Thermal stability of glucokinase (GK) as influenced by the substrate glucose, an allosteric glucokinase activator drug (GKA) and the osmolytes glycerol and urea.

We investigated how glycerol, urea, glucose and a GKA influence kinetics and stability of wild-type and mutant GK. Glycerol and glucose stabilized GK additively. Glycerol barely affected the TF spectra of all GKs but decreased k(cat), glucose S(0.5) and K(D) values and ATP K(M) while leaving cooperativity unchanged. Glycerol sensitized all GKs to GKA as shown by TF. Glucose increased TF of GKs without influence of glycerol on the effect. Glycerol and GKA affected kinetics and binding additively. The activation energies for thermal denaturation of GK were a function of glucose with K(D)s of 3 and 1mM without and with glycerol, respectively. High urea denatured wild type GK reversibly at 20 and 60°C and urea treatment of irreversibly heat denatured GK allowed refolding as demonstrated by TF including glucose response. We concluded: Glycerol stabilizes GK indirectly without changing the folding structure of the apoenzyme, by restructuring the surface water of the protein, whereas glucose stabilizes GK directly by binding to its substrate site and inducing a compact conformation. Glucose or glycerol (alone or combined) is unable to prevent irreversible heat denaturation above 40°C. However, urea denatures GK reversibly even at 60°C by binding to the protein backbone and directly interacting with hydrophobic side chains. It prevents irreversible aggregation allowing complete refolding when urea is removed. This study establishes the foundation for exploring numerous instability mutants among the more than 600 variant GKs causing diabetes in animals and humans.

Glucokinase and glucose homeostasis: proven concepts and new ideas.

The enzyme GK (glucokinase), which phosphorylates glucose to form glucose 6-phosphate, serves as the glucose sensor of insulin-producing beta-cells. GK has thermodynamic, kinetic, regulatory and molecular genetic characteristics that are ideal for its glucose sensor function and allow it to control glycolytic flux of the beta-cells as indicated by control-, elasticity- and response-coefficients close to or larger than 1.0. GK operates in tandem with the K(+) and Ca(2+) channels of the beta-cell membrane, resulting in a threshold for glucose-stimulated insulin release of approx. 5 mM, which is the set point of glucose homoeostasis for most laboratory animals and humans. Point mutations of GK cause 'glucokinase disease' in humans, which includes hypo- and hyper-glycaemia syndromes resulting from activating or inactivating mutations respectively. GK is allosterically activated by pharmacological agents (called GK activators), which lower blood glucose in normal animals and animal models of T2DM. On the basis of crystallographic studies that identified a ligand-free 'super-open' and a liganded closed structure of GK, on thermostability studies using glucose or mannoheptulose as ligands and studies showing that mannoheptulose alone or combined with GK activators induces expression of GK in pancreatic islets and partially preserves insulin secretory competency, a new hypothesis was developed that GK may function as a metabolic switch per se without involvement of enhanced glucose metabolism. Current research has the goal to find molecular targets of this putative 'GK-switch'. The case of GK research illustrates how basic science may culminate in therapeutic advances of human medicine.

Mutants of glucokinase cause hypoglycaemia- and hyperglycaemia syndromes and their analysis illuminates fundamental quantitative concepts of glucose homeostasis.

AIMS/HYPOTHESIS: Mutations of the glucokinase gene cause hyperglycaemia or hypoglycaemia. A quantitative understanding of these defects of glucose homeostasis linked to the glucokinase gene was lacking. Therefore a database of kinetic variables of wild-type and 20 missense mutants of glucokinase was developed and used in mathematical modelling to predict the thresholds for glucose-stimulated insulin release. METHODS: Recombinant human glucokinase was generated in E. coli. The k(cat), glucose S(0.5), ATP K(m), and Hill number of glucokinase were determined. Inhibition by Stearoyl CoA and glucokinase regulatory protein and thermal stability were assayed for all mutants kinetically similar to wild-type glucokinase. A mathematical model predicting the threshold for glucose-stimulated insulin release was constructed. This model is based on the two substrate kinetics of glucokinase and the kinetic variables of the database. It is assumed that both glucokinase gene alleles are equally expressed in beta-cells and that induction of glucokinase occurs as a function of basal blood glucose. RESULTS: Large changes, varying greatly between mutants were found in nearly all variables. Glucokinase flux at threshold for glucose-stimulated insulin release was about 25 % of total phosphorylating potential in the normal beta-cell and this was used to predict thresholds for the mutant heterozygotes. Clinical data for maturity onset diabetes of the young type linked to the glucokinase gene and familial hyperinsulinaemic hypoglycaemia linked to the glucokinase gene and the glucokinase kinetic data of this study were used to test the model. The model predicts fasting blood glucose between 3 and 7 mmol/l in these cases. CONCLUSION/INTERPRETATION: A kinetics database of wild-type and 20 mutants of glucokinase was developed. Many kinetic differences were found for the mutants. The mathematical model to calculate the threshold for glucose-stimulated insulin release predicts fasting blood glucose between 3 and 7 mmol/l in subjects with glucokinase gene mutations. [Diabetologia 42: 1175-1186]

Cell-biological assessment of human glucokinase mutants causing maturity-onset diabetes of the young type 2 (MODY-2) or glucokinase-linked hyperinsulinaemia (GK-HI).

Mutations in the glucokinase (GK) gene cause type-2 maturity-onset diabetes of the young type 2 (MODY-2) and GK-linked hyperinsulinaemia (GK-HI). Recombinant adenoviruses expressing the human wild-type islet GK or one of four mutant forms of GK, (the MODY-2 mutants E70K, E300K and V203A and the GK-HI mutant V455M) were transduced into glucose-responsive insulin-secreting beta-HC9 cells and tested functionally in order to initiate the first analysis in vivo of recombinant wild-type and mutant human islet GK. Kinetic analysis of wild-type human GK showed that the glucose S(0. 5) and Hill coefficient were similar to previously published data in vitro (S(0.5) is the glucose level at the half-maximal rate). E70K had half the glucose affinity of wild-type, but similar enzyme activity. V203A demonstrated decreased catalytic activity and an 8-fold increase in glucose S(0.5) when compared with wild-type human islet GK. E300K had a glucose S(0.5) similar to wild-type but a 10-fold reduction in enzyme activity. E300K mRNA levels were comparable with wild-type GK mRNA levels, but Western-blot analyses demonstrated markedly reduced levels of immunologically detectable protein, consistent with an instability mutation. V455M was just as active as wild-type GK, but with a markedly reduced S(0.5). The effects of the different GK mutants on glucose-stimulated insulin release support the kinetic and expression data. These experiments show the utility of a combined genetic, biochemical and cell-biological approach to the quantification of functional and structural changes of human GK that result from MODY-2 and GK-HI mutations.

Chronic effect of fatty acids on insulin release is not through the alteration of glucose metabolism in a pancreatic beta-cell line (beta HC9).

Hyperinsulinaemia in the fasting state and a blunted insulin secretory response to acute glucose stimulation are commonly observed in obesity associated non-insulin-dependent diabetes mellitus. Hyperlipidaemia is a hallmark of obesity and may play a role in the pathogenesis of this beta-cell dysfunction because glucose metabolism in pancreatic beta cells may be altered by the increased lipid load. We tested this hypothesis by assessing the chronic effect of oleic acid on glucose metabolism and its relationship with glucose-induced insulin release in beta HC9 cells in tissue culture. Our results show: (1) A 4-day treatment with oleic acid caused an enhancement of insulin release at 0-5 mmol/l glucose concentrations while a significant decrease in insulin release occurred when the glucose level was greater than 15 nmol/l; (2) Hexokinase activity was increased and a corresponding left shift of the dose-dependency curve of glucose usage was observed associated with inhibition of glucose oxidation in oleic acid treated beta HC9 cells, yet the presumed glucose-related ATP generation did not parallel the change in insulin release due to glucose; (3) The rate of cellular respiration was markedly increased in oleic acid treated beta HC9 cells both in the absence of glucose and at all glucose concentrations tested. This enhanced oxidative metabolism may explain the increased insulin release at a low glucose level but is clearly dissociated from the blunted insulin secretion at high glucose concentrations. We conclude that a reduction of oxidative metabolism in pancreatic beta cells is unlikely to be the cause of the dramatic effect that high levels of non-esterified fatty acids have on glucose-induced insulin release.

Glucose metabolism and insulin release in mouse beta HC9 cells, as model for wild-type pancreatic beta-cells.

Glucose metabolism and its relationship with glucose-induced insulin release were studied in beta HC9 and beta TC3 cells to identify and characterize key factors controlling the intermediary metabolism of glucose and glucose-induced insulin release. The beta HC9 cell line, derived from pancreatic islets with beta-cell hyperplasia, is characterized by a normal concentration-dependency curve for glucose-stimulated insulin release, whereas the beta TC3 cell line, derived from pancreatic beta-cell tumors, shows a marked leftward shift of this curve. Maximum velocity and the Michaelis-Menten constant of glucose uptake in beta HC9 and beta TC3 cells were similar, even though GLUT-2 expression in these two cell lines differed. In both cell lines, the kinetic characteristics of glucose usage, glucose oxidation, and glucose-induced oxygen consumption were similar to those of glucose phosphorylation, indicating that the kinetics of glucose metabolism from the glucose phosphorylation step in the cytosol to the mitochondrial process of oxidative phosphorylation are determined by the glucose-phosphorylating enzyme, that is, by glucokinase in beta HC9 cells and by hexokinase in beta TC3 cells. Thus beta HC9 cells provide an opportunity for the quantitative analysis of glucose metabolism, the associated generation of coupling factors, and other essential beta-cell functions involved in glucose sensing and insulin secretion.

Insulin family growth factors have specific effects on protein synthesis in preimplantation mouse embryos.

Previously constructed protein databases for two stages of preimplantation mouse embryogenesis, the compacted eight-cell stage and the fully expanded blastocyst stage, have been used to analyze the effects of insulin, IGF-I, and IGF-II on protein synthesis in these developmental stages. Proteins were labeled by placing, for 2 hr, synchronous cohorts of 35-50 embryos into human tubal fluid (HTF) medium containing L-[35S]-methionine (1 mCi/ml) in the presence or absence of one of the growth factors. The embryos were then washed with medium and lysed. Samples were processed for 2-D gel analysis. For each embryonic stage and each growth factor, four or five experimental replicates were done and the gel images were compared using the PDQUEST system. Using the computer-assisted analysis, we were able to identify proteins that showed a statistically significant (P < 0.05) change in synthesis. At the eight-cell stage of development insulin caused increased synthesis of two proteins and decreased synthesis in three proteins. Insulin-treated blastocyst stage embryos exhibited an increased synthesis in eight proteins and decreased synthesis for one protein. The effect of IGF-I at the eight-cell stage of development was mostly inhibitory; the synthesis of only one protein increased and the synthesis of five proteins showed a decrease. Similar results were obtained with blastocyst stage embryos; four proteins demonstrated an increase in synthesis while 14 proteins showed a decrease. Eight-cell stage embryos incubated with IGF-II had seven proteins with a decreased synthesis, although in blastocyst stage embryos, nine proteins showed increased synthesis.(ABSTRACT TRUNCATED AT 250 WORDS)

Protein databases for compacted eight-cell and blastocyst-stage mouse embryos.

High-resolution two-dimensional sodium dodecyl sulfate-polyacrylamide (2D-SDS) gel electrophoresis combined with computerized analysis of gel images was used to construct and analyze protein data-bases for two stages of preimplantation mouse embryogenesis, the compacted eight-cell stage and the fully expanded blastocyst stage. These stages were chosen for their ease in identification of multiple synchronous embryos. Synchronous cohorts of 30-50 embryos were labelled with L-[35S]methionine for 2 hr. The embryos were then lysed in 30 microliters hot SDS sample buffer, and the lysates were stored at -80 degrees C until the gels were run. Five replicates were run for eight-cell embryos, and four for blastocyst-stage embryos. The samples were processed for 2D gel electrophoresis and fluorography; multiple exposures were made. Gel images were analyzed using the PDQUEST system, and databases were constructed. Analysis of the databases for both developmental stages showed high reproducibility of protein spots in multiple gel images. Of 1,674 total spots in eight-cell embryo standards, > 79% of spots had a percentage error (S.E.M./average) < 50%, and > 45% had a percentage error < 30%. Similarly, of 1,653 total spots in blastocyst-stage embryo standards, 74% of spots had a percentage error < 50%, and approximately 47% of spots had a percentage error < 30%. Forty-three spots (approximately 3% of the total spots) were found to be detected only in the eight-cell stage, while 75 spots were detected solely in the blastocyst stage. Sixty-nine proteins showed a greater than threefold increase in isotope incorporation from the eight-cell to the blastocyst stage, with a percentage error < 50% in both the eight-cell and the blastocyst stages. In contrast, 41 of the proteins showed a decrease during this period. Analysis of the protein databases described in this study has allowed us to document the overall quantitative changes in proteins from the compacted eight-cell stage to the blastocyst stage of mouse preimplantation development. These data-bases provide a valuable tool for further detailed quantitative analysis of specific proteins associated with developmental events. In addition they will permit analysis of the effects of environmental factors, such as growth factors, on early embryo development.

Sulfonylurea-binding sites and ATP-sensitive K+ channels in alpha-TC glucagonoma and beta-TC insulinoma cells.

alpha-Cells secrete glucagon in a fuel-dependent fashion. We tested the hypothesis that alpha-cells contain sulfonylurea- and ATP-sensitive K+ channels. We studied two clonal lines of alpha-TC cells (simian virus 40 T-antigen induced glucagonoma cells) and for reference purposes, similarly transformed beta-TC insulinoma cells. alpha-TC cells each contained approximately 3000 high-affinity binding sites for the sulfonylurea [3H]glyburide. Whole-cell ATP- and tolbutamide-sensitive K+ currents of alpha-TC and beta-TC cells, relative to cell surface area, were comparable. In cell-attached membrane patches of alpha-TC cells, two types of K+ channels were observed. They had slope conductances of approximately 63 and 33 pS when the electrode contained 151 mM K+. Tolbutamide and diazoxide decreased and enhanced, respectively, the open probability of these channels. The membrane of alpha-TC cells depolarized periodically. This electrical activity was inhibited by diazoxide. A physiological mixture of amino acids enhanced glucagon release, and high glucose partially inhibited this release. Tolbutamide also enhanced glucagon release, whereas diazoxide inhibited it. Thus, alpha-TC glucagonoma cells contain ATP-sensitive K+ channels that regulate glucagon release, yet allow inhibition of hormone release by glucose.

Metabolic labelling and quantitation of proteins synthesized by single chick cochleas.

Molecular studies of the peripheral auditory system are made difficult by the small quantities of tissue available and by their relative inaccessibility. In addition, the cochlea and other hair cell-containing receptor organs are composed of both hair cells and supporting cells, as well as several other cell types. The identification of known proteins and the characterization of specific and novel protein molecules from these tissues require the use of sensitive techniques and a consideration of the complex histology. The chick cochlea was selected as an experimental system since the cochlea is relatively accessible in the bird, the receptor neuroepithelium contains a large number of hair cells compacted in a small area, and the physiology of the auditory periphery has been studied extensively. A general procedure is described for the metabolic radiolabelling of proteins from single cochleas followed by their solubilization, separation by high-resolution two-dimensional gel electrophoresis, and accurate quantitation. The method is highly reproducible and sensitive, and should prove useful in studies of proteins from the specialized cell types of the chick cochlea, including the identification of those whose rates of synthesis are modified in response to acoustic stimulation and sound damage or recovery.

Identification of glucose response proteins in two biological models of beta-cell adaptation to chronic high glucose exposure.

High resolution 2-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) combined with computerized analysis of gel images was used to search for proteins whose biosynthesis was induced or repressed in pancreatic islet cells chronically exposed to high glucose in an in situ and a tissue culture model of islet cell adaptation to excessive fuel load. The in situ model involved a 4-day intravenous infusion of either 50% glucose or 0.45% saline solution, followed by islet isolation, [35S]methionine labeling at 3 and 18 mM glucose for both groups, and protein analysis by 2-dimensional SDS-PAGE. The tissue culture model involved a 7-day culture of isolated rat islets in RPMI 1640 with 10% fetal calf serum containing either 3 or 30 mM glucose, followed by radiolabeling and 2-dimensional PAGE of proteins as in the in situ model. A small fraction of about 1.5% of the approximately 2000 identifiable proteins can be characterized as adaptive proteins. Of these altogether 58 proteins in the two models, 5 proteins were demonstrable in both models and two of these (proteins 1526 and 7622) are particularly noteworthy. Protein 1526 (Mr 57,000; pI 5.09) showed the same response pattern in both models and its expression was most enhanced when islets from chronically glucose-infused animals or those cultured for 7 days at 30 mM were radiolabeled at 18 mM glucose. Protein 7622 (Mr 68,000; pI 6.50) (also known as GSP-65; Collins, H.W., Buettger, C., and Matschinsky, F.M. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 5494-5498) showed a different labeling pattern in the two models: stimulation of [35S]methionine incorporation by 18 mM glucose both in control and experimental islets from the infusion study, but lack of such stimulation of radiolabeling in islets cultured for 7 days at 30 mM glucose in contrast to islets cultured at 3 mM. The experimental strategy and the methodology are evaluated and the significance of the results is discussed. Potentials of the approach and plans for future experiments are considered.

High-resolution two-dimensional polyacrylamide gel electrophoresis reveals a glucose-response protein of 65 kDa in pancreatic islet cells.

High-resolution two-dimensional PAGE was used to search for glucose-response proteins in isolated pancreatic islets that were labeled with [35S]methionine at ambient glucose concentrations of 0-18 mM. A 65-kDa protein, isoelectric focusing point of approximately 6.6-7.0, was discovered that showed at least a 20-fold stimulation of radiolabeling when glucose in the labeling medium was increased from 3 to 18 mM, in contrast to a 2.5-fold enhancement of label incorporation into total islet proteins. This 65-kDa protein is evident after 30 min of labeling with 18 mM glucose and is preferentially synthesized compared to its nearest neighbors after both 30 and 60 min of labeling. Glucose induction of the 65-kDa protein was virtually blocked by D-mannoheptulose. Glucose induction of this 65-kDa protein is in practically all aspects comparable to glucose induction of insulin and glucokinase in pancreatic beta cells. A working hypothesis is developed proposing that glucose-response proteins or "glucospondins" are pivotal constituents of pancreatic islet cells and that their discovery and exploration promise new insights into normal and pathological islet cell function.

The 28-kDa calbindin-D is a major calcium-binding protein in the basilar papilla of the chick.

In previous work we identified a basilar papilla protein (BPP23) that appears to be one of the most abundant soluble proteins in the basilar papilla of the chick cochlea. Here we report the purification of protein BPP23 from chick cochlea and the generation of a specific antiserum. Immunoblotting and immunoprecipitation experiments with this antiserum indicate that BPP23 is a calcium-binding protein very similar, if not identical, to avian calbindin, the 28-kDa vitamin D-dependent calcium-binding protein. Although the basilar papilla contains both receptor hair cells and supporting cells, immunocytochemical studies by others have localized calbindin-like immunoreactivity to the hair cells in the rat auditory receptor epithelium. Our estimates of the abundance of protein BPP23, assuming exclusive localization within the hair cell, indicate a concentration of at least 1 mM. Avian calbindin has four high-affinity (Kd = 0.5 X 10(-6)) calcium-binding sites. The presence of a specific calcium-binding activity at such high levels suggests an important function for cochlear calbindin (BPP23) in hair cell calcium homeostasis and auditory transduction.

Filamentous capsulated streptococci from the human respiratory tract: chemical and immunochemical characterization of a glycoprotein capsular antigen of provisional binary capsular type 87.

A filamentous alpha-hemolytic streptococcus of provisional capsular type 87 isolated from the human respiratory tract has been shown to be binary capsulated. One of the capsular antigens appears to be a glycoprotein; the other appears to be a polysaccharide. Transformation reactions with deoxyribonucleic acid from streptococcus type 87 and a number of noncapsulated pneumococci yielded transformed pneumococci with either a glycoprotein capsule or a polysaccharide capsule, but not with both. Capsular precipitin (quellung) reactions were observed when streptococcus type 87 was treated with homologous antiserum or with antisera to either of the two distinct capsular transformants. Each of the transformed pneumococci gave a quellung reaction with its homologous antiserum or with antiserum to streptococcus type 87, but neither reacted with antiserum to the heterologous transformant. Chemical analysis showed the glycoprotein antigen of streptococcus type 87 to contain, in addition to amino acids, glucose, galactose, glucosamine, and phosphate. The amino acid composition of the glycoprotein capsular antigens from streptococcus type 87 and of those from transformed pneumococci were similar, showing only minor differences. The glycoprotein capsular antigen from streptococcus type 87 gave two closely associated precipitin bands with homologous antiserum or antisera to transformed pneumococci with the glycoprotein capsule. That the two precipitin bands represent two unrelated proteins is precluded largely on the basis of the unlikely probability of 100% cotransformation of the genes coding for both proteins in the pneumococcal transformants that were isolated. Chemical analyses of the various fractions of the glycoprotein indicate that the two precipitin bands may represent a glycoprotein and its corresponding apoprotein.

Filamentous capsulated streptococci from the human respiratory tract. I. Antigenic attributes of provisional capsular type 83 and its relationship to streptococci of so-called group M.

Two immunologically reactive polysaccharides have been isolated from the cell walls and from culture filtrates of a filamentous alpha-hemolytic streptococcus provisionally designated capsular type 83. Both polysaccharides were purified by diethylaminoethyl-cellulose chromatography. Analysis indicates that the capsular polysaccharide consists of galactose and phosphorus, whereas the cell wall polysaccharide contains galactosamine, glucosamine, glucose, and phosphorus. On the basis of immunochemical experiments, it is suggested that the capsular polysaccharide is composed of galactose-phosphate units with terminal galactose residues at the nonreducing end. It has also been found that the capsular antigen of streptococcus type 83 is shared by a number of streptococcal strains classified in Lancefield's group M. The cell wall polysaccharide of streptococcus type 83 cross-reacts with antibody to the C(s), or cell wall-like capsular, polysaccharide of Diplococcus pneumoniae, and this cross-reactivity may be a reflection that the streptococcal antigen possesses certain structural features which are similar to those of pneumococcal C and C(s) polysaccharides.