approx. 2-fold. The antibody did not affect insulin-stimulated autophosphorylation, but increased the basal autophosphorylation rate approx. 2-fold. The amino acid residues contributing to the epitope for CT-1 were defined by construction and screening of an epitope library. Oligonucleotides containing 23 random bases were synthesized and ligated into the vector pCL627, and the corresponding peptide sequences expressed as fusion proteins in Escherichia coli were screened by colony blotting. Reactive peptides were identified by sequencing the oligonucleotide inserts in plasmids purified from positive colonies. Six different positive sequences were found after 900,000 colonies had been screened, and the consensus epitope was identified as GRVLTLPRS. Phosphorylation of the threonine residue within this sequence (corresponding to the known phosphorylation site Thr-1348 in the insulin receptor) decreased the affinity of antibody binding approx. 100-fold, as measured by competition in an e.l.i.s.a. Antibody CT-1 was used for immunoaffinity isolation of insulin receptor from detergent-solubilized human placental or rat liver microsomal membranes. Highly purified receptor was obtained in 60% yield by binding to CT-1-Sepharose immunoadsorbent and specific elution with a solution of peptide corresponding to the known epitope. This approach to purification under very mild conditions may in principle be used with any protein for which an antibody is available and for which a peptide epitope or 'mimotope' can be identified.

Phenylarsine oxide stimulates a cytosolic tyrosine kinase activity and glucose transport in mouse fibroblasts

In the present report we further approach the mechanism by which insulin and phenylarsine oxide (PAO), a trivalent arsenical compound, regulate glucose transport in mouse fibroblasts (NIH3T3). First, we show that PAO is a powerful stimulatory agent on glucose transport. Second, at least three series of observations indicate that this action of PAO is not mediated through the insulin receptor: (i) the same effect of PAO is observed in NIH3T3 and in transfected cells expressing 6 x 10(6) insulin receptors, while the effect of insulin is markedly increased in the transfected cells; (ii) PAO does not affect the tyrosine phosphorylation of the insulin receptor; (iii) the tyrosine kinase activity of the insulin receptor toward exogenous substrates is not increased by PAO. Since PAO appears to act on glucose transport by a different mechanism than insulin, we have compared the effect of PAO and insulin on tyrosine phosphorylation of cellular proteins. Using Western blot analysis we did not detect common substrates in PAO- and insulin-treated cells. However, we found in cell extracts from both PAO- and insulin-treated cells a 50-kDa protein that is immunoprecipitated by antiphosphotyrosine antibody. In addition, PAO activates a cytosolic tyrosine kinase capable of poly(Glu/Tyr) phosphorylation. As a whole, our data suggest that the 50-kDa protein found in cells incubated with PAO and insulin could be the convergence point of the insulin and PAO signaling pathways.

Control of insulin receptor autophosphorylation by polypeptide substrates: inhibition and stimulation by interaction with the catalytic subunit

Autophosphorylation of purified insulin receptor, in the absence of insulin, was stimulated by selected polypeptide substrates. In the presence of 1 microM insulin these peptides inhibited autophosphorylation. Stimulation was observed with reduced [S-(carboxamidomethyl)cysteinyl]lysozyme (RCAM-lysozyme) and three peptides generated by CNBr cleavage, V8 proteinase digestion and/or chemical modification. We also generated two peptide substrates from RCAM-lysozyme which did not stimulate receptor autophosphorylation and were very weak inhibitors. As a control peptide, the simple substrate angiotensin inhibited receptor autophosphorylation in the absence or presence of insulin. However, stimulatory peptide, but not insulin, significantly shifted the concentration dependence for inhibition by angiotensin. The stimulatory peptides also increased autophosphorylation of the cloned cytoplasmic domain of the kinase [R-BIRK; Villalba, M., Wente, S. R., Russell, D. S., Ahn, J., Reichelderfer, C. F., & Rosen, O. M. (1989) Proc. Natl. Acad. Sci. U.S.A. 86, 7848]. Therefore, stimulation occurs by interaction with the cytoplasmic process of the beta-subunit and not through interaction with the insulin binding alpha-subunit of the native receptor. Autophosphorylation was analyzed by mapping 32P-labeled tryptic phosphopeptides from the beta-subunit and from R-BIRK. Nearly identical phosphopeptide maps were found, comparing first, basal R-BIRK and basal native receptor, second, peptide- and insulin-stimulated native receptor, and third, peptide-stimulated R-BIRK and insulin-stimulated native receptor. Therefore, R-BIRK functions as a basal-state enzyme and can be stimulated in an insulin-like manner. On the basis of these observations, stimulation by insulin and by peptides yields similar functional results, but by apparently different mechanisms.

Insulin and insulin-like-growth-factor-I (IGF-I) receptors in Xenopus laevis oocytes. Comparison with insulin receptors from liver and muscle

Insulin and insulin-like-growth-factor-I (IGF-I) receptors were partially purified from full-grown (stages V-VI) Xenopus laevis oocytes by affinity chromatography on wheat-germ agglutinin-agarose. Competitive-binding assays revealed high-affinity binding sites for both insulin and IGF-I (Kd = 2.5 x 10(-10) M and 8 x 10(-10) M respectively). However, IGF-I receptors were about 15 times more abundant than insulin receptors (22.5 x 10(11) versus 1.5 x 10(11)/mg of protein). Moreover, comparison of intact and collagenase-treated oocytes showed that most of the insulin receptors were in the oocyte envelopes, whereas IGF-I receptors were essentially at the oocyte surface. Oocyte receptors were composed of alpha-subunits of approximately 130 kDa and a doublet of beta-subunits of 95 and 105 kDa, which both had ligand-induced phosphorylation patterns compatible with IGF-I receptor beta-subunits. Accordingly, the receptor tyrosine kinase was stimulated at low IGF-I concentrations [half-maximally effective concentration (EC50) approximately 0.5-1 nM], and at higher insulin concentrations (EC50 approximately 20-50 nM). Partially purified glycoproteins from Xenopus liver and muscle contained mainly receptors of the insulin-receptor type, with alpha-subunits of 140 kDa in liver and 125 kDa in muscle, and doublets of beta-subunits of 92-98 kDa in liver and 85-94 kDa in muscle. Immunoprecipitation of receptors from oocytes, liver and muscle by receptor-specific anti-peptide antibodies suggested that the beta-subunit heterogeneity resulted from the existence of two distinct IGF-I receptors in oocytes and of two distinct insulin receptors in both liver and muscle. In the different tissues, the two receptor subtypes differed at least by their beta-subunit C-terminal region.

Autoantibodies to the insulin receptor are infrequent findings in type 1 (insulin-dependent) diabetes mellitus of recent onset

To determine whether autoantibodies to the insulin receptor may represent markers of Type 1 (insulin-dependent) diabetes, the prevalence of such antibodies was investigated in sera of 60 newly diagnosed untreated Type 1 diabetic patients. A sensitive assay, based on enzyme linked immunosorbent assay has been set up which detects antibodies to the insulin receptor irrespective of their potentially inhibiting effect on insulin binding. Moreover, this method allows easy determination of the immunoglobulin class involved in the anti-receptor activity. Among the 60 sera examined, only one was found to contain anti-insulin receptor autoantibodies (IgG class). In view of our data, we conclude that autoantibodies to the insulin receptor are infrequent findings in Type 1 diabetes of recent onset.

Basal autophosphorylation of insulin receptor occurs preferentially on the receptor conformation exhibiting high affinity for insulin and stabilizes this conformation

Insulin signal transmission through the plasma membrane was studied in terms of relationship between basal autophosphorylation of the beta-subunit and the ability to bind insulin by the alpha-subunit of the insulin receptor. In a cell free system, receptors phosphorylated on tyrosine residues in the absence of insulin were separated from non-phosphorylated receptors using antiphosphotyrosine antibodies. Insulin binding assays were then performed on basally autophosphorylated and on non-phosphorylated receptors. We found that the tyrosine phosphorylated receptors, which corresponded to 25% of the total number of receptors, were accountable for 60-80% of insulin binding. Scatchard representation of binding data has shown that the plot corresponding to tyrosine phosphorylated receptors was localized above, and was steeper than the plot corresponding to non-phosphorylated receptors. These data make it likely that the conformation of alpha-subunit which favours ligand binding is connected to the conformation of beta-subunit which favours phosphate reception on tyrosine residues. Reciprocally, the high-affinity conformation of insulin receptor seems to become stabilized by basal autophosphorylation.

Intrinsic kinase activity of the insulin receptor

Since the identification of the insulin receptor by insulin-binding activity almost two decades ago, our understanding of the structure and function of the insulin receptor has progressed tremendously. The importance of the intrinsic tyrosine protein kinase activity of the insulin receptor is implied by the fact that the insulin receptor belongs to a family of receptor tyrosine kinases which play a role in growth control, by experiments demonstrating the intimate association of normal kinase activity and insulin action, and by evidence that the intrinsic kinase activity can be regulated under certain conditions. There are still some major gaps in our knowledge concerning the structure/function of the insulin receptor such as how activation of the intrinsic kinase activity of the receptor leads to altered cellular physiology. The kinase may phosphorylate endogenous substrates or autophosphorylation may simply alter beta subunit conformation so it can then interact with an effector system (i.e. a serine kinase) directly, or indirectly through a G-protein. The truth may lie somewhere between these two pathways.