Genetic variants of human erythrocyte glucose 6-phosphate dehydrogenase. I. Regulation of activity by oxidized and reduced nicotinamide-adenine dinucleotide phosphate

A Journey from Blood Cells to Genes and Back

I was attracted to hematology because by combining clinical findings with the use of a microscope and simple laboratory tests, one could often make a diagnosis. I was attracted to genetics when I learned about inherited blood disorders, at a time when we had only hints that somatic mutations were also important. It seemed clear that if we understood not only what genetic changes caused what diseases but also the mechanisms through which those genetic changes contribute to cause disease, we could improve management. Thus, I investigated many aspects of the glucose-6-phosphate dehydrogenase system, including cloning of the gene, and in the study of paroxysmal nocturnal hemoglobinuria (PNH), I found that it is a clonal disorder; subsequently, we were able to explain how a nonmalignant clone can expand, and I was involved in the first trial of PNH treatment by complement inhibition. I was fortunate to do clinical and research hematology in five countries; in all of them, I learned from mentors, from colleagues, and from patients.

Kinetic and thermodynamic properties of two electrophoretically similar genetic variants of human erythrocyte glucose-6-phosphate dehydrogenase

Glucose-6-phosphate dehydrogenase (G6PD) A(+) and G6PD A(-) were purified to electrophoretic homogeneity from human male erythrocytes. The steady state kinetics of the binding reaction of NADP+ to the two variants were studied as a function of pH and temperature at a constant ionic strength of 0.01. The interaction coefficient, obtained according to the Hill equation, rises with an increase in pH and temperature. The observed variation of cooperative interaction is interpreted in terms of an increase in the percentage of the dimeric form of the enzyme as pH or temperature increases. The more rapid increase in the interaction coefficient with increase in pH or temperature for G6PD A(-) as compared with G6PD A(+) shows that G6PD A(-) forms dimers at a lower pH or temperature than G6PD A(+). Analysis of the log Vmax versus pH curves and their temperature dependence for the two enzyme variants indicates the participation in the reaction mechanism of sulphydryl groups or imidazolium group of histidine. The two variants show very similar but minor specific and significant differences in kinetic and thermodynamic properties with respect to NADP+ binding. Thus the additional mutation in G6PD A(-) must be responsible for its increased affinity for NADP(+) when compared to G6PD A(+) from which it has been derived. These results are consistent with an earlier report (Babalola O et al (1972) Proc Natl Acad Sci USA 69, 946-950) that enzyme 'deficiency' in vivo is due mainly to a loss of active enzyme molecules, rather than to a decreased activity of each molecule.

Unfolding and trypsin inactivation studies reveal a conformation drift of glucose-6-phosphate dehydrogenase upon binding of NADP

Binding of NADP to glucose-6-phosphate dehydrogenase (G6PD) from Dicentrarchus labrax liver has stabilized its native structure against thermal inactivation, guanidine hydrochloride unfolding and inactivation by tryptic digestion. The time-course of G6PD inactivation by guanidine hydrochloride in the presence of NADP has provided experimental evidence in favor of a conformational drift upon NADP binding to the bass enzyme. Based on the inactivation patterns obtained when the enzyme was treated with guanidine hydrochloride and trypsin, it is proposed that the enzyme conformation induced upon NADP binding is in slow equilibrium with the conformation stabilized in the absence of NADP. FPLC studies have shown that micromolar concentrations of NADP induced oligomerization of G6PD. In addition, the different K0.5 values of NADP binding to the enzyme, ranging from 1-2 microM (from trypsin inactivation) to 90 microM (from titration of the intrinsic fluorescence), suggest a step-wise binding of NADP to the oligomer, with negative cooperativity in the saturation process.

Molecular basis of chronic non-spherocytic haemolytic anaemia: a new G6PD variant (393 Arg----His) with abnormal KmG6P and marked in vivo instability

More than 80 genetic variants of glucose-6-phosphate dehydrogenase (G6PD) are associated with chronic non-spherocytic haemolytic anaemia (CNSHA). In order to help clarify the molecular basis of this association, we have carried out a detailed biochemical and genetic characterization of two G6PD deficient brothers affected by CNSHA. The G6PD from the two patients has altered electrophoretic mobility, abnormally elevated Michaelis constant (Km) for G6P, and extreme instability in vivo and in vitro. By comparison with published information we found that this is a new G6PD variant which we have designated G6PD Portici. The entire coding region of the gene has been sequenced, and a single point mutation, a G----A transition, was found at position 1178 in exon X, causing a substitution of histidine for arginine at residue 393 in the polypeptide chain. By polymerase chain reaction (PCR) amplification followed by diagnostic restriction enzyme analysis and allele-specific oligonucleotide hybridization we have demonstrated the inheritance of this mutation in the patient's family. Our results support the notion of a causative link between this mutation in the G6PD gene and CNSHA. Our data, in combination with previous data in the literature, suggest that the three-dimensional structure of G6PD is such as to cause interaction in the binding of its two substrates, G6P and NADP.

Kinetic properties of normal human erythrocyte glucose-6-phosphate dehydrogenase dimers

The steady-state kinetics of normal human erythrocyte glucose-6-phosphate dehydrogenase (D-glucose-6-phosphate: NADP+ oxidoreductase, EC 1.1.1.49) dimers were studied as a function of pH and temperature. Inhibition studies using glucosamine 6-phosphate, NADPH and p-hydroxymercuribenzoate (P-OHMB) were also carried out at pH 8.0. The existence of two binding sites on the enzyme with a transition from low to high affinity for NADP+ when NADP+ concentration is increased is indicated by the nonlinear Lineweaver-Burk plots and sigmoid kinetic patterns. NADPH inhibition was found to be competitive with respect to NADP+ and non-competitive with respect to glucose-6-phosphate. Logarithmic plot of Vmax against pH and inactivation by P-OHMB indicate the participation in the reaction mechanism of imidazolium group of histidine and sulhydryl groups. The initial velocity and product inhibition data gave results which are consistent with the dimeric enzyme following an ordered sequential mechanism. A possible random mechanism is ruled out by the inhibition results of glucosamine 6-phosphate.

Long-term adaptive response to dietary protein of hexose monophosphate shunt dehydrogenases in rat kidney tubules

We have studied the effects of several different macronutrients on the kinetic behaviour of rat renal glucose 6-phosphate dehydrogenase (G6PDH) and 6-phosphogluconate dehydrogenase (6PGDH). Rats were meal-fed with high-carbohydrate/low-protein, high-protein/low-carbohydrate and high-fat diets. High-protein increased renal G6PDH and 6PDGH activities by 66 per cent and 70 per cent respectively, without significantly changing the Km values of either and each Hexose monophosphate dehydrogenase activity increased steadily, reaching a significant difference on day 4. A rise in carbohydrate or fat in the diets, produced no significant change in either the activity or the kinetic parameters, Vmax and Km of the two dehydrogenases. In addition, the administration of a high-protein diet for 8 days significantly increased both the pentose phosphate pathway flux (92.6 per cent) and the kidney weigth (35 per cent), whereas no significant changes in these parameters were found when the animals were treated with the other diets. Our results suggest that an increase in the levels of dietary protein induces a rise in the intracellular levels of these enzymes. The possible role of this metabolic pathway in the kidneys under these nutritional conditions is also discussed.

Role of hexokinase in the regulation of erythrocyte hexose monophosphate pathway under oxidative stress

Human erythrocytes overloaded with homogeneous human hexokinase (up to 15-times the activity of normal RBC) show almost unmodified rates of glucose metabolized in the HMP, however hexokinase-loaded RBC are able to metabolize 1.5 fold more glucose than controls through the HMP when an oxidizing agent like methylene blue (5 to 100 microM) is present. Similarly, RBC loaded with inactivating anti-hexokinase IgG (12 +/- 3% residual hexokinase activity) show HMP rates unchanged under resting conditions, but only 12% of the HMP rate found in normal controls under oxidative stress. These data provide clear evidence that the HMP rate under conditions of oxidative stress is controlled by hexokinase activity and suggest that RBC from patients with hexokinase deficiency are not able to increase the HMP rate under oxidative stress like erythrocytes from individuals with G6PD deficiency.

Sigmoid kinetics of human erythrocyte glucose-6-phosphate dehydrogenase

Several disagreements and inconsistencies have appeared regarding whether human erythrocyte glucose-6-phosphate dehydrogenase exhibits sigmoid or classical kinetics with respect to NADP+ binding. The latest report is that the purified enzyme exhibits classical kinetics while the intracellular enzyme exhibits sigmoid kinetics (H. N. Kirkman, and G. F. Gaetani (1986) J. Biol. Chem. 261, 4033-4038). The various investigations were carried out at fixed pH, ionic strength, and temperature. The steady-state kinetics of crude and purified erythrocyte glucose-6-phosphate dehydrogenase are reported here at various temperatures, ionic strengths, and pH values and as a function of glucose 6-phosphate concentration. Sigmoid kinetics were observed for both purified and crude enzyme samples at high pH, temperature, ionic strength, and concentration of glucose 6-phosphate with Hill coefficients varying between 1.40 and 1.90. In contrast, at low pH, temperature, and ionic strength, the crude enzyme samples exhibit sigmoid kinetics while the purified samples exhibit classical kinetics despite the high concentration of glucose 6-phosphate. High concentrations of glucose 6-phosphate and factors favoring the enzyme in the dimeric form are necessary conditions for the observation of sigmoid kinetics in human erythrocyte glucose-6-phosphate dehydrogenase. These factors are high pH, ionic strength, and temperature. The observed sigmoid kinetics in this enzyme is explained as arising from tetramer-dimer transitions.

Comparative effects of insulin and refeeding on DNA synthesis, HMP shunt and cholesterogenesis in diabetic and fasted rats

DNA synthesis, cholesterogenesis and the enzymes of the hexosemonophosphate (HMP) shunt pathway were investigated in liver of diabetic rats treated with insulin and in fasted/re-fed rats. Both insulin and refeeding were found to induce liver cell proliferation, accompanied by a remarkable increase in cholesterogenesis. An enhancement of glucose-6-phosphate dehydrogenase (G6PD) and 6-phosphogluconate dehydrogenase (G6PD) activities was also found in insulin-treated diabetic rats and in re-fed rats, supporting the concept that these two enzymes are involved in the proliferative process. Since insulin did not exert the same biochemical effects in a non replicating cell population, such as in insulin-treated normal rats, these studies provide new evidence of a close correlation between DNA, cholesterol synthesis and HMP shunt enzymes during cell proliferation.

Gd(+)Cuiabá, a new rare glucose-6-phosphate dehydrogenase variant presenting normal activity

A 33-year-old Brazilian male of Portuguese extraction was found to have a new glucose-6-phosphate dehydrogenase variant, herein named Gd(+)Cuiabá. The enzyme variant is characterized by normal activity, normal electrophoretic mobility, high Km for glucose-6-phosphate, high Ki for NADPH, decreased thermal stability, normal utilization of substrate analogues and normal pH curve.

Improved biotinylation of glucose-6-phosphate dehydrogenase using active-site-blocking agents

Characteristics of the biotinylation of glucose-6-phosphate dehydrogenase are presented. The enzyme is inactivated in the presence of N-hydroxysuccinimido biotin but can be protected by an appropriate concentration of NADPH used as an active-site blocker. A Ki of 1.6 +/- 1.0 microM calculated for NADPH for this protection shows it to be an active-site phenomena. Enzyme inactivation is irreversible with consistent kinetic results requiring the presence of 10 mM EDTA. An improved methodology was developed for biotinylation allowing 100% protection of the enzyme with loading factors up to 30.8 mol of biotin per enzyme.

Genetic variants of human erythrocyte glucose-6-phosphate dehydrogenase: new characterization data obtained by multivariate analysis

Thirteen variables used in the course of characterization of G6PD variants from forty-one individuals have been submitted to multivariate factorial and cluster analysis. Because of the diagrammatic representation of the analysis, two main findings were made possible. Firstly, the three-dimensional plot of the experimental data gives the advantage of a model of classification which is closely related to the ethnogeographical origin of the subjects, and to the clinical and haematological incidence of the G6PD variants. Secondly, cluster analysis visualizes the distance between the G6PDs examined. In this respect, it was determined that three local original variants associated with acute haemolytic anaemia showed a close relationship and belonged to the same cluster (Gd(-) Toulouse, Gd(-) Muret and Gd(-) Colomiers). Conversely, two non-haemolytic variants (G6PD Luz Saint Sauveur and Lozère) were found to be linked in another remote cluster. The procedure developed in this work promotes a new approach to G6PD characteristics in human genetic studies.

Relationship between the rate of erythrocyte hexose monophosphate pathway and the glucose 6-phosphate concentration

Erythrocytes of individuals with increased (+ 50%) or reduced (-35%) hexokinase activity contain respectively 70 and 17 nmole/ml RBC of glucose-6-phosphate (normal concentration 30 +/- 5nmole/ml RBC) and show comparable rates of the HMP (60 +/- 5nmole/hr/ml RBC). Similarly, in RBC of different ages, obtained by density gradient ultracentrifugation, the glucose-6-phosphate concentration range from 57 (young cells) to 18 (old cells) nmole/ml RBC but the rate at which glucose is utilized in the HMP is unchanged. These data exclude a regulatory role of glucose 6-phosphate in the HMP even if its concentration is under that required for maximal G6PD activity.

Genetic polymorphism of G6PD in a Bulgarian population

Considerable genetic heterogeneity in G6PD was found in the Bulgarian population-14 G6PD variants isolated from 117 hemizygous carriers of G6PD deficiency. Of these, G6PD Mediterranean type was a polymorphic variant and G6PD Corinth occurred with high frequency. Two new variants were identified-G6PD Rudosem and G6PD Nedelino. In a selected group of 78 subjects with clinical manifestations, four variants were established: G6PD Mediterranian, G6PD Corinth, G6PD Seattle and G6PD Ohut II.

G6PD Ciudad de la Habana: a new slow variant with deficiency found in a Cuban family

A new G6PD variant has been detected in a Cuban male and there is no evidence of associated hematological abnormalities. The main characteristics of this variant, moderate deficiency, slow electrophoretic mobility, increased utilization of the substrate analogues, and a different chromatographic behavior, indicate that it is a variant that has not been previously described.

Purification and properties of glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase from a methanol-utilizing yeast, Candida boidinii

Glucose-6-phosphate dehydrogenase (D-glucose-6-phosphate: NADP oxidoreductase, EC 1.1.1.49) and 6-phosphogluconate dehydrogenase (6-phospho-D-gluconate: NADP oxidoreductase, EC 1.1.1.44) were purified approx. 1700 fold and 330 fold, respectively, from Candida boidinii grown on methanol. The final enzyme preparations were homogeneous as judged by polyacrylamide gel electrophoresis. The molecular weights of the enzymes were estimated to be 118 000 and 110 000, respectively. Both enzymes are composed of two probably identical subunits and the molecular weights of the polypeptide chains were calculated to be 61 000 and 58 000, respectively. From a consideration of enzyme activities and types of inhibition by different metabolites the role of these two enzymes in glucose- and methanol-metabolism is discussed.

Parallel occurrence of oxidant-sensitivity and decreased inhibition by NADPH in G-6-PD Lublin and G-6-PD Poxnań

In two studied variants of G-6-PD without chronic hemolysis in probands, sensitivity of enzymes to inhibition by NADPH was decreased. Ki for NADPH was 28 micronM in Gd Lublin and 19 micronM in Gd Poznań. Susceptibility to the oxidant-induced hemolysis was described in probands, as well as in patients hemizygous for two other variants of G-6-PD with increased Ki for NADPH. It is suggested that in these cases, the oxidant-induced hemolysis is aggravated by their inability to counteract the drop in NADPH concentration with an increase in G-6-PD activity.

Genetic variants of human erythrocyte glucose 6-phosphate dehydrogenase: new variants in West Africa characterized by column chromatography

Five electrophoretically slow-moving genetic variants of glucose 6-phosphate dehydrogenase are described: four are from Nigeria and one is from Togo. All variants have normal or moderately reduced activity, and they are not associated with adverse clinical or haematological manifestations. Three variants have been fully characterized and are different from all previously described ones. Two variants have been partially characterized and at least one of them is also probably new. The overall population incidence of sporadic variants of G6PD in the Nigerian population is 0-3%. In the course of this study a previously described ion-exchange chromatographic technique for the characterization of G6PD variants has been extensively evaluated. Data are given on ten different variants to demonstrate the high resolving power of this technique.

Hemolytic anemia and G6PD deficiency

The severity of enzyme deficiency often does not correlate well with the clinical severity of genetic diseases. Thus, some G6PD variants associated with severe enzyme deficiency, such as Union and Markham, cause no hemolytic problem, while some variants associated with less severe deficiency, such as Manchester, Alhambra, and Tripler, cause chronic hemolytic anemia. The kinetic characteristics of these variant enzymes have not explained the discrepancy. However, examination of the normal and variant enzymes under simulated physiologic conditions, with the effects of various intermediate metabolites and co-enzymes in red cells being taken into consideration, reveal that the G6PD's from hemolytic variant subjects are strongly inhibited by a physiologic concentration of NADPH because of their high Michaelis constant for NADP or low inhibition constant for NADPH, and they are more sensitive to inhibition by ATP. These variant enzymes cannot generate enough NADPH in red cells to maintain an adequate concentration of reduced glutathione. The nonhemolytic variant enzymes are far less sensitive to the inhibition by NADPH because of their low Michaelis constant for NADP and high inhibition constant for NADPH. The physiologic activity of these nonhemolytic variant enzymes is estimated to be more than 30 percent of the activity of the normal G6PD, and this activity is adequate to maintain the red cells unhemolyzed.

Kinetic studies on the two common inherited forms of human erythrocyte adenylate kinase

1. The kinetic properties of two genetic variants of human erythrocyte adenylate kinase were studied at limiting concentrations of both ADP and MgADP(-) in the forward direction and at limiting concentrations of both AMP and MgATP(2-) in the reverse direction. 2. Primary reciprocal plots rule out the possibility of a Ping Pong mechanism for both forms of the enzyme. 3. Analysis of the kinetic data by an appropriate computer program gave the following K(m) values for the type 1 enzyme: AMP, 0.33mm+/-0.1; MgATP(2-), 0.95mm+/-0.13; ADP, 0.12mm+/-0.03; MgADP(-), 0.22mm+/-0.04. Values for the type 2 enzyme were: AMP, 0.27mm+/-0.03; MgATP(2-), 0.40mm+/-0.05; ADP, 0.08mm+/-0.07; MgADP(-), 0.20mm+/-0.04. 4. Product inhibition studies were done by studying the reverse reaction. With ADP as product inhibitor competitive inhibition patterns were obtained with AMP and/or MgATP(2-) as variable substrate. Similar results were obtained for product inhibition by MgADP(-) with AMP as variable substrate. The results are consistent with a Rapid Equilibrium Random mechanism. 5. Secondary plots of slope versus product concentration were linear. The data were fitted to the appropriate equation and analysed by computer to give values for the product inhibition constants. 6. Differences between the values of certain kinetic constants for the two forms of the enzyme were observed.

Genetic variants of glucose 6-phosphate dehydrogenase from human erythrocytes: unique properties of the A - variant isolated from "deficient" cells

The A(-) type of glucose 6-phosphate dehydrogenase (EC 1.1.1.49) has been isolated from human erythrocytes deficient in this enzyme. The specific activity of the purified protein is similar to that previously reported for the enzyme isolated from normal, nondeficient erythrocytes. During the purification procedure, a portion of the A(-) enzyme converts spontaneously, from the native "fraction I", to a "fraction II" having different kinetic and chromatographic properties. The conversion of fraction I to II can be reproduced freely by treatment with iodosobenzoate, and fraction II can be converted back to fraction I by treatment with dithioglycol. We suggest that fraction II is an enzyme species in which one or more sulfhydryl groups have been oxidized to disulfide(s). The tendency to oxidation appears to be a property specific to the A(-) variant and may represent the basis for its rapid rate of inactivation and consequent deficiency in vivo.