The mechanism by which the binding of epidermal growth factor (EGF) to specific cell surface receptors induces a range of biological responses remains poorly understood. An important part of the study of signal transduction in this system involves the production of sufficient native and mutant EGF receptor species for X‐ray crystallographic and spectroscopic analysis. Baculovirus vectors containing the cDNA encoding the human EGF receptor protein have here been utilized to infect insect cells. This results in expression of a 155‐kb transmembrane protein which is recognized by four antibodies against different regions of the human EGF receptor. Studies with tunicamycin, monensen and endoglycosidase H show the difference in size between the recombinant and the native receptor is due to alterations in glycocsylation. Studies of [125I] EGF binding shows a Kd of 2 X 10(‐9) M in intact infected insect cells which falls to 2 X 10(‐7) M upon detergent solubilization. The recombinant protein exhibits an EGF‐stimulated tyrosine protein kinase activity and an analysis of tryptic peptides shows that the phosphate acceptor sites are similar to those of the EGF receptor isolated from A431 cells. These observations indicate that functional EGF receptor can be expressed in insect cells, and furthermore, this system can be used for large‐scale production.

Leptin is secreted from adipose tissue, and is thought to act as a 'lipostat', signalling the body fat levels to the hypothalamus resulting in adjustments to food intake and energy expenditure to maintain body weight homeostasis. In addition, plasma leptin concentrations have been shown to be related to insulin sensitivity independent of body fat content, suggesting that the hyperleptinemia found in obesity could contribute to the insulin resistance. We investigated the effects of leptin on insulin binding by isolated adipocytes. Adipocytes isolated from Sprague-Dawley rats exhibited a dose-dependent reduction in the uptake of 125I-labelled insulin when incubated with various concentrations of exogenous leptin. For example, addition of 50 nM leptin reduced total insulin binding in isolated adipocytes by 19% (P < 0.05). Analysis of displacement curve binding data suggested that leptin reduced maximal insulin binding in a dose-dependent manner, but had no significant effect on the affinity of insulin for its binding site. We conclude that leptin directly inhibited insulin binding by adipocytes, and the role of leptin in the development of insulin resistance in obese individuals requires further investigation.

Sulphydryl reagents have been shown to produce a variety of effects on insulin-receptor structure and function. However, localization of these effects to specific receptor domains has not been attempted. We have investigated this question with insulin- and epidermal growth factor (EGF)-receptors (both are receptor tyrosine kinases but have different sulphydryl/disulphide structures within the external domain), and the insulin receptor kinase (IRK) protein consisting solely of the insulin-receptor cytoplasmic domain and exhibiting constitutive kinase activity. Results showed a differential response between basal and activated receptors. The physiological reductant GSH stimulated basal receptor autophosphorylation, but was either without effect (EGF) or inhibited (insulin) activated receptors, and occurred without visible reduction of receptor structure. These results contrast with those obtained with dithiothreitol which appears to activate phosphorylation in association with reduction of the extracellular insulin-receptor disulphides, but is without effect on the EGF receptor or the IRK protein. Alkylating agents N-ethylmaleimide (NEM) and iodoacetamide (IAM) had opposing effects on receptor autophosphorylation. However, only in the basal state was IAM able to protect receptors from the inhibitory effect of NEM. Our results suggest that complex sulphydryl interactions can occur within the cytoplasmic domain of insulin- and EGF-receptors to alter receptor kinase activity. The basal and activated state of receptors is not the same with respect to sulphydryl reagent action, possibly due to conformational change in the receptor induced by ligand (insulin, EGF) or constitutive (IRK) activation.

The insulin receptor for human placental membranes has been solubilized in Triton X-100 and its properties have been examined in detail. Binding of [125 I]iodoinsulin to the soluble receptor is markedly inhibited by increas-ng concentrations of Triton X-100, due to a fall in receptor affinity. In 0.02--0.10% Triton X-100, the soluble receptor exhibits all the essential characteristics of the intact or particulate receptor. These include strict specificity for insulin and its analogues, increase in steady state binding with decrease in temperature, a pH optimum at 7.8--8.0, and negatively cooperative site-site interactions. The initial association rate of [125 I]iodoinsulin and the soluble receptor is a direct function of temperature, but the level of steady-state binding is lower at higher temperatures due to a marked increase in dissociation rate. Scatchard binding plots are curvilinear and show a large increase in affinity at 4 C with no change in total binding capacity (R0); increased binding to the particulate placental membrane at 4 C is due chiefly to an increase in R3. Negative cooperatively in the soluble receptor has been confirmed by kinetic experiments; thus, the dissociation of [125I]iodoinsulin from the receptor in the presence of "infinite" dilution is accelerated in the presence of 10(-8) M unlabeled insulin. The apparent molecular weight of the placental receptor, determined by gel filtration on 6% agarose, is approximately 300,000. These studies show that the basic properties of the insulin receptor do not depend on it being an integral conponent of the cell membrane.

The hexosamine biosynthesis pathway and protein kinase C (PKC) activation mediate hyperglycaemia-induced impaired glucose transport, but the relative role of each pathway is unknown. Following a 2 h preincubation of rat adipocytes in the presence of either high glucose (30 mM) plus insulin (0.7 nM) or glucosamine (3 mM), both high glucose and glucosamine inhibited subsequent basal and insulin-stimulated glucose transport, measured at 5.0 mM glucose. Azaserine, an inhibitor of the enzyme glutamine:fructose-6-phosphate aminotransferase, abolished the effect of high glucose, but not that of glucosamine. Ro-31-8220, an inhibitor of PKC, reversed the effects of both high glucose and glucosamine, suggesting that flux through the hexosamine biosynthesis pathway impaired glucose transport acutely by activating PKC. Both high glucose and glucosamine caused a 3-fold increase in PKC activity; this effect of high glucose, but not that of glucosamine, was partially decreased by azaserine. Neither high glucose nor glucosamine altered basal or insulin-stimulated plasma membrane GLUT1 levels, whereas both treatments decreased basal, but not insulin-stimulated, GLUT4 levels. Azaserine abolished the effect of high glucose, but not that of glucosamine, on basal plasma membrane GLUT4 levels. Ro-31-8220, which returned glucose transport to control values, caused a further decrease in plasma membrane GLUT4 levels. It is concluded that, in rat adipocytes, an acute increase in flux through the hexosamine biosynthesis pathway inhibits glucose transport by activation of PKC.

The tyrosine kinase of the insulin receptor can be activated by trypsin treatment. The concomitant abolition of insulin binding has been postulated to result from proteolytic destruction of the receptor. A discrepancy between the decrease in insulin binding and receptor immunoreactivity after trypsin treatment led us to investigate more closely the structure of the trypsin-treated receptor. After trypsin treatment of the CHOT cell line, which over-expresses transfected human insulin receptors, insulin binding was significantly decreased, but reactivity with five alpha-subunit monoclonal antibodies was either unaffected or only moderately decreased, indicating that the alpha-subunit was substantially intact. Examination of receptor structure after trypsin treatment, receptor autophosphorylation and gel electrophoresis revealed a single band at 110 kDa in non-reduced gels, comprising a small fragment (21 kDa) of the alpha-subunit linked to the beta-subunit by class II disulphides. When the receptor was radio-labelled with 125I, two additional alpha-subunit bands of 142 kDa and 81 kDa (composed of identical reduced bands) were observed on non-reduced gels, which contained disulphide-linked (class I) fragments. All fragments could be precipitated by antibodies to both alpha- and beta-subunits. However, only antibodies directed towards the N-terminus of the receptor could immunoblot trypsin-treated fragments. Thus activation of the receptor tyrosine kinase by trypsin occurs after cleavage, but not loss of the alpha-subunit. This finding has implications for the mechanism of transmembrane activation of the receptor kinase by insulin.

No abstract

The EGF receptor cDNA has been transfected into receptor-negative Chinese hamster ovary (CHO) cells. A mutant cell line (CHO 11) was isolated that expresses a receptor of lower molecular weight than the EGF receptor from A431 cells (150,000 daltons compared to 170,000 daltons) and which appeared as a doublet on SDS-PAGE. By digestion of the receptor with endoglycosidase F it was shown that an altered pattern of glycosylation could not account for the smaller size of the protein, although it could explain the appearance of the CHO 11 receptor as a doublet protein. A deletion was located to the transfected cDNA and shown to involve the removal of coding sequences for the most C-terminal 20,000 daltons of the EGF receptor, which contains the three major autophosphorylation sites. Despite the loss of these sites the EGF receptor from CHO 11 cells binds EGF, demonstrates protein tyrosine kinase activity in response to EGF, and transduces a mitogenic signal. The CHO 11 receptor protein is still autophosphorylated on alternative tyrosine residues. We conclude that phosphorylation of the three tyrosines (P1, P2, and P3) in the C-terminal domain of the receptor is not required for signal transduction by the EGF receptor in these cells.