Glucose-6-phosphate dehydrogenase cytochemistry using a copper ferrocyanide method and its application to rapidly frozen cells

We describe an improved copper ferrocyanide-based method for cytochemical detection of glucose-6-phosphate dehydrogenase (G6PD), which was used to localize the enzyme within the ultrastructure of rat hepatocytes and adrenocortical cells. With this method, glutaraldehyde fixation and the addition of exogenous electron carriers (for example, phenazine methosulfate) to the cytochemical reaction medium were essential. Copper ferrocyanide reaction product showing the distribution of G6PD was readily recognized at the light microscopic level as Hatchett’s brown staining and at the electron microscopic level as electron-dense deposits. Within stained regions, enzyme cytochemical G6PD activity was found to be associated with ribosome-like structures. Because G6PD is a soluble, cytosolic enzyme, its displacement or extraction may occur during conventional fixation. We, therefore, combined a rapid-freezing technique with G6PD enzyme cytochemistry. The resultant rapid-freezing enzyme cytochemistry enabled us to show the subcellular distribution of G6PD in a more life-like state; the localization of G6PD in rapidly frozen cells was in substantial agreement with that in conventionally fixed cells. Accepted: 14 July 1999     

Enhanced resolution of histochemical distribution of glucose-6-phosphate dehydrogenase activity in rat neural tissue by use of a semipermeable membrane

We examined the histochemical distribution of glucose-6-phosphate dehydrogenase (G6PD) activity in neural tissue using different diffusion barriers. Although polyvinyl alcohol and agar overlays permitted regional localization of G6PD, a semipermeable membrane revealed cellular differences in G6PD activity within populations of neurons. Distribution of G6PD activity in selected regions of the nervous system was examined using the membrane technique. White matter usually exhibited strong G6PD activity. The neuronal somata of the dorsal root ganglia (L4-L6) and anterior horns of the spinal lumbar enlargement demonstrated a variation in activity which was independent of somal size. Satellite cells showed intense activity when the membrane technique was used. Hippocampal pyramidal and granular cells of the dentate gyrus exhibited moderate, uniform G6PD activity, but only weak activity was seen in hippocampal and dentate molecular layers. High levels of activity were observed in the vascular endothelial cells of the brain, spinal cord, and choroid plexus, and in the ependymal cells of the spinal central canal and ventricles of the brain. The superior vestibular nucleus appeared to have little G6PD activity in either the neuron cell bodies or the surrounding parenchyma. The use of a semipermeable membrane for localization of G6PD activity in neural tissues permits enhanced resolution of neuron elements and may provide a more accurate assessment of G6PD activity in histological preparations.     

Association of red cell glucose-6-phosphate dehydrogenase with haemoglobinopathies

A total of 1,112 randomly selected Saudi Arabs, of both sexes, living in Jeddah and the surrounding areas were screened for the phenotypic distribution of red cell glucose-6-phosphate dehydrogenase (G6PD) and 6-phosphogluconate dehydrogenase (6PGD). They were also investigated for haemoglobin and for thalassaemia. Phenotyping of the haemoglobins and the red cell enzymes was carried out by starch gel electrophoresis and the dye-decolouration screening test, while the investigation for thalassaemia was carried out by globin-chain biosynthesis, followed by column chromatography. The red cell Gd- alleles were significantly associated with the sickle-cell gene in both the males (chi 2(1): AS-28.80; SS-4.89) and females (chi 2(1): AS-10.99; SS-13.16). A similar association was also observed between G6PD deficiency and thalassaemias in males (chi 2(1): alpha-thalassaemia - 3.13; beta-thalassaemia - 11.06) and females (chi 2(1): alpha-thalassaemia - 6.63). However, no such association was detected between red cell 6PGD types and haemoglobin genes. The results suggest that the red cell G6PD deficiency, sickle-cell and thalassaemia genes might have evolved as a result of the same ecological factor, probably malaria.     

Identification of glucose 6-phosphate dehydrogenase deficiency in a population with a high frequency of thalassemia

High frequencies of both thalassemia trait (5.2%) and glucose 6-phosphate dehydrogenase (G6PD) deficiency for only males (1.3%) have been observed in the Calabrian population. The G6PD activity measurement was carried out on 1239 samples of whole blood from Calabrian subjects of both sexes (age range 10-55) by a differential pH-metry technique which was quite suitable to determine the G6PD deficiency in mass screenings. The analyzed subjects showed: only the thalassemia trait; or only the G6PD deficiency; or only the total iron serum deficiency; or G6PD deficiency associated with the thalassemia trait or with the total iron serum deficiency. The G6PD heterozygous subjects have an enzymatic activity which is masked by both the thalassemia trait and the total iron serum deficiency. In a population showing high frequencies of both thalassemia trait and G6PD deficiency, the comparison of G6PD activity of heterozygous subjects also affected with the thalassemia trait is more reliable if referred to the enzymatic activity of the carriers of the latter inherited anomaly rather than to G6PD activity of normal subjects.     

Alternative targeting of Arabidopsis plastidic glucose-6-phosphate dehydrogenase G6PD1 involves cysteine-dependent interaction with G6PD4 in the cytosol

Arabidopsis peroxisomes contain an incomplete oxidative pentose-phosphate pathway (OPPP), consisting of 6-phosphogluconolactonase and 6-phosphogluconate dehydrogenase isoforms with peroxisomal targeting signals (PTS). To start the pathway, glucose-6-phosphate dehydrogenase (G6PD) is required; however, G6PD isoforms with obvious C-terminal PTS1 or N-terminal PTS2 motifs are lacking. We used fluorescent reporter fusions to explore possibly hidden peroxisomal targeting information. Among the six Arabidopsis G6PD isoforms only plastid-predicted G6PD1 with free C-terminal end localized to peroxisomes. Detailed analyses identified SKY as an internal PTS1-like signal; however, in a medial G6PD1 reporter fusion with free N- and C-terminal ends this cryptic information was overruled by the transit peptide. Yeast two-hybrid analyses revealed selective protein-protein interactions of G6PD1 with catalytically inactive G6PD4, and of both G6PD isoforms with plastid-destined thioredoxin m2 (Trx(m2) ). Serine replacement of redox-sensitive cysteines conserved in G6PD4 abolished the G6PD4-G6PD1 interaction, albeit analogous changes in G6PD1 did not. In planta bimolecular fluorescence complementation (BiFC) demonstrated that the G6PD4-G6PD1 interaction results in peroxisomal import. BiFC also confirmed the interaction of Trx(m2) with G6PD4 (or G6PD1) in plastids, but co-expression analyses revealed Trx(m2) -mediated retention of medial G6PD4 (but not G6PD1) reporter fusions in the cytosol that was stabilized by CxxC113S exchange in Trx(m2) . Based on preliminary findings with plastid-predicted rice G6PD isoforms, we dismiss Arabidopsis G6PD4 as non-functional. G6PD4 orthologs (new P0 class) apparently evolved to become cytosolic redox switches that confer thioredoxin-relayed alternative targeting to peroxisomes.     

Two commonly occurring nucleotide base substitutions in Chinese G6PD variants

Using a direct PCR sequencing technique, we have identified two DNA base substitutions in 8 different biochemical G6PD variants of Chinese origin. Neither one of these abnormalities has been reported in other ethnic groups. An abnormality (C1) of G to T substitution at cDNA 1376 causing an amino acid change from Arg to Leu has been found in 3 variants. Another abnormality (C2) of G to A substitution at cDNA 1388 causing an amino acid change from Arg to His has been found in 5 variants. Both C1 and C2 are located in exon 12 of the G6PD gene and are only 12 base pairs apart. However, C1 is associated with a significant increase in the deamino-NADP utilization rate, whereas C2 is not. Taken together, our data suggest that C1 and C2 are very common among Chinese with a G6PD deficiency and exon 12 may define an important functional domain of the human G6PD.     

云南阿昌族G6PD基因突变G487A在DF213中的表达

为获得阿昌族G6PDWT和G6PDG487A重组蛋白,研究G6PDG487A的结构和功能改变,从云南省德宏州梁河县杞木寨乡湾中村阿昌族聚集地的G6PD缺陷家系先证者和正常阿昌族个体全血提取RNA,经RT-巢式PCR得cDNA,将cDNA克隆至pMD18-Tsimple载体中并测序;错配碱基经定点突变修复后,目的基因亚克隆至pThioHis(A)载体,构建了阿昌族G6PD基因野生型和G487A突变型原核表达载体:pThioHis(A)-AChang-G6PDWT和pThioHis(A)-AChang-G6PDG487A。用重组质粒转化E.coli Competent Cells DF213(G6PD defeciency),经IPTG诱导G6PD表达、10%SDS-PAGE电泳检测表达蛋白和紫外340nm定量测定G6PD活性的分析表明,pThioHis(A)-AChang-G6PDWT和pThioHis(A)-AChang-G6PDG487A在DF213中成功表达,分子量约为59kDa。IPTG诱导0、3、6、9、和12h后,G6PD活性逐渐增高,G6PD基因WT表达的酶活性约是G487A的20-25倍。表达载体的构建以及G6PDcDNA在DF213中成功表达,为重组酶G6PDG487A的进一步研究奠定了基础。     

Functional expression of human glucose-6-phosphate dehydrogenase in Escherichia coli

The complete coding sequence for human glucose-6-phosphate-dehydrogenase (G6PD) was inserted downstream from the tac promoter of a plasmid, pJF118EH, which also carries the lacIq repressor gene. When Escherichia coli strains (that are unable to grow on glucose due to the absence of functional zwf (G6PD-) and pgi genes) were transformed with this plasmid (pAC1), they were able to grow on glucose as sole carbon source. The rate of growth on glucose was faster in the presence of the inducer of the tac promoter, isopropyl-beta-D-thiogalactopyranoside (IPTG). Extracts of the transformed cells contained a G6PD activity that was not detectable in the parental strains and that was inducible by IPTG. The G6PD activities from normal E. coli and from pAC1-transformed cells comigrated with human G6PD when subjected to electrophoresis on agarose gels. However, when denatured, the G6PD produced by pAC1 was, like the human enzyme, distinguishable from the E. coli-encoded enzyme on the basis of its immunoreactivity with antibody specific for human G6PD. Therefore, human G6PD can be expressed in E. coli and can function to complement the bacterial enzyme deficiency.     

The prevalence of factor V Leiden,prothrombin G20210A and methylenetetrahydrofolate reductase polymorphism C677T among G6PD deficient individuals from Western Iran

It has been suggested that the allele frequency of thrombophilic mutations is affected by glucose-6-phosphate dehydrogenase (G6PD) deficiency. The prevalence of thrombophilic mutations were studied in sixty G6PD deficient individuals including 57 males and three females with the mean age of 15 ± 3.08 and 110 age and sex matched healthy individuals consisted of 95 males and 15 females with the mean age of 16.19 ± 2.17 from the Kermanshah Province of Iran. Using a combination of PCR-RFLP technique, single strand conformation polymorphism (SSCP) analysis and DNA sequencing polymorphic G6PD mutations were identified. The factor V Leiden, prothrombin G20210A and methylenetetrahydrofolate reductase (MTHFR) C677T were detected by PCR-RFLP method using MnlI, HindIII and HinfI restriction enzymes, respectively. Three mutations, G6PD Mediterranean, G6PD Chatham and G6PD Cosenza were identified in 60 G6PD deficient individuals with highest prevalence of G6PD Mediterranean (91.6%). In G6PD deficient individuals the prevalence of factor V Leiden tended to be higher (5%) compared to healthy individuals (2.7%). The prevalence of prothrombin G20210A mutation in G6PD deficient individuals was 1.7%. However, in normal subjects the prevalence of this mutation was 2.7%. The frequency of T allele in G6PD deficient individuals were insignificantly higher (29.16%) than those in healthy individuals (26.8%). Our finding indicates that the prevalence of factor V Leiden, prothrombin G20210A and MTHFR C677T in G6PD deficient individuals is not statistically different compared to normal subjects and G6PD deficiency is not associated with these thrombophilic mutations in Western Iran.     

Suppression of interleukin-1 beta-induced nitric oxide production in RINm5F cells by inhibition of glucose-6-phosphate dehydrogenase

In rat pancreatic islets and insulin-producing cell lines, IL-1beta induces expression of inducible nitric oxide synthase and NO production leading to impairment of glucose-stimulated insulin release and decreased cell survival. NADPH is an obligatory cosubstrate for iNOS synthesis of NO. We hypothesized that IL-1beta stimulates an increase in activity of NADPH-producing enzyme(s) prior to NO production and that this increase is necessary for NO production. Using rat insulin-secreting RINm5F cells, we found that (1) IL-1beta caused a biphasic change in the NADPH level (increased by 6 h and decreased after prolonged incubation in the presence of 2 ng/mL IL-1beta); (2) IL-1beta stimulated increased activity of glucose-6-phosphate dehydrogenase (G6PD) in a time- and dose-dependent manner, and G6PD expression was increased by about 80% after exposure to 2 ng/mL IL-1beta for 18 h: (3) IL-1beta-stimulated NO production was positively correlated with increased G6PD activity; (4) IL-1beta did not cause any significant change in enzyme activity of another NADPH-producing enzyme, malic enzyme; (5) IL-1beta-induced NO production was significantly reduced either by inhibiting G6PD activity using an inhibitor of G6PD (dehydroepiandrosterone) or by inhibiting G6PD expression using an antisense oligonucleotide to G6PD mRNA; and (6) IL-1beta stimulated a decrease in the cAMP level. 8-Bromo-cAMP caused decreased G6PD activity, and the protein kinase A inhibitor H89 led to a increase in G6PD activity in RINm5F cells. In conclusion, our data show that IL-1beta stimulated G6PD activity and expression level, providing NADPH that is required by iNOS for NO production in RINm5F cells. Also, inhibition of the cAMP-dependent PKA signal pathway is involved in an IL-1beta-stimulated increase in G6PD activity.     

Genetic heterogeneity of glucose-6-phosphate dehydrogenase deficiency revealed by single-strand conformation and sequence analysis.

We have carried out a systematic study of the molecular basis of glucose-6-phosphate dehydrogenase (G6PD) deficiency on a sample of 53 male subjects from Calabria, in southern Italy. Our sequential approach consisted of the following steps: (1) Partial biochemical characterization was used to pinpoint candidate known variants. The identity of these was then verified by restriction-enzyme or allele-specific oligonucleotide hybridization analysis of the appropriate PCR-amplified fragment. (2) On samples for which there was no obvious candidate mutation, we proceeded to amplify the entire coding region in eight fragments, followed by single-strand conformation polymorphism (SSCP) analysis of each fragment. (3) The next step was M13 phage cloning and sequencing of those individual fragments that were found to be abnormal by SSCP. Through this approach we have identified the molecular lesion in 51 of the 53 samples. In these we found a total of nine different G6PD-deficient variants, five of which (G6PD Mediterranean, G6PD A-, G6PD Coimbra, G6PD Seattle, and G6PD Montalbano) were already known, whereas four are new (G6PD Cassano, G6PD Cosenza, G6PD Sibari, and G6PD Maewo). G6PD Mediterranean is the commonest variant, followed by G6PD Seattle. At least seven of the variants are present, at polymorphic frequencies, in the Calabria region, and some have a nonrandom distribution within the region. This study shows that the genetic heterogeneity of G6PD deficiency in Calabria, when analyzed at the DNA level, is even greater than had been anticipated from biochemical characterization. The sequential approach that we have followed is fast and efficient and could be applied to other populations.     

葡萄糖6-磷酸脱氢酶调控细胞周期蛋白E1和细胞周期蛋白依赖性激酶2促进细胞增殖及其在肾透明细胞癌中的预后价值

肾透明细胞癌(clear cell renal cell carcinoma,ccRCC)是一种转移率高、预后差的细胞代谢性疾病,对其有效诊疗及预后分子标志物的研究十分重要。葡萄糖6-磷酸脱氢酶(glucose 6-phosphatedehydrogenase, G6PD)在ccRCC中高表达,并提示患者不良预后,其促进ccRCC细胞增殖的分子机制有待进一步揭示。本研究发现,降低G6PD可抑制细胞周期G₁/S期转化并显著抑制ccRCC细胞增殖。G6PD可在细胞水平调控G₁/S期转化及增殖相关因子Cyclin D1,CDK4,CDK6,Cyclin E1和CDK2基因表达。TCGA数据库分析结果表明,ccRCC 中Cyclin D1,Cyclin E1 和 CDK2的mRNA 水平显著升高,而CDK4表达无明显差异,CDK6表达却显著降低。相关性分析结果显示,G6PD与Cyclin D1呈显著负相关(P<0.0001),G6PD与CDK4,CDK6之间无显著相关性(P>0.05),G6PD与Cyclin E1(P<0.0001)以及CDK2(P<0.05)显著正相关。进一步免疫组化检测结果表明,Cyclin E1和 CDK2在ccRCC肿瘤组织中表达显著升高。生存预后分析结果显示,Cyclin D1高表达提示ccRCC患者整体预后更为良好,CDK4和CDK6表达水平在ccRCC患者总生存率预测中无意义;而Cyclin E1和CDK2高表达均可提示ccRCC患者预后不良。进一步细胞水平检测发现,Cyclin E1、CDK2表达降低可显著逆转G6PD促进ccRCC细胞增殖的能力。综上,与增殖相关因子Cyclin D1,CDK4和CDK6相比,G6PD有可能通过促进Cyclin E1和CDK2表达升高而发挥促进 ccRCC肿瘤细胞增殖的作用,并且这3者的异常高表达有望成为ccRCC患者不良预后的独立生存预测因素。     

Purification and some properties of human placental glucose-6-phosphate dehydrogenase

Glucose-6-phosphate dehydrogenase was purified from human placenta using DEAE-Sepharose fast flow, 2',5'-ADP Sepharose 4B column chromatography, and chromatofocusing on PBE 94 with PB 74. The enzyme was purified with 62% yield and had a specific activity of 87 units per milligram protein. The pH optimum was determined to be 7.8, using zero buffer extrapolation method. The purified placental glucose-6-phosphate dehydrogenase gave two activity bands on native PAGE: one band, constituting about 3--5% of total activity, moved slower than the remaining 95%. Among the activity bands only the faster moving band gave a band on protein staining. The slower moving band, which probably corresponded to the higher polymeric form of the G6PD with high specific activity, was not seen on native PAGE due to insufficient protein for Coomassie brilliant blue staining. The observation of one band on SDS--PAGE with an M(r) of 54 kDa and a specific activity lower than expected, suggests the presence of both forms of the G6PD, the high polymeric form at low concentration and the inactive form at high concentration, in our preparation. Measuring the activities of placental glucose-6-phosphate dehydrogenase between 20 and 50 degrees C, the activation energy, activation enthalpy, and Q(10) were calculated to be 8.16 kcal/mol, 7.55 kcal/mol, and 1.57, respectively. It was found that human placental G6PD obeys Michaelis-Menten kinetics. K(m) values were determined using the concentration ranges of 20--300 microM for G6P and 10--200 microM for NADP(+). The K(m) value for G6P was 40 microM; the K(m) value NADP(+) was found to be 20 microM. Double-reciprocal plots of 1/Vm vs 1/G6P (at constant [NADP(+)]) and of 1/Vm vs 1/NADP(+) (at constant [G6P]) intersected at the same point on the 1/V(m) axis to give V(m) = 87 U/mg protein.     

Development of allelic discrimination assay to detect Mediterranean G6PD mutation and its linked inheritance with normal vision and/colorblindness loci for 4 generations among Egyptian and Emirati families

G6PD deficiency c563T is the most common inherent blood disease among the Mediterranean populations and its molecular diagnosis is critical as the enzyme assay fails for heterozygous individuals. The purpose of the study is to estimate the ubiquity of the heterozygous G6PD Med (c563T) variants among Egyptians and UAE nationals living in Dubai. We validated two molecular methods, polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) and qPCR allelic discrimination assay for detection of G6PD Med variants. Among 100 screened individuals, G6PD c563T variants are 30% of whom 15% are carriers. Sanger sequencing validated the qPCR discrimination assays. In search of a phenotypic marker to detect G6PD heterozygous variants, inheritance of G6PD locus and red-green color vision genes is studied in 1 Egyptian and 2 Emirati families. Among the 3 families, G6PD is polymorphic, displaying 4 phenotypes: in phenotype-1, person is normal, in phenotype-2 the person has no G6PD deficiency but with deuteranopia/deuteranomaly, in phenotype-3 the person is G6PD Med variant with deuteranopia/deuteranomaly and in phenotype 4 the person is G6PD Med variant has normal vision. Based on the molecular analysis of G6PD and Ishihara vision test it can be concluded that the two mutations at the two loci arose independent of each other without any interaction (epistatic effect) between them. Following the pedigree analysis of the two genes for 4 generations it is presumed that it is infeasible to use “deuteranopia /deuteranomaly” as a phenotypic marker to detect G6PD c563T heterozygous individuals among the Egyptian populations.     

Glucose-6-phosphate dehydrogenase (G6PD) Iserlohn and G6PD Regensburg: two new severe enzyme defects in German families

Two new inheritable variants of glucose-6-phosphate dehydrogenase have been found in two unrelated German families. Patients with one variant (G6PD Iserlohn, also referred to as G6PD I) suffered from intermittent hemolytic crises caused by fava beans; patients with the other variant (G6PD Regensburg, G6PD II) disclosed chronic nonspherocytic hemolytic anemia aggravated by drug treatment. Due to their unusual biochemical characteristics, the new variants were designated G6PD Iserlohn and G6PD Regensburg. Both variants showed a reduction of enzyme activity to about 6% of the normal in erythrocytes, normal electrophoretic mobility, increased affinity for glucose-6-phosphate, a reduced affinity for NADP and a pH optimum in the neutral region (7.0 and 7.5). G6PD Iserlohn had a decreased affinity for the inhibitor NADPH; G6PD Regensburg had a normal inhibitor constant. Deamino NADP was utilized at an increased rate by G6PD Regensburg. G6PD Iserlohn was thermostable, G6PD Regensburg mildly instable. G6PD activity in leukocytes was normal in G6PD Iserlohn and reduced to the same degree as in erythrocytets in G6PD Regensburg. The cause of the decreased activity of G6PD Iserlohn appears to be in vivo instability; in G6PD Regensburg further mechanisms might include reduced specific activity or reduced synthesis of the variant enzyme.     

Maternally transmitted severe glucose 6-phosphate dehydrogenase deficiency is an embryonic lethal

Mouse chimeras from embryonic stem cells in which the X-linked glucose 6-phosphate dehydrogenase (G6PD) gene had been targeted were crossed with normal females. First-generation (F(1)) G6PD(+/-) heterozygotes born from this cross were essentially normal; analysis of their tissues demonstrated strong selection for cells with the targeted G6PD allele on the inactive X chromosome. When these F(1) G6PD(+/-) females were bred to normal males, only normal G6PD mice were born, because: (i) hemizygous G6PD(-) male embryos died by E10.5 and their development was arrested from E7.5, the time of onset of blood circulation; (ii) heterozygous G6PD(+/-) females showed abnormalities from E8.5, and died by E11.5; and (iii) severe pathological changes were present in the placenta of both G6PD(-) and G6PD(+/-) embryos. Thus, G6PD is not indispensable for early embryo development; however, severe G6PD deficiency in the extraembryonic tissues (consequent on selective inactivation of the normal paternal G6PD allele) impairs the development of the placenta and causes death of the embryo. Most importantly, G6PD is indispensable for survival when the embryo is exposed to oxygen through its blood supply.     

A new improved electrophoretic method of glucose-6-phosphate dehydrogenase with discontinuous buffer and gel system

A new improved electrophoretic method of glucose-6-phosphate dehydrogenase (G6PD) was devised by using a slab disc electrophoresis with a discontinuous gel of 1-mm thickness. This was suitable for electrophoresis of thermolabile G6PD variants with low activities. Using this system, sharp G6PD bands and acceptable separations of G6PD Konan, G6PD Tokyo, and G6PD Yokohama, were obtained and compared with those of starch gel electrophoresis.     

Identification of HeLa cell glucose 6-phosphate dehydrogenase

Although the electrophoretic mobility of HeLa G6PD is similar to that of the common Negro variant G6PD A+, several reports have suggested slight differences between HeLa G6PD and G6PD A+. This study, carried out using the pure homogeneous B+, A+, and HeLa G6PD, showed that (1) the electrophoretic mobility of HeLa G6PD is identical to that of G6PD A+, (2) the enzymatic properties and thermostability of HeLa G6PD are indistinguishable from those of G6PD A+, and (3) the peptide map of the tryptic digest of HeLa G6PD is identical to that of G6PD A+, with one peptide spot of HeLa G6PD different from the corresponding spot of G6PD B+. These results indicate that the structure of HeLa G6PD is identical to that of G6PD A+, and that the amino acid substitution in HeLa G6PD is from one asparagine residue in the wild-type G6PD B+ to an aspartic acid residue in HeLa G6PD.This research was supported by research grant GM 15253 from the National Institutes of Health.