Purification and characterization of the human brain insulin receptor

The insulin receptor from human brain cortex was purified by a combination monoclonal antibody affinity column and a wheat germ agglutinin column. This purified receptor preparation exhibited major protein bands of apparent Mr = 135,000 and 95,000, molecular weights comparable to those for the alpha and beta subunits of the purified human placental and rat liver receptors. A minor protein band of apparent Mr = 120,000 was also observed in the brain receptor preparation. Crosslinking of 125I-insulin to all three receptor preparations was found to preferentially label a protein of apparent Mr = 135,000. In contrast, cross-linking of 125I-labeled insulin-like growth factor I to the brain preparation preferentially labeled the protein of apparent Mr = 120,000. The purified brain insulin receptor was found to be identical with the placental insulin receptor in the amount of neuraminidase-sensitive sialic acid and reaction with three monoclonal antibodies to the beta subunit of the placental receptor. In contrast, a monoclonal antibody to the insulin binding site recognized the placental receptor approximately 300 times better than the brain receptor. These results indicate that the brain insulin receptor differs from the receptor in other tissues and suggests that this difference is not simply due to the amount of sialic acid on the receptor.     

Structural differences between insulin receptors in the brain and peripheral target tissues

Insulin receptors in various brain regions (olfactory tubercle, hippocampus, and hypothalamus) were photoaffinity labeled using the photoreactive analogue of insulin B2(2-nitro,4-azidophenylacetyl)-des-PheB1-insulin (NAPA-DP-insulin). A protein with an apparent Mr of 400,000 was specifically labeled with 125I-NAPA-DP-insulin in all three brain regions. When radiolabeled proteins were reduced with dithiothreitol prior to electrophoresis, specific labeling occurred predominantly in a protein with an apparent Mr of 115,000 and to a much lesser extent in a protein with an apparent Mr of 83,000. The size of these receptor proteins, based on their electrophoretic mobilities, was consistently smaller than insulin receptor proteins in adipocytes. The covalent labeling of insulin receptors in brain by 125I-NAPA-DP-insulin was not blocked by anti-insulin receptor antiserum. Additionally, in contrast to effects observed in peripheral target tissues, this antisera did not inhibit the binding of 125I-insulin to brain membranes. Neuraminidase treatment resulted in an increase in the electrophoretic mobilities of insulin receptor subunits in adipocytes, but, had no effect on receptor subunits in brain. Solubilized insulin receptors from adipocytes were retained by wheat germ agglutinin columns and specifically eluted with N-acetylglucosamine. In contrast, solubilized insulin receptors from brain did not bind to these columns. The results from this study indicate that structural differences, including molecular weight, antigenicity, and carbohydrate composition exist between insulin receptors in brain and peripheral target tissues.     

Evidence for a subtype of insulin-like growth factor I receptor in brain

We examined the structure of receptors for insulin-like growth factor I (IGF-I), insulin, and epidermal growth factor (EGF) in human brain and human placenta using affinity cross-linking procedures and sodium dodecyl sulfate-polyacrylamide gel electrophoresis. In human brain, proteins specifically cross-linked to 125I-IGF-I, 125I-insulin, and 125I-EGF had apparent molecular weights of 120,000, 115,000 and 170,000, respectively. In human placenta, proteins cross-linked to 125I-IGF-I and 125I-insulin were 10 kDa larger than the corresponding subunits in brain. The receptor labeled by 125I-EGF in placenta was indistinguishable from the EGF receptor in brain. The size discrepancy of IGF-I receptors in brain and placenta was no longer apparent after removing the carbohydrate moieties of the proteins with endo-beta-N-acetylglucosaminidase F (EndoF). Furthermore, the brain IGF-I receptor was not cleaved by neuraminidase, whereas, the placental IGF-I receptor had increased mobility on SDS gels following neuraminidase treatment. The results indicate that receptors for IGF-I and insulin in human brain are structurally distinct from the corresponding receptors in human placenta, the structural heterogeneity of the receptors involves differences in N-linked glycosylation, particularly the terminal processing steps, and EGF receptors are present in human brain and human placenta but are structurally similar in these tissues. We conclude that there is a selective modification in the glycosylation of receptors for IGF-I and insulin in brain.     

Fasting and refeeding alter the insulin receptor tyrosine kinase in chicken liver but fail to affect brain insulin receptors

Insulin receptors from chicken liver and brain were studied following alterations in the nutritional state. Chickens were either fasted for 48 h, fasted for 48 h and then refed for 24 h, or fed a regular diet ad libitum. 125I-Porcine insulin binding was significantly elevated in liver membranes from the fasted animals and lowered in refed chickens when compared to preparations from ad libitum fed chickens. These changes in 125I-insulin binding were inversely related to the levels of plasma insulin and since receptor affinities for insulin were similar in each group, they probably represent alterations in receptor number. Apparent Mr of alpha subunits of the insulin receptors was unaffected by alterations in the nutritional states. The presence of ATPase-like activities that co-eluted with liver insulin receptors from wheat germ agglutinin lectin columns but not from pea lectin columns necessitated the use of both pea and wheat germ agglutinin for liver insulin receptor purification. The insulin receptors purified from both lectin columns were recognized by anti-insulin receptor antiserum and had similar affinities for insulin which were unaltered by the nutritional state. Insulin-stimulatable autophosphorylation of the beta subunit of the insulin receptor was lower in livers from fasted chickens and intermediate in refed chickens. Furthermore, basal and insulin-induced phosphorylation of the artificial substrate poly(Glu,Tyr) 4:1 was significantly less in the fasting state and intermediate in the refed state compared to the ad libitum fed state. Insulin sensitivity (measured as the dose of insulin required for 50% maximal stimulation of kinase activity) was similar in all three states suggesting that the differences in insulin-induced phosphorylation are due to a change in maximal stimulation and not a change in insulin sensitivity. In contrast to the alterations seen with liver receptors, brain insulin receptors were unaffected by these alterations in nutritional state. These findings suggest that: liver insulin receptors are affected by altering the nutritional state; insulin binding to liver membranes is inversely related to plasma insulin levels; and tyrosine kinase is decreased both in fasted and refed animals suggesting an uncoupling of the normal interaction between alpha subunit and beta subunit in liver insulin receptors.     

Insulin receptors in the mammalian central nervous system: binding characteristics and subunit structure

Insulin receptors in rat and human central nervous system have been identified by binding of 125I-insulin on purified synaptic plasma membranes; affinity labelling of receptors by chemical cross-linking 125I-insulin; or phosphorylation of receptors with [gamma-32P]ATP. Brain insulin receptors showed significant differences in their binding characteristics and subunit structure when compared with receptors in other tissues like adipose and liver cells: absence of negatively cooperative interactions; a distinct binding specificity i.e. porcine proinsulin, coypu insulin and insulin-like growth factor I and II showed 2-5 times higher binding affinity in brain than in other cell types; a smaller molecular size of the brain receptor alpha-subunit than in other tissues (Mr approximately 115,000 instead of 130,000). In contrast, the size (Mr approximately 94,000) and function of the insulin receptor beta-subunit kinase was identical with that described in other cells. We conclude, that insulin receptors in mammalian brain represent a receptor subtype which may mediate growth rather than metabolic activity of insulin.     

Purification and immunochemical interrelationship of insulin receptors from rat brain and liver.

Insulin receptors from rat brain and liver were purified. Brain purified receptor exhibited protein bands of apparent Mr = 135,000 and 95,000 molecular weight corresponding to alpha- and beta-subunits, retained a tyrosine specific protein kinase activity and demonstrated phosphorylation that is hormonally sensitive. Antisera were raised against both insulin receptor preparations and enzyme-linked immunosorbent assay was developed. The comparison of two insulin receptors was based on a displacement enzyme-linked immunosorbent assay where antisera were interchanged on predetermined optimal dilutions. This indicated that both insulin receptors possess some unique antigenic determinants thereby implying a structural difference.     

Cell-surface changes during mitogenic stimulation of lymphocytes assessed by the binding of wheat-germ agglutinin and other plant lectins

Lymphocytes from murine lymph node, cultured in the presence of an optimally mitogenic dose of phytohaemagglutinin, were stained with fluoresceinated lectins and analysed by flow cytometry. A marked increase in the ability of lymphocytes to bind wheat-germ agglutinin was observed that is particularly pronounced for the blast cells, reaching a maximum at about 40 h, when they are 5.5-times brighter than cells at zero time. The corresponding intensification of the small cells is 2-fold. Much smaller increases in binding accompanying blast transformation were observed when fluoresceinated concanavalin A or Lens culinaris haemagglutinin were used. Polyacrylamide gel electrophoresis of plasma membranes followed by treatment of the gels with radioactively labelled lectins and autoradiography also showed a very distinct increase in the binding of wheat-germ agglutinin to membranes from mitogen-stimulated porcine lymphocytes. Less marked changes in the binding of concanavalin A Lens culinaris heamagglutin and Ricinus communis agglutinin 120 were also noted. The apparent multiplicity of glycoproteins that bind each lectin, suggests that in each case the sites are heterogeneous. We conclude that lymphocytes stimulated by the T-cell mitogen phytohaemagglutinin expose new glycoprotein receptors for wheat-germ agglutinin that are most abundant on blast cells at 40 h. Attempts to characterize the receptor biochemically suggest that the carbohydrate moiety recognised by wheat-germ agglutinin is present on a glycoprotein of approx. 120 kDa molecular mass and also possibly on glycoproteins of 170–190 kDa.     

Distance between the propylbenzilylcholine mustard attachment site and carbohydrates and thiol groups in muscarinic acetylcholine receptor protein from rat cerebral cortex.

When rat cerebral-cortex membranes were labelled with [3H]propylbenzilylcholine mustard ([3H]PrBCM), a single protein of Mr 68,000 was found to carry the atropine-sensitive covalent label. After trypsinolysis of the receptors solubilized in 0.075% SDS, the resulting fragments were submitted to size analysis in combination with wheat-germ agglutinin (WGA)-Sepharose and organomercurial-agarose chromatography. Peptides of Mr 75,000, 50,000, 30,000, 18,000 and 8000 were specifically released from the receptor. All fragments above Mr 8000 were able to bind WGA-Sepharose and therefore the peptide of Mr 18,000 was taken as the upper limit of the distance between the antagonist and the glycan moieties. The limit fragment of Mr 8000 carried chemical groups which were modified by N-ethylmaleimide and reacted with an immobilized organomercurial. About 65-80% of the labelled receptors were adsorbed on concanavalin A-Sepharose with low affinity, generating two further components after sequential application to WGA-Sepharose. About 50% of the receptors were susceptible to neuraminidase treatment, with a concomitant slight modification of the SDS/polyacrylamide-gel-electrophoretic pattern.     

Variation in insulin receptor messenger ribonucleic acid expression in human and rodent tissues

The expression of insulin receptor mRNA was studied in human and rodent tissues by Northern analysis. Human EBV-transformed lymphocytes contained four receptor mRNA species of sufficient length to encode the entire proreceptor: 9.5, 7.9, 7.1, and 5.7 kb. In human fibroblasts, the same four species were observed; however, the 7.9 and 5.7 kb mRNAs were markedly decreased. In mouse liver, rat hepatoma cells, and normal rat brain, kidney, liver, and muscle only two mRNA species (7.4 and 9.6 kb) were detected. Each of these human and rodent mRNAs hybridized equally well with cDNA sequences encoding the binding and kinase domains of the insulin receptor. Several smaller polyadenylated mRNAs (approximately 1.8 to 3.3 kb) were also identified in human cell lines that appeared to separately encode either alpha- or beta-subunit sequences of the receptor. In rats, liver had the highest content of insulin receptor mRNA, followed by kidney, brain, and muscle. The relative amount of the two mRNA species also varied among the rat tissues. The ratio of the 9.6-7.4 kb species was 2.7 in brain but only 1.0 to 1.6 in the other tissues (P less than 0.025). Dexamethasone treatment increased the content of the two insulin receptor mRNAs in rat liver by 2-fold. The half-life of both mRNA species was 70 min in rat hepatoma cells. These findings indicate that insulin receptor gene expression is complex and regulated with differential expression of insulin receptor mRNA and/or alterations in mRNA processing among various tissues.     

Post-translational modification of the protein-synthesis initiation factor eIF-4D by spermidine in rat hepatoma cells.

The structural characteristics and glycoprotein nature of the human growth hormone (hGH) receptor in cultured lymphocytes (IM-9 cell line) were studied with the use of a bifunctional reagent (disuccinimidyl suberate) to couple 125I-hGH covalently to intact cells. After cross-linking, the hormone-receptor complexes were analysed by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. A single band of Mr 140,000 was identified under reducing conditions. The labelling of this band was blocked by unlabelled hGH but not by insulin, ovine prolactin, bovine or ovine growth hormones. The Mr 140,000 band was immunoprecipitated by either anti-hGH antibody or by a monoclonal antibody against rat liver growth hormone receptor. In the absence of reductant two major bands of Mr 270,000 and 140,000 were found. On two-dimensional gel electrophoresis, with the first dimension in the absence of reductant and the second in its presence, the Mr 270,000 complex generated the Mr 140,000 band. The nature of the oligosaccharide chains of the receptor was studied by treatment with different glycosidases. The electrophoretic mobility of the Mr 140,000 receptor complex was markedly increased after digestion with endoglycosidase F but showed no or little change after digestion with endoglycosidase H. The Mr 140,000 band was also sensitive to neuraminidase treatment. In addition the 125I-hGH-receptor complex was adsorbed by immobilized wheat germ agglutinin and to a smaller extent by immobilized concanavalin A, lentil lectin, ricin I and ricin II. In conclusion, taking into account that hGH is a Mr 22,000 polypeptide, the binding subunit of the GH receptor in human IM-9 lymphocytes has an Mr of approx. 120,000. The native receptor may exist as a homodimer of the binding subunit formed by disulphide bonds. Furthermore, the GH receptor subunit contains asparagine N-linked type of oligosaccharide chains. Most, if not all, of these chains are of the complex type and appear to be sialylated whereas no high-mannose type chains are detectable in the mature form of the receptor.     

Peanut Agglutinin Receptor Is a Marker of Myelin in Rat Brain. Developmental Changes in Its Distribution

Peanut agglutinin receptor, which is a glycoconjugate with terminal Gal(beta 1-3)GalNAc residues, was found to be a specific biochemical marker of the myelin in rat brain. The localization of peanut agglutinin receptor in rat brain was studied by use of fluorescein-conjugated peanut agglutinin. In adult rat brain, receptors were found only on myelin. In contrast, the receptors in fetal brain were not found in the white matter, but only in the ependyma. Peanut agglutinin receptors were extracted from adult and fetal rat brains with 2% Triton X-100, subjected to electrophoresis on 10% sodium dodecyl sulfate-polyacrylamide gel, blotted onto nitrocellulose filters, and detected by their reaction with peroxidase-conjugated peanut agglutinin. A major receptor of Mr 130,000 and a minor one of Mr 200,000 were found in both adult and fetal rat brains.     

Peanut agglutinin, a new mitogen that binds to galactosyl sites exposed after neuraminidase treatment.

Peanut agglutinin, purified by affinity chromatography, agglutinates lymphocytes from mouse, rat, guinea pig, and man only after their treatment with neuraminidase. However, it stimulates only neuraminidase-treated rat and human cells. A similar number cell surface receptors for peanut agglutinin was found on neuraminidase-treated rat and mouse lymphocytes although the latter cells were not stimulated by the lectin. Galactose specifically inhibited the agglutination and stimulation of lymphocytes by peanut agglutinin. Sequential treatment of lymphocytes with neuraminidase and beta-galactosidase markedly reduced the response of the cells to stimulation by peanut agglutinin, soybean agglutinin, and galactose oxidase. It is suggested that the same galactosyl residue may be the target for the initial step in triggering lymphocytes by the above mentioned mitogens.     

Deglycosylation and proteolysis of photolabeled D2 dopamine receptors of the porcine anterior pituitary

Dopamine D2 receptor binding subunits of the porcine anterior pituitary were visualized by autoradiography following photoaffinity labeling with [125I]N-azidophenethylspiperone and sodium dodecyl-sulfate polyacrylamide gel electrophoresis. The ligand binding subunit comprising the pituitary D2 dopamine receptor migrated as two distinct bands of apparent Mr approximately equal to 150,000 and 118,000, substantially higher than neuronal D2 receptor subunits from porcine or canine brain. The glycoprotein nature of pituitary D2 receptor binding subunits was investigated by the use of exo- and endo-glycosidase treatments and peptide mapping experiments. Photoaffinity labeled polypeptides of the anterior pituitary were susceptible to both neuraminidase and alpha-mannosidase digestion as indexed by their increased electrophoretic mobility on sodium dodecyl-sulfate polyacrylamide gels, and suggests the presence of both complex type and terminal mannose carbohydrate residues. Moreover, the additive effects of sequential treatment with these enzymes suggests that both types of carbohydrate chains are present on each receptor peptide. N-linked deglycosylation of pituitary D2 photolabeled receptors with glycopeptidase-F produced a further increase in the mobility of the labeled protein to apparent Mr approximately equal to 44,000, similar to that of deglycosylated D2 binding subunits of porcine and canine brain. Peptide mapping experiments following limited proteolysis with Staphylococcus aureus V8 proteinase and papain demonstrated that deglycosylated D2 dopamine receptors (Mr = 44,000), in different tissues and species, were homologous. Taken together, these data suggest that despite the differences in the overall molecular weight and tissue specific glycosylation pattern of pituitary D2 dopamine receptors, the primary structure of mammalian D2 receptors appears to be conserved.     

Identification and characterization of glucose transport proteins in plasma membrane- and Golgi vesicle-enriched fractions prepared from lactating rat mammary gland.

Previous studies have indicated that turkey erythrocyte and rat liver membranes contain endogenous alpha beta heterodimeric insulin receptors in addition to the disulphide-linked alpha 2 beta 2 heterotetrameric complexes characteristic of most cell types. We utilized 125I-insulin affinity cross-linking to examine the structural properties of insulin receptors from rat liver and turkey erythrocyte membranes prepared in the absence and presence of sulphydryl alkylating agents. Rat liver membranes prepared in the absence of sulphydryl alkylating agents displayed specific labelling of Mr 400,000 and 200,000 bands, corresponding to the alpha 2 beta 2 heterotetrameric and alpha beta heterodimeric insulin receptor complexes respectively. In contrast, affinity cross-linking of membranes prepared with iodoacetamide (IAN) or N-ethylmaleimide identified predominantly the alpha 2 beta 2 heterotetrameric insulin receptor complex. Similarly, affinity cross-linking and solubilization of intact turkey erythrocytes in the presence of IAN resulted in exclusive labelling of the alpha 2 beta 2 heterotetrameric insulin receptor complex, whereas in the absence of IAN both alpha 2 beta 2 and alpha beta species were observed. Turkey erythrocyte alpha 2 beta 2 heterotetrameric insulin receptors from IAN-protected membranes displayed a 3-4-fold stimulation of beta subunit autophosphorylation and substrate phosphorylation by insulin, equivalent to that observed in intact human placenta insulin receptors. Turkey erythrocyte alpha beta heterodimeric insulin receptors, prepared by defined pH/dithiothreitol treatment of IAN-protected membranes, were also fully competent in insulin-stimulated protein kinase activity compared with alpha beta heterodimeric human placenta receptors. In contrast, endogenous turkey erythrocyte alpha beta heterodimeric insulin receptors displayed basal protein kinase activity which was insulin-insensitive. These data indicate that native turkey erythrocyte and rat liver insulin receptors are structurally and functionally similar to alpha 2 beta 2 heterotetrameric human placenta insulin receptors. The alpha beta heterodimeric insulin receptors previously identified in these tissues most likely resulted from disulphide bond reduction and denaturation of the alpha 2 beta 2 holoreceptor complexes during membrane preparation.     

Demonstration and structural comparison of receptors for insulin-like growth factor-I and -II (IGF-I and -II) in brain and blood-brain barrier

Specific receptors for insulin-like growth factors (IGF) I and II on microvessel-free rat brain cell membranes (RBCM) and in the microvessels that constitute the blood-brain barrier (BBB) were identified and characterized by means of affinity cross-linking techniques and specific anti-receptor antibodies. Two different models of BBB were examined: isolated rat brain capillaries and cultured bovine brain microvessel endothelial cells. Cross-linking with 125-I-IGF-I, followed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), revealed an alpha subunit of apparent Mr = 138,000 in both BBB preparations, compared to 120,000 in RBCM. Cross-linking was inhibited by unlabeled IGF and insulin, but not by antibody directed against the IGF-II receptor. When 125-I-IGF-II was cross-linked, followed by SDS-PAGE under reducing conditions, a major band of apparent Mr = 250,000 was identified in RBCM and both BBB preparations. This band, which migrated with an approximately equivalent Mr in both brain and BBB membranes, was inhibited by unlabeled IGF and by antibody specific for the IGF-II receptor. Thus, both rat and bovine brain microvessels possess classical Type I and II IGF receptors. While the alpha subunit of the Type I receptor of brain is smaller than that of the BBB, the Type II receptor of brain and BBB appear to be structurally and immunologically identical.     

Photoaffinity cross-linked A1 adenosine receptor-binding subunits. Homologous glycoprotein expression by different tissues

Mammalian A1 adenosine receptor-binding peptides can be visualized by covalently labeling them with the photoaffinity cross-linking ligand N6-2-(4-amino-3-[125I] iodophenyl)ethyladenosine followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis/autoradiography. The proteins comprising the A1 adenosine receptor-binding subunit of rat brain and fat migrate with Mr 38,000. In this study, the glycoproteins representing the radiolabeled A1 adenosine receptor-binding subunit expressed in each of these tissues (brain and fat) were compared through the use of peptide mapping and exo- and endoglycosidase treatments. Peptide mapping studies with several enzymes demonstrate that the protein component of the radiolabeled A1 adenosine receptor-binding subunit is conserved between different tissues. Both labeled receptor peptides demonstrate a sensitivity to neuraminidase as evidenced by increased mobility on sodium dodecyl sulfate-polyacrylamide gel electrophoresis suggesting that the receptors contain complex-type carbohydrate chains. Insensitivity to alpha-mannosidase suggests a lack of high mannose-type carbohydrate chains. Deglycosylation of the labeled receptor-binding subunits with endoglycosidase F results in a single labeled polypeptide of Mr 32,000 for both systems. These data suggest that the A1 adenosine receptor-binding subunits expressed in the rat brain and fat are similar glycoproteins as evidenced by similar overall molecular weights, identical peptide maps, and equivalent responses to endo- and exoglycosidase treatment.     

Receptors for insulin-like growth factors in the central nervous system: structure and function

Insulin-like growth factors (IGFs) I and II are homologous peptides, which stimulate growth of several vertebrate tissues. Expression of IGF I and IGF II genes and production of IGFs have recently been demonstrated in rat and human brain. In search for the function of IGF I and IGF II in the central nervous system, we have studied IGF receptors in fetal and adult mammalian brain and growth effects of IGFs on primary cultures of fetal rat astrocytes. Two types of IGF receptor are present on adult rat brain cortical plasma membranes, on fetal rat astrocytes and on human glioma cells. Type I IGF receptor is composed of 2 types of subunits: alpha-subunits which bind IGF I and IGF II with high affinity and insulin weakly, and beta-subunits which show tyrosine kinase activity and autophosphorylation stimulated by IGF I and IGF II with almost similar potency. The molecular size of the type I IGF receptor alpha-subunit is larger in cultured fetal rat astrocytes and human glioma cells than in normal adult brain (Mr 130,000 versus 115,000), whereas the beta-subunit has the same electrophoretic mobility (Mr 94,000). The type II IGF receptor is a monomeric protein (Mr 250,000), which binds IGF II 5 times better than IGF I, and does not recognize insulin. The amounts of type II IGF receptor are significantly higher in fetal and malignant cells than in adult brain. Based on these findings we suggest that IGF receptors in brain undergo changes during fetal development and malignant transformation.(ABSTRACT TRUNCATED AT 250 WORDS)     

Structural differences between liver- and muscle-derived insulin receptors in rats

The structure of insulin receptors, solubilized from rat skeletal muscle and liver, was studied. The alpha subunit was identified by specific cross-linking to A14 125I-insulin with disuccinimidyl suberate. Muscle- and liver-derived alpha subunits migrated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with a Mr of 131,000 and 135,000, respectively. There was no significant difference in insulin binding affinity. Treatment of cross-linked, immunoprecipitated receptors with either neuraminidase or endoglycosidase H decreased the Mr of muscle- and liver-derived alpha subunits but did not affect the difference in Mr. Autophosphorylated beta subunits migrated with a Mr of 98,000 for muscle and 101,000 for liver. After partial V8 digestion of autophosphorylated, immunoprecipitated receptors the major phosphopeptide fragment migrated on SDS-PAGE at Mr 57,000 from muscle and 60,000 from liver. Glycosidase digestion of autophosphorylated receptors suggested that Mr heterogeneity was due in part to differences in the sialic acid content of beta subunits. Muscle and liver are the major target organs of insulin; the apparent heterogeneity of insulin receptor structure may be relevant to tissue-specific differences in insulin action.     

Structural analysis and functional role of the carbohydrate component of somatostatin receptors.

SRIF receptors are membrane-bound glycoproteins. To structurally identify the carbohydrate components of SRIF receptors, solubilized rat brain SRIF receptors were subjected to lectin affinity chromatography. Solubilized SRIF receptors specifically bound to wheat germ agglutinin-lectin affinity columns but not to succinylated wheat germ agglutinin. This finding, as well as the ability of the solubilized receptor to interact with a Sambucus nigra L. lectin affinity column suggested that sialic acid residues are associated with SRIF receptors. The inability of the receptor to bind to concanavalin A, Dolichus biflorus agglutinin, Ulex europeaus I, and Jacalin lectin affinity columns suggests that high mannose, N-acetylgalactosamine, fucose, and O-linked carbohydrates are not associated with receptor. To investigate the functional role of the carbohydrate groups in brain SRIF receptors, specific sugars were selectively cleaved from SRIF receptors and the subsequent effect on the specific high affinity binding of the agonist [125I]MK 678 to SRIF receptors was determined. Treatment of the receptor with endoglycosidase D did not affect the specific binding of [125I] MK 678 to the solubilized SRIF receptors, consistent with the finding from lectin affinity chromatography that high mannose-type carbohydrate structures were not associated with SRIF receptors. Treatment of solubilized SRIF receptors with peptide-N-glycosidase F and endoglycosidases H and F reduced [125I]MK 678 binding to SRIF receptors indicating that either hybrid, or a combination of hybrid and complex N-linked carbohydrate structures, have a role in maintaining the receptor in a high affinity state for agonists. Treatment of solubilized SRIF receptors with neuraminidase from Vibrio cholera abolished high affinity agonist binding to the receptors, whereas treatment of the receptor with neuraminidase from Newcastle disease virus did not affect [125I]MK 678 binding to the receptor. These findings suggest that sialic acid residues in an alpha 2,6-configuration have a role in maintaining the SRIF receptor in a high affinity conformation for agonists. This is further indicated by studies on SRIF receptors in the pituitary tumor cell line, AtT-20. Treatment of AtT-20 cells in culture with neuraminidase (V. cholera) greatly reduces high affinity [125I] MK 678 binding sites, but did not alter the maximal ability of SRIF to inhibit forskolin-stimulated cAMP accumulation in intact AtT-20 cells. This finding suggests that the desialylated SRIF receptor is functionally active and remains coupled to GTP-binding proteins, but exhibits a reduced affinity for agonists. Treatment of AtT-20 cell membranes with neuraminidase from V. cholera was also able to greatly reduce the affinity of SRIF receptors for [125I]MK 678.(ABSTRACT TRUNCATED AT 400 WORDS)     

Demonstration of two subtypes of insulin-like growth factor receptors by affinity cross-linking

The structure of receptors for insulin-like growth factors in rat liver plasma membranes and the BRL 3A2 rat liver cell line has been examined by chemical cross-linking with disuccinimidyl suberate and sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing and nonreducing conditions. Two receptor subtypes have been identified: (i) 125I multiplication-stimulating activity cross-linked to liver membranes or intact cells appeared in a complex of Mr = 260,000 (reduced) and 220,000 (nonreduced) and (ii) 125I-insulin-like growth factor I cross-linked to BRL 3A2 cells appeared predominantly in two bands of Mr greater than 300,000 without disulfide reduction and in a Mr = 130,000 complex following reduction. The two subtypes of insulin-like growth factor receptors identified by structural analysis correspond to previously observed differences in their specificity for insulin and insulin-like growth factors.