NADPH, not glutathione, status modulates oxidant sensitivity in normal and glucose-6-phosphate dehydrogenase-deficient erythrocytes

Glucose-6-phosphate dehydrogenase (G6PD) deficiency is characterized by the loss of NADPH and enhanced erythrocyte oxidant sensitivity. Historically, it has been theorized that the elevated oxidant sensitivity of G6PD-deficient erythrocytes arises as the direct consequence of decreased intracellular glutathione (GSH) concentrations. To directly investigate the basis of G6PD deficiency oxidant sensitivity, the effects of altered GSH and NADPH concentrations were examined in normal and G6PD-deficient erythrocytes. The results of this study demonstrated that GSH depletion, by 1-chloro-2,4-dinitrobenzene (CDNB), had no effect on hemoglobin oxidation in response to hydrogen peroxide (H2O2) generating systems (phenazine methosulfate and menadione bisulfite) in either normal or G6PD-deficient cells. Furthermore, a fourfold to sixfold increase in intracellular GSH concentration also did not protect against H2O2-generating systems in the normal or G6PD-deficient erythrocytes. Conversely, introduction of an NADPH-generating system (purified G6PD) into G6PD-deficient cells resulted in a significant decrease in oxidant sensitivity and an ability to cycle GSH. Further experiments demonstrated that the reduced oxidant sensitivity of the G6PD-reconstituted erythrocytes was not due to the maintenance of GSH levels because CDNB-mediated depletion of GSH did not alter this protective effect. Analysis of these results demonstrated a direct correlation between NADPH, but not GSH, concentration and hemoglobin oxidant sensitivity.     

Intracellular ferriprotoporphyrin IX is a lytic agent

Human erythrocytes were treated with menadione to oxidatively denature hemoglobin and release ferriprotoporphyrin IX (ferriheme, FP) intracellularly. The high affinity of FP for chloroquine was used to detect its release. After incubation for 1 hr at 37 degrees C and pH 7.4 with 0.5 mM menadione, erythrocytes bound 14C-chloroquine with an apparent dissociation constant of 10(-6)M. Untreated erythrocytes did not bind chloroquine with high affinity. At a chloroquine concentration in the medium of 2 microM, for example, menadione-treated erythrocytes bound 70 mumole chloroquine/kg and untreated erythrocytes bound 13.4 mumole/kg. The intracellular location of FP released by menadione was verified by finding that Tween 80 did not prevent chloroquine binding. By contrast, Tween 80 inhibited the binding of chloroquine to erythrocytes treated with extracellular FP. The hemolytic response to menadione was characteristic of the hemolytic response to FP. Thus, 5 microM chloroquine caused hemolysis to increase to 60% from baseline values of 5% in experiments using erythrocytes treated either with 0.5 mM menadione or with 5 microM FP; and, in both cases, the potentiating effect of chloroquine was inhibited by 1 microM mefloquine or 10 microM quinine. Higher concentrations of menadione caused hemolysis in the absence of chloroquine. We conclude that FP released by menadione exists intracellularly in a form that is accessible to bind chloroquine and to express its lytic activity.     

Haemolytic Effect of Two Sulphonamides Evaluated by a New Method

A technique to investigate drugs which could cause haemolysis in subjects deficient in glucose-6-phosphate dehydrogenase (D-glucose-6-phosphate: NADP oxidoreductase; G6PD) has been developed. The method is based on the technique of 14CO2 evolution during the incubation of normal erythrocytes in the presence of [I-14C]glucose and their own serum, the latter containing the active metabolites of the drugs received by normal subjects. By this method agents causing a stimulation of the hexosemonophosphate pathway of normal erythrocytes should be regarded as potentially haemolytic for G6PD-deficient subjects. Two sulphonamides, sulphormethoxine and sulphalene, of which until now no haemolytic effects have been reported, together with chloroquine, have been investigated. While chloroquine does not affect the hexosemonophosphate shunt of normal erythrocytes, the two sulphonamides stimulate this pathway. THE RESULTS ARE CONFIrmed by the reduction of the half-life of 51Cr-labelled G6PD-deficient red cells (Mediterranean variant), after administration of the two sulphonamides.     

Favism: disordered erythrocyte calcium homeostasis

The biochemical events that take place during acute hemolysis of G6PD- deficient subjects in favism are far from being elucidated. Evidence is here reported for a constantly and heavily disordered calcium homeostasis in the erythrocytes from seven favic patients. The abnormality, ie, a significantly impaired calcium ATPase activity and a parallel marked increase of intracellular calcium levels, was characteristic of the acute hemolytic crisis although unrelated to the attendant reticulocytosis. Concomitantly, a remarkable decrease of intracellular potassium was also observed. The mean +/- SD Ca2+-ATPase activity in the favic patients was 20.8 +/- 7.8 mumol Pi/g Hb/h compared with 37.2 +/- 8.5 in the matched controls represented by 12 healthy G6PD-deficient subjects (P less than .001). The mean +/- SD intraerythrocytic calcium content was 288 +/- 158 mumol/L of erythrocytes in the favic patients as compared with 22.0 +/- 8.2 in the G6PD-deficient controls (P less than .001). The intraerythrocytic potassium content was 76.6 +/- 19.3 mmol/L of erythrocytes in the favic patients and 106.6 +/- 8.2 in the G6PD-deficient controls (P less than .001). In vitro incubation of normal and G6PD-deficient erythrocytes with divicine, a pyrimidine aglycone present in fava beans and strongly implicated in the pathogenesis of favism, reproduces most of these events, including drop of calcium ATPase, increased intracellular calcium, and leakage of erythrocyte potassium.     

Mediterranean glucose 6-phosphate dehydrogenase (G6PD) deficiency--near normal decay of the mutant enzyme protein in circulating erythrocytes

Complete removal of leucocytes and platelets from erythrocytes and the development of a sensitized procedure for the assay of G6PD activity allowed the biochemical mechanisms of the Mediterranean variety of G6PD deficiency to be re-evaluated. Activity in the young erythrocytes from 9 G6PD-deficient subjects averaged 0.1% of the levels observed in the corresponding erythrocyte fraction from normal individuals: moreover, the decline of activity during aging of the G6PD-deficient erythrocytes was comparable with that observed for the normal enzyme. Mutant G6PD purified from granulocytes of a G6PD-deficient subject and entrapped within the corresponding erythrocytes was remarkably stable. Exposure of native erythrocytes to an oxidative stress (divicine plus ascorbate) resulted in a decrease of G6PD activity that was significantly more rapid and extensive in control than in G6PD-deficient cells. These results seem to exclude enhanced intracellular breakdown of the mutant protein within the circulating erythrocytes.     

Mediterranean glucose 6-phosphate dehydrogenase (G6PD) deficiency – Near normal decay of the mutant enzyme protein in circulating erythrocytes

Complete removal of leucocytes and platelets from erythrocytes and the development of a sensitized procedure for the assay of G6PD activity allowed the biochemical mechanisms of the Mediterranean variety of G6PD deficiency to be re-evaluated. Activity in the young erythrocytes from 9 G6PD-deficient subjects averaged 0.1% of the levels observed in the corresponding erythrocyte fraction from normal individuals: moreover, the decline of activity during aging of the G6PD-deficient erythrocytes was comparable with that observed for the normal enzyme. Mutant G6PD purified from granulocytes of a G6PD-deficient subject and entrapped within the corresponding erythrocytes was remarkably stable. Exposure of native erythrocytes to an oxidative stress (divicine plus ascorbate) resulted in a decrease of G6PD activity that was significantly more rapid and extensive in control than in G6PD-deficient cells. These results seem to exclude enhanced intracellular breakdown of the mutant protein within the circulating erythrocytes.     

Oxidant damage to erythrocyte membrane in glucose-6-phosphate dehydrogenase deficiency: correlation with in vivo reduced glutathione concentration and membrane protein oxidation

Chronic nonspherocytic hemolytic anemia has been observed in a recently described glucose-6-phosphate dehydrogenase (G6PD) variant, G6PDWayne. The mechanical properties of these erythrocytes and other G6PD variants were examined. The deformability of G6PD-deficient erythrocytes was normal, as determined by osmotic scan ektacytometry, and was not significantly affected by hemolytic crisis. In the common varieties of G6PD deficiency, the mechanical stability of the red blood cell (RBC) membrane was greater than normal, but G6PDWayne membranes were abnormally susceptible to shear-induced fragmentation. There was no evidence for a concurrent genetic defect in spectrin, because self- association constants and tryptic digests were normal. The fragility of G6PDWayne membranes appeared to be a consequence of oxidative damage to membrane thiol groups associated with a low glutathione (GSH) level in these RBCs. Associations among GSH level, thiol oxidation, and membrane instability were also found when a larger group of G6PD-deficient RBCs were examined. In normal erythrocytes, 1-chloro-2,4-dinitrobenzene was used to reduce GSH levels by 50%. Membrane thiol oxidation and membrane fragility both increased when these cells were kept at 4 degrees C for 3 to 5 days. Our findings suggest that chronic depletion of GSH leads to the destabilization of membrane skeleton through oxidation of membrane protein thiols.     

Erythrocyte membrane ATPase activity of G6PD-deficient individuals and the effect of primaquine metabolite(s) on membrane ATPase enzymes

Erythrocyte membrane Na+/K+, Ca2+/Mg2+ and Mg2+ ATPase activities in addition to the calmodulin-activated Ca2+/Mg2+ ATPase enzyme were measured in both G6PD-deficient and normal individuals. Although all three membrane ATPase activities were somewhat higher in the G6PD-deficient erythrocytes, only activated Ca2+/Mg2+ ATPase activity was significantly increased. The effect of primaquine on the membrane ATPases was also compared with other ATPase inhibitors. Primaquine was ineffective on erythrocyte membrane ATPase in-vitro. However, sera containing primaquine metabolite(s) were inhibitory to Ca2+/Mg2+ and Mg2+ ATPase systems of only G6PD-deficient erythrocytes. Other ATPase inhibitors showed a similar inhibitory effect in G6PD-deficient and normal erythrocytes, indicating a specific influence of primaquine on ATPase system in G6PD deficiency. It is suggested that this effect of primaquine may be an additional factor for haemolysis observed in the people with GdB- type of G6PD deficiency among the Mediterranean populations.     

Low membrane protein sulfhydrils but not G6PD deficiency predict ribavirin-induced hemolysis in hepatitis C

Hemolysis is a frequent adverse effect of ribavirin (RBV). It has been suggested that oxidative stress plays a role, but mechanisms and predictive risk factors for severe forms remain unknown. Markers of redox status were determined in erythrocytes of 34 patients with hepatitis C-four of them with glucose-6-phosphate-dehydrogenase (G6PD) deficiency-before and during treatment with RBV and interferon (IFN) and were compared with 10 healthy control subjects. In addition, erythrocytes were incubated with RBV, and the effects of dipyridamole (DPD), diethylmaleate (DEM), and glutathione ester (GSHE) were studied in vitro. Of the 30 patients without G6PD deficiency who were treated with RBV and IFN-alpha, five developed major hemolysis (Delta hemoglobin > 6 g/dL) and 25 developed minor hemolysis (Delta hemoglobin < 2.5 g/dL). Patients with major hemolysis had lower median pretreatment values of membrane protein sulfhydrils than patients with minor hemolysis (28.4 vs. 36.7 nmol/mg, P <.001). Erythrocytes of G6PD-deficient patients were not more susceptible to RBV-induced hemolysis. In in vitro incubations of erythrocytes, DEM enhanced the RBV-induced decrease of glutathione, protein sulfhydrils, and osmotic resistance. Supplementation of GSHE and DPD prevented the RBV-induced decrease in osmotic resistance, adenosyl triphosphate (ATP), and 2,3-diphosphoglycerate (DPG), the loss of glutathione and protein sulfhydrils, and the formation of thiobarbituric acid reactive substances (TBARs). In conclusion, the data indicate that low membrane protein sulfhydrils prior to therapy but not G6PD deficiency are predictive of RBV-induced major hemolysis. In vitro, GSHE and DPD reduce the RBV-associated oxidative stress in erythrocytes and prevent the increase in osmotic fragility, suggesting that these compounds might decrease the risk of hemolysis in patients.     

Expression of Plasmodium falciparum G6PD-6PGL in laboratory parasites and in patient isolates in G6PD-deficient and normal Nigerian children

As the production of NADPH in the pentose phosphate pathway is the main antioxidant defence mechanism available to the Plasmodium falciparum, we have studied the expression of P. falciparum glucose 6-phosphate dehydrogenase-6-phosphogluconolactonase (PfG6PD-6PGL) in G6PD-deficient and normal erythrocyte host cells. Both erythrocytes infected in vitro with a laboratory isolate and erythrocytes from natural human infections were used. Total RNA was prepared from parasites collected from five G6PD-deficient and nine G6PD-normal children in Ibadan, Nigeria, selected after screening 189 rural schoolchildren and 68 clinical malaria patients, and was subjected to Northern blot analysis. The probe was a cDNA fragment of the G6PD domain of the PfG6PD-6PGL gene, with an internal control probe of P. falciparum 18S ribosomal RNA. Quantification was performed using a phosphoimager. Relative to internal control, the abundance of PfG6PD-6PGL mRNA (mean +/- standard deviation) was lower in parasites from G6PD-deficient children (0.29 +/- 0.27) than in G6PD-normal control subjects (0.74 +/- 0.26) (P = 0.014, Mann-Whitney U-test). Although confirmation in a larger study is required, our results suggest a lower relative abundance of PfG6PD-6PGL, and presumably antioxidant activity, in malaria parasites from G6PD-deficient hosts, thus extending the current knowledge of the mechanism of G6PD-deficiency related host protection.     

The effect of BCNU and adriamycin on normal and G6PD deficient erythrocytes

In cell free systems Adriamycin induces oxygen radicals. We have shown previously that Adriamycin generates peroxide in human erythrocytes. BCNU, by inhibiting glutathione reductase, interferes with the major erythrocyte pathway to degrade peroxide. In this investigation we looked at interactions of these drugs with normal and abnormal erythrocytes. In G6PD-deficient erythrocytes Adriamycin posed a significant oxidant stress as demonstrated by hexose monophosphate shunt (HMPS) activity, hydrogen peroxide (H2O2) production, and glutathione depletion. At similar molar concentrations Adriamycin was a stronger oxidant than acetylphenylhydrazine. BCNU=treated normal erythrocytes showed an enhanced susceptibility to oxidant stress as demonstrated by a lack of HMPS response to H2O2 and glutathione depletion during incubations with Adriamycin. The HMPS shunt of BCNU treated RBC was intact as shown by their nearly normal response to methylene blue stimulation. These BCNU studies also demonstrated the inability of H2O2 to react directly with NADPH. In conclusion Adriamycin poses a potent oxident stress to G6PD-deficient erythrocytes. BCNU promotes enhanced susceptibility of normal RBC to oxidant stress and BCNU can act as a probe to define drug interactions with the HMPS. These studies add to a growing body of evidence postulating the importance of oxygen radicals in the therapeutic and/or effects of Adriamycin.     

Haemolysis in a G6PD-deficient child induced by eating unripe peaches

A child suffering from G6PD deficiency developed a severe haemolytic crisis without an apparent trigger. The possible pathogenetic role of the ingestion of unripe peaches was studied biochemically in this anaemia. We show that an extract from the unripe peach exerts an oxidative challenge on normal as well as on asymptomatic G6PD-deficient erythrocytes. This effect is analogous to that of the favism-inducing agents. The effect of the extract on the patient's red blood cells was more pronounced than on other asymptomatic G6PD-deficient erythrocytes, particularly during his haemolytic crisis. The chemical nature of the deleterious component was not identified. It is suggested that unripe peaches be added to the list of hazards for G6PD-deficient subjects in combination with other factors.     

Favism: Erythrocyte Metabolism during Haemolysis and Reticulocytosis

The reduced activity of glucose-6-phosphate dehydrogenase (D-glucose-6-phosphate; NADP+ 1-oxidoreductase; G6PF) in Mediterranean erythrocytes explains the precarious equilibrium of the hexose monophosphate pathway (HMP) and the susceptibility of these cells to haemolytic agents. G6PD-deficient erythrocytes, in steady-state conditions, have a low NADPH/NADP+ ratio, thus allowing the HMP to operate at its maximal intracellular rate and to compensate the intrinsic erythrocyte enzyme deficiency. Studies started soon after accidental intake of fava beans by sensitive G6PD-deficient subjects demonstrate a decrease of both NADPH/NADP+ ratio and reduced glutathione. The metabolic effects induced by fava beans may be attributed to oxidative stress probably associated with an inhibitor effect of some unknown metabolite on the HMP. The availability of erythrocytes from subjects recovering from haemolysis with high reticulocyte counts and increased G6PD activity, provides new information on the rate of synthesis as well as on the in vivo decay of the mutant enzyme. Correlation of G6PD activity to reticulocyte count and extrapolation to an ideally homogenous population of reticulocytes reveal that the mutant enzyme is synthesized at a nearly normal rate. Furthermore, the present correlation allows an estimate of the in vivo half-life of Mediterranean G6PD. The rate of decline of about 8 d observed in this study well correlates to the intracellular metabolic aspects of G6PD Mediterranean erythrocytes.     

Posttransfusional Hemolysis in Recipients of Glucose-6-Phoiphate Dehydrogenase-Deficient Erythrocytes

To test the hypothesis that transfusion of blood donated by individuals with glucose-6-phosphate dehydrogenease (G6PD) deficiency may result in a hemolytic reaction, we conducted a prospective longitudinal study in which 10 patients transfused with 1 unit of G6PD-deficient and 1 unit of normal red blood cells (RBC) were compared with 10 patients transfused with 2 units of age-matched normal RBC. We found that 24 h after transfusion serum bilirubin (μmol/1) in the recipients of G6PD-deficient RBC was significantly higher than in the recipients of normal RBC (36±14 vs. 18±5, respectively, p>0.004). A parallel increase was found in the serum lactate dehydrogenase (LDH; IU/1) between the two groups (378±151 vs. 264±56, p<0.001). The difference in serum bilirubin(26±10 vs. 15±5, p<0.03) was still noted 48 h after transfusion, with only a marginal difference (p<0.08) in LDH. We conclude that an immediate posttransfusional hemolytic reaction can occur in recipients of G6PD-deficient RBC and therefore suggest that the differential diagnosis of posttransfusional hemolysis, particularly in populations where G6PD deficiency is prevalent, includes transfusion of erythrocytes from G6PD-deficient donors.     

Posttransfusional Hemolysis in Recipients of Glucose-6-Phoiphate Dehydrogenase-Deficient Erythrocytes

To test the hypothesis that transfusion of blood donated by individuals with glucose-6-phosphate dehydrogenease (G6PD) deficiency may result in a hemolytic reaction, we conducted a prospective longitudinal study in which 10 patients transfused with 1 unit of G6PD-deficient and 1 unit of normal red blood cells (RBC) were compared with 10 patients transfused with 2 units of age-matched normal RBC. We found that 24 h after transfusion serum bilirubin (μmol/1) in the recipients of G6PD-deficient RBC was significantly higher than in the recipients of normal RBC (36±14 vs. 18±5, respectively, p>0.004). A parallel increase was found in the serum lactate dehydrogenase (LDH; IU/1) between the two groups (378±151 vs. 264±56, p<0.001). The difference in serum bilirubin(26±10 vs. 15±5, p<0.03) was still noted 48 h after transfusion, with only a marginal difference (p<0.08) in LDH. We conclude that an immediate posttransfusional hemolytic reaction can occur in recipients of G6PD-deficient RBC and therefore suggest that the differential diagnosis of posttransfusional hemolysis, particularly in populations where G6PD deficiency is prevalent, includes transfusion of erythrocytes from G6PD-deficient donors.     

Toxic heme in sickle cells: an explanation for death of malaria parasites

In an effort to elucidate a mechanism of genetic resistance to malaria, we asked whether a toxic form of heme is included in the excess of ferriprotoporphyrin IX (FP) which has been reported to accumulate as hemichromes in sickle cells. When FP is bound to certain erythrocytic elements, such as native hemoglobin, it is inaccessible to bind chloroquine with high affinity and is nontoxic. However, when FP is accessible to bind chloroquine with high affinity, it has been demonstrated to be sufficiently free to have membrane toxicity and, under certain conditions, to lyse Plasmodium falciparum parasites. [14C]-chloroquine was used, therefore, as a reporter molecule to evaluate the quantity, accessibility, and potential toxicity of FP released from hemoglobin. Intact erythrocytes from subjects with sickle cell anemia bound approximately 71 mumoles of chloroquine per kg with an apparent Kd of 10(-6) M. Erythrocytes from normal subjects or subjects with sickle trait bound little or no chloroquine with high affinity. Since the oxidant stress introduced by malaria parasites would increase the tendency for denaturation of hemoglobin S with additional release of FP, we suggest that FP toxicity accounts for the death of malaria parasites in sickle cells.     

Glucose-6-phosphate dehydrogenase deficiency protects against coronary heart disease

Summary  Previous studies suggest a reduction in cardiovascular risk among subjects expressing the glucose-6-phosphate dehydrogenase (G6PD, EC 1.1.1.49) deficient phenotype. We aimed to test this hypothesis in male subjects expressing the G6PD-deficient phenotype vs wild type G6PD. In a case–control study we examined consecutive patients admitted for acute myocardial infarction or unstable angina, and controls admitted for diagnoses other than coronary heart disease (CHD). The G6PD phenotype was determined by measuring the enzyme activity in erythrocytes, as the absorbance rate change due to NADPH reduction. The CHD risk associated with the G6PD phenotype was assessed with unconditional logistic regression. G6PD-deficient subjects were less frequently represented among cases (11.8%) than among controls (18.6%, p=0.002). The genetic condition of G6PD deficiency conveyed a significant reduction in CHD risk (OR=0.6; 95% CI 0.4 to 0.9). We confirm the hypothesis that subjects with the G6PD-deficient phenotype are less prone to CHD. We suggest that such a protective effect may be ascribable to a reduced 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMG-CoA R) activity, a statin-like effect, as well as to a downregulation in NADPH oxidase activity with a consequent reduction in oxygen-free radical production. Competing interests: None declared References to electronic databases: Glucose-6-phosphate dehydrogenase, EC 1.1.1.49.     

Synergistic enhancement of a copper chelator,bathocuproine disulphonate,and cysteine on in vitro growth of Plasmodium falciparum in glucose-6-phosphate dehydrogenase-deficient erythrocytes

In vitro growth of Plasmodium falciparum is restricted in glucose-6-phosphate dehydrogenase (G6PD)-deficient erythrocytes (RBC), as a result of oxidative stress. Bathocuproine disulphonate (BCS), a copper chelator, as well as cysteine have been shown to synergistically stimulate the in vitro growth of various mammalian cells and Trypanosoma under oxygenated conditions. We examined the effects of these two chemicals on the in vitro growth of P. falciparum in G6PD-deficient RBC, and found that addition of BCS and cysteine synergistically enhanced the growth of the P. falciparum FCR-3 strain in these RBC to the same level as in normal RBC. However, BCS or cysteine alone had no stimulatory effect. To explain this synergistic enhancement, changes in thiol, NADPH and glutathione contents were investigated. After addition of cysteine alone, thiol content in the medium decreased rapidly, but when BCS was added, it was maintained at about 35% at 24 hours after incubation, suggesting that BCS stimulates parasite growth in G6PD-deficient RBC by inhibiting copper-mediated oxidation of cysteine in the medium. In these RBC, no increase in NADPH level, but a slight increase in glutathione, was observed in the presence of both BCS and cysteine. The slight increase of glutathione, was probably due to incorporation of cysteine from the medium, although this could not fully explain the synergistic growth enhancement. These findings taken together suggest that cysteine incorporated into G6PD-deficient RBC may help maintain the thiol groups in many proteins, such as membrane proteins, hemoglobin and enzymes, and plays an important role in maintaining an appropriate culture state necessary for parasite growth. We also examined the effects of BCS and cysteine on adaptation of wild isolates of P. falciparum to in vitro cultivation using the candle jar method. Although there was no drastic effect on growth enhancement, the presence of BCS and cysteine accelerated the appearance of schizonts in many isolates.     

Erythrocyte glucose-6-phosphate dehydrogenase and pyruvate kinase activities in hemoglobin H disease.

Erythrocyte glucose-6-phosphate dehydrogenase (G6PD) and pyruvate kinase (PK) activities were studied in hemoglobin H (HbH) patients by spectrophotometric method, cytochemical method and the methemoglobin reduction (MR) test for the detection of heterozygous G6PD deficiency. G6PD deficiency was found in 7 of 64 cases (10.9%), including 3 cases of genotype alpha 1/alpha 2 and 4 cases of genotype alpha 1/CS. None of the HbH patients was found to be PK-deficient. Spectrophotometrically determined G6PD and PK activities were significantly higher in HbH patients than in normals (p less than 0.001), whereas the MR test yielded a significantly lower percentage of residual methemoglobin in HbH patients than in normals (p less than 0.05). All three methods were efficient in the detection of hemizygous G6PD deficiency in HbH patients, but not in G6PD-deficient females.