Insulin activation of NADH ferricyanide reductase in human erythrocytes is mediated by the insulin receptor tyrosine kinase: a comparative study in normal and diabetic states

Summary

We have previously reported that NADH ferricyanide reductase in human erythrocytes is stimulated by insulin. Hormone-stimulated activities are attenuated in the presence of glycolytic inhibitors like vanadate, indicating the involvement of glycolysis in the mechanism by which insulin stimulates ferricyanide reduction. Activation of erythrocyte metabolism in response to insulin could be a result of hormone binding to its receptor, inducing phosphorylation of band 3 (at a site for reversible association of glycolytic enzymes) and/or other membrane proteins like the Na⁺/H⁺ antiport. Activation of the antiporter protein by insulin can stimulate glycolysis by an increase in intracellular pH, an effect which is prevented by amiloride. Evidence for a role for tyrosine phosphorylation in triggering the reductase activation came from studies with protein kinase inhibitors. Genistein, sphingosine and acridine orange have been shown to prevent insulin-stimulated ferricyanide reduction, implicating tyrosine phosphorylation as an important signal for activation of the enzyme by insulin. To evaluate activation of the enzyme by insulin stimulated phosphorylation, a comparative study was done using erythrocytes from healthy and diabetic humans. We measured ferricyanide reductase activities in basal and insulin stimulated states. Basal activities were lower in diabetics than in normal humans. Nevertheless, hormone stimulated activities were similar, despite earlier reports of decreased receptor phosphorylation of exogenous substrates in type 2 diabetics. These observations, together with previous ones, suggest that insulin-receptor kinase interaction may mediate the action of insulin on human erythrocytes by phosphorylation of cellular proteins like band 3 and/or the Na⁺/H⁺ antiport.     

Mucosal surface ferricyanide reductase activity in mouse duodenum

Mouse duodenum possesses mucosal surface ferricyanide reductase activity. The reducing activity, determined in vitro by measuring ferrocyanide production from ferricyanide, was found to be greater in duodenal fragments when compared with ileal fragments. Experiments with right-side out tied-off duodenal sacs show that reduction occurs mainly on the mucosal side and indicates that the reducing activity is associated with the brush border membrane. Experiments using mice with increased levels of iron absorption (hypoxic, iron-deficient) showed corresponding increases in reducing activity. The increase was present in duodenal but not ileal fragments. Inhibitor studies showed no effect of several compounds which inhibit other, more characterized, transplasma membrane reductases. In particular, doxorubicin (10 m) and quinacrine (1 mm) were without effect on duodenal mucosal transplasma membrane reducing activity. Depolarization of the membrane potential with high medium K ⁺ inhibited reducing activity. N-ethyl malemide (1 mm) was a potent inhibitor, but iodoacetate was found to be less inhibitory. Comparision with inhibitory effects on glyceraldehyde-3-phosphate dehydrogenase (GAPDH) demonstrated that the effect of N-ethyl malemide on reducing activity was not secondary to GAPDH. Collectively these results indicate that mouse duodenum possesses mucosal surface transplasma membrane ferricyanide reductase activity and that the activity is correlated with the process of intestinal iron absorption. Furthermore, the reducing activity appears to be distinct from other reported transplasma membrane reductases.     

NADH: Nitrate reductase activity restoration by rhodanese

The NADH: nitrate reductase from durum wheat leaves was inactivated by cyanide and its activity restored by thiosulphate and beef kidney rhodanese. Rhodanese and thiosulphate, added to NADH-nitrate reductase before cyanide treatment protected NADH-nitrate reductase activity. No oxidizing agent was required for the protection or restoration of cyanide treated NADH-nitrate reductase.     

Transmembrane ferricyanide reductase activity in Ehrlich ascites tumor cells

A transmembrane ferricyanide reductase activity was assayed in intact Ehrlich ascites tumor cells. Kinetic measurements gave a Km of 0.14 mM and a Vmax of 0.31 mumol/min per 10(6) cells. In short-term batch experiments, this activity was enhanced in the presence of 10 mM lactate, a source of cytosolic NADH. The transmembrane redox activity was accompanied by alkalinization of the cytosol. Both ferricyanide reduction and proton extrusion were diminished in the presence of 0.2 mM amiloride. Several cytotoxic drugs significantly inhibited the ferricyanide reductase activity at concentrations at which they show antineoplastic activity.     

The peroxidase and peroxynitrite reductase activity of human erythrocyte peroxiredoxin 2

Peroxiredoxin 2 (Prx2) is a 2-Cys peroxiredoxin extremely abundant in the erythrocyte. The peroxidase activity was studied in a steady-state approach yielding an apparent K_M of 2.4 μM for human thioredoxin and a very low K_M for H₂O₂ (?0.7 μM). Rate constants for the reaction of peroxidatic cysteine with the peroxide substrate, H₂O₂ or peroxynitrite, were determined by competition kinetics, k₂ = 1.0 × 10⁸ and 1.4 × 10⁷ M⁻¹ s⁻¹ at 25 °C and pH 7.4, respectively. Excess of both oxidants inactivated the enzyme by overoxidation and also tyrosine nitration and dityrosine were observed with peroxynitrite treatment. Prx2 associates into decamers (5 homodimers) and we estimated a dissociation constant K_d < 10⁻²³ M⁴ which confirms the enzyme exists as a decamer in vivo. Our kinetic results indicate Prx2 is a key antioxidant enzyme for the erythrocyte and reveal red blood cells as active oxidant scrubbers in the bloodstream.     

Control of erythrocyte membrane microviscosity by insulin

The human erythrocyte membrane binds insulin through high-affinity, low-capacity binding sites (dissociation constant Kd1 2.45 X 10(-9)M; capacity n1 207 fmol/mg protein) and low-affinity, high-capacity binding sites (Kd2 0.63 X 10(-6) M; n2 37 pmol/mg protein). Treatment of the erythrocyte membrane or the intact cells with the physiological concentration of insulin, which is within the range of Kd value of the high-affinity sites, results in a significant reduction of the membrane microviscosity and the filtration time of the intact cells. Use of supraphysiological concentrations of the hormone reverses the effect of the lower concentration of insulin on the membrane microviscosity and the filtration time.     

Regulation of insulin receptor activity of human erythrocyte membrane by prostaglandin E1

Incubation of human erythrocyte membrane with low concentration of prostaglandin E1 or prostacyclin increased the binding of 125I-labeled insulin to the membrane. The binding of the radioiodinated hormone was maximally stimulated at 3 nM prostaglandin E1 and the use of higher concentrations (above 8 nM) of the autacoid tended to reverse its own effect at lower concentrations. While prostaglandins A1, A2, B1, B2, D2, F1 alpha, F2 alpha or 6-keto-prostaglandin F1 alpha had no effect on the binding of insulin to the erythrocyte membrane, prostaglandin E2 at similar concentrations decreased the binding of the hormone. The effect of prostaglandin E1 on the increased binding of the insulin was found to be reversible and depended on the occupancy of the autacoid molecules on the membrane and showed positive cooperativity. Scatchard analysis of the binding of 125I-labeled insulin to the erythrocyte ghosts indicated that in the presence of the autacoid, the binding capacity of the insulin receptor increased 2-fold (from 207 to 424 fmol/mg protein) without any change in the ghosts affinity for the ligand (Kd 2.4 X 10(-9) versus 2.49 X 10(-9) M). As a consequence of increased binding of insulin to the erythrocyte membrane in the presence of prostaglandin E1 (3.0 nM), the optimal concentration of the peptide hormone for the maximal reduction of the membrane microviscosity decreased from approx. 1.6 to approx. 0.4 nM. Addition of prostaglandin E1 alone at the above concentration to the assay mixture had no effect on the membrane microviscosity.     

Reduction of extracellular potassium ferricyanide by transmembrane NADH: (acceptor) oxidoreductase of human erythrocytes

Reduction of extracellular ferricyanide by intact erythrocytes proceeds by a membrane bound, NADH-dependent reaction. It is depressed by a glycolysis inhibitor and a non penetrable sulfhydryl reagent, and activated by dehydroascorbate. Dehydroascorbate activation cannot be accounted for by release of reducing equivalents from the cells. It is concluded that the observed reaction is brought about by transmembrane NADH-acceptor oxidoreductase with donor binding at the inner and acceptor binding at the outer cell surface.     

Population study of erythrocyte glutathione reductase activity

Summary The activity of erythrocyte glutathione reductase (GR) was determined in a group of 87 prisoners from northern Thailand (65 with normal, 22 with deficient erythrocyte G-6-PD) without and with added FAD. The amount of stimulation by FAD was inversely related to the original activity suggesting that FAD stimulation in vivo is one of the main determinants of GR activity. 4 subjects showed insufficient stimulation by FAD. The binding of FAD to GR seemed to be closer to saturation in G-6-PD deficient subjects but the maximal stimulated activity of these subjects was higher than in the group with normal G-6-PD. This suggests that the marked increase of GR activity in G-6-PD deficient erythrocytes is due to increased binding of FAD and to a larger amount of stimulable enzyme. Original GR activity was positively correlated with the socio-economic status, and subjects who had taken riboflavin in the period prior to examination had a higher mean GR activity than those without vitamin intake. GR activity was markedly raised by administration of riboflavin and in G-6-PD deficient subjects glutathione stability was improved. In comparison to central European subjects more than 50 percent of the examined population are GR deficient. This seems to be due to a low dietary supply of riboflavin in most cases. The findings in 4 subjects with insufficient stimulation by added FAD raise the question whether hereditary forms of GR deficiency exist in this population.

---

Zusammenfassung In einer Gruppe von 87 Strafgefangenen in Nordthailand (65 mit normaler, 22 mit defizienter G-6-PD) wurde die Aktivität der Glutathion-Reductase (GR) mit und ohne Zusatz von FAD bestimmt. Das Ausmaß der Stimulierung des Enzyms durch FAD verhielt sich reziprok zur ausgangsaktivität. Bei 4 Versuchspersonen war die Stimulierung durch FAD insuffizient. Die Bindung von FAD an GR war bei Versuchspersonen mit G-6-PD-Mangel näher am Sättigungspunkt. Aber auch die gesamte stimulierbare Aktivität der GR war bei Versuchspersonen mit G-6-PD-Mangel größer als bei Gesunden. Die deutliche Erhöhung der GR-Aktivität bei G-6-PD-Mangel scheint zwei Ursachen zu haben: vermehrte Bindung von FAD an das Enzym und Vermehrung der Gesamtmenge an stimulierbarem Enzym. In einer weiteren Gruppe war die Ausgangsaktivität der GR vom sozio-ökonomischen Status der Versuchsperson abhängig. Personen, die vor der Untersuchung Riboflavin-haltige Medikamente eingenommen hatten, hatten höhere Aktivitäten. Die GR-Aktivität wurde durch Einnahme von Riboflavin unter Kontrolle stark erhöht, und bei Versuchspersonen mit G-6-PD-Mangel wurde die Glutathion-Stabilität der Erythrocyten verbessert. Im Vergleich zu Mitteleuropäern sind mehr als die Hälfte der thailändischen Versuchspersonen als GR defizient einzustufen. Dies ist wahrscheinlich auf eine geringere Zufuhr an Riboflavin in der Nahrung zurückzuführen. Die Befunde bei 4 Versuchspersonen mit insuffizienter Stimulierung der GR durch FAD könnte durch einen erblichen Mangel an GR verursacht sein.

---

---

Established and supported by Stiftung Volkswagenwerk.     

NADH nitrate reductase and related NADH cytochrome c reductase species in barley

Nitrate and nitrate-less barley (Hordeum vulgare cv Golden Promise) shoot extracts were examined by Sephadex G200 gel filtration and sucrose density gradient analysis and the MWs of NR and CR species present were determined from their Stokes radii and sedimentation coefficients by the method of Siegel and Monty. Nitrate-less plant extracts possessed a CR species of MW 27 800 whilst nitrate-plant extracts possessed CR species of MW 203 000, 61 000, 40 000 and 27 800. The MW 203 000 CR species was associated with NADH-NR, FMNH-NR and MV°-NR activities and represents the NR complex. The MW 40 000 and 61 000 CR species were shown to be derived from the NR complex. We suggest that the MW 40 000 and 61 000 CR species represent either subunits of the NR complex or domains cleaved from the intact NR complex by endogenous proteinases.     

Purification and properties of aldose reductase and aldehyde reductase II from human erythrocyte

Aldose reductase (EC 1.1.1.21) and aldehyde reductase II (L-hexonate dehydrogenase, EC 1.1.1.2) have been purified to homogeneity from human erythrocytes by using ion-exchange chromatography, chromatofocusing, affinity chromatography, and Sephadex gel filtration. Both enzymes are monomeric, Mr 32,500, by the criteria of the Sephadex gel filtration and polyacrylamide slab gel electrophoresis under denaturing conditions. The isoelectric pH's for aldose reductase and aldehyde reductase II were determined to be 5.47 and 5.06, respectively. Substrate specificity studies showed that aldose reductase, besides catalyzing the reduction of various aldehydes such as propionaldehyde, pyridine-3-aldehyde and glyceraldehyde, utilizes aldo-sugars such as glucose and galactose. Aldehyde reductase II, however, did not use aldo-sugars as substrate. Aldose reductase activity is expressed with either NADH or NADPH as cofactors, whereas aldehyde reductase II can utilize only NADPH. The pH optima for aldose reductase and aldehyde reductase II are 6.2 and 7.0, respectively. Both enzymes are susceptible to the inhibition by p-hydroxymercuribenzoate and N-ethylmaleimide. They are also inhibited to varying degrees by aldose reductase inhibitors such as sorbinil, alrestatin, quercetrin, tetramethylene glutaric acid, and sodium phenobarbital. The presence of 0.4 M lithium sulfate in the assay mixture is essential for the full expression of aldose reductase activity whereas it completely inhibits aldehyde reductase II. Amino acid compositions and immunological studies further show that erythrocyte aldose reductase is similar to human and bovine lens aldose reductase, and that aldehyde reductase II is similar to human liver and brain aldehyde reductase II.     

Decrease of NADH in HeLa cells in the presence of transferrin or ferricyanide

The short-term incubation of HeLa cells in the presence of diferric transferrin or ferricyanide, which are reduced externally by the transplasma membrane reductase, produces a stoichiometric decrease in NADH and increase in NAD+, which is stimulated by insulin. The NADP/NADPH ratio does not change during 15 min incubation with the oxidants. The total pyridine nucleotide pool of HeLa cells is not affected. Incubation with apotransferrin and ferrocyanide, which cannot act as oxidants for transmembrane electron transport, does not change the pyridine nucleotide concentrations in the cells. Our results show that NADH can act as the internal electron donor for the reduction of external oxidants by the transmembrane reductase. It appears that oxidation of NADH by the transmembrane electron transport using ferricyanide or iron transferrin as external electron acceptors is sufficient to stimulate growth in HeLa cells.     

Leucine aminopeptidase and ferricyanide reductase activities in radish microsomes

A NADH-ferricyanide reductase activity found in radish microsomes isolated from germinated seeds has been shown to be stimulated by pCMB and pCMBS which are both strong nactivators of many plant proteolytic enzymes. In the same preparation a leucine aminopeptidase was found while endoprotease and carboxypeptidase activities were not detected using exogenous substrates. The aminopeptidase, highly active at the same optimal pH-condition of FeCN reductase, was stimulated by CoCl₂ and non-polar detergents (Triton X-100 and Brij 35). It was inhibited by sulphydryl reagents. By gel filtration of microsomal detergent extract two peaks of activity were separated: red I coeluted with LeuAPase and red II, free of aminopeptidase. Red I, a protein, was inhibited by sulphydral reagents and stimulated by duroquinone. Red II, stimulated by pCMB, is not a protein because of the small size and the noninfluence of heating treatment on catalytic activity.     

Dependence of human erythrocyte methemoglobin reductase on temperature

The effect of temperature on the activities of cytoplasmic and membrane-bound fractions of NADH-cytochrome beta 5 reductase on the total activity of methemoglobin reductase in intact human erythrocytes was studied within the temperature range of 20-50 degrees C. The above three activities showed a break in the Arrhenius plots at 42 degrees C which was attributed to irreversible inactivation of the enzymes. Thermal inactivation of methemoglobin reductase in erythrocytes was found to increase the methemoglobin content concomitantly with a decrease in the osmotic stability and activation of spontaneous cell hemolysis.     

Low resolution structure of human erythrocyte glutathione reductase

Glutathione reductase from human erythrocytes is a dimeric flavoenzyme with a molecular weight of 100,000. X-ray diffraction analysis using the isomorphous replacement technique with four heavy-atom derivatives yielded an electron density map at 6 Å resolution with a figure of merit of 0.88. Only minor cuts had to be made in the electron density map to isolate one molecule. The dimer interface is on a crystallographic 2-fold axis. Each subunit can be subdivided into three domains: I, II and III, which are aggregated in such a way that deep clefts are formed on opposite sides of the subunit. These clefts accommodate the substrate glutathione, binding to domain III, and the oxidized cofactor NADP, binding to domain I in a similar extended conformation as NAD binds to the dehydrogenases. The shortest connection between the centres of the nicotinamide ring of NADP and the cystine of oxidized glutathione is 18 Å long and goes along the interface between domains II and III right through the centre of the subunit. Presumably, FAD binds to domain II and its isoalloxazine ring bridges the gap between NADP and glutathione.     

Effects of domain-specific erythrocyte membrane modulators on acetylcholinesterase and NADH:cytochrome b5 reductase activities

Previously, we showed using electron paramagnetic resonance that the physical state of one side of erythrocyte membranes could be modulated by agents which interact with the opposite side (reviewed in Butterfield, 1989, Biological and Synthetic Membranes, A. R. Liss, Inc., New York). The present study was undertaken to determine whether membrane-bound enzymes would exhibit a similar transmembrane modulation effect. The effects of known, domain-specific modulators of the physical state of erythrocyte membranes on the activity of two membrane-bound enzymes were investigated. Acetylcholinesterase, an enzyme having its active site situated on the extracellular side of the membrane, seemed to be unaffected by most of the modulators employed in this study, with the exception of reversible inhibition by benzyl alcohol. Conversely, the activity of NADH:cytochrome b5 reductase, an enzyme whose active site is located on the cytoplasmic side of the erythrocyte membrane, was increased by those agents that interact primarily with skeletal proteins to increase skeletal protein-protein interactions; however, those agents which interact primarily with the skeleton to decrease protein-protein interactions decreased the activity of NADH:cytochrome b5 reductase. This enzyme's activity was also significantly altered by lectins which bind specifically to the external face of glycophorin A on the opposite side of the membrane, but it's activity was unaffected by concanavalin A, a lectin which binds to the external face of band 3. The results of these biochemical studies suggested that NADH:cytochrome b5 reductase can interact with and its activity can be modulated by skeletal or transmembrane proteins. In addition, these results support the hypothesis that in transmembrane signaling processes, biophysical and biochemical changes are correlated.     

Isolation and characterization of human erythrocyte insulin receptors

The noncovalently associated 125I-insulin-receptor complex was isolated from human erythrocyte membranes after allowing 125I-insulin to interact with the membranes followed by extraction of the 125I-insulin-receptor complex with Triton X-102 or, alternatively, by complete solubilization of the membranes with sodium dodecyl sulfate (SDS), removal of SDS, and then treatment of the solubilized sample with 125I-insulin. Sepharose CL-6B column chromatography of the 125I-insulin-receptor complex obtained by both of the above procedures yielded a highly radioactive 140,000-Da complex which was dissociated into small peptides when subjected to SDS-polyacrylamide gel electrophoresis. In contrast, when the 125I-insulin-treated membrane sample was extracted with Triton X-102, purified by DEAE-Sephacel ion exchange chromatography, covalently cross-linked with disuccinimidyl suberate, and then subjected to SDS-polyacrylamide gel electrophoresis, a highly radioactive component with Mr = 53,000 was obtained. On the other hand, when the Triton X-102-solubilized membrane receptor sample was fractionated by DEAE-Sephacel ion exchange chromatography, complexed with 125I-insulin, covalently crosslinked, and then applied to a Sepharose CL-6B column, a 95,000-Da complex with high specific radioactivity was obtained. Upon SDS-polyacrylamide gel electrophoresis, the 95,000-Da complex was dissociated into a 53,000-Da component which appeared identical with that obtained from the receptor complex described above which was obtained by direct interaction of the membranes with 125I-insulin.