Individual variations in arsenic metabolism in Vietnamese: the association with arsenic exposure and GSTP1 genetic polymorphism

Abstract

We investigated the association of As exposure and genetic polymorphism in glutathione S-transferase π1 (GSTP1) with As metabolism in 190 local residents from the As contaminated groundwater areas in the Red River Delta, Vietnam. Total As concentrations in groundwater ranged from <0.1 to 502 μg l⁻¹. Concentrations of dimethylarsinic acid (DMA^V), monomethylarsonic acid (MMA^V), and arsenite (As^III) in human urine were positively correlated with total As levels in the groundwater, suggesting that people in these areas may be exposed to As through the groundwater. The concentration ratios of urinary As^III/arsenate (As^V) and MMA^V/inorganic As (IA; As^III + As^V)(M/I), which are indicators of As metabolism, increased with the urinary As level. Concentration and proportion of As^III were high in the wild type of GSTP1 Ile105Val compared with the hetero type, and these trends were more pronounced in the higher As exposure group (>56 μg l⁻¹creatinine in urine), but not in the lower exposure group. In the high As exposure group, As^III/As^V ratios in the urine of wild type of GSTP1 Ile105Val were significantly higher than those of the hetero type, while the opposite trend was observed for M/I. These results suggest that the excretion and metabolism of IA may depend on both the As exposure level and the GSTP1 Ile105Val genotype.

---

1. Introduction

It is well known that inorganic As (IA) is one of the human carcinogenic chemicals. Contamination by naturally derived IA in groundwater has been reported in certain areas and has caused a large number of serious health issues.^1–3 In such contaminated areas, skin pigmentation, hyperkeratosis, cancers, and resultant high mortalities have been caused by the chronic IA exposure in the local people.^4–6 On the other hand, large differences in the sensitivity to IA-related diseases among individuals have been reported, suggesting its association with individual variations in IA metabolism.⁷

Ingested IA is metabolized to methylated arsenicals in the body and then mainly excreted through urine. There are two hypotheses regarding IA metabolic pathways;⁸ oxidative methylation^9,10 and reductive methylation.^11,12 In these metabolic pathways, two enzymes, arsenic (+3 oxidation state) methyltransferase (AS3MT) and glutathione S-transferase ω (GSTO), participate in the methylation and reduction of As compounds, respectively, in a variety of animals including the human.¹³

GSTs are a family of enzymes that play an important role in detoxification of various xenobiotics by catalyzing the conjugation of hydrophobic and electrophilic compounds with reduced glutathione. There are seven classes of GSTs including α, μ, ω, π, θ, σ, and ζ. GSTO1 is involved in the reduction activities of arsenate (As^V), monomethylarsonic acid (MMA^V), and dimethylarsinic acid (DMA^V).^14–16DMA^V reductase activity of GSTO2 is much lower than that of GSTO1.¹⁷ Some researchers have reported the relevance of genetic polymorphisms of GSTO1 and O2 to As metabolism by in vitro assays ^17–19 and in human studies.^20–22

It has been suggested that GST π1 (GSTP1) plays a role in the reduction of IA toxicity. An in vitro study using SA7 cells (As-resistant Chinese hamster ovary cells) revealed GSTP1 level-dependent resistance of IA.²³ Zhou et al. (2005) found that GSTP1 prevented IA-induced apoptosis in human lymphoma cell lines by reducing intracellular H₂O₂ levels.²⁴ There are several single nucleotide polymorphisms (SNPs) in GSTP1 (http://www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?locusId = 2950). For the GST activity, the Ile (AA) type of GSTP1 Ile105Val (rs1695; A to G substitution at nucleotide base 6624 and isoleucine to valine substitution at amino acid base 105) was higher than the Val (GG) type in the erythrocyte.²⁵

Several epidemiological studies have recently investigated the relationships between genetic polymorphisms in GSTP1 and IA-related diseases (Table 1), but the results were not consistent. In the study in As-contaminated areas of West Bengal, no association of GSTP1 Ile105Val with skin lesion was observed.²⁶ McCarty et al. (2007) have reported that there was no significant difference in the genotype distribution of GSTP1 Ile105Val between Bangladeshi people with and without skin lesion.²⁷ On the other hand, several studies have suggested that the Val type of GSTP1 Ile105Val was associated with increased risks of As-induced skin lesion in Chinese²⁸ and Bangladeshi,²⁹ and of atherosclerosis in Taiwanese.³⁰ Wang et al. (2007) also reported that the risk increased in Val type GSTP1 with high As exposure.³⁰ However, Hsu et al. (2008) evaluated the interaction of GSTP1 polymorphism with urinary transitional cell carcinoma in southwestern Taiwan and found that the Val type of GSTP1 Ile105Val showed a significantly lower cancer risk than the Ile type, suggesting that the wild type of GSTP1 Ile105Val may be sensitive to urinary transitional cell carcinoma.³¹

Table 1 Associations of GSTP1 Ile105Val polymorphism with endpoints related to As

Location	Endpoint	Association with genotype	References
West Bengal, India	Skin lesion	No association	26
Bangladesh	Skin lesion	No association	27
China	Skin lesion	High risk in Val type	28
Bangladesh	Skin lesion	High risk in Val type	29
Taiwan	Atherosclerosis	High risk in Val type under high As exposure	30
Taiwan	Bladder cancer	High risk in Ile type	31
Chile	Methylation of DMA^V	No association	32
Vietnam	Methylation of IA	High in Ile/Val type	40
Vietnam	As^V reduction	High in Ile/Ile type	40

---

There are few studies evaluating whether GSTP1 Ile105Val polymorphism influences As metabolism. Marcos et al. (2006) revealed that the Val type had higher %DMA^V than the Ile type, but it was not significant in Chilean.³² Since 2001, we have investigated As pollution in groundwater and its human exposure in Vietnam.^8,33–43 Recently, we have focused on the association of genetic polymorphisms in As metabolic enzymes including GSTP1 with As metabolism and suggested that the heterozygote of GSTP1 Ile105Val had a higher metabolic capacity from IA to monomethyl As, while the opposite trend was observed for the metabolism from As^V to As^III.^40,41 However, the metabolic capacity of As has not been evaluated in association with GSTP1 Ile105Val polymorphisms and As exposure levels. Inconsistent findings among previous studies on association of GSTP1 Ile105Val with As metabolic capacity as well as As-related diseases may partly result from the joint effects of human As exposure levels and genetic polymorphisms in GSTP1.

The aim of this study is to clarify the association of As exposure and genetic polymorphism in GSTP1 Ile105Val with As metabolism as well as their co-interaction. Therefore, we analyzed concentrations of As compounds in the urine and genotyped GSTP1 Ile105Val in residents from As-contaminated groundwater areas in Vietnam. Biological factors such as sex, age, body mass index (BMI), and habits of alcohol consumption and smoking were also incorporated to evaluate their contributions to individual variations in IA metabolism.

2. Materials and methods

2.1 Samples

Samples of well water (n= 64) and human urine (n= 190) and blood (n= 190) were randomly collected from Hoa Hau (HH) and Liem Thuan (LT) in Ha Nam Province in March 2006³⁹ and from Thanh Vanh (TV) and THach Hoa (TH) in Ha Tay Province in September 2007. All communes are located in the rural area of the Red River Delta, Vietnam and there are no other significant sources of As such as mining sites or industries. Several houses in HH, LT, and TV had wells equipped with a sand filter system and thus the sand-filtrated groundwater samples (n= 40) were also collected along with unfiltered water samples. Usage period of the wells and their depth are shown in Table 2. The informed consent was obtained from all the participants. This study has been approved by the Ethical Committee of Ehime University, Japan. Information on sex, age, residential years, body height and weight, body mass index (BMI), and smoking and alcohol habits of donors is summarized in Table 2. All samples were kept at −25 °C in a freezer of the Environmental Specimen Bank (es-BANK) in Ehime University⁴⁴ until chemical analyses and genotyping.

Table 2 Information on water and human samples from Hoa Hau (HH), Liem Thuan (LT), Thanh Vanh (TV), and Thach Hoa (TH) in Vietnam

Location	Hoa Hau	Liem Thuan	Thanh Vanh	Thach Hoa	p value
	(HH)	(LT)	(TV)	(TH)
a Arithmetic mean and range. b Geometric mean and range. c Tukey–Kramer test. d χ ^2 test. e In a house equipped with sand filter, filtered water instead of raw groundwater is assumed to be consumed.
Sampling season	Mar-06	Mar-06	Sep-07	Sep-07
Groundwater
No.	15	13	21	15
Used period (years)^a	9 (5.5–13)	6 (1–16)	7 (3–12)	6 (3–10)	0.015^c
Well depth (m)^a	14 (8–16)	15 (12–24)	38 (20–60)	33 (24–50)	<0.001^c
TA (μg l^−1)^b	368 (163–502, and 2120 (an outlier))	1.4 (0.7–6.8)	36.0 (5.5–145)	0.1 (<0.1–0.5)	<0.001^c
Filtered water
No.	10	9	21	0
TA (μg l^−1)^b	18.9 (3.2–143)	2.0 (1.0–4.9)	5.4 (1.5–50.7)	—	<0.001^c
Drinking water ^e
No.	15	13	21	15
TA (μg l^−1)^b	50.1 (3.2–486)	1.7 (0.9–4.9)	5.4 (1.5–50.7)	0.1 (<0.1–0.5)	<0.001^c
Subjects
No.	51	49	50	40
No. of male/female	22/29	22/27	21/29	21/19	>0.05^d
Age (years)^a	37 (11–60)	34 (11–70)	32 (13–71)	35 (15–60)	>0.05^c
Residential time (years)^a	33 (3–60)	31 (6–65)	30 (13–71)	17 (3–45)	<0.001^c
Height (cm)^a	156 (137–173)	150 (121–169)	155 (142–170)	158 (137–171)	<0.001^c
Weight (kg)^a	48 (27–66)	44 (22–67)	46 (32–65)	52 (38–72)	0.001^c
BMI^a	20 (14–26)	19 (12–29)	19 (15–25)	21 (16–28)	0.027^c
No. of smoker/non-smoker	14/37	6/43	8/42	7/33	>0.05^d
No. of alcohol drinker/non-alcohol drinker	14/37	10/39	13/37	13/27	>0.05^d
Urinary SA (μg g^¬1creatinine)^b	92.6 (45.2–365)	97.9 (38.6–397)	63.5 (28.7–115)	43.2 (20.0–96.0)	<0.001^c
Urinary AB (%)^a	22.7 (4.0–56.8)	19.6 (3.1–58.6)	16.0 (0–63.5)	28.3 (3.0–78.1)	0.001^c
Urinary DMA^V (%)^a	55.9 (32.6–77.2)	59.0 (29.1–78.9)	51.8 (26.1–68.6)	44.4 (13.1–68.8)	<0.001^c
Urinary MMA^V (%)^a	10.6 (2.9–17.8)	10.0 (4.8–20.9)	11.5 (3.3–20.1)	7.2 (0–15.6)	<0.001^c
Urinary As^III (%)^a	8.5 (0–20.3)	8.7 (0–19.8)	9.7 (0–30.0)	6.6 (0–16.6)	0.013^c
Urinary As^V (%)^a	2.3 (0–11.1)	2.7 (0–11.3)	11.0 (3.1–34.4)	13.5 (0–37.4)	<0.001^c

---

2.2 Analyses of As

Analytical methods for samples collected in 2006 have already been reported in our previous study.³⁹ The methods of As analysis of water, and human urine and blood samples collected in 2007 were briefly summarized below. After acidification with concentrated HNO₃, total As (TA) in water samples was analyzed with an inductively coupled plasma-mass spectrometer (ICP-MS; HP-4500, Hewlett-Packard, Avondale, PA, USA). Rhodium was used as an internal standard to correct matrix effects and instrumental drift.³⁹ Accuracy of the analytical method was confirmed in good agreement (91–95%) with certified TA concentration by analyzing a certified reference material, SLRS-4 River Water from the National Research Council Canada (NRCC). In addition, we have participated in an inter-calibration exercise program organized by the Swiss Federal Institute of Aquatic Science and Technology (Eawag) in the frame of the ongoing cooperation of Vietnam and Switzerland for As-related researches. Concentration of TA in water samples is expressed in μg per l.

Urinary As compounds including arsenobetaine (AB), DMA^V, MMA^V, As^III, and As^V were separated by a high-performance liquid chromatograph (HPLC; Shimadzu, LC10A Series, Kyoto, Japan) equipped with an Inertsil AS column (15 cm, 2.1 mm i.d.; GL Sciences Inc., Japan). The column was equilibrated with the mobile phase (10 mM sodium 1-butanesulfonate, 4 mM tetramethylammonium hydroxide, 4 mM malonic acid, and 0.5% methanol; pH 3.0 was adjusted with nitric acid) at a flow rate of 0.5 ml min⁻¹ at 45 °C. As internal standard, Rh was added into the buffer. The injection volume was 10 μl. Five arsenicals separated by the column were determined with ICP-MS. Ion intensities at m/z 75 (⁷⁵As), 77 (⁴⁰Ar³⁷Cl and ⁷⁷Se), and 103 (¹⁰³Rh) were monitored and there was no interference during HPLC/ICP-MS analysis. A certified reference sample, NIES No. 18 human urine that was provided by the National Institute for Environmental Studies (NIES), Japan, was analyzed to assure the methodological accuracy. Analyzed concentrations of AB and DMA^V were in good agreement with the certified values (90–106%). In the present study, sum of all As compounds, inorganic As (As^III + As^V), and As^III + As^V + MMA^V + DMA^V detected in urine sample are denoted as SA, IA, and IMDA, respectively. Percentages of AB, As^III, As^V, MMA^V, DMA^V, IA, and IMDA to SA in the human urine were denoted as %AB, %As^III, %As^V, %MMA^V, %DMA^V, %IA, and %IMDA, respectively. Urinary creatinine was determined at SRL, Inc. (Tokyo, Japan) and concentrations of As compounds in the urine were expressed as μg As per g on a creatinine basis. Because As^V, IA, and MMA^V are metabolized to As^III, MMA^V, and DMA^V, respectively, in the human body, concentration ratios of As^III/As^V (III/V), MMA^V/IA (M/I), and DMA^V/MMA^V (D/M) were used as an index for each metabolic process of As^V, IA, and MMA^V.

2.3 Genotyping of GSTP1 polymorphisms

Genotyping of GSTP1 followed the methods described in our previous study.⁴⁰ A QIAamp DNA mini kit (Qiagen, Chatworth, CA) was used to extract DNA from whole blood sample. The reference sequence of GSTP1 (accession number, AY324387) was based on the DNA Data Bank of Japan (DDBJ). Forward and reverse primers of GSTP1 Ile105Val were 5′-ACCCCAGGGCTCTATGGGAA-3′ and 5′-TGAGGGCACAAGAAGCCCCT-3′, respectively. DNA was amplified with PCR in a 10 μl reaction mixture containing GoTaq® Green Master Mix (Promega, Madison WI, USA) at 55 °C of annealing temperature and then treated with Bsm AI at 37 °C. The PCR products were separated in 8% polyacrylamide gel by electrophoresis (300 V, 15 min) and were detected by silver staining. The genotyping was carried out in duplicate. The representativeness of nucleotide sequences for the genotype was confirmed with a Genetic Analyzer (model 310, Applied Biosystems).

2.4 Statistical analyses

Together with the data in the present study, analytical data in the water and human urine from HH and TL provided in our previous study^39,40 were used for statistical analyses. StatView (version 5.0, SAS® Institute, Cary, NC, USA), PASW Statistics (v. 18.0J, SPSS Inc., Chicago, IL, USA), and EXCEL Toukei (Version 6.05, Esumi Co., Ltd., Tokyo, Japan) were used for the statistical analyses. One half of the value of the respective limits of detection was substituted for those values below the limits of detection and used in the statistical analysis. Normality of the distribution of all variables was checked by Kolmogorov–Smirnov's one sample test. To adapt parametric analyses, the data, which showed non-normal distribution, were log-transformed. Student's t-test and the Tukey–Kramer test were conducted to find differences in As levels and compositions in human urine among locations and the genotype of GSTP1. A χ² test was employed for checking sample size distribution in each group category. Relationships between variables were assessed by the Pearson correlation coefficient. To assess the factors affecting As levels in the urine and metabolic capacity of As, a stepwise multiple regression analysis was executed. In the regression models, nominal variables such as As exposure status, sex, alcohol and smoking habits, and genotype of GSTP1 Ile105Val were transformed to dummy variables (0 and 1). The multicollinearity of independent variables was assessed by calculating the variance inflation factor (VIF). p< 0.05 was considered to be statistically significant.

3. Results and discussion

Concentration of TA in groundwater

Concentrations of TA in groundwater are shown in Table 2. The range of concentration was <0.1–502 μg l⁻¹ (one sample that had 2120 μg l⁻¹ was regarded as an outlier and removed from further statistical analysis because the water had large amounts of particles). A significant regional difference in TA concentration was observed; HH (geometric mean (GM), 368 μg l⁻¹) > TV (GM, 36.0 μg l⁻¹) > LT (GM, 1.4 μg l⁻¹) > TH (GM, 0.1 μg l⁻¹) (p< 0.001). 54.7% of all groundwater samples exceeded the drinking water guideline (10 μg l⁻¹) established by WHO.⁵ Remarkably, 100% and 95.2% of groundwater samples from HH and TV had TA concentration over the guideline value.⁵ These results indicate that groundwaters from HH and TV are not suitable for drinking. Analyses of relationships between TA concentration and well depth showed a significant positive correlation in the groundwater from TV (r= 0.693, p= 0.003), indicating that the concentration of TA may be higher in the deeper layer of the aquifer in TV.

Several residents have consumed sand-filtered groundwater in these sampling areas except in TH. Concentrations of TA in the filtered water were in the range of 1.0–143 μg l⁻¹ (Table 2). There was a significant (p< 0.001) regional difference in TA concentrations in the filtered water; HH (GM, 18.9 μg l⁻¹) > TV (GM, 5.4 μg l⁻¹) > LT (GM, 2.0 μg l⁻¹). Through the sand filtration, TA concentrations in the raw groundwater from HH and TV significantly reduced (p< 0.001) and the removal efficiencies were 93% in HH and 82% in TV on arithmetic mean (AM) (Fig. 1). However, 80% and 29% of filtered-water samples from HH and TV were still higher than the WHO guideline value.⁵ This result suggests that safe drinking water is not always obtained by only a sand filter system and thus, further removal techniques of As from groundwater are required in highly As-contaminated groundwater areas.

Fig. 1 Concentrations of TAs in raw and sand-filtrated groundwater from Hoa Hau (HH), Thanh Vanh (TV), and Liem Thuan (LT) in Vietnam. Bar indicates each concentration of TA in raw and sand-filtered groundwater.

To evaluate the As exposure status of the residents, we considered the well water, which local people are drinking, as the major source of As. Concentration of TA in drinking water was regarded as those in raw groundwater for the houses without a sand-filter system, and as those in filtered water for the houses with the filter system. Concentrations of TA in drinking water from HH, TV, LT, and TH are shown in Table 2. The highest As concentration in drinking water was observed in HH (GM, 50.1 μg l⁻¹), followed by TV (GM, 5.4 μg l⁻¹), LT (GM, 1.7 μg l⁻¹), and TH (GM, 0.1 μg l⁻¹) and a significant difference was detected among all the four locations (p< 0.001). In HH and TV, samples with TA concentrations exceeding the guideline value for drinking water⁵ were 88% and 29%, respectively. Considered that As concentration in the drinking water represents close to the real exposure status in local residents, potential health risk of people drinking those As-contaminated water is of great concern.

Concentration and composition of As compounds in human urine

Concentrations of SA and composition of As compounds in the urine of people from HH, TV, TL, and TH are summarized in Table 2. Urinary As was detected in all samples and the range of urinary SA concentrations was from 20.0 to 397 μg g⁻¹creatinine.

To understand the exposure level of As in local people through drinking water, relationships between As concentrations in drinking water and human urine were assessed. As shown in Fig. 2, concentrations of DMA^V (R² = 0.118, p< 0.001), MMA^V (R² = 0.141, p< 0.001), and As^III (R² = 0.068, p< 0.001) in human urine were positively correlated with that of TA in drinking water. Significant positive correlations between concentrations of TA in drinking water and urinary As^V (R² = 0.028, p = 0.036), IMDA (R² = 0.114, p< 0.001), and SA (R² = 0.088, p< 0.001) were also observed (data not shown). These results suggest that the residents in these areas are exposed to As through the consumption of drinking water and ingested As are metabolized to MMA^V and DMA^V in the body. On the other hand, concentration of urinary AB, which is probably derived from fish and shellfish ingestion, showed no association with the TA level in drinking water (p> 0.05) (Fig. 2).

Fig. 2 Relationships between concentrations of TA in drinking water and As compounds (DMA^V, MMA^V, As^III, and AB) in human urine from Hoa Hau (HH), Thanh Vanh (TV), Liem Thuan (LT), and Thach Hoa (TH) in Vietnam. Dashed line indicates WHO guideline value (10 μg l⁻¹) for drinking water (WHO, 2004).

Among the As compounds detected, DMA^V (AM, 53%) was the most predominant species, followed by AB (AM, 21%), MMA^V (AM, 10%), As^III (AM, 8%), and As^V (AM, 7%). Because As compounds are transformed by reduction and methylation processes in the human body,^9–12 concentration ratios of As^III/As^V (III/V), MMA^V/IA (M/I), and DMA^V/MMA^V (D/M) in human urine were defined as metabolic indices for the reduction, first methylation, and second methylation, respectively. In the present study, GM for III/V, M/I, and D/M were 1.2, 0.7, and 5.4, respectively.

Genotype distribution of GSTP1 Ile105Val

Genotyping results of GSTP1 Ile105Val showed no mutation of the homo type (Val/Val) in this population. Genotype frequency in all donors was 0.68 for AA (Ile/Ile) and 0.32 for AG (Ile/Val), whereas A and G allele frequencies were 0.84 and 0.16, respectively. However, the GSTP1 Ile105Val genotype did not follow the Hardy–Weinberg principle in this study (p= 0.010). Although the reason remains unclear, we could confirm no genotyping error by conducting duplicate analyses of all DNA samples and by sequence analyses of some representative samples. Compared with the allele frequency of GSTP1 Ile105Val in 11 populations published in the HapMap database (HapMap Data Rel 28 PhaseII +^III, August 10, on NBCI B36 assembly, dvSNP b 126; http://hapmap.ncbi.nlm.nih.gov/cgi-perl/snp_details_phase3?name = rs1695&source = hapmap28_B36&tmpl = snp_details_phase3), the A allele frequency (0.84) in Vietnamese detected in this study was similar to those in Chinese populations (0.816 in Han Chinese in Beijing, China groups (CHB (H)) and 0.812 in Chinese in Metropolitan Denver, Colorado (CHD (D)). On the contrary, allele distribution in Vietnamese in the present study was largely different from Africans, Europeans, and Americans (0.448–0.701 for A allele frequencies).

Factors influencing As concentration and metabolism in humans

To understand which factors can affect the concentration of As and its metabolic capacity, a stepwise multiple regression analysis was performed. As potential factors, As exposure status, genotype of GSTP1 Ile105Val, sex, age, BMI, alcohol consumption, and smoking habit were taken into consideration. Before the analyses, the As exposure level was defined by dividing all donors into two categories, high (HA) and low (LA) As exposure groups, based on GM for urinary IMD concentration (56 μg g⁻¹creatinine). No significant bias in sample numbers among As exposure level, genotype of GSTP1 Ile105Val, and sex was validated by a χ² test. Sex ratios were significantly different in both smoking and alcohol habits, because only a few females had these habits. The calculated VIF values of explanatory variables were less than 10, and thus multi-collinearity in the multiple regression analysis was rejected.

Results of the multiple regression analyses are listed in Table 3. When all donors were evaluated, As exposure level, genotype of GSTP1 Ile105Val, sex, and BMI were significantly correlated with urinary As concentration and metabolic capacity, with the influence of exposure status being the strongest. These results were similar to our previous study.⁴⁰ Remarkably, the exposure level of As was significantly associated with not only urinary concentrations of As compounds as expected, but also metabolic indices except for D/M. Indicators of As metabolism such as %DMA^V, %MMA^V, %As^III, %IMDA, III/V, and M/I of HA were higher than those of LA, while the opposite results were observed for %AB, %As^V, and %IA (Table 3). Comparisons of III/V and M/I between HA and LA are shown in Fig. 3. These results indicate that the metabolism from As^V to As^III and from IA to MMA^V may be facilitated by high As exposure level. No significant increase in D/M with the As exposure level implies that 2nd methylation may not be facilitated by high exposure. Although decreased %AB could be explained by increased %DMA^V, %MMA^V, and %IA with As exposure, it was not clear why higher concentration of urinary AB was observed.

Table 3 Stepwise multiple regression analysis of As concentrations and compositions in urine against sex^a, age, BMI, alcohol and smoking habits^a, As exposure level, and polymorphism of GSTP1 Ile105Val^a

Dependent variable	R ^2adj	p	Independent variable	β	p	Dependent variable	R ^2adj	p	Independent variable	β	p
a These nominal variables were transformed to dummy variables (0 or 1).
%AB	0.031	0.015	Exposure (0 = low, 1 = high)	−0.193	0.015	AB	0.085	<0.001	Exposure (0 = low, 1 = high)	0.254	<0.001
									GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.156	0.027
%DMA^V	0.094	<0.001	Exposure (0 = low, 1 = high)	0.287	<0.001	DMA^V	0.564	<0.001	Exposure (0 = low, 1 = high)	0.753	<0.001
			Sex (0 = female, 1 = male)	−0.158	0.039
%MMA^V	0.028	0.021	Exposure (0 = low, 1 = high)	0.183	0.021	MMA^V	0.410	<0.001	Exposure (0 = low, 1 = high)	0.613	<0.001
									BMI	−0.141	0.013
%As^III	0.016	0.045	Exposure (0 = low, 1 = high)	0.146	0.045	As^III	0.258	<0.001	Exposure (0 = low, 1 = high)	0.472	<0.001
									GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.173	0.007
%As^V	0.080	<0.001	Exposure (0 = low, 1 = high)	−0.293	<0.001	As^V	0.039	0.009	BMI	−0.188	0.010
									Exposure (0 = low, 1 = high)	−0.143	0.049
%IA	0.025	0.026	Exposure (0 = low, 1 = high)	−0.177	0.026	IA	0.250	<0.001	Exposure (0 = low, 1 = high)	0.393	<0.001
									GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.217	<0.001
									BMI	−0.176	0.006
%IMDA	0.031	0.015	Exposure (0 = low, 1 = high)	0.193	0.015	IMDA	0.627	<0.001	Exposure (0 = low, 1 = high)	0.757	<0.001
									BMI	−0.123	0.007
									GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.104	0.021
III/V	0.196	<0.001	Exposure (0 = low, 1 = high)	0.402	<0.001	SA	0.510	<0.001	Exposure (0 = low, 1 = high)	0.693	<0.001
			GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.183	0.019				GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.139	0.007
M/I	0.144	<0.001	Exposure (0 = low, 1 = high)	0.370	<0.001
			GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	0.159	0.021
D/M	0.057	<0.001	Sex (0 = female, 1 = male)	−0.250	<0.001

---

Fig. 3 Comparison of III/V and M/I between low (LA) and high As (HA) exposure groups from Hoa Hau (HH), Thanh Vanh (TV), Liem Thuan (LT), and Thach Hoa (TH) in Vietnam. Data are given as geometric mean and geometric standard deviation. *** indicates statistical significance at p < 0.001.

In all participants, GSTP1 Ile105Val was associated with III/V, M/I, and concentrations of AB, As^III, IA, IMDA, and SA. Negative correlations between BMI and concentrations of MMA^V, As^V, IA, and IMDA in human urine were observed. Females had higher %DMA^V and D/M than males.

Because it was found that the As exposure status significantly influenced many parameters (Table 3), we repeated the stepwise regression analysis by dividing all donors into HA and LA to better understand the difference in factors associated with As excretion and metabolic capacity between those groups (Table 4).

Table 4 Stepwise multiple regression analysis of As concentrations and compositions in urine against sex^a, age, BMI, alcohol and smoking habits^a, As exposure status^a, and polymorphism of GSTP1 Ile105Val^a for each As exposure level

Dependent variable	R ^2adj	p	Independent variable	β	p	Dependent variable	R ^2adj	p	Independent variable	β	p
a These nominal variables were transformed to dummy variables (0 or 1).
High As exposure
						DMA^V	0.085	0.005	BMI	−0.397	0.001
									Age	0.252	0.037
%MMA^V	0.057	0.009	Sex (0 = female, 1 = male)	0.258	0.009	MMA^V	0.048	0.016	BMI	−0.240	0.016
%As^III	0.066	0.006	GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.274	0.006	As^III	0.126	<0.001	GSTP1 (0 = Ile/Ile, 1 = Ile/Val)	−0.367	<0.001
						IA	0.179	<0.001	GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.377	<0.001
									BMI	−0.235	0.011
%IMDA						IMDA	0.160	<0.001	BMI	−0.455	<0.001
									GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.242	0.010
									Age	0.251	0.031
III/V	0.068	0.019	GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.287	0.019	SA	0.051	0.014	BMI	−0.245	0.014
M/I	0.121	0.001	Age	0.264	0.007
			GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	0.249	0.010
D/M	0.042	0.024	Sex (0 = female, 1 = male)	−0.227	0.024
Low As exposure
						AB	0.054	0.016	GSTP1 Ile105Val (0 = Ile/Ile, 1 = Ile/Val)	−0.255	0.016
%DMA^V	0.066	0.008	Sex (0 = female, 1 = male)	−0.277	0.008	DMA^V	0.057	0.013	Sex (0 = female, 1 = male)	−0.261	0.013
%As^V	0.035	0.044	Sex (0 = female, 1 = male)	0.214	0.044
%IA	0.038	0.038	BMI	−0.220	0.038
D/M	0.133	0.001	Sex (0 = female, 1 = male)	−0.298	0.004
			Age	0.265	0.011

---

Interestingly, it was found that the factors, which could relate to As concentration and metabolism, were different between HA and LA except relationships between sex and D/M (Table 4); D/M in females was significantly higher than that in males regardless of the As exposure level, suggesting a higher methylation capacity from MMA^V to DMA^V in females regardless of the As exposure level. It has been reported that the 2nd methylation capacity is higher in females than in males in most studies.^39,45 Our results support these previous reports. The present study also showed that %MMA^V in males was significantly higher compared with females in HA. A similar trend was observed for %As^V in LA, although the significant level was weak. The sexual difference in methylation capacity may be partly associated with an estrogen-related metabolic pathway.⁴⁵ The mechanism of sexual difference in As metabolism needs more attention in future studies.

The GSTP1 Ile105Val hetero type had lower concentrations of As^III, IA, and IMDA, and III/V and higher M/I than the wild type in HA, but not in LA (Table 4). For urinary concentration of As^III (Fig. 4), it is suggested that the GSTP1 genotype may be linked with the excretion of As^III into the urine. Leslie et al. (2004) investigated a transport mechanism of As^III by a multidrug resistance protein 1 (MRP1/ABCC1) using a specific cell line, H69AR over-expressing MRP1 and found that MRP1 can transport As^III only in the presence of GSH and expression of GSTP1 in the plasma membrane is required for the transportation of As^III(+GSH).⁴⁶ Zhong et al. (2006) reported that in the erythrocyte of the healthy Chinese, the GSTP1 Ile105Val wild type showed a higher catalytic activity than the mutation type.²⁵ Considering these reports together, in the higher As exposure group, the GSTP1 Ile105Val wild type might accelerate the conjugation of GSH to As^III more than the mutation type and the conjugate may be more efficiently excreted from the cell through the MRP1 transporter. Further in vivo and human case studies are needed to verify this hypothesis.

Fig. 4 Comparison of As^III concentration between the wild and hetero types of GSTP1 Ile105Val among all participants, and high (HA) and low As (LA) exposure groups from Hoa Hau (HH), Thanh Vanh (TV), Liem Thuan (LT), and Thach Hoa (TH) in Vietnam. Data are given as geometric mean and geometric standard deviation. ** and *** indicate statistical significance at p< 0.01 and p< 0.001, respectively.

A lower III/V in the hetero type of GSTP1 Ile105Val than the wild type in all participants and HA (Table 3 and 4, and Fig. 5) suggest that the heterozygote of GSTP1 Ile105Val might have a lower reduction capacity from As^V to As^III and this reduction capacity may depend on the As exposure level. Although the reductase activity of As^V by GSTP1 was not measured, the wild GSTP1 Ile105Val protein has a high catalytic GST activity compared with the mutation type.²⁵

Fig. 5 Comparison of III/V and M/I between the wild and hetero types of GSTP1 Ile105Val among all participants, and high (HA) and low As (LA) exposure groups from Hoa Hau (HH), Thanh Vanh (TV), Liem Thuan (LT), and Thach Hoa (TH) in Vietnam. Data are given as geometric mean and geometric standard deviation. * and ** indicate statistical significance at p< 0.05 and p< 0.01, respectively.

The hetero type of GSTP1 Ile105Val had higher M/I than the wild type in all participants and HA (Fig. 5). This result may be associated with the capacity of As transport from cells, because GSTP1 has no function of As methylation. Considering that the GSTP1 Ile105Val hetero type may have less function to excrete As^III (IA) (Fig. 4), the pathway of methylation from IA to MMA^V might be more dominant than the excretion in the hetero type. In addition, this may be more likely when people are exposed to high As. In the copper mine workers from Chile, %DMA^V in the Val type of GSTP1 Ile105Val was higher than that in the Ile type, although the result was not statistically significant³² (Table 1). Similarly, this study revealed no association of %DMA^V in the GSTP1 Ile105Val genotype.

BMI has been used as an indicator of nutritional status or obesity. By using the stepwise regression analyses, BMI had negative correlations with DMA^V, MMA^V, IA, IMDA, and SA concentrations in HA (Table 4). These results suggest two hypotheses; the exacerbation of nutritional status by As exposure and the effect of increased body fat on As accumulation in the high As exposure group. Similar results were obtained in Vietnam in our previous studies.^39,40 Other studies^47–50 have reported the interaction between BMI and metabolic capacity of As, which was not observed in our studies. In the present study, a negative correlation between BMI and %IA was found only in LA. Increased %DMA^V and decreased %MMA^V with an increase in BMI in local residents were reported from blackfoot disease-hyperendemic areas in Taiwan⁴⁷ and in European males.⁴⁸ On the other hand, there are some contradictory reports, indicating no significant association of BMI with As metabolism.^49,50

Age was positively correlated with M/I and concentrations of DMA^V and IMDA only in HA (Table 4). Similar findings were reported in Vietnamese.^39,40 Kurttio et al. (1998) found a slight increase of DMA^V with age in adults from Finland.⁵¹ In a study of Argentina, %IA decreased with age, but there were no age-dependent variations in %MMA^V, %DMA^V, and D/M.⁵² It has been suggested that children may have a higher 2nd methylation capacity compared to adults.^37,53,54 However, no clear associations were detected between age and urinary D/M or %DMA^V in the present study, probably due to small sample size of children (n= 21 for < 15 years old).

One should notice that adjusted determination coefficients (R²_adj) in the multiple regression equations were moderate (0.016–0.627), even though the p values were less than 0.001. This suggests that there are other factors that are involved in As concentration and metabolism of the participants. Genetic polymorphisms of other As metabolic enzymes such as AS3MT^8,39,40,41 and methylenetetrahydrofolate reductase (MTHFR)^48,52,55 may be one of the potential factors. In addition, several SNPs are known in MRP1.⁵⁶ Further studies are necessary to assess potential effects of these genetic variations on the metabolism and toxicity of As.

4. Conclusions

This study revealed that both environmental (As exposure status) and genetic factors (GSTP1 Ile105Val polymorphism) are significantly associated with the concentration and metabolism of As in humans. Furthermore, it can be suggested that the association of GSTP1 Ile105Val polymorphism with As is enhanced under high As exposure. This means that it can be important when association of genetic polymorphisms in As metabolic enzymes is evaluated in some populations.

A proposed mechanism of As metabolism and excretion by GSTP1 Ile105Val is summarized in Fig. 6. The wild type (Ile type) of GSTP1 Ile105Val may have a high reductive capacity from As^V to As^III. GSTP1 may conjugate GSH to As^III on the membrane and then As^III-GS may be excreted through MRP1. Since the GSTP1 Ile105Val wild type may have a higher activity than the hetero type, the wild type may efficiently excrete As^III-GSH compared with the hetero type. On the contrary, the hetero type (Val type) of GSTP1 Ile105Val may have a relatively lower reduction activity and excretion and thus the metabolism to MMA^V may be more facilitated. These pathways may be accelerated with an increase in the As exposure status in humans. To verify these hypotheses, further studies are required to determine (i) whether GSTP1 can reduce from As^III to As^V like GSTO1; (ii) whether GSTP1 can facilitate excretion of As^III through MRP1; (iii) whether these functions vary among the GSTP1 Ile105Val genotype; and (iv) whether interaction of GSTP1 polymorphism with As excretion and metabolism is influenced by the As exposure level.

Fig. 6 Suspected pathways of As metabolism and excretion by GSTP1 Ile105Val polymorphism. Solid and dashed arrows indicate strong and weak pathways, respectively.

Abbreviations

AB	arsenobetaine
AS3MT	As (+3 oxidation state) methyltransferase
As	arsenic
As^III	arsenite
As^V	arsenate
BMI	body mass index
D/M	DMA^V/MMA^V
DMA^V	dimethylarsinic acid
Eawag	Swiss Federal Institute of Aquatic Science and Technology
GST	glutathione-S-transferase
GSTO1	glutathione-S-transferase ω 1
GSTO2	glutathione-S-transferase ω 2
GSTP1	glutathione-S-transferase π 1
HA	high As exposure group
HH	Hoa Hau
HPLC	high performance liquid chromatograph
IA	inorganic As. Here, As^[V] + As^[III]
ICP-MS	inductively coupled plasma mass spectrometer
III/V	As^III/As^V
IMDA	As^III + As^V + MMA^V + DMA^V
LA	low As exposure group
LT	Liem Thuan
M/I	MMA^V/IA
MMA^V	monomethylarsonic acid
MRP1	multidrug resistance protein 1
MTHFR	methylenetetrahydrofolate reductase
NRCC	National Research Council Canada
PCR-RFLP	PCR restriction fragment length polymorphism
SNP	single nucleotide polymorphism
TA	total As
TH	Thach Hoa
TV	Thanh Vanh
VIF	variance inflation factor

Acknowledgements

We wish to thank Dr A. Subramanian, CMES, Ehime University, Japan for critical reading of the manuscript. The authors express their thankfulness to the staff of the CETASD, Hanoi University of Science and Dr H. Sakai (current affiliation; Department of Pharmacology, Yamaguchi University Graduate School of Medicine, Japan), Dr D. Imaeda (current affiliation; IDEA Consultants, Inc.), and Ms. H. Mizukawa from CMES for their help in sample collection. We also acknowledge Ms. H. Touma, Ms. N. Tsunehiro, and Dr Ogawa, staff of the es-BANK, CMES for their support in sample management and Ms. Y. Fujii, Department of Legal Medicine, Shimane University Faculty of Medicine, Japan for her technical assistance. This study was supported by Japan Society for the Promotion of Science (JSPS) for the cooperative research program under the Core University Program between JSPS and Vietnamese Academy of Science and Technology (VAST). Financial support was also provided by grants from Research Revolution 2002 (RR2002) Project for Sustainable Coexistence of Human, Nature and the Earth (FY2002), Grants-in-Aid for Scientific Research (S) (No. 20221003 and 21221004) and (A) (No. 19209025) from JSPS, and 21st Century and Global COE Programs from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan and JSPS. The award of the JSPS Post Doctoral Fellowship for Researchers in Japan to T. Agusa (No. 207871) is also acknowledged.

References

1. B. K. Mandal and K. T. Suzuki, Arsenic round the world: a review, Talanta, 2002, 58, 201–235.
2. D. K. Nordstrom, Public health. Worldwide occurrences of arsenic in ground water, Science, 2002, 296, 2143–2145.
3. P. L. Smedley and D. G. Kinniburgh, A review of the source, behaviour and distribution of arsenic in natural waters, Appl. Geochem., 2002, 17, 517–568.
4. M. Tondel, M. Rahman, A. Magnuson, I. A. Chowdhury, M. H. Faruquee and S. A. Ahmad, The relationship of arsenic levels in drinking water and the prevalence rate of skin lesions in Bangladesh, Environ. Health Perspect., 1999, 107, 727–729.
5. WHO, Guidelines for Drinking Water Quality, World Health Organization, Geneva, Switzerland, 3rd edn, 2004.
6. M. M. Wu, T. L. Kuo, Y. H. Hwang and C. J. Chen, Dose–response relation between arsenic concentration in well water and mortality from cancers and vascular diseases, Am. J. Epidemiol., 1989, 130, 1123–1132.
7. M. Vahter, Mechanisms of arsenic biotransformation, Toxicology, 2001, 164, 17.
8. T. Agusa, J. Fujihara, H. Takeshita and H. Iwata, Individual variations in inorganic arsenic metabolism associated with AS3MT genetic polymorphisms, Int. J. Mol. Sci., 2011, 12, 2351–2382.
9. F. Challenger, Biological methylation, Chem. Rev., 1945, 36, 315–361.
10. W. R. Cullen and K. J. Reimer, Arsenic speciation in the environment, Chem. Rev., 1989, 89, 713–764.
11. T. Hayakawa, Y. Kobayashi, X. Cui and S. Hirano, A new metabolic pathway of arsenite: arsenic-glutathione complexes are substrates for human arsenic methyltransferase Cyt19, Arch. Toxicol., 2005, 79, 183–191.
12. H. Naranmandura, N. Suzuki and K. T. Suzuki, Trivalent arsenicals are bound to proteins during reductive methylation, Chem. Res. Toxicol., 2006, 19, 1010–1018.
13. H. V. Aposhian and M. M. Aposhian, Arsenic toxicology: five questions, Chem. Res. Toxicol., 2006, 19, 1–15.
14. R. A. Zakharyan and H. V. Aposhian, Enzymatic reduction of arsenic compounds in mammalian systems: the rate-limiting enzyme of rabbit liver arsenic biotransformation is MMA(V) reductase, Chem. Res. Toxicol., 1999, 12, 1278–1283.
15. R. A. Zakharyan, A. Sampayo-Reyes, S. M. Healy, G. Tsaprailis, P. G. Board, D. C. Liebler and H. V. Aposhian, Human monomethylarsonic acid (MMA(V)) reductase is a member of the glutathione-S-transferase superfamily, Chem. Res. Toxicol., 2001, 14, 1051–1057.
16. R. A. Zakharyan, G. Tsaprailis, U. K. Chowdhury, A. Hernandez and H. V. Aposhian, Interactions of sodium selenite, glutathione, arsenic species, and omega class human glutathione transferase, Chem. Res. Toxicol., 2005, 18, 1287–1295.
17. E. M. Schmuck, P. G. Board, A. K. Whitbread, N. Tetlow, J. A. Cavanaugh, A. C. Blackburn and A. Masoumi, Characterization of the monomethylarsonate reductase and dehydroascorbate reductase activities of Omega class glutathione transferase variants: implications for arsenic metabolism and the age-at-onset of Alzheimer's and Parkinson's diseases, Pharmacogenet. Genomics, 2005, 15, 493–501.
18. T. Tanaka-Kagawa, H. Jinno, T. Hasegawa, Y. Makino, Y. Seko, N. Hanioka and M. Ando, Functional characterization of two variant human GSTO 1-1s (Ala140Asp and Thr217Asn), Biochem. Biophys. Res. Commun., 2003, 301, 516–520.
19. (a) A. K. Whitbread, N. Tetlow, H. J. Eyre, G. R. Sutherland and P. G. Board, Characterization of the human Omega class glutathione transferase genes and associated polymorphisms, Pharmacogenetics, 2003, 13, 131–144,  DOI:10.1097/01.fpc.0000054062.98065.6e.
20. L. L. Marnell, G. G. Garcia-Vargas, U. K. Chowdhury, R. A. Zakharyan, B. Walsh, M. D. Avram, M. J. Kopplin, M. E. Cebrian, E. K. Silbergeld and H. V. Aposhian, Polymorphisms in the human monomethylarsonic acid (MMA V) reductase/hGSTO1 gene and changes in urinary arsenic profiles, Chem. Res. Toxicol., 2003, 16, 1507–1513.
21. M. M. Meza, L. Yu, Y. Y. Rodriguez, M. Guild, D. Thompson, A. J. Gandolfi and W. T. Klimecki, Developmentally restricted genetic determinants of human arsenic metabolism: association between urinary methylated arsenic and CYT19 polymorphisms in children, Environ. Health Perspect., 2005, 113, 775–781.
22. L. Paiva, R. Marcos, A. Creus, M. Coggan, A. J. Oakley and P. G. Board, Polymorphism of glutathione transferase Omega 1 in a population exposed to a high environmental arsenic burden, Pharmacogenet. Genomics, 2008, 18, 1–10.
23. J. F. Lo, H. F. Wang, M. F. Tam and T. C. Lee, Glutathione S-transferase pai in an arsenic-resistant Chinese hamster ovary cell line, Biochem. J., 1992, 288, 977–982.
24. L. Zhou, Y. Jing, M. Styblo, Z. Chen and S. Waxman, Glutathione-S-transferase inhibits As₂O₃ induced apoptosis in lymphoma cells: involvement of hydrogen peroxide catabolism, Blood, 2005, 105, 1198–1203.
25. S. L. Zhong, S. F. Zhou, X. Chen, S. Y. Chan, E. Chan, K. Y. Ng, W. Duan and M. Huang, Relationship between genotype and enzyme activity of glutathione S-transferases M1 and P1 in Chinese, Eur. J. Pharm. Sci., 2006, 28, 77–85.
26. P. Ghosh, A. Basu, J. Mahata, S. Basu, M. Sengupta, J. K. Das, A. Mukherjee, A. K. Sarkar, L. Mondal, K. Ray and A. K. Giri, Cytogenetic damage and genetic variants in the individuals susceptible to arsenic-induced cancer through drinking water, Int. J. Cancer, 2006, 118, 2470–2478,  DOI:10.1002/ijc.21640.
27. K. M. McCarty, Y. C. Chen, Q. Quamruzzaman, M. Rahman, G. Mahiuddin, Y. M. Hsueh, L. Su, T. Smith, L. Ryan and D. C. Christiani, Arsenic methylation, GSTT1, GSTM1, GSTP1 polymorphisms, and skin lesions, Environ. Health Perspect., 2007, 115, 341–345.
28. G. F. Lin, H. Du, J. G. Chen, H. C. Lu, W. C. Guo, H. Meng, T. B. Zhang, X. J. Zhang, D. R. Lu, K. Golka and J. H. Shen, Arsenic-related skin lesions and glutathione S-transferase P1 A1578G (Ile105Val) polymorphism in two ethnic clans exposed to indoor combustion of high arsenic coal in one village, Pharmacogenet. Genomics, 2006, 16, 863–871.
29. K. M. McCarty, L. Ryan, E. A. Houseman, P. L. Williams, D. P. Miller, Q. Quamruzzaman, M. Rahman, G. Mahiuddin, T. Smith, E. Gonzalez, L. Su and D. C. Christiani, A case-control study of GST polymorphisms and arsenic related skin lesions, Environ. Health Global Access Sci. Source, 2007, 6, 1–10.
30. Y. H. Wang, M. M. Wu, C. T. Hong, L. M. Lien, Y. C. Hsieh, H. P. Tseng, S. F. Chang, C. L. Su, H. Y. Chiou and C. J. Chen, Effects of arsenic exposure and genetic polymorphisms of p53, glutathione S-transferase M1, T1, and P1 on the risk of carotid atherosclerosis in Taiwan, Atherosclerosis, 2007, 192, 305–312.
31. L. I. Hsu, A. W. Chiu, S. K. Huan, C. L. Chen, Y. H. Wang, F. I. Hsieh, W. L. Chou, L. H. Wang and C. J. Chen, SNPs of GSTM1, T1, P1, epoxide hydrolase and DNA repair enzyme XRCC1 and risk of urinary transitional cell carcinoma in southwestern Taiwan, Toxicol. Appl. Pharmacol., 2008, 228, 144–155.
32. R. Marcos, V. Martinez, A. Hernandez, A. Creus, C. Sekaran, H. Tokunaga and D. Quinteros, Metabolic profile in workers occupationally exposed to arsenic: role of GST polymorphisms, J. Occup. Environ. Med., 2006, 48, 334–341.
33. T. Agusa, T. Kunito, J. Fujihara, R. Kubota, T. B. Minh, P. T. K. Trang, A. Subramanian, H. Iwata, P. H. Viet and S. Tanabe, Contamination by trace elements in groundwater of Vietnam, Biomed. Res. Trace Elem., 2004, 15, 339–341.
34. T. Agusa, S. Inoue, T. Kunito, R. Kubota, T. B. Minh, P. T. K. Trang, A. Subramanian, H. Iwata, P. H. Viet and S. Tanabe, Widely-distributed arsenic pollution in groundwater in the Red River Delta, Vietnam, Biomed. Res. Trace Elem., 2005, 16, 296–298.
35. T. Agusa, T. Kunito, J. Fujihara, R. Kubota, T. B. Minh, P. T. K. Trang, H. Iwata, A. Subramanian, P. H. Viet and S. Tanabe, Contamination by arsenic and other trace elements in tube-well water and its risk assessment to humans in Hanoi, Vietnam, Environ. Pollut., 2006, 139, 95–106.
36. T. Agusa, R. Kubota, T. Kunito, T. B. Minh, P. T. K. Trang, C. Chamnan, H. Iwata, P. H. Viet, T. S. Tana and S. Tanabe, Arsenic pollution in groundwater of Vietnam and Cambodia: a review, Biomed. Res. Trace Elem., 2007, 18, 35–47.
37. T. Agusa, T. Kunito, T. B. Minh, P. T. K. Trang, H. Iwata, P. H. Viet and S. Tanabe, Relationship of urinary arsenic metabolites to intake estimates in residents of the Red River Delta, Vietnam, Environ. Pollut., 2009, 157, 396–403.
38. T. Agusa, S. Inoue, T. Kunito, T. B. Minh, N. N. Ha, N. P. C. Tu, P. T. K. Trang, H. Iwata, P. H. Viet, B. C. Tuyen and S. Tanabe, Human exposure to arsenic from groundwater in the Red River and the Mekong River Deltas in Vietnam, Int. J. Environ. Stud., 2009, 66, 49–57.
39. T. Agusa, H. Iwata, J. Fujihara, T. Kunito, H. Takeshita, T. B. Minh, P. T. K. Trang, P. H. Viet and S. Tanabe, Genetic polymorphisms in AS3MT and arsenic metabolism in residents of the Red River Delta, Vietnam, Toxicol. Appl. Pharmacol., 2009, 236, 131–141.
40. T. Agusa, H. Iwata, J. Fujihara, T. Kunito, H. Takeshita, T. B. Minh, P. T. K. Trang, P. H. Viet and S. Tanabe, Genetic polymorphisms in glutathione S-transferase (GST) superfamily and arsenic metabolism in residents of the Red River Delta, Vietnam, Toxicol. Appl. Pharmacol., 2010, 242, 352–362.
41. T. Agusa, T. Kunito, R. Kubota, S. Inoue, J. Fujihara, T. B. Minh, N. N. Ha, N. P. C. Tu, P. T. K. Trang, C. Chamnan, H. Takeshita, H. Iwata, B. C. Tuyen, P. H. Viet, T. S. Tana and S. Tanabe, Exposure, metabolism, and health effects of arsenic in residents from arsenic-contaminated groundwater areas of Vietnam and Cambodia: A review, Rev. Environ. Health, 2010, 25, 193–220.
42. H. Iwata, E. Y. Kim, M. Yamauchi, S. Inoue, T. Agusa and S. Tanabe, Chemical contamination in aquatic ecosystems, Yakugaku Zasshi—J. Pharm. Soc. Jpn, 2007, 127, 417–428.
43. R. Kubota, T. Kunito, T. Agusa, J. Fujihara, I. Monirith, H. Iwata, A. Subramanian, T. S. Tana and S. Tanabe, Urinary 8-hydroxy-2′-deoxyguanosine in inhabitants chronically exposed to arsenic in groundwater in Cambodia, J. Environ. Monit., 2006, 8, 293–299.
44. S. Tanabe, Environmental Specimen Bank in Ehime University (es-BANK), Japan for global monitoring, J. Environ. Monit., 2006, 8, 782–790.
45. C. H. Tseng, A review on environmental factors regulating arsenic methylation in humans, Toxicol. Appl. Pharmacol., 2009, 235, 338–350.
46. E. M. Leslie, A. Haimeur and M. P. Waalkes, Arsenic transport by the human multidrug resistance protein 1 (MRP1/ABCC1): evidence that a tri-glutathione conjugate is required, J. Biol. Chem., 2004, 279, 32700–32708.
47. C.-H. Tseng, Y.-K. Huang, Y.-L. Huang, C.-J. Chung, M.-H. Yang, C.-J. Chen and Y.-M. Hsueh, Arsenic exposure, urinary arsenic speciation, and peripheral vascular disease in blackfoot disease-hyperendemic villages in Taiwan, Toxicol. Appl. Pharmacol., 2005, 206, 299–308.
48. A. L. Lindberg, R. Kumar, W. Goessler, R. Thirumaran, E. Gurzau, K. Koppova, P. Rudnai, G. Leonardi, T. Fletcher and M. Vahter, Metabolism of low-dose inorganic arsenic in a central European population: influence of sex and genetic polymorphisms, Environ. Health Perspect., 2007, 115, 1081–1086.
49. L. Li, E. C. Ekström, W. Goessler, B. Lönnerdal, B. Nermell, M. Yunus, A. Rahman, S. El Arifeen, L. Å. Persson and M. Vahter, Nutritional status has marginal influence on the metabolism of inorganic arsenic in pregnant Bangladeshi women, Environ. Health Perspect., 2008, 116, 315–321.
50. A. L. Lindberg, E. C. Ekström, B. Nermell, M. Rahman, B. Lönnerdal, L. A. Persson and M. Vahter, Gender and age differences in the metabolism of inorganic arsenic in a highly exposed population in Bangladesh, Environ. Res., 2008, 106, 110–120.
51. P. Kurttio, H. Komulainen, E. Hakala, H. Kahelin and J. Pekkanen, Urinary excretion of arsenic species after exposure to arsenic present in drinking water, Arch. Environ. Contam. Toxicol., 1998, 34, 297–305.
52. C. Steinmaus, L. E. Moore, M. Shipp, D. Kalman, O. A. Rey, M. L. Biggs, C. Hopenhayn, M. N. Bates, S. C. Zheng, J. K. Wiencke and A. H. Smith, Genetic polymorphisms in MTHFR 677 and 1298, GSTM1 and T1, and metabolism of arsenic, J. Toxicol. Environ. Health, Part A, 2007, 70, 159–170,  DOI:10.1080/15287390600755240.
53. U. K. Chowdhury, M. M. Rahman, M. K. Sengupta, D. Lodh, C. R. Chanda, S. Roy, Q. Quamruzzaman, H. Tokunaga, M. Ando and D. Chakraborti, Pattern of excretion of arsenic compounds [arsenite, arsenate, MMA(V), DMA(V)] in urine of children compared to adults from an arsenic exposed area in Bangladesh, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng., 2003, 38, 87–113.
54. J. S. Chung, D. A. Kalman, L. E. Moore, M. J. Kosnett, A. P. Arroyo, M. Beeris, D. N. Guha Mazumder, A. L. Hernandez and A. H. Smith, Family correlations of arsenic methylation patterns in children and parents exposed to high concentrations of arsenic in drinking water, Environ. Health Perspect., 2002, 110, 729–733.
55. K. Schläwicke Engström, K. Broberg, G. Concha, B. Nermell, M. Warholm and M. Vahter, Genetic polymorphisms influencing arsenic metabolism: evidence from Argentina, Environ. Health Perspect., 2007, 115, 599–605,  DOI:10.1289/Ehp.9734.
56. I. J. Letourneau, R. G. Deeley and S. P. C. Cole, Functional characterization of non-synonymous single nucleotide polymorphisms in the gene encoding human multidrug resistance protein 1 (MRP1/ABCC1), Pharmacogenet. Genomics, 2005, 15, 647–657.

---