Development of a novel mouse model of severe glucose-6-phosphate dehydrogenase (G6PD)-deficiency for in vitro and in vivo assessment of hemolytic toxicity to red blood cells

Blood Cells, Molecules, and Diseases 47 (2011) 176–181

Contents lists available at ScienceDirect

Blood Cells, Molecules, and Diseases j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / y b c m d

Development of a novel mouse model of severe glucose-6-phosphate dehydrogenase (G6PD)-deficiency for in vitro and in vivo assessment of hemolytic toxicity to red blood cells Chun Hay Ko a, b, Karen Li b, Chung Leung Li c, Pak Cheung Ng b, Kwok Pui Fung a, Anthony Edward James d, Raymond Pui-On Wong b, Goldie Jia-Shi Gu b, Tai Fai Fok a,⁎ a

Institute of Chinese Medicine, The Chinese University of Hong Kong, Shatin, Hong Kong, SAR, China Department of Paediatrics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong, SAR, China c Institute of Cellular and Organismic Biology/Genomic Research Center, Academia Sinica, Taipei, Taiwan d Laboratory Animal Services Centre, The Chinese University of Hong Kong, Shatin, Hong Kong, SAR, China b

a r t i c l e

i n f o

Article history: Submitted 24 May 2011 Revised 29 June 2011 Available online 11 August 2011 (Communicated by Sir D. Weatherall, F.R.S., 12 July 2011) Keywords: G6PD-deficiency Mouse Oxidative stress Glutathione Methemoglobin

a b s t r a c t Studies of hemolytic agents on G6PD-deficient subjects have been extensively performed on red blood cells obtained from donors, only using in vitro methods. However, there has been no adequate G6PD-deficient animal model for in vivo assessment of potentially hemolytic agents. The objective of this study is to establish a novel mouse model of severe G6PD-deficiency, with high susceptibility to hemolytic damage upon oxidative agents. To create this model, G6PD mutant Gpdx allele was introduced into the C57L/J mouse strain background by breeding program. The hemolytic toxicity of naphthalene and its metabolite α-naphthol on G6PD-deficient red blood cells was evaluated. Our data showed that the F2 homozygous Gpdx mutant with C57L/J background exhibiting the G6PD activity was 0.9 ± 0.1 U/g Hb, level similar to those of G6PD deficiency in human. A significantly negative correlation was demonstrated between GSH percentage reduction and G6PD activity (r = − 0.51, p b 0.001) upon challenge of the red blood cells with alpha-naphthol in vitro. Similar correlation was also found between GSSG elevation and G6PD activity. Our in vivo studies showed that the administration of naphthalene at 250 mg/kg inflicted significant oxidative damage to the G6PD-deficient mice, as illustrated by the decrease of the GSH-to-GSSG ratio (by 34.2%, p = 0.005) and the increase of the methemoglobin level (by 1.9 fold, p b 0.001). Hemolytic anemia was also found in G6PD-deficient mice at this dosage of naphthalene. In summary, this novel mouse model could be utilized as a screening platform to more accurately determine the hemolytic toxicity of pharmacological agents on G6PD-deficient subjects. © 2011 Published by Elsevier Inc.

Introduction Glucose-6-phosphate dehydrogenase (G6PD) deficiency is the most common human enzymopathy [1]. It is estimated that over 400 million people are affected by this disorder worldwide [2]. G6PDdeficient subjects are vulnerable to oxidative stress. This predisposes them to chemical-induced hemolysis if exposed to pro-oxidative agents. Studies of hemolytic agents on G6PD-deficient subjects have been essentially performed on red blood cells obtained from donors, using only in vitro methods [3,4]. These investigations are limited by sample size and cannot accurately predict the in vivo response of G6PD-deficient subjects to metabolites of the test agents. However, there has been no G6PD-deficient animal model suitable for in vivo

⁎ Corresponding author at: Department of Paediatrics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong, SAR, China. Fax: + 852 26360020. E-mail address: [email protected] (T.F. Fok). 1079-9796/$ – see front matter © 2011 Published by Elsevier Inc. doi:10.1016/j.bcmd.2011.07.003

assessment of potentially hemolytic agents. Existing G6PD-deficient mutant animals are either insensitive to hemolytic agents due to residual G6PD activity [5], or exhibit severe/complete G6PD deficiency that are incompatible with survival during embryonic development [6]. The objective of this study was to establish a novel mouse model of severe G6PD-deficiency, with high susceptibility to hemolytic damage upon exposure to oxidative agents. We hypothesized that to create this model, we introduced mutant Gpdx allele (a severe ENU-induced mutation that results in 13–15% G6PD activities of wild type littermates) into the C57L/J background (a strain that constitutively exhibits low G6PD activity) to generate mice with G6PD activity b2 U/ g Hb, level similar to those of severe G6PD deficiency in human. The efficacy of this model would be validated by challenging their red blood cells (RBCs) with known hemolytic agents, naphthalene and its metabolite α-naphthol [4], in vitro and in vivo, respectively. The oxidative status of RBCs (reduced and oxidized glutathione and methemoglobin levels) would be monitored.

C.H. Ko et al. / Blood Cells, Molecules, and Diseases 47 (2011) 176–181

177

glutathione level was studied upon oxidative challenge by α-naphthol in vitro.

Materials and methods This study protocol was approved by Animal Experimentation Ethics Committee of the Chinese University of Hong Kong.

Preparation of red blood cells and G6PD activity determination Animals were anesthetized and their blood samples were collected in acid citrate-dextrose (Sigma; St. Louis, MO, USA). The blood samples were centrifuged at 1500 g for 5 min and the plasma together with the buffy coat layer was removed. Then, the RBCs were washed thrice with 9 g/L sodium chloride and resuspended into 40% suspension (v/v, in PBS), and kept at −20 °C for enzyme assays. All analysis was carried within a day of blood collection. G6PD levels of the RBCs were determined using a standard diagnostic kit (Trinity Biotech Plc, Co Wicklow, Ireland). This enzyme activity assay is based on the oxidation rate of glucose-6-phosphate to 6-phosphogluconate and the concurrent production of NADPH. The G6PD activities were normalized by hemoglobin per gram.

Animal The moderately G6PD-deficient mouse in the C3H strain was reconstituted from frozen embryos from the Medical Research Council (Harwell, UK). This mouse was originally created by Pretsch and Charles [7] and showed decreased translation of protein caused by a single mutation in the untranslated region in the splice site of the X-linked G6PD gene [7]. Female homozygotes (XmXm) with a C3H background were bred with male mice with C57L/J background obtained from Jackson Laboratories (Bar Harbor, ME). C57L/J mice naturally exhibited a relatively low G6PD activity (5.3–6.0 U/g Hb) because of the autosomal genetic regulation of Gdr-1 and Gdr-2 [8]. The erythrocytic G6PD activities and genotype of F1 littermates were determined after 8 weeks after birth. Among offspring of this breeding, 3 pairs of male and female mice with the lowest G6PD activities were selected as breeders for the F2 generation. Schematic diagram was shown in Fig. 1A. For the F2 littermates, according to the sex and their genotype, we divided them into 4 groups: homozygous mutant female (Xm/Xm), heterozygous mutant female (Xm/X), hemizygous mutant male (Xm/Y) and the male without mutant allele (X/Y). Each group consisted of 10 animals. All animals used in this study were 3 months old in order to allow full maturation of RBCs. The relationship between RBCs G6PD level and their

G6PD genotyping The presence of the G6PD mutant allele was identified by PCR method followed by restriction fragment length polymorphism (PCRRFLP) [5]. Briefly, DNA was prepared from biopsies of animal tails, and a 269-bp fragment of the G6PD gene was amplified. A 20 μL reaction mixture contained 1.5 U of Taq Polymerase, 1X Reaction buffer, 0.2 mM ddNTPs, 0.4 μM each primer (sense: GGAAACTGGCTGTGCGCTAC, antisense: TCAGCTCCGGCTCTCTTCTG), and 100 ng of genomic DNA. PCR was performed on a PTC-200 MJ Research Thermal Cycler (Bio-rad,

A

B

C

MW

1

2

3

4

5

500 200

100

Fig. 1. Establishing a novel severe G6PD-deficient mouse model. (A) Breeding program for model establishment. C57L/J male mice were intercrossed with G6PD homozygous female mutant (C3H background) to give F1 progeny. F2 progeny was obtained brother–sister cross of F1 littermates as described in Materials and methods. (B) RBCs G6PD activities of parental (C57L/J male: wild type, XY; G6PD mutant female: homozygotes, XmXm), F1 (male: hemizygotes, XmY; female: heterozygotes, XmX) and F2 (male: hemizygotes, XmY and wild type, XY; female; heterozygotes, XmX and homozygotes, XmXm) littermates were determined. RBCs G6PD activity of F2 homozygous female was the lowest among all studied groups. Data are the means from 10 animals in each group. *** P b 0.001 for difference in G6PD levels from the F2 homozygous female. (C) A typical PCR-RFLP genotyping result using allele specific endonuclease (Ddel) digestion. Lane 1: non-digested PCR amplicon control; lane 2: wild type male, XY; lane 3: hemizygous male, XmY; Lane 4: heterozygous female, XmX; lane 5: homozygous female XmXm. Size standards are shown in the left panel and the molecular weight (MW) was represented in base pair.

178

C.H. Ko et al. / Blood Cells, Molecules, and Diseases 47 (2011) 176–181

CA, USA) with the following profile: 94 °C for 1 min; 25 cycles of 95 °C for 30 s, 58 °C for 30 s and 72 °C for 30 s; and a final extension at 72 °C for 5 min. Twenty microlitres of PCR fragment was subsequently digested by adding 1 U of DdeI restriction endonuclease (Invitrogen; Carlsbad, CA, USA) at 37 °C for 16 h. The digested products were analyzed on 3% (w/v) agarose gel (NuSieve 3:1 Agarose; Cambrex Bio Science, Rockland, ME, USA) with ethidium bromide and visualized under UV. DdeI enzyme produced bands of 55-bp and 214-bp in wildtype mice but did not cleave the G6PD mutant sequence. Challenge of RBCs with α-naphthol in vitro and in vivo treatment The RBC suspension was pretreated with 1 mM sodium azide at 37 °C for 10 min to inhibit the catalase activity. After that, 200 μL of RBC suspension was incubated with 1 mg/mL of α-naphthol (precoated on incubation tubes) at 37 °C with shaking for 2 h. For in vivo treatment, F2 homozygous female littermates (Xm/Xm) with G6PD activity lower than 2 U/g Hb were employed as G6PD-deficient animals for drug challenge. Its genetically similar counterparts, F2 heterozygous female littermates (Xm/X) were used as control experiment. Our in vitro studies indicated that the oxidative responses between F2 heterozygous female and F2 male without mutant allele (X/Y) were similar, therefore, in order to simplify the screening process and breeding program, only F2 female groups were chosen for subsequent in vivo drug testing. Both groups of female animals were orally fed with two dosage of hemolytic agent, naphthalene (in corn oil; 125 mg/kg and 250 mg/kg). These two doses were shown to be non-toxic to F2 heterozygous female in our pilot studies. Control animals were administrated with same volume of corn oil. Animals were anesthetized and blood was collected and analyzed 24 h after drug treatment. Determination of GSH and GSSG by high performance liquid chromatography (HPLC) GSH and GSSG levels in blood samples were determined by HPLC according to a described method [9]. Each treatment sample (175 μL) was reacted with 17.5 μL N-ethylmaleimide (NEM, 310 mmol/L) for 5 s. The mixture was diluted with an equal volume (192.5 μL) of 5% (w/v) trichloroacetic acid (TCA) and centrifuged at 15,000 xg for 2 min. The excess NEM was extracted from the clear supernatant with 10 volumes of dichloromethane. The supernatant (200 μL) was mixed with 100 μL Tris–Cl 1 mol/L, pH 10.0 for alkalization. Each mixture was reacted with an equal volume of 1.5% (v/v) 2,4-dinitrofluorobenzene (FDNB) solution for 3 h at room temperature in the dark. After acidification with 20 μL (v/v) 37% HCl, the mixture was filtered and loaded onto the Prevail amino (−NH2) HPLC column (250 × 4.6 mm i. d., particle size 5 μm) (Alltech, Deerfield, IL, USA) with pre-running into a guard column (7.5 × 4.6 mm). A gradient elution was carried out using solvent A (double distilled water/methanol, 20/80; v/v) and solvent B (sodium acetate/solvent A, 4.1/100; w/v, pH 4.6). The elution system was 30% solvent A for 10 min, and followed by a linear gradient of 30–95% solvent B for 10–35 min. A constant flow rate of 1 mL/min was applied. Detection was performed with a UV detector set at 355 nm. Each treatment sample was prepared in duplicate and analyzed by HPLC. By comparing the retention time and area under curve of the peaks with those of standard GSH and GSSG in different concentrations, the concentrations of GSH and GSSG in the blood sample were determined. Hematocrit, hemoglobin (Hb), methemoglobin (MetHb) determination Before blood collection, the mouse tail was cut and 40 to 50 μL whole blood was obtained using heparinized capillary tube. The hematocrit and hemoglobin levels were determined by centrifugation method and commercial dialogistic kit (Stanbio, TX, USA) respectively. For methe-

Fig. 2. Linear regression analysis between G6PD activity and change in RBCs glutathione content upon challenge by α-naphthol in vitro in F2 littermates. The percentage changes of (A) GSH and (B) GSSG levels on F2 mice with different G6PD activities were investigated in response to challenge with 1 mg/mL α-naphthol. The percentage reduction of GSH and elevation of GSSG were negatively correlated with increasing G6PD levels.

moglobin (MetHb) determination, each treatment sample (10 μL) was diluted with ddH2O (1 mL) and divided into two equal volumes (A and B). For component A, the OD change (630 nm) was determined before and after adding of 20 μL of potassium cyanide (KCN). For component B, 20 μL of potassium ferricyanide, K3[Fe(CN)6] was added. The OD change was determined before and after the addition of 20 μL of KCN. The percentage of MetHb was calculated by dividing the OD change of component A with that of component B. Statistical analysis For multiple group comparison of read-out parameters, the significance of the differences was tested by one-way ANOVA, followed by post-hoc Dunnet's test. Correlation of G6PD with their response to oxidative agent was investigated by linear regression analysis. All statistical analyses were performed by using the Statistical Package of Social Science (SPSS) version 16.0 for Windows (SPSS Inc., Chicago, IL, USA). All statistical tests were carried out at 5% level of significance (p b 0.05). Result Establishing severe G6PD-deficient mice We have created the in vivo mouse model of G6PD deficiency (with G6PD activity b2 U/g Hb) following the breeding program shown in Fig. 1A. Two pairs of homozygous G6PD mutant females and C57L/J males were crossed to give the F1 progeny. The G6PD levels of the male C57L/J were 5.0 and 5.4 U/g Hb, whereas the female homozygous G6PD mutants were 2.0 and 2.1 U/g Hb. In the F1 generation, the G6PD levels of hemizygous male and heterozygous female were 1.6 ± 0.1 U/g Hb and 3.7 ± 0.2 U/g Hb, respectively. Three pairs of F1 littermates with low

C.H. Ko et al. / Blood Cells, Molecules, and Diseases 47 (2011) 176–181

179

8.3 ± 0.7, 0.9 ± 0.1 and 4.6 ± 0.3 U/g Hb respectively (Fig. 1B). The genotype of all individuals was determined by PCR-RFLP as illustrated in Fig. 1C.

Challenge of G6PD-deficient erythrocytes with α-naphthol in vitro The exposure of G6PD-deficient RBCs to α-naphthol led to the decrease of GSH and increase of GSSG levels. A significant negative correlation was demonstrated between GSH percentage reduction and G6PD activity upon challenge with α-naphthol at 1 mg/mL (r = − 0.51, p = 0.0009, Fig. 2A). Similar correlation was also found between GSSG percentage elevation and G6PD level (r = − 0.32, p = 0.04, Fig. 2B). Our results suggested that G6PD plays an important role in protecting RBCs against exogenous oxidative stress in mice RBCs, similar to human RBCs as shown in our previous report [10].

Effects of naphthalene on glutathione levels in vivo Similar levels of RBCs GSH (Fig. 3A) and GSSG (Fig. 3B) were observed in G6PD-deficient group when compared with non-deficient group without treatment. In Fig. 3A, slight but insignificant decreases of GSH were observed in both G6PD-deficient and non-deficient mice upon naphthalene treatment at increasing dose. In contrast, our data illustrated that GSSG level was increased significantly by 42% in G6PDdeficient mice at 250 mg/kg of naphthalene (P = 0.004) when compared with the respective control group without treatment (Fig. 3B). However, no significant result was found in non-deficient group. Similar results were demonstrated in GSH-to-GSSG ratio as shown in Fig. 3C. The administration of naphthalene at 250 mg/kg resulted in a decrease of GSH-to-GSSG ratio in G6PD-deficient group by 34.2% (P = 0.005). However, our data did not show any significant change in GSH-to-GSSG ratio when the non-deficient mice were treated with all concentrations of naphthalene.

Fig. 3. Effect of naphthalene on RBC glutathione content in G6PD-deficient (F2 homozygous female with G6PD activity b 2 U/g Hb) and control animals (F2 heterozygous female with G6PD activity N 2 U/g Hb). (A) GSH levels, (B) GSSG levels and (C) GSH-toGSSG ratio in G6PD-deficient and control RBCs were determined by HPLC after treatment with naphthalene at ( )250, ( )125 and (□)0 mg/kg. Data are the means (S.E.M; error bars) from 15 animals in each group.

G6PD activities were selected as the breeders for the F2 production. The G6PD levels of the male F1 breeders were 1.33, 1.37 and 1.44 U/g Hb, whereas the female were 3.13, 3.16 and 3.17 U/g Hb. For the F2 generation, four groups of mice were produced: hemizygous mutant male, wild type male (with mixed background), homozygous mutant female heterozygous mutant female. Their G6PD levels were 1.9 ± 0.1,

Effect of naphthalene on hematocrit, hemoglobin and methemoglobin levels As shown in Table 1, the basal levels of hematocrit, hemoglobin methemoglobin in both G6PD-deficient and control groups were similar. At 250 mg/kg of naphthalene, the G6PD-deficient mice were found to develop hemolytic anemia. The hematocrit and hemoglobin levels in model mice were significantly decreased by 24% and 30%, respectively, when compared with respective G6PD-deficient mice without treatment. In addition, the methemoglobin level was increased significantly in G6PD-deficient group by 1.9 fold (p b 0.001). On the other hand, the administration of naphthalene at all concentrations had no effect on hematocrit, hemoglobin and methemoglobin levels in control group.

Table 1 The hematocrit, hemoglobin and methemoglobin levels in G6PD-deficient and control mice after naphthalene treatment. Naphthalene (mg/kg)

Hematocrit (%) Hb (g/dL) MetHb (% of total Hb)

G6PD-deficient

Control

250

125

0

250

125

0

38.3 ± 2.0 (P = 0.004)# 8.9 ± 0.1 (P = 0.002)# 0.8 ± 0.1 (P b 0.001)#

48.4 ± 1.2

50.1 ± 1.5

49.0 ± 2.6

50.2 ± 1.6

49.5 ± 2.5

12.1 ± 0.5

12.7 ± 0.3

12.1 ± 0.3

12.4 ± 0.5

12.3 ± 0.4

0.3 ± 0.1

0.3 ± 0.1

0.4 ± 0.1

0.3 ± 0.1

0.3 ± 0.1

Data are the means (S.E.M; error bars) from 10 to 15 animals in each group. # P b 0.05 vs G6PD-deficient control group without naphthalene treatment.

180

C.H. Ko et al. / Blood Cells, Molecules, and Diseases 47 (2011) 176–181

Discussion Drug-induced anemia is a potentially life-threatening implication for G6PD-deficient individuals. Researchers have attempted to find a suitable in vivo model for drug screening. Lin et al. used acetyphenylhydrazine-treated rats as a G6PD-deficient model for drug screening [11]. However, this model could only demonstrate the GSH-depleted condition of G6PD-deficiency. The GSH regeneration ability of this model was not impaired and the G6PD enzyme was not defective. Therefore, results obtained in this model should be interpreted with caution. Over the decades, there has been little success on developing in vivo predictive animal models for assessment of pro-oxidative agents on G6PD deficiency. Four G6PD-mutant mice have been developed through various technologies. However, none of them is suitable for systemic analysis of G6PD-deficient red cell hemolysis, either because of embryonic lethality or insufficient reduction in G6PD activity. Following are data of these G6PD mutant mouse strains: 1. N-ethyl-N-nitrosourea (ENU)-induced mutant G6PD mouse 1581 (Gpd2e 1Neu): Heterozygotes display 150% of G6PD activity of the wild type. Homozygotes are lethal [8]. 2. ENU-induced mutant G6PD mouse 10168 (Gpdxa-m1Neu): hemizygous, heterozygous and homozygous mutants have 15%, 60% and 13% G6PD activities in their RBCs, respectively compared with the wild type animal [7,8]. 3. Mutant G6PD mouse (Gpdx a-m2Neu): apart from two mutant mice, which were induced by ENU mutagenesis, this mutant mouse arose from ionization radiation. Hemizygous, heterozygous and homozygous mutants have 33%, 65% and 29% G6PD activities in their RBCs, respectively compared with the wild type. This mutant allele caused a less dramatic reduction in residual enzyme activities than the a-m1Neu mutation [12,13]. 4. G6PD-knockout mutant mice: these animals were generated from embryonic stem cells with targeted inactivation of the G6PD gene. They are highly sensitive to oxidative stress in vitro [14]. However, G6PD male embryos derived from chimeric mice died at E10.5 and the maternally transmitted G6PD allele led to embryonic lethality in heterozygous animals due to severe placental damage [6]. Nevertheless, heterozygous females with paternal G6PD allele are viable with normal G6PD activities in RBCs as a result of ‘schewed’ somatic cell selection, favoring survival and growth of G6PDnormal red cells after X chromosome inactivation. The above studies support the notion that either severe/complete G6PD deficiency or excessive G6PD activities are incompatible with survival during embryonic development. Unlike in human RBCs, G6PD levels in mouse RBCs vary markedly among normal inbred strains: with high (e.g. A/J; 14.3–14.7 U/g Hb); intermediate (e.g. 129/J and C57BL/6J; 9.4–12.5 U/g/Hb) and low (e.g. C57L/J and C57BR/cdJ; 5.3–6.7 U/g Hb) G6PD activities [15]. Two autosomal regulatory foci: Gdr-1 and Gdr-2, have been proposed in modulating G6PD activities in mouse RBCs [8]. However, their respective gene locations and sequences were not determined. We hypothesized that there may be a threshold G6PD activity that allows embryonic survival but the mouse carrying such genotype would associate with drug-induced anemia as seen in human G6PD-deficiency. To create this novel animal model, we bred a known G6PD-deficient mutation into a ‘normal’ mouse strain with constitutive low G6PD activity. In this study, we have developed a G6PD-deficient mouse model with low residual G6PD activities (b2 U/g Hb), by breeding the homozygous mutant Gpdx a-m1Neu allele (from mutant G6PD mouse 10168) into the C57L/J background. Our results demonstrated that all hemizygous and homozygous F1 male and F2 female had the G6PD levels lower than their parental G6PD mutants although they had the same genotype harboring the Gpdx a-m1Neu allele on X chromosome. It implicated those G6PD regulators: Gdr-1 and Gdr-2 can down-

regulate the G6PD activity. However, their regulatory actions on G6PD gene/protein were not well understood. As mentioned before, there is lack of basic information about these two regulators. Therefore, further work is needed to map out the locations of these regulators and analyze their corresponding protein sequences, before any indepth mechanistic studies. Recently, mutant G6PD mouse 10168 (Gpdx a-m1Neu), the parental strain of our model, has been used as a model of G6PD-deficiency in studies of myocardial dysfunction [16], vascular disease [17], renal failure [18] and sepsis [5]. The reduced G6PD level of this strain was similar to the class 3 G6PD-deficient variants (moderate). This strain of mice demonstrated a limited number of pathophysiological conditions. However, some characteristics such as favism occurs mainly in class 2 G6PD-deficient variants (severe), and only occasionally in class 3 variants [3]. Most drug-induced hemolysis was reported in class 2 variants. Compared with the parental strain (1.97 ± 0.10 U/g Hb), our new model had lower G6PD activity by 57% (0.85 ± 0.08 U/g Hb) and fell in the range of class 2 G6PD-deficiency of human. To assess the efficacy of our model, we have tested this model in vitro and in vivo by the known pro-oxidant α-naphthol and its metabolic precursor naphthalene, respectively. Our in vitro data are in accordance with the reports on the pro-oxidative activities of αnaphthol on erythrocytes with low G6PD activities, as indicated by negative correlation between G6PD activities with (i) the reduction in GSH and (ii) the elevation in GSSG levels. However, when compared with our previous studies on human erythrocytes, the dose imposing significant oxidative effects on mice G6PD-deficient erythrocytes was 10 times higher (1 mg/mL). No significant observation was found when the cells were treated with 0.1 mg/mL of α-naphthol. Our result suggested that the anti-oxidative role of G6PD in mice erythrocytes may be different from human. Further investigation is needed to compare the species difference in total anti-oxidant profile of erythrocytes with low G6PD activities and their specific response to oxidative stress. More importantly, our in vivo studies demonstrated that naphthalene could inflict significant oxidative stress in G6PD-deficient animals and lead to hemolytic anemia. Oxidative injuries from naphthalene (at 250 mg/kg) significantly decreased GSH, increased GSSG content, and resulted to the overall reduction of GSH/GSSG ratio. The methemoglobin levels were also increased. Our findings are in concordance to the reported in vitro assay that naphthalene decreased the GSH levels of G6PD-deficient erythrocytes after the addition of rat liver microsome [19]. The derivatives of naphthalene, such as αnaphthol and α-naphthoquinone, lowered GSH levels of G6PDdeficient erythrocytes [19,20], whereas naphthalene itself did not pose any oxidative injuries to these cells in vitro. In the current study, the pro-oxidative activities of naphthalene not only cause oxidative damage but also led to hemolytic toxicity to red blood cells, as indicated by the decrease of hematocrit and hemoglobin levels. Hemolytic anemia is a condition in which blood has a lower number of RBCs (as reflected by hematocrit) or contains low hemoglobin concentrations in RBCs. Naphthalene was long regarded as a hemolytic agent to particular groups of people well before the discovery of G6PD-deficiency. A number of acute hemolytic poisoning cases were reported in G6PD-deficient subjects who consumed naphthalene-containing mothball [21]. The metabolic activities of the liver would change the pro-oxidative nature of chemicals or vice versa. In summary, we have presented a novel G6PD-deficient model by breeding the mutant Gpdx a-m1Neu allele into the C57L/J background. The efficacy of the model was verified by the known hemolytic agents, α-naphthol and naphthalene. Our data illustrated that this model was responsive to oxidative stress, although they did not present the same pattern as human G6PD-deficiency. Further study on this mouse model is required to confirm its suitability for mimicking human pathophysiological conditions related to G6PD-deficient related

C.H. Ko et al. / Blood Cells, Molecules, and Diseases 47 (2011) 176–181

hemolysis. In addition, the efficacy of this animal model will be further validated using other agents with proven hemolytic properties such as anti-malarial drug (i.e. primaquine) and anti-bacterial drug (i.e. sulphonamides). If successful, this mouse colony would represent the first severely G6PD-deficient mouse model with which the hemolytic toxicity of pharmacological agents on G6PD-deficient red cells can be validated.

[10]

[11]

[12]

Acknowledgments [13]

This work was supported by a direct grant from the Chinese University of Hong Kong.

[14]

References [1] E. Beutler, T.J. Vulliamy, Hematologically important mutations: glucose-6phosphate dehydrogenase, Blood Cells Mol Dis. 28 (2002) 93–103. [2] S.T. Lam, M.L. Cheng, Neonatal screening in Hong Kong and Macau, Southeast Asian J Trop Med Public Health. 34 (Suppl 3) (2003) 73–75. [3] E. Beutler, G6PD deficiency, Blood. 84 (1994) 3613–3636. [4] N. Bashan, O. Makover, A. Livne, S. Moses, Effect of oxidant agents on normal and G6PD-deficient erythrocytes, Isr J Med Sci. 16 (1980) 351–356. [5] J. Wilmanski, E. Villanueva, E.A. Deitch, Z. Spolarics, Glucose-6-phosphate dehydrogenase deficiency and the inflammatory response to endotoxin and polymicrobial sepsis, Crit Care Med. 35 (2007) 510–518. [6] L. Longo, O.C. Vanegas, M. Patel, V. Rosti, H. Li, J. Waka, T. Merghoub, P.P. Pandolfi, R. Notaro, K. Manova, L. Luzzatto, Maternally transmitted severe glucose 6-phosphate dehydrogenase deficiency is an embryonic lethal, EMBO J. 21 (2002) 4229–4239. [7] W. Pretsch, D.J. Charles, S. Merkle, X-linked glucose-6-phosphate dehydrogenase deficiency in Mus musculus, Biochem Genet. 26 (1988) 89–103. [8] W. Pretsch, Enzyme-activity mutants in Mus musculus. I. Phenotypic description and genetic characterization of ethylnitrosourea-induced mutations, Mamm Genome. 11 (2000) 537–542. [9] C.H. Ko, K. Li, P.C. Ng, K.P. Fung, C.L. Li, R.P. Wong, K.M. Chui, G.J. Gu, E. Yung, C.C. Wang, T.F. Fok, Pro-oxidative effects of tea and polyphenols, epigallocatechin-3-

[15] [16]

[17]

[18]

[19]

[20]

[21]

181

gallate and epigallocatechin, on G6PD-deficient erythrocytes in vitro, Int J Mol Med. 18 (2006) 987–994. C.H. Ko, E. Yung, K. Li, C.L. Li, P.C. Ng, K.P. Fung, R.P. Wong, K.M. Chui, G.J. Gu, T.F. Fok, Multiplex primer extension reaction screening and oxidative challenge of glucose-6-phosphate dehydrogenase mutants in hemizygous and heterozygous subjects, Blood Cells Mol Dis. 37 (2006) 21–26. N. Lin, X. Gao, J. Li, J. Zhu, Influence of Huanglian and berberine on the erythrocytic osmotic fragilitas of experimental glucose-6-phosphate dehydrogenase deficiency in rats, Zhongguo Zhong Yao Za Zhi. 23 (1998) 562–564. S.L. Baader, M.L. Schilling, B. Rosengarten, W. Pretsch, H.F. Teutsch, J. Oberdick, K. Schilling, Purkinje cell lineage and the topographic organization of the cerebellar cortex: a view from X inactivation mosaics, Dev Biol. 174 (1996) 393–406. K. Felix, L.D. Rockwood, W. Pretsch, J. Nair, H. Bartsch, G.W. Bornkamm, S. Janz, Moderate G6PD deficiency increases mutation rates in the brain of mice, Free Radic Biol Med. 32 (2002) 663–673. P.P. Pandolfi, F. Sonati, R. Rivi, P. Mason, F. Grosveld, L. Luzzatto, Targeted disruption of the housekeeping gene encoding glucose 6-phosphate dehydrogenase (G6PD): G6PD is dispensable for pentose synthesis but essential for defense against oxidative stress, EMBO J. 14 (1995) 5209–5215. J.J. Hutton, Genetic regulation of glucose 6-phosphate dehydrogenase activity in the inbred mouse, Biochem Genet. 5 (1971) 315–331. M. Jain, L. Cui, D.A. Brenner, B. Wang, D.E. Handy, J.A. Leopold, J. Loscalzo, C.S. Apstein, R. Liao, Increased myocardial dysfunction after ischemia-reperfusion in mice lacking glucose-6-phosphate dehydrogenase, Circulation 109 (2004) 898–903. R. Matsui, S. Xu, K.A. Maitland, A. Hayes, J.A. Leopold, D.E. Handy, J. Loscalzo, R.A. Cohen, Glucose-6 phosphate dehydrogenase deficiency decreases the vascular response to angiotensin II, Circulation. 112 (2005) 257–263. Y. Xu, Z. Zhang, J. Hu, I.E. Stillman, J.A. Leopold, D.E. Handy, J. Loscalzo, R.C. Stanton, Glucose-6-phosphate dehydrogenase-deficient mice have increased renal oxidative stress and increased albuminuria, FASEB J. 24 (2010) 609–616. K.E. Bloom, G.J. Brewer, A.M. Magon, N. Wetterstroem, Microsomal incubation test of potentially hemolytic drugs for glucose-6-phosphate dehydrogenase deficiency, Clin Pharmacol Ther. 33 (1983) 403–409. N. Bashan, N. Peleg, S.W. Moses, Attempts to predict the hemolytic potential of drugs in glucose-6-phosphate dehydrogenase deficiency of the Mediterranean type by an in vitro test, Isr J Med Sci. 24 (1988) 61–64. K. Santucci, B. Shah, Association of naphthalene with acute hemolytic anemia, Acad Emerg Med 7 (2000) 42–47.