Maria G. Buse

The Inhibition of Islet Superoxide Dismutase by Diabetogenic Drugs

The enzyme superoxide dismutase (SOD) is a scavenger of superoxide radicals and protects the integrity of cell membranes. We previously reported that streptozotocin inhibits the SOD activity of erythrocytes and retinae in vivo and in vitro. We now report that the three major diabetogenic drugs, i.e., alloxan, streptozotocin, and Vacor (in order of increasing potency), interact with erythrocyte Cu-Zn SOD in vitro. Maximum inhibition of erythrocyte SOD of man, dog, rat, and cow is 40% and is achieved within 10 min. At submaximal doses the effect of the three drugs is additive. Inhibition induced by streptozotocin, but not by the other two agents, is reversible by dialysis. Alloxan (80 mg/kg), like streptozotocin, also inhibits erythrocyte SOD activity when injected i.v. into rats. Glucose or 3- o-methylglucose does not prevent SOD inhibition by alloxan or streptozotocin in vitro. Injection of glucose before alloxan prevents the development of diabetes, but does not prevent alloxan-lnduced inhibition of erythrocyte SOD. SOD is present in the islets of Langerhans of rats and dogs, as demonstrated by biochemical assay of isolated islets and by immunofluorescent staining of frozen pancreases. The specific activity of SOD in the islets exceeds that of the exocrine pancreas more than 100-fold. The islet SOD is of the Cu-Zn type, since it is inhibited by KCN, and not by chloroformethanol. The mobility of the islet enzyme on polyacrylamlde disc gel electrophoresis is different from that of erythrocyte SOD. Streptozotocin, alloxan, and Vacor Inhibit the activity of islet SOD in vitro; the rate and magnitude of inhibition is the same as that observed with erythrocyte SOD. The Mn SOD of liver mitochondria is not affected by the diabetogenic drugs. It is suggested that the inhibitory effect of the diabeto- genie drugs on islet cell SOD may contribute to their cytotoxicity, and that changes in the amount or activity of this protective enzyme in β-cells may play a role in determining their vulnerability to noxious agents.

Protective Role of Superoxide Dismutase against Diabetogenic Drugs

Copper-zinc superoxide dismutase (SOD) is present in relatively high concentrations in the beta-cells of human islets. The activity of the extracted enzyme is partially inhibited upon incubation with the diabetogenic drugs alloxan, streptozotocin, or Vacor. The role of this enzyme in protecting beta-cells against chemically induced diabetes was further investigated. Incubation of intact canine islets with alloxan (0.2 mg/ml) and 4 mM glucose decreased the insulin secretory response by 87% during subsequent exposure to 28 mM glucose. Concomitantly the SOD-specific activity (units of enzyme activity per milligram immunoreactive SOD) decreased 50% in alloxan-exposed islets. When islets were protected from alloxan toxicity by including 28 mM glucose with alloxan, the insulin secretory response and SOD specific activity remained identical to controls. Thus, SOD specific activity correlates with maintenance of beta-cell function. To test the effectiveness of SOD against streptozotocin in vitro, canine islets were incubated 10 min with or without streptozotocin (0.1 mg/ml) with 4 mM glucose; their functional integrity was tested subsequently as the insulin secretory response to 28 mM glucose. Exposure to streptozotocin alone decreased the response by 70%; inclusion of SOD (1.5 mg/ml) before and during exposure to streptozotocin completely prevented this effect. Cyanide-inactivated SOD was not effective. The potential of SOD to prevent streptozotocin-induced diabetes was tested in rats in vivo. SOD injected 10 s or 50 min before streptozotocin prevented or significantly attenuated diabetes. Injection of SOD and streptozotocin simultaneously was much less effective, and cyanide-inactivated SOD was ineffective. No protection was afforded by injection of SOD 12 or 24 h before streptozotocin. Our results support hypotheses that (a) oxygen radicals mediate the beta-cell toxicity of both alloxan and streptozotocin, and (b) beta-cells may be particularly vulnerable to oxygen radical damage.

Pre-Exposure to Glucosamine Induces Insulin Resistance of Glucose Transport and Glycogen Synthesis in Isolated Rat Skeletal Muscles: Study of Mechanisms in Muscle and in Rat-1 Fibroblasts Overexpressing the Human Insulin Receptor

Increased routing of glucose through the hexosamine-biosynthetic pathway has been implicated in the development of glucose-induced insulin resistance of glucose transport in cultured adipocytes. Because both glucosamine and glucose enter this pathway as glucosamine-6-phosphate, we examined the effects of preincubation with glucosamine in isolated rat diaphragms and in fibroblasts overexpressing the human insulin receptor (HIR-cells). In muscles, pre-exposure to glucosamine inhibited subsequent basal and, to a greater extent, insulin-stimulated glucose transport in a time- and dose-dependent manner and abolished the stimulation by insulin of glycogen synthesis. Insulin receptor number, activation of the insulin receptor tyrosine kinase in situ and after solubilization, and the total pool of glucose transporters (GLUT4) were unaffected, and glycogen synthase was activated by glucosamine pretreatment. In HIR-cells, which express GLUT1 and not GLUT4, basal and insulin-stimulated glucose transport were unaffected by glucosamine, but glycogen synthesis was markedly inhibited. Insulin-stimulated activation of protein kinases (MAP and S6) was unaffected, and the fractional velocity and apparent total activity of glycogen synthase was increased in glucosamine-treated HIR-cells. In pulse-labeling studies, addition of glucosamine during the chase prolonged processing of insulin proreceptors to receptors and altered the electrophoretic mobility of proreceptors and processed α-subunits, consistent with altered glycosylation. Glucosamine-induced insulin resistance of glucose transport appears to be restricted to GLUT4-expressing cells, i.e., skeletal muscle and adipocytes; it may reflect impaired translocation of GLUT4 to the plasmalemma. The glucosamine-induced imbalance in UDP sugars, i.e., increased UDP-N-acetylhexosamines and decreased UDP-glucose, may alter glycosylation of critical proteins and limit the flux of glucose into glycogen.

Effects of Diabetes and Hyperglycemia on the Hexosamine Synthesis Pathway in Rat Muscle and Liver

In vitro studies suggested that increased flux of glucose through the hexosamine biosynthesis pathway (HexNSP) contributes to glucose-induced insulin resistance. Glutamine:fructose-6- phosphateamidotransferase (GFAT) catalyzes glucose flux via HexSNP; its major products are uridine diphosphate (UDP)-N-acetyl hexosamines (UDP-HexNAc). We examined whether streptozotocin (STZ)-induced diabetes (4–10 days) or sustained hyperglycemia (1–2 h) in normal rats alters absolute or relative concentrations of nucleotide-linked sugars in skeletal muscle and liver in vivo. UDP-HexNAc and UDP-hexoses (UDP-Hex) were increased and decreased, respectively, in muscles of diabetic rats, resulting in an ∼ 50% increase in the UDP-HexNAc:UDPHex ratio (P < 0.01). No significant changes in nucleotide sugars were observed in livers of diabetic rats. In muscles of normal rats, UDP-HexNAc concentrations increased (P < 0.01) and UDP-Hex decreased (P < 0.01) during hyperglycemia. The UDP-HexNAc:UDP-Hex ratio increased ∼40% (P < 0.01) and correlated strongly with plasma glucose concentrations. Changes in liver were similar to muscle but were less marked. GFAT activity in muscle and liver was unaffected by 1–2 h of hyperglycemia. GFAT activity decreased 30–50% in muscle, liver, and epididymal fat of diabetic rats, and this was reversible with insulin therapy. No significant change in GFAT mRNA expression was detected, suggesting post-transcriptional regulation. The data suggest that glucose flux via HexNSP increases in muscle during hyperglycemic hyperinsulinemia and that the relative flux of glucose via HexNSP is increased in muscle in STZ-induced diabetes. Since nucleotide sugars are essential substrates for glycoprotein synthesis, changes in their absolute or relative concentrations may affect signal transduction and contribute to insulin resistance.

Plasma lipids and lipoproteins in young insulin-dependent diabetic patients: relationship with control.

SummaryPlasma and lipoprotein cholesterol and triglycerides, glucose and haemoglobin A1c concentrations were measured in 106 patients (56 males) with insulin-dependent diabetes mellitus (age range 2–22 years) and 36 normal volunteers (19 males) with similar age and sex distribution. The diabetic patients were further divided into three subgroups: “good”, fair and poor control, based on 24 h glycosuria and haemoglobin A1c concentrations. Total, low density lipoprotein and very low density lipoprotein cholesterol levels were significantly increased in male patients in poor control when compared with the group in “good” control and with normal subjects. Triglyceride and very low density lipoprotein triglyceride levels were also significantly increased in poorly controlled males. The most significant difference however was a decrease of high density lipoprotein cholesterol in male patients in poor control. There was a significant inverse correlation between haemoglobin A1c and high density lipoprotein cholesterol (r = -0.63) and a direct correlation between haemoglobin A1c and low density lipoprotein cholesterol (r = 0.35) and triglycerides (r = 0.62) in the male diabetics. The findings were similar in females. The most striking change was observed in low density lipoprotein cholesterol levels, which were more markedly increased in poorly controlled females than in poorly controlled males. No statistical singificant differences were found between the groups in good and fair control for any of the plasma and lipoprotein lipids studied. A significant difference however was found between the groups in poor and fair control. There was a significant correlation in females between haemoglobin A1c and low density lipoprotein cholesterol levels (r = 0.43), haemoglobin A1c and triglycerides levels (r = 0.54) and an inverse correlation between haemoglobin A1c and high density lipoprotein cholesterol (r = -0.59).

Diabetes-induced functional and structural changes in insulin receptors from rat skeletal muscle.

The effect of diabetes on the structure and function of insulin receptors was studied in rats 7 d after streptozotocin injection, using solubilized, partially purified receptors from rat hindlimb muscles. Diabetes increased the number of insulin receptors per gram of muscle 60-70% without apparent change in insulin binding affinity. Incubation of receptors at 4 degrees C with [gamma-32P]ATP and insulin resulted in dose-dependent autophosphorylation of the beta-subunit on tyrosine residues; receptors from diabetic rats showed decreased base-line phosphorylation, as well as a decrease in autophosphorylation at maximally stimulating insulin concentrations. These receptors also showed diminished exogenous substrate kinase activity using histone H2b and angiotensin II as phosphoacceptors. The electrophoretic mobility (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) of a subpopulation of beta-subunits derived from diabetics was slightly decreased; differences in electrophoretic mobility between control- and diabetic-derived beta-subunits were enhanced by generating fragments by partial Staphylococcus aureus V8 protease digestion. Endoglycosidase-H or neuraminidase treatment increased the electrophoretic mobility of beta-subunits in both groups, but only neuraminidase appeared to decrease or abolish differences in electrophoretic mobility between controls and diabetics, suggesting that excess sialilation may account, in part, for the altered mobility of diabetic derived beta-subunits. All structural and functional alterations in insulin receptors were prevented by treating diabetic rats with insulin for 60 h. Peripheral insulin resistance associated with insulinopenic diabetes may be related to modifications in insulin receptor structure, resulting in impaired signal transmission.

Studies concerning the specificity of the effect of leucine on the turnover of proteins in muscles of control and diabetic rats.

The protein anabolic effect of branched chain amino acids was studied in isolated quarter diaphragms of rats. Protein synthesis was estimated by measuring tyrosine incorporation into muscle proteins in vitro. Tyrosine release during incubation with cycloheximide served as an index of protein degradation. In muscles from normal rats the addition of 0.5 mM leucine stimulated protein synthesis 36--38% (P less than 0.01), while equimolar isoleucine or valine, singly or in combination were ineffective. The three branched chain amino acids together stimulated no more than leucine alone. The product of leucine transamination, alpha-keto-isocaproate, did not stmino norborane-2-carboxylic acid (a leucine analogue) were ineffective. Leucine and isoleucine stimulated protein synthesis in muscles from diabetic rats.Leucine, isoleucine, valine and the norbornane amino acid but not alpha-ketoisocaproate or beta-hydroxybutyrate decreased the concentration of free tyrosine in tissues during incubation with cycloheximide; tyrosine release into the medium did not decrease significantly. Leucine caused a small decrease in total tyrosine release, (measured as the sum of free tyrosine in tissues and media), suggesting inhibition of protein degradation. The data suggest that leucine may be rate limiting for protein synthesis in muscles. The branched chain amino acids may exert a restraining effect on muscle protein catabolism during prolonged fasting and diabetes.

Effect of denervation on the expression of two glucose transporter isoforms in rat hindlimb muscle.

Denervation rapidly (within 24 h) induces insulin resistance of several insulin-responsive pathways in skeletal muscle, including glucose transport; resistance is usually maximal by 3 d. We examined the effect of denervation on the expression of two glucose transporter isoforms (GLUT-1 and GLUT-4) in rat hindlimb muscle; GLUT-4 is the predominant species in muscle. 1 d postdenervation, GLUT-1 and GLUT-4 mRNA and protein concentrations were unchanged. 3 and 7 d postdenervation, GLUT-4 mRNA and protein (per microgram DNA) were decreased by 50%. The minor isoform, GLUT-1 mRNA increased by approximately 500 and approximately 100%, respectively, on days 3 and 7 while GLUT-1 protein increased by approximately 60 and approximately 100%. The data suggest that the insulin resistance of glucose transport early after denervation does not reflect a decrease in total glucose transporter number; however, decreased GLUT-4 expression may contribute to its increased severity after 3 d. Parallel decreases in GLUT-4 mRNA and GLUT-4 protein postdenervation are consistent with pretranslational regulation; GLUT-1 expression may be regulated pre- and posttranslationally. The cell type(s) which overexpress GLUT-1 postdenervation need to be identified. Nervous stimuli and/or contractile activity may modulate the expression of GLUT-1 and GLUT-4 in skeletal muscle tissue.

Identification of the Major Site of O-Linked β-N-Acetylglucosamine Modification in the C Terminus of Insulin Receptor Substrate-1

Signal transduction from the insulin receptor to downstream effectors is attenuated by phosphorylation at a number of Ser/Thr residues of insulin receptor substrate-1 (IRS-1) resulting in resistance to insulin action, the hallmark of type II diabetes. Ser/Thr residues can also be reversibly glycosylated by O-linked β-N-acetylglucosamine (O-GlcNAc) monosaccharide, a dynamic posttranslational modification that offers an alternative means of protein regulation to phosphorylation. To identify sites of O-GlcNAc modification in IRS-1, recombinant rat IRS-1 isolated from HEK293 cells was analyzed by two complementary mass spectrometric methods. Using data-dependent neutral loss MS3 mass spectrometry, MS/MS data were scanned for peptides that exhibited a neutral loss corresponding to the mass of N-acetylglucosamine upon dissociation in an ion trap. This methodology provided sequence coverage of 84% of the protein, permitted identification of a novel site of phosphorylation at Thr-1045, and facilitated the detection of an O-GlcNAc-modified peptide of IRS-1 at residues 1027–1073. The level of O-GlcNAc modification of this peptide increased when cells were grown under conditions of high glucose with or without chronic insulin stimulation or in the presence of an inhibitor of the O-GlcNAcase enzyme. To map the exact site of O-GlcNAc modification, IRS-1 peptides were chemically derivatized with dithiothreitol following β-elimination and Michael addition prior to LC-MS/MS. This approach revealed Ser-1036 as the site of O-GlcNAc modification. Site-directed mutagenesis and Western blotting with an anti-O-GlcNAc antibody suggested that Ser-1036 is the major site of O-GlcNAc modification of IRS-1. Identification of this site will facilitate exploring the biological significance of the O-GlcNAc modification.

A Simplified Assay of Hemoglobin AIc in Diabetic Patients by Use of Isoelectric Focusing and Quantitative Microdensitometry

Isoelectric focusing (IEF) of human erythrocyte hemolysates on poly aery la mide slab gels over a pH gradient of 6 to 8 provides sufficient resolution of hemoglobin Ale (HbAIc) from other hemoglobin components (AIa, AIb, AII, S, and F) to allow quantification by high-resolution microdensitometry. Twice-chromatographed HbAIc alone and in admixtures with normal and diabetic hemolysates were employed to verify the identification and quantification of the Aie component. Relative concentration was determined as a per cent of the total hemoglobin absorption at 556 nm on acid-fixed, unstained gels. A total of 60 patient samples were examined by IEF, and, for 35 samples, both column chromatography and IEF determination were obtained, revealing excellent correlation between these two methods.