Receptor-mediated degradation of insulin in isolated rat adipocytes. Formation of a degradation product slightly smaller than insulin

More than 90% of the radioactivity associated with isolated rat adipocytes incubated with [TyrA14-125I]monoiodoinsulin represented at steady state iodoinsulin possessing full binding affinity. In contrast, about half of the radioactivity dissociating from the cells was [125I]monoiodotyrosine. The other half was of a molecular size similar to that of iodoinsulin as judged from gel-filtration chromatography. However, the descending limb of the 'insulin' peak (i.e., the smaller molecules) possessed a reduced binding activity compared with native iodoinsulin, material from the ascending limb, or a similar fraction isolated from dissociation medium from IM-9 lymphocytes, a cell type devoid of receptor-mediated insulin degradation. The cells, thus, release an intermediary degradation product.     

Receptor binding properties and insulin-like effects of human growth hormone and its 20 kDa-variant in rat adipocytes

The natural 20 kDa-variant of human growth hormone (hGH) binds with high affinity to IM-9 human lymphocyte receptors, in agreement with its potency in biological assays for growth promoting and lactogenic activities. In contrast, 20 kDa-hGH has only 3% of the potency of 22 kDa-hGH in binding to the receptors of normal and hypophysectomized rat adipocytes. In agreement with the binding potency, 20 kDa-hGH is only 3% as potent as 22 kDa-hGH in stimulating lipogenesis in normal rat adipocytes preincubated for a few hours in hGH-free medium. The 20 kDa-hGH is also much weaker than 22 kDa-hGH in stimulating lipogenesis in adipocytes from hypophysectomized rats. These data strongly support the concept that the rat adipocyte receptor, which mediates the insulin-like effects of growth hormone, is different from the receptor found on human IM-9 lymphocytes. Preincubation of rat adipocytes with hGH induces a refractoriness to subsequent activation of lipogenesis by hGH but does not abolish the response to insulin, while preincubation with insulin slightly potentiates the hGH response and does not change the insulin response. Additivity studies and a detailed comparison of the lipogenic effects of insulin and hGH suggest that hGH shares only a subset of the metabolic pathways activated by insulin.     

Receptor-mediated degradation of insulin in isolated rat adipocytes: Formation of a degradation product slightly smaller than insulin

More than 90% of the radioactivity associated with isolated rat adipocytes incubated with [Tyr^A14-¹²⁵I]monoiodoinsulin represented at steady state iodoinsulin possessing full binding affinity. In contrast, about half of the radioactivity dissociating from the cells was [¹²⁵I]monoiodotyrosine. The other half was of a molecular size similar to that of iodoinsulin as judged from gel-filtration chromatography. However, the descending limb of the ‘insulin’ peak (i.e., the smaller molecules) possessed a reduced binding activity compared with native iodoinsulin, material from the ascending limb, or a similar fraction isolated from dissociation medium from IM-9 lymphocytes, a cell type devoid of receptor-mediated insulin degradation. The cells, thus, release an intermediary degradation product.     

Rabbit brain glucose transporter responds to insulin when expressed in insulin-sensitive Chinese hamster ovary cells

Transfection of Chinese hamster ovary cells with the expression vector containing rabbit brain HepG2-type glucose transporter cDNA resulted in a dramatic over-expression (approximately 10-fold) of glucose transporter as assessed by either immunoblotting with antipeptide antibody against rabbit brain glucose transporter or photoaffinity labeling with [3H]cytochalasin B. 2-Deoxyglucose uptake was also increased 4-fold in the transfected cells, while no increase in transport activity or transporter amount was observed in cells that were transfected with the expression vector alone without glucose transporter cDNA. Significantly, insulin (10(-7) M) increased 2-deoxyglucose uptake in both control and transfected cells, but the increased amount of the transported 2-deoxyglucose by insulin in the transfected cells was 4.2-fold greater than that in control cells, indicating that the expressed rabbit brain HepG2-type glucose transporter responded to insulin. In addition, we have recently demonstrated that the HepG2-type glucose transporter exists in rat adipocytes and responds to insulin in a fashion similar to a majority of other types of glucose transporters (Oka, Y., Asano, T., Shibasaki, Y., Kasuga, M., Kanazawa, Y., and Takaku, F. (1988) J. Biol. Chem. 263, 13432-13439). In contrast, insulin did not stimulate glucose transport activity in HepG2 cells or IM-9 lymphocytes that have a significant amount of the HepG2-type glucose transporter. Thus, the results in this study further support the notion that insulin regulation of glucose transport activity depends on a tissue-specific signaling mechanism.     

Characterization of a somatomedin (insulin-like growth factor) synthesized by fetal rat liver organ cultures.

Explants of 19- to 20-day fetal rat liver synthesize polypeptides biochemically and immunologically related to the well characterized somatomedin (insulin-like growth factor) BRL-MSA, multiplication-stimulating activity. Fetal MSA was purified from media conditioned by fetal liver explants by chromatography on Sephadex G-75 under acid conditions. Partially purified fetal MSA: 1) inhibited the binding of BRL-MSA to the MSA receptor of rat liver plasma membranes, to somatomedin-binding proteins from rat serum, and to rabbit anti-BRL-MSA serum; 2) had a molecular weight of 4,500 to 12,500 determined by polyacrylamide gel electrophoresis in sodium dodecyl sulfate; 3) stimulated the incorporation of [3H]thymidine into the DNA of chick embryo fibroblasts and induced cell multiplication; 4) stimulated glucose oxidation in rat adipocytes and weakly inhibited the binding of insulin to the insulin receptors of IM-9 lymphocytes; and 5) stimulated sulfate uptake in costal cartilage from hypophysectomized rats. These activities were associated with the same molecular species in fetal MSA preparations following disc acrylamide electrophoresis and co-migrated with active BRL-MSA peptides.     

Cross-linking of insulin receptors to MHC antigens in human B lymphocytes: evidence for selective molecular interactions

Molecular interactions between insulin receptors and MHC antigens were investigated in human B cells. Two B lymphoblastoid cell lines, IM-9 and 526, chosen for their high insulin binding capacity, were found to express 15,000 and 25,000 insulin receptors per cell, respectively. Insulin receptors were labeled with a 125I-photoreactive insulin analogue, and all other surface proteins by lactoperoxidase-catalyzed radioiodination. Neighbor proteins were cross-linked with a cleavable homobifunctional reagent dithio-bis-(succinimidyl propionate) (DSP) and solubilized before immunoprecipitation by anti-HLA monoclonal antibodies. Gel analysis of the precipitated proteins showed that 90% of insulin receptors precipitable by anti-insulin receptor antibodies were precipitated by anti-class I antibodies (anti-heavy chain and anti-beta 2-microglobulin) after cross-linking with 2 mM DSP. In neither IM-9- nor 526 cells could HLA antigens be precipitated by anti-insulin receptor antibodies, suggesting that the concentration of class I antigens largely exceeds the concentration of insulin receptors at the cell surface. In 526 lymphocytes, class I MHC antigens were also found to adjoin class II antigens, since both molecules could be coprecipitated with anti-HLA A, B, C and with anti-HLA-DR antibodies after chemical cross-linking. Down-regulation of insulin receptors by chronic exposure of IM-9 cells to insulin did not affect the amount of MHC molecules present on the cell surface, and conversely, class I MHC molecules were internalized in 526 cells irrespective of the presence of insulin. These results thus show that insulin receptors and MHC antigens form multimolecular complexes in the plasma membrane of cultured human B cells. These interactions, which do not appear to influence the regulation of these proteins on the cell surface, may be involved in the mechanism of hormone signaling.     

Regulation of human leukocyte microsomal hydroxymethylglutaryl-CoA reductase activity by a phosphorylation and dephosphorylation mechanism

The activity of microsomal 3-hydroxy-3-methylglutaryl coenzyme A reductase (EC 1.1.1.34), obtained from cultured human IM-9 lymphoid cells or freshly isolated human peripheral blood leukocytes, is modulated by a phosphorylation/dephosphorylation mechanism. Addition of MgATP + ADP to IM-9 cell microsomal reductase leads to a time-dependent loss of enzyme activity. Inactivated reductase is reactivated by rat liver reductase phosphatase. Kinase-dependent IM-9 cell microsomal reductase, prepared by heating IM-9 microsomes for 15 min at 50°C, is inactivated in the presence of MgATP and ADP only after addition of cytosolic reductase kinase from either IM-9 cells, freshly isolated leukocytes or rat liver. Inactivation is time-dependent and dependent on the cytosolic protein concentration. Inactivated reductase is reactivated by rat liver reductase phosphatase. For cultured IM-9 cells and freshly isolated leukocytes incubated with culture medium for 2 h, the ratios of active (unphosphorylated) to total (phosphorylated + unphosphorylated) reductase activity are 0.22 and 0.43, respectively. Thus, in addition to its regulation by changes in the amount of total enzyme protein, human leukocyte reductase activity is also modulated by a phosphorylation/dephosphorylation mechanism.     

AM1, a glycoprotein from the submandibular glands of the mouse, has esterolytic and amidolytic activities

The activity of microsomal 3-hydroxy-3-methylglutaryl coenzyme A reductase (EC 1.1.1.34), obtained from cultured human IM-9 lymphoid cells or freshly isolated human peripheral blood leukocytes, is modulated by a phosphorylation/dephosphorylation mechanism. Addition of MgATP + ADP to IM-9 cell microsomal reductase leads to a time-dependent loss of enzyme activity. Inactivated reductase is reactivated by rat liver reductase phosphatase. Kinase-dependent IM-9 cell microsomal reductase, prepared by heating IM-9 microsomes for 15 min at 50 degrees C, is inactivated in the presence of MgATP and ADP only after addition of cytosolic reductase kinase from either IM-9 cells, freshly isolated leukocytes or rat liver. Inactivation is time-dependent and dependent on the cytosolic protein concentration. Inactivated reductase is reactivated by rat liver reductase phosphatase. For cultured IM-9 cells and freshly isolated leukocytes incubated with culture medium for 2 h, the ratios of active (unphosphorylated) to total (phosphorylated + unphosphorylated) reductase activity are 0.22 and 0.43, respectively. Thus, in addition to its regulation by changes in the amount of total enzyme protein, human leukocyte reductase activity is also modulated by a phosphorylation/dephosphorylation mechanism.     

The effect of phenformin and other adenosine triphosphate (ATP)-lowering agents on insulin binding to IM-9 human cultured lymphocytes

In the present study, we investigated the mechanism by which the antidiabetic drug phenformin increases insulin binding to its receptors in IM-9 human cultured lymphocytes. After a 24-hr preincubation, phenformin induced a twofold increase in specific 125I-insulin binding, and removal of phenformin was followed 6 hr later by a return in binding to control levels. This effect of phenformin on insulin binding was not a consequence of either inhibition of cell growth, changes in cellular cyclic adenosine monophosphate (AMP) levels, or changes in guanosine triphosphate (GTP) content. Since phenformin is known to inhibit various aspects of cellular energy metabolism, the relationship between 125I-insulin binding and energy metabolism in IM-9 cells was investigated. The phenformin-induced increase in insulin binding to IM-9 cells was related to a time- and dose-dependent decrease in ATP levels. Other agents that lowered ATP levels, including antimycin, dinitrophenol, and 2-deoxyglucose, also raised insulin binding. These studies indicated, therefore, that phenformin enhances insulin binding to receptors on IM-9 cells and that this effect on insulin receptors may be related to alterations in metabolic functions that are reflected by a lowering of ATP levels.     

Insulin receptor expression in the Burkitt lymphoma cells Daudi and Raji

The specific binding of insulin to either intact or Triton-solubilized Daudi cells (a Burkitt lymphoma cell line) was reduced by over 95% compared to that to control IM-9 lymphocytes due to a decrease in receptor number without a change in affinity. Analysis by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography revealed that 125I-labeled Daudi cells had reduced amounts (approximately 1/20th) of immunoprecipitable binding (alpha) subunit [mol wt (Mr), 130,000] of the receptor and a relatively abundant 210,000 Mr form not seen in IM-9 cells. The transmembranous (beta) subunit (Mr, 90,000) of the receptor, although not detected by 125I surface labeling, could be phosphorylated and, together with the 210,000 Mr form, exhibited the same 2-fold stimulation of phosphorylation by insulin as that in IM-9 cells. Northern blot hybridization revealed a decrease in Daudi cells of all four major species of insulin receptor mRNA. The Raji cell, another Burkitt lymphoma cell line, also exhibited reduced protein and genetic expression of the insulin receptor, indicating that reduced insulin receptor expression may be representative of other Burkitt lymphoma cell lines.     

Partial purification and characterization of an insulin-like material from spinach and Lemna gibba G3

The existence in invertebrates, unicellular eukaryotes, and prokaryotes of materials that resemble several vertebrate peptide hormones led to the suggestion that these peptide messengers may have arisen earlier in evolution than had previously been thought. Consistent with this hypothesis, we describe here material in two plants, spinach and Lemma gibba G3, that is very similar to mammalian insulin, yet distinctive. In each of the early purification steps, which consisted of acidic methanol chloroform extraction and sequential chromatography on C-18 hydrophobic resin, Sephadex G-50, CM-Sepharose, and a short C-3 high performance liquid chromatography column, the immunoactive material from plants resembled the common vertebrate insulins. The protein nature of the material was suggested by its destruction by Pronase but not by the inactivated enzyme. In addition, on TSK chromatography it eluted in a position similar to that of insulin, i.e. equivalent to a protein of 6000 daltons. Using an isocratic high performance liquid chromatography system, the plant immunoactivity eluted earlier, and thus was more hydrophilic, than most of the common mammalian insulins, including pork insulin. The interaction of the plant material with anti-insulin antibodies in a radioimmunoassay was confirmed by using an affinity column of anti-insulin antibodies which adsorbed the plant immunoactivity at neutral pH, and released the material with acid elution. Using a quantitative double radioimmunoassay, the plant insulin-like material was distinguished immunologically from chicken insulin. Although the plant insulin-like material is clearly distinct from pork insulin chromatographically, and from chicken insulin immunologically, it resembles vertebrate insulins in its overall configuration. The plant insulin-like material bound to insulin receptors on IM-9 lymphocytes and stimulated glucose oxidation and lipogenesis in isolated adipocytes from young rats. The bioactivity was neutralized in the presence of anti-insulin antibodies, but not in the presence of normal guinea pig IgG. The role of this insulin-like material in plants is unknown but its existence is consistent with an early evolutionary origin of the insulin messenger peptide family. Alternatively we cannot exclude a later convergent development of this family or introduction of vertebrate DNA into plants.     

Down-regulation and recycling of insulin receptors. Effect of monensin on IM-9 lymphocytes and U-937 monocyte-like cells

Receptor down-regulation is the result of various cellular processes including receptor internalization, new synthesis, and recycling. Monensin, a monocarboxylic acid ionophore, has been used to characterize the role of recycling in the metabolism of insulin receptors on two cultured human cell lines, U-937 and IM-9, which have different rates of internalization. The U-937 monocyte-like cell internalizes insulin receptors readily. Incubation with monensin at low doses (10(-6) to 10(-7) M) for 2 h did not affect subsequent surface insulin binding. However, the drug markedly enhanced insulin-induced down-regulation. Monensin had little effect on ligand internalization in this cell line as demonstrated by quantitative morphometric analysis. The IM-9 lymphocyte, a slow internalizer, was less sensitive to monensin exposure. Prolonged exposure (12 h) to this compound of either cell line resulted in apparent inhibition of insertion into the surface membrane of both newly synthesized and recycled receptors. When solubilization was used to quantitate total cell receptors, there was essentially no difference in intact cell binding (i.e. surface receptors) and total cell binding in IM-9 cells when insulin-induced down regulation alone was compared to insulin and monensin. By contrast for the U-937 cells there was only a small further decrease in binding when monensin was added to insulin in the solubilized cells compared to the marked augmentation of down-regulation when monensin was added to insulin in intact cells. These data demonstrate that cells with a rapid internalization rate have an associated active recycling process. By contrast cells with a slow internalization rate have a similarly slow recycling rate. This is consistent with relatively equal rates of receptor biosynthesis and plasma membrane insertion in both cell types.     

Structure-function relationships of shortened [LeuB25]insulins, semisynthetic analogues of a mutant human insulin

Replacement of B25-phenylalanine by leucine in the insulin sequence causes marked inactivation. The effect of this sequence variation was studied here in des-(B26-30)-insulin. [LeuB25]des-(B26-30)-insulin and its B25-amide were prepared by trypsin-mediated semisynthesis from N-terminally protected des-(B23-30)-insulin and synthetic tripeptides. The relative lipogenic potency in isolated rat adipocytes was 8.0% for the truncated analogue with a free B25-carboxyl function, and 18.1% for the amidated analogue. Binding to cultured human IM-9 lymphocytes was 4% and 9%, respectively. Thus, both shortened insulins are markedly more active than [LeuB25]insulin. The PheB25----LeuB25 substitution in both the shortened and the full sequence has a moderate effect on the CD spectrum, indicating that the gross main chain conformation is largely retained in both molecules. Independent of the substitution an absolute increase of the circular dichroism is observed upon amidation of the B25-carboxyl group.     

Glucocorticoids increase insulin binding and the amount of insulin-receptor mRNA in human cultured lymphocytes.

The effect of steroid hormones on insulin binding and the amount of insulin-receptor mRNA was examined in IM-9 lymphocytes. Cortisol and cortexolone, but not oestrogen, increased both the binding of insulin and the amount of insulin-receptor mRNA in a time- and dose-dependent manner. Cortisol was most potent, and induced a 2-fold increase in insulin binding and a 4-fold increase in mRNA. The elevation in binding was due to an increased number of insulin receptors at the cell surface. The increase in mRNA involved all four of the insulin-receptor mRNAs and could not be inhibited by cycloheximide. The cortisol-induced increase in mRNA was associated with a 3-4-fold increase in the synthesis of pro-receptor. The relative potency of the three steroids indicated that these effects were mediated by an interaction with the glucocorticoid receptor. The results of this study suggest that cortisol can increase the number of insulin receptors at the cell surface by increasing the amounts of insulin-receptor mRNA and the synthesis de novo of insulin receptors.     

Antipeptide antibodies that recognize a lymphocyte substance P receptor

In an effort to investigate the presence of substance P (SP) receptors on lymphocytes, polyclonal antibodies against SP receptors were developed. The immunogen used to generate these antibodies was a peptide encoded by an RNA complementary to the mRNA for SP. The rationale for using this SP complementary peptide (termed SP CP) as an immunogen resulted from the observation that 3H-SP bound to microtiter wells coated with SP CP in a dose dependent and saturable fashion. Furthermore, binding was blocked with excess unlabeled SP or SP antagonist, D-Pro2-D-Phe7-D-Trp9-SP. Inasmuch as the peptide, SP CP, specifically bound 3H-SP, we hypothesized that antibodies against this peptide might recognize a SP receptor binding site. Using the SP receptor positive lymphoblast cell line, IM-9, affinity-purified antibodies against SP CP but not antibodies against keyhole limpet hemocyanin recognized a molecule on the surface of IM-9 cells. Anti-SP CP binding to IM-9 cells was blocked with excess SP antagonist, suggesting that the antibody and the SP antagonist were competing for the same binding site. In support of this possibility, anti-SP CP antibodies blocked 3H-SP binding to IM-9 cells. An immunoaffinity column coupled with antibodies against SP CP bound protein from solubilized IM-9 cells. This isolated protein bound 125I-Tyr8-SP and binding was specifically blocked with SP as well as by SP antagonist, neurokinin A, and eledoisin. Passthrough material did not bind SP suggesting that a SP receptor had been purified. Western blot analysis of solubilized IM-9 cell proteins using anti-SP CP antibodies but not preimmune IgG recognized a single protein of 58,000 D. Taken together, these results demonstrate that antibodies against SP CP recognize a SP receptor present on the lymphocyte cell line, IM-9.     

Regulation of the insulin receptor by a monoclonal anti-receptor antibody. Evidence that receptor down regulation can be independent of insulin action

In the present study, we investigated the ability of a monoclonal antibody to the insulin receptor to regulate the expression of the insulin receptor of IM-9 lymphocytes. Previously, this antibody was shown to be a competitive antagonist of insulin action on severe metabolic functions. In the present study, we report that preincubation of IM-9 cells with the monoclonal antibody caused a dose- and time-dependent decrease in the subsequent ability of these cells to bind 125I-insulin, a phenomenon termed down regulation. The antibody was approximately 100 times more potent than insulin at down regulating the receptor. In contrast, the antibody was 5 times less potent than insulin in competing for binding to insulin receptors and dissociated 4 times more rapidly than insulin from IM-9 cells. Three lines of evidence suggested that the mechanism of down regulation by the antibody was the same as the one used by insulin. First, both agents caused a rapid initial decrease in insulin binding to cells followed by a slower, gradual decrease in binding. Second, the down regulation caused by both was reversible, and this reversibility required new protein synthesis. Third, the antibody, like insulin, accelerated receptor degradation. Since the antibody does not mimic the other effects of insulin on metabolic processes, these results suggest that the mechanism of insulin receptor down regulation is different from the mechanism of insulin action on other cellular functions.     

Insulin-like growth factor binding to the atypical insulin receptors of a human lymphoid-derived cell line (IM-9).

The cells of the IM-9 human lymphocyte-derived line contain a sub-population of insulin-binding sites whose immunological and hormone-binding characteristics closely resemble those of the atypical insulin-binding sites of human placenta. These binding sites, which have moderately high affinity for multiplication-stimulating activity [MSA, the rat homologue of insulin-like growth factor (IGF) II] and IGF-I, are identified on IM-9 cells by 125I-MSA binding. They account for approximately 30% of the total insulin-receptor population, and do not react with a monoclonal antibody to the type I IGF receptor (alpha IR-3). The relative concentrations of unlabelled insulin, MSA and IGF-I required to displace 50% of 125I-MSA from these binding sites (1:4.7:29 respectively) are maintained for cells, particulate membranes, Triton-solubilized membranes precipitated either by poly(ethylene glycol) or a polyclonal antibody (B-10) to the insulin receptor, and receptors purified by insulin affinity chromatography. Because the atypical insulin/MSA-binding sites outnumber the type I IGF receptors in IM-9 cells by approximately 10-fold, they also compete with the latter receptors for 125I-IGF-I binding. Thus 125I-IGF-I binding to IM-9 cells is inhibited by moderately low concentrations of insulin (relative potency ratios for insulin compared with IGF-I are approx. 1/14 to 1/4) and is partially displaced (65-80%) by alpha IR-3. When type I IGF receptors are blocked by alpha IR-3 or removed by B-10 immunoprecipitation or insulin affinity chromatography, the hormone-displacement patterns for 125I-IGF-I binding resemble those of the atypical insulin/MSA-binding sites.     

Characterization of a monoclonal antibody against the lymphoblast substance P receptor

A murine monoclonal antibody to the IM-9 lymphoblast substance P (SP) receptor has been produced which recognizes the membrane-associated proteins of the SP receptor as demonstrated by immunoprecipitation of [125I]SP affinity-labeled and [35S]methionine biosynthetically labeled IM-9 soluble membranes. SP and anti-SP receptor binding to [35S]methionine-labeled IM-9 cell proteins were directly compared by attachment of each to affinity supports. Eluants from these affinity columns were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and revealed an equivalent 33-kDa protein in both cases. This protein corresponds to one of the previously described [125I]SP specifically affinity-labeled membrane-associated proteins. In addition, two-color fluorescence-activated cell sorter analysis with human peripheral blood T lymphocytes with fluorescein-SP and rhodamine-labeled antireceptor antibody revealed a distinct population of cells (20 to 30%) that were equally labeled by both the fluorescent peptide and antibody. This result indicates that the anti-SP receptor antibody recognizes an epitope of the receptor that is common to both human peripheral blood T lymphocytes and IM-9 lymphoblast cells.     

Characterization of insulin receptor carbohydrate by comparison of chemical and enzymatic deglycosylation

To characterize the carbohydrate moieties of the insulin receptor on IM-9 lymphocytes, the cells were surface iodinated and solubilized, and the insulin receptors were precipitated with anti-receptor antibody. The precipitates were resuspended, subjected to either enzymatic digestion or chemical treatment with trifluoromethanesulfonic acid and the relative mobilities of the alpha and beta subunits before and after treatment were analyzed by polyacrylamide gel electrophoresis and autoradiography. The results indicate that the alpha subunit possesses primarily N-linked carbohydrate which is both complex (Endoglycosidase F sensitive) and polymannose (Endoglycosidase H sensitive). The beta subunit also contains polymannose oligosaccharide units and has, in addition, a substantial amount of carbohydrate which is removed by chemical treatment but is not susceptible to Endoglycosidase F, suggesting the presence of O-linked saccharides. The apparent molecular weights of the core protein of the mature alpha and beta subunits as determined by gel electrophoresis following complete deglycosylation are 98 kDa and 80 kDa, respectively.