Appraising the mitogenicity of insulin analogues relative to human insulin-response to: Weinstein D, Simon M, Yehezkel E, Laron Z, Werner H. Insulin analogues display IGF-I-like mitogenic and anti-apoptotic activity in cultured cancer cells. Diabetes Metab Res Rev 2009; 25(1): 41-9.

Interest in mitogenic and potentially carcinogenic effects of insulin and insulin analogues has been renewed by several recent publications that have examined the relationship between cancer and insulin analogues. Actions mediated through the insulin-like growth factor-I receptor in a hyperinsulinaemic state have been implicated mechanistically. Both type 2 diabetes and endogenously elevated insulin-like growth factor-I have been epidemiologically linked to malignancies. Therefore, in vitro mitogenic effects and binding affinities of the various analogues have been analysed. A recent publication by Weinstein et al. studied the in vitro mitogenic and anti-apoptotic activities of insulin analogues, and their conclusion asserts that insulins glargine, detemir, and lispro displayed proliferative and anti-apoptotic effects in a number of malignant cell lines. However, their conclusions are not supported by the data which are not complete and lack clear statistical significance. This data should be interpreted cautiously in light of all other presently available scientific evidence. Prospective, randomized clinical trials will best address any direct relationship between insulin analogues and cancer. Until those studies are designed and completed, clinicians should consider the demonstrated strong benefit of glycaemic control in balance with any alleged risk.

Uncoupling protein 3 (UCP3) stimulates glucose uptake in muscle cells through a phosphoinositide 3-kinase-dependent mechanism.

UCP3 is a mitochondrial membrane protein expressed in humans selectively in skeletal muscle. To determine the mechanisms by which UCP3 plays a role in regulating glucose metabolism, we expressed human UCP3 in L6 myotubes by adenovirus-mediated gene transfer and in H(9)C(2) cardiomyoblasts by stable transfection with a tetracycline-repressible UCP3 construct. Expression of UCP3 in L6 myotubes increased 2-deoxyglucose uptake 2-fold and cell surface GLUT4 2.3-fold, thereby reaching maximally insulin-stimulated levels in control myotubes. Wortmannin, LY 294002, or the tyrosine kinase inhibitor genistein abolished the effect of UCP3 on glucose uptake, and wortmannin inhibited UCP3-induced GLUT4 cell surface recruitment. UCP3 overexpression increased phosphotyrosine-associated phosphoinositide 3-kinase (PI3K) activity 2.2-fold compared with control cells (p < 0.05). UCP3 overexpression increased lactate release 1.5- to 2-fold above control cells, indicating increased glucose metabolism. In H(9)C(2) cardiomyoblasts stably transfected with UCP3 under control of a tetracycline-repressible promotor, removal of doxycycline resulted in detectable levels of UCP3 at 12 h and 2.2-fold induction at 7 days compared with 12 h. In parallel, glucose transport increased 1.3- and 2-fold at 12 h and 7 days, respectively, and the stimulation was inhibited by wortmannin or genistein. p85 association with membranes was increased 5.5-fold and phosphotyrosine-associated PI3K activity 3.8-fold. In contrast, overexpression of UCP3 in 3T3-L1 adipocytes did not alter glucose uptake, suggesting tissue-specific effects of human UCP3. Thus, UCP3 stimulates glucose transport and GLUT4 translocation to the cell surface in cardiac and skeletal muscle cells by activating a PI3K dependent pathway.

Melanin-concentrating hormone receptor is a target of leptin action in the mouse brain.

Melanin-concentrating hormone (MCH) is a hypothalamic neuropeptide that is important in the regulation of energy homeostasis. MCH signals via a seven-transmembrane G protein-coupled receptor, which is coupled to Galpha(i). This receptor was initially cloned in rat and human and designated SLC-1 because of its homology to the somatostatin receptor. In rat brain, it is expressed in a pattern that mirrors the previously described pattern of projections of MCH-immunoreactive fibers. In the present study we cloned the mouse MCH receptor (MCH-R) ortholog by a rapid amplification of 5'- and 3'-cDNA ends approach and have found it to be 98% homologous with the rat sequence. We have characterized MCH-R messenger RNA distribution in the mouse brain by in situ hybridization and have shown MCH-R to be expressed in diverse brain areas implicated in the regulation of feeding, body adiposity, and sensory integration of smell and gustatory inputs, including the hypothalamus [paraventricular nucleus (magnocellular part) and dorsomedial, ventromedial, and arcuate nucleus], areas of the olfactory pathway, and the nucleus of the solitary tract. We also studied MCH-R regulation and found that MCH-R expression is increased 7-fold by 48-h fasting or genetic leptin deficiency (ob/ob mice) and is completely blunted by leptin administration. In contrast, MCH-R messenger RNA expression remains unaltered in genetic MCH deficiency. Our findings suggest that MCH-R constitutes a central target of leptin action in the mammalian brain.

The effect of weight reduction on skeletal muscle UCP2 and UCP3 mRNA expression and UCP3 protein content in Type II diabetic subjects.

AIMS/HYPOTHESIS: The aim of this study was to examine the effect of weight loss on UCP2/UCP3 mRNA expression and UCP3 protein content in subjects with Type II (non-insulin-dependent) diabetes mellitus. METHODS: We studied seven Type II diabetic subjects who followed a 10-week very low calorie diet. Expression of skeletal muscle UCP2 and UCP3 mRNA was measured using RT-competitive PCR and UCP3 protein content by western blotting, before and after the diet. Total and plasma fatty acid oxidation was measured using infusion of 13C labelled palmitate. RESULTS: Body weight decreased from 105.5 +/- 8.2 kg to 91.6 +/- 7.2 kg (p < 0.001), after 10 weeks of diet intervention. Expression of UCP2 and UCP3 mRNA were significantly reduced after 10 weeks of diet (p < 0.05) but UCP3 protein contents were not significantly altered. Notably, the change in UCP3L mRNA expression and UCP3 protein content after the very low calorie diet were negatively associated with changes in body weight (r = -0.97, p = 0.006 and r = -0.83, p = 0.043, respectively) and BMI (r = -0.99, p = 0.0007 and r = -0.9, p = 0.016, respectively). Furthermore, changes in UCP3L mRNA expression and UCP3 protein content induced by the diet were positively correlated with changes in cytosolic fatty acid-binding protein content (r = 0.93, p = 0.023 and r = 0.84, p = 0.039, respectively). No correlation between diet-induced changes in UCP3 protein and resting energy expenditure or plasma non-esterified fatty acid concentrations were found. CONCLUSION/INTERPRETATION: The negative correlation between the change in UCP3 protein content after weight loss and the change in BMI, suggests that the decrease in UCP3 during weight loss could prevent further weight loss. The finding that the change in UCP3 protein content correlates with the change in skeletal muscle fatty acid-binding protein content, suggests a role for UCPs in the handling of lipids as a fuel.

Characterization of [Phe(13), Tyr(19)]-MCH analog binding activity to the MCH receptor.

Melanin concentrating hormone (MCH), a hypothalamic neuropeptide, is an important regulator of energy homeostasis in mammals. Characterization of an MCH specific receptor has been hampered by the lack of a suitable radioligand. The [Phe(13), Tyr(19)]-MCH analog has been shown by different investigators to bind specifically to cell lines of epithelial or pigment cell origin. Recently, using functional assays, the MCH receptor has been characterized as a seven transmembrane G-coupled protein initially identified as SLC-1. In the present study, we used tyrosine iodinated [Phe(13), Tyr(19)]-MCH analog, which stimulates food intake in a manner similar to that of MCH, as well as native MCH to conduct binding studies. Specific binding could not be demonstrated in intact cells of several cell lines, including A431 and B16. Specific binding associated with membranes localized to the microsomal, not the plasma membrane, fraction. Message for SLC-1 was absent in these cell lines, as assessed by Northern blot analysis. We conclude that cells previously reported to express the MCH receptor do not express SLC-1 and that both iodinated MCH and the [Phe(13), Tyr(19)]-MCH have a large component of non-specific binding. These ligands may be useful for binding studies in transfected cells with high levels of SLC-1 expression. However they do not appear to be suitable for screening for the MCH receptor as most cells demonstrate significant low affinity non-specific binding.

Assessment of uncoupling activity of the human uncoupling protein 3 short form and three mutants of the uncoupling protein gene using a yeast heterologous expression system.

The human uncoupling protein 3 gene generates two mRNA transcripts, uncoupling protein 3L and uncoupling protein 3S, which are predicted to encode long and short forms of the uncoupling protein 3 protein, respectively. While uncoupling protein 3L is similar in length to the other known uncoupling proteins, uncoupling protein 3S lacks the last 37 C-terminal residues. A splice site mutation in the human uncoupling protein 3 gene, resulting in the exclusive expression of uncoupling protein 3S, and a number of point mutations in the uncoupling protein 3 gene have been described. This study compares the biochemical activity of uncoupling protein 3S as well as three mutants of the uncoupling protein 3 gene (V9M, V102I, R282C) with that of uncoupling protein 3L utilizing a yeast expression system. All proteins were expressed at similar levels and had qualitatively similar effects on parameters related to the uncoupling function. Both uncoupling protein 3S and uncoupling protein 3L decreased the yeast growth rate by 35 and 52%, increased the whole yeast basal O2 consumption by 26 and 48%, respectively, and decreased the mitochondrial membrane potential as measured in whole yeast by uptake of the fluorescent potential-sensitive dye 3'3-dihexyloxacarbocyanine iodide. In isolated mitochondria, uncoupling protein 3S and uncoupling protein 3L caused a similar (33 and 35%, respectively) increase in state 4 respiration, which was relatively small compared to uncoupling protein 1 (102% increase). A truncated version of uncoupling protein 3S, lacking the last three C-terminal residues, Tyr, Lys and Gly, that are part of a carrier motif that is highly conserved among all mitochondrial carriers, had a greatly reduced uncoupling activity. The two naturally occurring uncoupling protein 3 mutants, V9M and V102I, were similar to uncoupling protein 3L with respect to effects on the yeast growth and whole yeast O2 consumption. The R282C mutant had a reduced effect compared to uncoupling protein 3L. In summary, uncoupling protein 3S and the three mutants of uncoupling protein 3 appear to be functional proteins with biochemical activities similar to uncoupling protein 3L, although uncoupling protein 3S and the R282C mutant have a modestly reduced function.

Assessment of uncoupling activity of uncoupling protein 3 using a yeast heterologous expression system.

Uncoupling protein 3L, uncoupling protein 1 and the mitochondrial oxoglutarate carrier were expressed in Saccharomyces cerevisae. Effects on different parameters related to the energy expenditure were studied. Both uncoupling protein 3L and uncoupling protein 1 reduced the growth rate by 49% and 32% and increased the whole yeast O2 consumption by 31% and 19%, respectively. In isolated mitochondria, uncoupling protein 1 increased the state 4 respiration by 1.8-fold, while uncoupling protein 3L increased the state 4 respiration by 1.2-fold. Interestingly, mutant uncoupling protein 1 carrying the H145Q and H147N mutations, previously shown to markedly decrease the H+ transport activity of uncoupling protein 1 when assessed using a proteoliposome system (Bienengraeber et al. (1998) Biochem. 37, 3-8), uncoupled the mitochondrial respiration to almost the same degree as wild-type uncoupling protein 1. Thus, absence of this histidine pair in uncoupling protein 2 and uncoupling protein 3 does not by itself rule out the possibility that these carriers have an uncoupling function. The oxoglutarate carrier had no effect on any of the studied parameters. In summary, a discordance exists between the magnitude of effects of uncoupling protein 3L and uncoupling protein 1 in whole yeast versus isolated mitochondria, with uncoupling protein 3L having greater effects in whole yeast and a smaller effect on the state 4 respiration in isolated mitochondria. These findings suggest that uncoupling protein 3L, like uncoupling protein 1, has an uncoupling activity. However, the mechanism of action and/or regulation of the activity of uncoupling protein 3L is likely to be different.

Changes in leptin expression are not associated with corresponding changes in CCAAT/enhancer binding protein-alpha.

C/EBP-alpha binds a C/EBP consensus site in the leptin promoter and activates transcription in vitro. We assessed adipose tissue expression of C/EBP-alpha, leptin and beta-actin in Sprague Dawley rats under conditions that modulate leptin mRNA abundance in order to study the relationship between leptin and C/EBP-alpha expression patterns. During acute fasting, which decreased the level of leptin and beta-actin mRNA, C/EBP-alpha mRNA expression was unaltered. In leptin-treated and pair-fed animals, C/EBP-alpha mRNA was unaltered compared to ad libitum fed controls, while leptin and beta-actin mRNA expression was again decreased. These results indicate that changes in the level of leptin gene expression are not directly associated with changes in the level of C/EBP-alpha abundance.

Differentiation method-dependent expression of leptin in adipocyte cell lines.

Leptin, the product of the ob gene, is expressed exclusively in adipose tissue. However, adipocyte cell lines, such as 3T3-L1 adipocytes, have generally been reported to express extremely low levels of leptin mRNA. We compared 3T3-L1's to the closely related line 3T3-F442A, and to another murine adipocyte line, TA1. TA1 cells, when differentiated by indomethacin/insulin treatment, express leptin at levels greater than those of 3T3-L1 adipocytes differentiated by the traditional methylisobutylxanthine/dexamethasone/insulin protocol. However, when 3T3-L1's are differentiated in the presence of indomethacin/insulin their expression levels of leptin increase dramatically. 3T3-F442A preadipocytes also express high levels of leptin when differentiated in the presence of T3 and insulin, but when differentiated in the presence of indomethacin/insulin, expression levels drop precipitously. These changes in leptin mRNA and protein expression are not reflected by changes in CCAAT/enhancer binding protein-alpha (c/EBPalpha), peroxizomal proliferator activated receptor-gamma (PPARgamma), lipoprotein lipase (LPL), fatty-acid binding protein aP2 or uncoupling protein-2 (UCP2) mRNA levels, and suggest a mechanism unique to the leptin gene.

Modifications in the B10 and B26-30 regions of the B chain of human insulin alter affinity for the human IGF-I receptor more than for the insulin receptor.

Inversion of the natural sequence of the B chain of human insulin (HI) from ProB28LysB29 to LysB28ProB29 generates an insulin analogue with reduced tendency to self-associate. Since this substitution increases the homology of insulin to insulin-like growth factor-I (IGF-I), we have examined the affinity of a series of insulin analogues with the general modified structure XaaB28ProB29 HI for binding to both human placental insulin and IGF-I receptors. The XaaB28ProB29 HI series is approximately equipotent to HI in binding to the insulin receptor with the exception of when Xaa = Phe, Trp, Leu, Ile, and Gly (40-60% relative to HI). Substitution with basic residues in the B28 position increased the relative affinity to the IGF-I receptor approximately 1.5-2-fold (ArgB28ProB29 > OrnB28ProB29 = LysB28ProB29). Substitution with acidic residues reduced relative affinity for the IGF-I receptor approximately 2-fold (CyaB28ProB29 = GluB28ProB29 > AspB28ProB29). Combination of AspB10 substitution in conjunction with a modification in the B28-29 position (e.g. AspB10LysB28ProB29 HI) showed an additional 2-fold selective increase in affinity for the IGF-I receptor, suggesting that these two effects are additive. Addition of Arg residues at B31-32, on the backbone of either HI or AspB10 HI, increased affinity for the IGF-I receptor 10 and 28 fold, respectively, compared to HI, confirming the significance of enhanced positive charge at the C-terminal end of the insulin B-chain in increasing selectivity for the IGF-I receptor. This relative increase in IGF-I receptor affinity correlated largely, but not completely, with enhanced growth promoting activity in human mammary epithelial cells. In the case of LysB28ProB29 HI, growth activity correlated with dissociation kinetics from the insulin receptor which were shown to be identical with those of human insulin.

Regulation of expression of ob mRNA and protein by glucocorticoids and cAMP.

Regulation of obese gene (ob) expression in ob/ob and db/db mice and in cultured rat adipocytes was examined. It has been demonstrated that exogenous human OB protein (leptin) treatment reduces food intake and weight gain, as well as insulin, glucose, and corticosterone levels in ob/ob mice. In the present report we show that leptin treatment down-regulates endogenous adipose ob mRNA. However, treatment of isolated rat adipocytes with 100 ng/ml human or murine leptin had no direct effect on expression of endogenous ob mRNA, suggesting that leptin may be able to down-regulate its own expression by an indirect, non-autocrine mechanism. Glucocorticoids increased both ob mRNA levels and secreted leptin levels in vitro. Conversely, agents that increase intracellular cAMP, such as beta-adrenergic agonists or Bt2cAMP itself, decreased ob mRNA expression and leptin secretion. Therefore, increased glucocorticoid levels and decreased sympathetic neural activity may contribute to the elevated ob mRNA expression observed in genetically obese, hyperglucocorticoid rodents. Furthermore, leptin might regulate its own expression through a feedback mechanism involving the hypothalamic pituitary axis.

Preparation of an insulin with improved pharmacokinetics relative to human insulin through consideration of structural homology with insulin-like growth factor I.

The Diabetes Control and Complications Trial has emphasized the need for improved control of blood glucose as a means to diminish long-term complications of diabetes. LysPro-insulin is an analog of human insulin whose design was modeled on structural homology with insulin-like growth factor I. An analysis of the structural conformation of insulin suggested that an inversion of amino acids B28 and B29 in the C-terminus of the B chain could yield an insulin analog with a faster onset of biological action. This insulin analog has proved to be virtually identical to human insulin in action, with one important exception. LysPro-insulin has demonstrated an improved time course of action in control of a mealtime glucose elevation. This offers the opportunity for improved convenience and safety for patients with insulin-dependent diabetes mellitus.

Glucose transporter levels in tissues of spontaneously diabetic Zucker fa/fa rat (ZDF/drt) and viable yellow mouse (Avy/a).

We used antibodies to the fat/muscle glucose transporter (GLUT4) and the liver glucose transporter (GLUT2) to measure levels of these proteins in various tissues of two rodent models of non-insulin-dependent (type II) diabetes mellitus: the obese spontaneously diabetic male Zucker fa/fa rat (ZDF/drt) and the male viable yellow Avy/a obese diabetic mouse. The ZDF/drt strain generally develops overt diabetes associated with decreased plasma insulin levels. Depending on the age of the animals, the ZDF/drt rats can be arbitrarily segregated into age-matched obese, mildly diabetic (blood glucose less than 11 mM) and obese, and severely diabetic (blood glucose greater than 20 mM) groups. Avy/a mice are comparably hyperglycemic but unlike the ZDF/drt rats are severely hyperinsulinemic. In both groups of diabetic animals, GLUT4 in adipose tissue, heart, and skeletal muscle was reduced 25-55%, and GLUT2 in liver was increased 30-40%, relative to lean, age-matched controls. However, when the mildly diabetic ZDF/drt rats were compared to the lean controls, the only significant difference was a 25% reduction of GLUT4 in heart. Within all of the ZDF/drt rats (excluding the lean controls), GLUT2 in liver and GLUT4 in adipose tissue, heart, and skeletal muscle correlated significantly with glycemia. These data suggest that, in these two models of type II diabetes, glucose transporter levels in muscle, adipose tissue, and liver are regulated in a tissue-selective manner in response to changes in insulin and glucose. Furthermore, at least in the ZDF/drt rat, alterations in GLUT2 and/or GLUT4 protein levels appear not to be associated with obesity per se but appear to be secondary to the severely diabetic state.(ABSTRACT TRUNCATED AT 250 WORDS)

(A-C-B) human proinsulin, a novel insulin agonist and intermediate in the synthesis of biosynthetic human insulin.

The hormone insulin is synthesized in the beta cell of the pancreas as the precursor, proinsulin, where the carboxyl terminus of the B-chain is connected to the amino terminus of the A-chain by a connecting or C-peptide. Proinsulin is a weak insulin agonist that possesses a longer in vivo half-life than does insulin. A form of proinsulin clipped at the Arg65-Gly66 bond has been shown to be more potent than the parent molecule with protracted in vivo activity, presumably as a result of freeing the amino terminal residue of the A-chain. To generate a more active proinsulin-like molecule, we have constructed an "inverted" proinsulin molecule where the carboxyl terminus of the A-chain is connected to the amino terminus of the B-chain by the C-peptide, leaving the critical Gly1 residue free. Transformation of Escherichia coli with a plasmid coding for A-C-B human proinsulin led to the stable production of the protein. By a process of cell disruption, sulfitolysis, anion-exchange chromatography, refolding, and reversed-phase high-performance liquid chromatography, two forms of the inverted proinsulin differing at their amino termini as Gly1 and Met0-Gly1 were identified and purified to homogeneity. Both proteins were shown by a number of analytical techniques to be of the inverted sequence, with insulin-like disulfide bonding. Biological analyses by in vitro techniques revealed A-C-B human proinsulin to be intermediate in potency when compared to human insulin and proinsulin. The time to maximal lowering of blood glucose in the fasted normal rat appeared comparable to that of proinsulin. Additionally, we were able to generate fully active, native insulin from A-C-B human proinsulin by proteolytic transformation. The results of this study lend themselves to the generation of novel insulin-like peptides while providing a simplified route to the biosynthetic production of insulin.

Effect of streptozocin-induced diabetes on insulin-receptor tyrosine kinase activity in obese Zucker rats.

We examined insulin binding, insulin-stimulated autophosphorylation, and phosphorylation of poly(Glu.Na,Tyr)4:1 by liver and skeletal muscle insulin receptor from lean, obese, and obese streptozocin-induced diabetic Zucker rats. Induction of diabetes with streptozocin (30 mg/kg) lowered the lasting insulin level from 11.4 to 3.8 ng/ml, which was not significantly greater than the lean control level. Autophosphorylation and tyrosine kinase activity of liver insulin receptors were increased 70-100% in the obese control group (relative to lean rats), but diabetes reversed this hyperresponsiveness to insulin. In muscle, obesity was associated with a 40-50% decrease in autophosphorylation and tyrosine kinase activity, which was also reversed in the diabetic state. Autophosphorylation and tyrosine kinase activity were significantly correlated in liver and muscle and were also correlated with fasting insulin levels. These data suggest that insulin-receptor tyrosine kinase activity is regulated differently in liver and muscle and that the abnormalities in kinase activity associated with the obese Zucker rat are at least partly secondary to hyperinsulinemia.

Biosynthesis of the epidermal growth factor receptor: post-translational glycosylation-independent acquisition of tyrosine kinase autophosphorylation activity.

We previously showed that the epidermal growth factor (EGF) receptor in human A431 epidermoid carcinoma cells undergoes a slow post-translational modification whereby it acquires (t1/2 = 30-40 min) EGF binding capacity (Slieker, L.J., et. al. (1986) J. Biol. Chem., 261, 15233-15241). This activation occurs in the endoplasmic reticulum and requires core N-linked glycosylation. By employing both anti-EGF receptor and anti-phosphotyrosine antibodies to immunoprecipitate receptor pulse-labeled with [35S]methionine, we demonstrate here that the EGF receptor also acquires tyrosine kinase autophosphorylation activity post-translationally (t1/2 = 10-15 min). The acquisition of tyrosine kinase activity is independent of the acquisition of EGF binding capacity, since it precedes the latter process and does not require N-linked glycosylation.

Post-translational acquisition of ligand binding- and tyrosine kinase-domain function by the epidermal growth factor and insulin receptors.

The epidermal growth factor receptor (EGFR) and insulin receptor undergo slow post-translational modification by which they acquire hormone binding and tyrosine kinase (EGFR) function. The half-time for acquisition of EGF or insulin binding activity is 30-40 min and of tyrosine kinase activity (EGFR), is 10-15 min. Tunicamycin, an inhibitor of N-linked oligosaccharide addition, blocks acquisition of both EGF and insulin binding activity. With EGFR, activation precedes acquisition of resistance to endoglucosaminidase H (t1/2 approximately equal to 75 min), a medial Golgi event. Treatment of active high mannose receptor with endo H generates fully active aglyco-receptor; thus, core oligosaccharide addition is a prerequisite for activation, but not for EGF binding per se. EGFR is activated in and translocated from the endoplasmic reticulum (ER) slowly (t1/2 approximately equal to 75 min). Since translocation rate equals the rate for acquisition of endo H resistance, transit from the ER is rate limiting for EGFR maturation. Tunicamycin inhibits exit from the ER parallel to its effect on acquisition of binding activity. Insulin proreceptor, a 210 kDa high-mannose glycopolypeptide, acquires insulin binding function (t1/2 approximately equal to 45 min) then is proteolytically cleaved (t1/2 approximately equal to 3 hr) into subunits of the mature alpha 2 beta 2 receptor. Modification giving rise to insulin binding activity is due to a conformational change in the binding domain, since human autoimmune antibody recognizes only the active species, while rabbit polyclonal antibody recognizes all forms. Newly-translated EGF proreceptor lacks a functional tyrosine domain capable of autophosphorylation; 30-40 min after translation, while still in the ER, tyrosine kinase activity is acquired. Since the kinase domain is cytoplasmic, the receptor may become phosphorylated on tyrosine before reaching the plasma membrane.

Synthesis of epidermal growth factor receptor in human A431 cells. Glycosylation-dependent acquisition of ligand binding activity occurs post-translationally in the endoplasmic reticulum.

It was previously demonstrated that the epidermal growth factor (EGF) receptor in human A431 cells undergoes a slow post-translational modification by which it acquires EGF binding capacity (Slieker, L.J., and Lane, M.D. (1985) J. Biol. Chem. 260, 687-690). In this report, the role of glycosylation in the acquisition of ligand binding activity and in the intracellular translocation of the receptor precursor is characterized. Human A431 cells were incubated with [35S]methionine, and 35S-labeled EGF receptors were purified either by immunoprecipitation (total receptor) or by adsorption to an EGF affinity matrix (high affinity, or active receptor). The half-time for receptor activation is approximately 30 min and precedes its acquisition of resistance to endo-beta-N-acetylglucosaminidase H (t 1/2 = 75 min), a medial Golgi event. Activation is blocked by tunicamycin and is markedly slowed (t 1/2 = 120 min) by 1-deoxynojirimycin, an inhibitor of glucosidase I. In the latter case, the oligosaccharide chains are not further processed to complex forms. Treatment of the active high mannose receptor with endo-beta-N-acetylglucosaminidase H generates a fully active aglycoreceptor polypeptide, indicating that core oligosaccharide addition is a prerequisite for activation but that oligosaccharide chains are not intrinsically required for EGF binding. Subcellular fractionation studies showed that the EGF receptor is activated in the endoplasmic reticulum and that translocation from that organelle is extremely slow (t 1/2 = 75 min). Since the latter translocation rate approximates that for the acquisition of the resistance to endoglycosidase H, transit from the endoplasmic reticulum appears to be rate-limiting for the maturation of the receptor. Both tunicamycin and 1-deoxynojirimycin inhibit exit from the endoplasmic reticulum in parallel with their effects on the acquisition of binding activity. Immunoprecipitation of 35S-labeled EGF receptor with antiphosphotyrosine antibody in the presence of ATP suggested that the autophosphorylation activity of the receptor is also acquired post-translationally. The possible correlation of this to EGF binding activity is discussed.