DSpace at VNU: Genetic polymorphisms in AS3MT and arsenic metabolism in residents of the Red River Delta, Vietnam

DSpace at VNU: Genetic polymorphisms in AS3MT and arsenic metabolism in residents of the Red River Delta, Vietnam tài li...

Genetic polymorphisms in AS3MT and arsenic metabolism in residents of the Red River Delta, Vietnam

Tetsuro Agusaa,b, Hisato Iwataa,⁎ , Junko Fujiharab, Takashi Kunitoc, Haruo Takeshitab, Tu Binh Minha,d,

a

Center for Marine Environmental Studies (CMES), Ehime University, Bunkyo-cho 2-5, Matsuyama 790-8577, Japan

b

Department of Legal Medicine, Shimane University Faculty of Medicine, Enya 89-1, Izumo 693-8501, Japan

c

Department of Environmental Sciences, Faculty of Science, Shinshu University, 3-1-1 Asahi, Matsumoto 390-8621, Japan

d

Center for Environmental Technology and Sustainable Development (CETASD), Hanoi University of Science, Vietnam National University, T3 Building, 334 Nguyen Trai Street, Thanh Xuan District, Hanoi, Vietnam

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 9 December 2008

Revised 23 January 2009

Accepted 25 January 2009

Available online 31 January 2009

Keywords:

Arsenic

AS3MT

Polymorphism

SNP

Vietnam

Groundwater

Sand-filtered water

Human urine

Human hair

To elucidate the role of genetic factors in arsenic (As) metabolism, we studied associations of single nucleotide polymorphisms (SNPs) in As (+3 oxidation state) methyltransferase (AS3MT) with the As concentrations in hair and urine, and urinary As profile in residents in the Red River Delta, Vietnam Concentrations of total As in groundwater were 0.7–502 μg/l Total As levels in groundwater drastically decreased by using sand filter, indicating that thefilter could be effective to remove As from raw groundwater Concentrations of inorganic As (IAs) in urine and total As in hair of males were higher than those of females A significant positive correlation between monomethylarsonic acid (MMA)/IAs and age in females indicates that older females have higher methylation capacity from IAs to MMA Body mass index negatively correlated with urinary As concentrations in males Homozygote for SNPs 4602AA, 35991GG, and 37853GG, which showed strong linkage disequilibrium (LD), had higher percentage (%) of dimethylarsinic acid (DMA) in urine SNPs 4740 and 12590 had strong LD and associated with urinary %DMA Although SNPs 6144,12390,14215, and 35587 comprised LD cluster, homozygotes

in SNPs 12390GG and 35587CC had lower DMA/MMA in urine, suggesting low methylation capacity from MMA

to DMA in homo types for these SNPs SNPs 5913 and 8973 correlated with %MMA and %DMA, respectively Heterozygote for SNP 14458TC had higher MMA/IAs in urine than TT homozygote, indicating that the heterozygote may have stronger methylation ability of IAs To our knowledge, this is thefirst study on the association of genetic factors with As metabolism in Vietnamese

© 2009 Elsevier Inc All rights reserved

Introduction

Consumption of arsenic (As)-polluted groundwater has adversely

affected human health in certain areas of the world (Mandal and

Suzuki, 2002; Nordstrom, 2002; Smedley and Kinniburgh, 2002)

Recently,Berg et al (2001) and Agusa et al (2004, 2005, 2006, 2007,

2009) reported elevated As contamination (up to 3150 μg/l) in

groundwater of the Red River Delta in Northern Vietnam; many

groundwater samples contained As over the WHO drinking water

guideline (10μg/l) (WHO, 2004)

It is known that As exposure causes lung and skin cancers and also

birth defects (WHO, 2004) There seems to be a wide variation in the

susceptibility to As toxicity among individuals and populations, which

is probably related to genetic factors in metabolism of As (Vahter,

2002) It has been generally accepted that inorganic As is oxidatively

methylated in the body (Challenger, 1945; Cullen and Reimer, 1989),

but recently reductive methylation pathway has also been proposed

(Hayakawa et al., 2005; Naranmandura et al., 2006) In either case, methylation is a critical metabolic pathway for As biotransformation, since the toxicity of organic As is generally lower than that of inorganic forms, and also the methylated As would be readily excreted into the urine In these processes, As (+ 3 oxidation state)

enzyme, catalyzes the methylation of arsenite (AsIII) and monomethyl

As (Lin et al., 2002; Wood et al., 2006) It is known that human AS3MT gene is approximately 32-kb long and is composed of 11 exons (Wood

et al., 2006)

It has been reported that there are some single nucleotide

Recombinant 173AlaNTrp (Ala to Trp substitution at amino acid base 173), 287MetNThr, and 306ThrNIle variants in AS3MT significantly altered levels of the enzyme activity and immunoreactive protein (Wood et al., 2006) 287MetNThr heterozygote was linked with increased percentage of monomethylarsonic acid (MMA) in urine of central European population (Lindberg et al., 2007) and miners in Chile (Hernandez et al., 2008) For SNPs in intron of AS3MT,Meza et al (2005) reported association between intronic SNPs 7395GNA,

⁎ Corresponding author Fax: +81 89 927 8172.

E-mail address: iwatah@agr.ehime-u.ac.jp (H Iwata).

0041-008X/$ – see front matter © 2009 Elsevier Inc All rights reserved.

Contents lists available atScienceDirect Toxicology and Applied Pharmacology

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 ta a p

12390GNC, and 35587TNC in AS3MT, and urinary dimethylarsinic acid

(DMA)/MMA values in the Mexican children Three intronic

were found to be associated with a lower percentage of MMA and a

Engström et al., 2007) Hence, SNPs in AS3MT may be responsible

for inter-individual variation in the As metabolism Given these

results, information on the genotyping of polymorphism in AS3MT

may help in understanding genetic susceptibility to As toxicity

However, data on distribution of AS3MT polymorphisms and their

relations to As methylation ability are limited, especially among Asian

populations (Fujihara et al., 2007)

To understand the importance of certain biological and

environ-mental factors in inter-individual variation in As metabolism, we

initially investigated the effects of sex, age, body mass index (BMI),

occupation, residential years, and alcohol and smoking habits on the

As metabolites in urine of residents from the Red River Delta, Vietnam

Moreover, to comprehend whether or not the genetic factor can affect

As metabolism, we studied the relationships between 13 SNPs in

AS3MT and urinary metabolite pattern of As

Materials and methods

(n = 28) were randomly collected at each home in rural areas of Hoa

Hau (HH) and Liem Thuan (LT) in Ha Nam Province located in the Red

River Delta, Vietnam during March (dry season), 2006 There are no

significant anthropogenic As pollution sources such as industrial sites

and mining regions in both locations Because some houses equipped

a sand filter system for the well, filtrated groundwater samples

(n = 19) were also collected along with unfiltered water samples from

both locations

Human hair (n = 99), urine (n = 100), and blood (n = 100) were

correspondingly collected from the residents of each house equipped

with the tube well in an ethical manner All participants were

randomly selected without an arbitrary manner We obtained the

informed consent from all the subjects The study was approved by the

Ethical Committee of Ehime University, Japan For donors who

participated in this study, information on age (mean, 35.8 years;

range, 11–70 years), sex (male, n=44; female, n=56), height (mean,

153 cm; range, 121–173 cm), weight (mean, 46 kg; range, 22–67 kg),

body mass index (BMI; mean, 19.6; range, 12.1–30.0), occupation

(farmer, n = 32; farmer with weaver, n = 4; weaver, n = 32; student,

n = 29; retired worker, n = 2; bricklayer, n = 1), residential years

(mean, 32 years; range, 3–65 years), and alcohol (yes, n=24; no,

n = 76) and smoking habits (yes, n = 20; no, n = 80) were obtained

All samples were preserved in the Environmental Specimen Bank (es-BANK), Center for Marine Environmental Studies (CMES), Ehime University, Japan at−25 °C (Tanabe, 2006) until chemical analysis and genotyping were conducted The details of wells and subjects are shown inTable 1

Analyses of total As and As compounds Water sample was acidified with nitric acid for total As (TAs) analysis Human hair sample washed

by sonication with 0.3% polyoxyethylene lauryl ether was dried for

12 h at 80 °C (Agusa et al., 2006) The dried hair sample was digested

by nitric acid with a microwave oven (Agusa et al., 2006) Thawed urine wasfiltrated with a syringe filter (Millex-HV, 0.45 μm syringe-drivenfilter unit, Millipore) and then diluted by Milli-Q water Concentrations of TAs in groundwater and human hair were measured with an inductively coupled plasma-mass spectrometer (ICP-MS; HP-4500, Hewlett-Packard, Avondale, PA, USA) Yttrium was used as an internal standard for ICP-MS measurement Six arsenic compounds including arsenocholine (AC), arsenobetaine (AB), DMA, MMA, AsIII, and arsenate (AsV) were determined in urine samples with

a high performance liquid chromatograph (HPLC; LC10A Series, Shimadzu, Kyoto, Japan) coupled with ICP-MS using an anion

2001; Agusa et al., 2009) The column was equilibrated with a mobile phase (15 mM citric acid, pH 2.0 with nitric acid) at aflow rate of 1.0 ml/min at 30 °C for more than 2 h before analysis The injection

standards for AB and other arsenicals, respectively Sum of all As compounds, AsIII+ AsV, and AsIII+ AsV+ MMA + DMA detected in urine are represented as SAs, IAs, and IMDAs, respectively Percentages

of AB, AsIII, AsV, MMA, DMA, IAs, and IMDAs to SAs in human urine were denoted as %AB, %AsIII, %AsV, %MMA, %DMA, %IAs, and %IMDAs, respectively Urinary creatinine was measured by SRL, Inc (Tokyo, Japan), and concentrations of As compounds in urine were expressed

asμg As/g on a creatinine basis Detection limits of TAs in water and hair, and As compounds (AC, AB, AsIII, AsV, MMA, and DMA) in urine were 0.1μg/l, 0.01 μg/g dry wt, and 1 μg/g creatinine, respectively in our methods

reference materials, SLRS-4 River Water from the National Research Council Canada (NRCC) and NIES No.13 Human Hair and NIES No.18 Human Urine provided by the National Institute for Environmental Studies (NIES), Japan were analyzed for TAs and As compounds (AB and DMA), respectively Results of TAs in the water and hair and AB and DMA in the urine were in very good agreement with the certified values, and the recoveries were in the range of 90–106% The analytical precision for these samples (n = 3) were within 4% In addition, we have participated in an inter-calibration exercise organized by the Swiss Federal Institute of Aquatic Science and Technology (Eawag) in the frame of the ongoing cooperation of Vietnam and Switzerland in

As related surveys and researches for analytical quality assurance and control

AS3MT was conducted by polymerase chain reaction and restriction fragment length polymorphism (PCR-RFLP) technique (Fujihara et al

2007) First, genomic DNA was extracted from blood sample using QIAamp DNA mini kit (Qiagen, Hilden, Germany), and genotyped for

(Met to Thr substitution at amino acid base)), 35587TNC, 35991GNA, and 37853GNA Primers for AS3MT were designed based on the DDBJ Sequence Database under accession no AY817668 (Table 2) The mismatched PCR method (Kumar and Dunn, 1989) was employed for identifying a new restriction enzyme site for the detection of SNPs One microgram of DNA was subjected to PCR amplification in 10 μl

Table 1

Information on water and human samples from Hoa Hau and Liem Thuan in Vietnam

Location Hoa Hau Liem Thuan

Groundwater

Used period (years) a 9 (5.5–13) 6 (1–16)

Well depth (m) a 14 (8–16) 15 (12–24)

Filtered water

Subjects

No of male/female 22/29 22/27

Age (years) a 37 (11–60) 34 (11–70)

Residential time (years) a

33 (3–60) 31 (6–65) Height (cm) a

156 (137–173) 150 (121–169) Weight (kg) a 48 (27–66) 44 (22–67)

No of smokers/non smokers 14/37 6/43

No of drinkers/non drinkers 14/37 10/39

BMI a , b 20 (14–26) 19 (12–29)

a

Arithmetic mean and range.

b

Body mass index (weight (kg)/height (m) 2

).

reaction mixture containing GoTaq® Green Master Mix (Promega,

Madison WI, USA) and each primer pair corresponding to each SNP

Information on primer sequence, annealing temperature, restriction

enzyme, and PCR product size is listed inTable 2 The PCR product was

digested with restriction enzyme and then was subjected to

electrophoresis on 8% polyacrylamide gel

Statistical analyses All statistical analyses were performed with

StatView (version 5.0, SAS®Institute, Cary, NC, USA) and SPSS (version

12, SPSS, Chicago, IL, USA) One half of the value of the respective limit

of detection was substituted for those values below the limit of

detection and used in statistical analysis All data were tested for

goodness offit to a normal distribution with Kolmogorov–Smirnov's

one sample test Because some variables were not normally

distributed, log transformation was conducted for parametric

analyses Outlier (As concentration of 2120 μg/l in groundwater)

was checked by Thompson test Relationships among concentrations

of As in water, hair, and urine, composition of As compounds in urine,

age, residential period, and BMI were examined by Pearson's

correlation coefficient test Student's t-test was used to detect

influences of regions, sexes, and smoking and drinking habits on As

levels in water, hair and urine, and urinary metabolites Differences in

As concentrations between genetic polymorphisms in AS3MT were

Relationships between genetic polymorphisms in AS3MT, and hair and

urinary As concentrations and As compositions were also examined by multiple regression analysis including sex, age, and BMI as independent variables The residual values were obtained from the difference between the actual value and the predicted value from the regressions These residual values reflect the variation after the effects

of the regression variables have been removed Linkage disequilibrium (LD) and haplotype of SNPs in AS3MT were calculated using Haploview (version 4.0, Day Lab at the Broad Institute Cambridge,

MA, USA) A p value of less than 0.05 was considered to indicate statistical significance Geometric mean value was represented as GM

in this study

Results and discussion Concentration of total As in groundwater Arsenic was detected in all groundwater samples and the levels

concentration of TAs (2120μg/l) was removed for further statistics,

extraordinary high As level of the outlier might be due to the presence

of large amount of particulate matters found in this sample In the data set without this outlier, the range was 0.7–502 μg/l (Table 3) Arsenic

significantly higher (pb0.001) than that from LT (GM, 1.4 μg/l)

Table 2

Information on primer sequences, annealing temperatures, restriction enzymes, and fragment sizes of the amplified products used for PCR-RFLP

SNP ID a rs number b Functional region Nucleotide

change

Amino acid change

Primer sequence c Annealing

temp (°C)

Restriction enzyme

Fragment size (bp)

4602 rs7085104 5′ upstream region A NG F: 5′-CGAAGAAACTTGTGGGCCAGA-3′ 60 MspI AA: 261

R: 5′-TCGCTCCACTGCGATTTTCAC-3′ AG: 261, 200, 61

GG: 200, 61

4740 rs12416687 5′ upstream region T NC F: 5′-CGAAGAAACTTGTGGGCCAGA-3′ 56.5 MslI TT: 224

R: 5′-CTGATTTAAATGAACACTCAC(CNG)T-3′ TC: 224, 205, 19

CC: 205, 19

5913 rs4917986 Intron 3 T NC F: 5′-GGTCACTAGGGAATTAACCCG-3′ 61 BglI TT: 251

R: 5′-TGGCTATGTTGACCAAGCTGG-3′ TC: 251, 151, 100

CC: 151, 100

6144 rs17878846 Intron 3 A NT F: 5′-GGTCACTAGGGAATTAACCCG-3′ 61 XbaI AA: 415

R: 5′-GGTTCCAACTAATCACCCACG-3′ AT: 415, 376, 39

TT: 376, 39

7395 rs12767543 Intron 3 GNA F: 5′-CGCCTATGGGACAGAAACCTT-3′ 55 AlwNI GG: 155

R: 5′-CTAAGGGACAGAGT(GNC)AGACTC-3′ GA: 155, 138, 17

AA: 138, 17

8979 rs7920657 Intron 5 T NA F: 5′-AGAGTGCAGTGGCCCAATGTC-3′ 63 NlaIII TT: 160, 94

R: 5′-TGAGCACAGTGCCTCACACCT-3′ TA: 254, 160, 94

AA: 254

12390 rs3740393 Intron 6 GNC F: 5′-GTTCCCCTATTCCTTTC(TNA)TTG-3` 51 MslI GG: 286

R: 5′-AACCTTGGCCTCATGGCCTAA-3′ GC: 286, 265, 21

CC: 265, 21

12590 rs3740392 Intron 7 T NC F: 5′-GTTTCAGCATGGTGGGGAGTT-3′ 51 BslI TT: 164

R: 5′-CTG(GNC)CTATTAGC-3′ TC: 164, 157, 7

CC: 157, 7

14215 rs3740390 Intron 8 CNT F: 5′-CTGTACAATGGTAACCCCCCA-3′ 63 Hpyl88I CC: 320, 81

R: 5′-GCAAGGGCAAGAGCAGAAAGA-3′ CT: 401, 320, 81

TT: 401

14458 rs11191439 Exon 9 T NC Met NThr F: 5′-GTGCTGGAGATGAACCGTGAA-3′ 59 HpyCH4IV TT: 233

R: 5′-GCAAGGGCAAGAGCAGAAAGA-3′ TC: 233, 154, 79

CC: 154, 79

35587 rs11191453 Intron 10 T NC F: 5′-CAGCAGTCTTGTCTTTTAAAT(ANT)AA-3′ 58 AseI TT: 194, 21

R: 5′-CCTCTTTGGAACTGAGATACGG-3′ TC: 215, 194, 21

CC: 215

35991 rs10748835 Intron 10 GNA F: 5′-CACGTGCAAATGCAACCCCA-3′ 51 ApaLI GG: 205, 22

R: 5′-GTTTGATTTAGGTTGAC(TNG)T(ANG)CA-3′ GA: 227, 205, 22

AA: 227

37853 rs11191459 3′ downstream region GNA F: 5′-CATGGTGAGACCCCCATCTCT-3′ 60 MspI GG: 415, 42

R: 5′-CCTGATGATAATGACC-3′ GA: 457, 415, 42

AA: 457

a SNP ID indicates the SNP identification number relative to the location in the consensus sequence, with the first base of the consensus numbered 1.

b

Rs numbers were cited from NCBI SNP Database ( http://www.ncbi.nlm.nih.gov/projects/SNP/ ).

c NNN in parenthesis indicates substitution of nucleotide for constriction of mismatched nucleotide.

(Table 3) Furthermore, all the samples (n = 15) from HH contained

TAs concentrations (range, 163–502 μg/l) exceeding 10 μg/l of WHO

drinking water guideline (WHO, 2004) This result suggests that the

groundwater in HH is heavily contaminated by As and the health of

this population is at risk On the contrary, As concentrations in all

concentration in groundwater was not related to depth and usage

history of the wells

Efficiency of sand filter for As removal

Concentration ranges of TAs in sand-filtered groundwater were

3.2–143 μg/l in HH and 1.0–4.9 μg/l in LT (Table 3) In thefiltered

water, regional difference in TAs concentration was observed

18.9μg/l) was significantly low (pb0.001) compared with the raw groundwater (GM, 368μg/l), and GM of the removal efficiency of TAs from raw groundwater was estimated to be 93% (Fig 1).Berg et al (2006) also reported that about 80% of As was removed from groundwater by sand filtration in Vietnam Therefore, the simple sandfilter system can be useful to remove TAs from contaminated groundwater in these regions However, severalfiltered water samples

in HH still contained TAs concentrations over 10μg/l of the WHO guideline value (WHO, 2004) (Fig 1), indicating that sandfiltration is not enough to remove excessive As and the water is not suitable for drinking in some cases

Concentration of total As in human hair Human hair could be a useful indicator of contamination status of trace elements including As, because it is easy for non-invasive collection, transportation, and preservation (Matsubara and Machida,

1985) Total As concentrations in hair of residents from HH and LT were in the range of 0.028–2.94 μg/g dry wt (Table 3) Similar to the results of groundwater, TAs level in human hair of HH (GM, 0.351μg/g dry wt) was significantly higher (pb0.001) than that of LT (GM, 0.232μg/g dry wt) Total As concentrations in hair of three individuals (1.00μg/g dry wt, 2.67 μg/g dry wt, and 2.94 μg/g dry wt) from HH exceeded the level of 1μg/g dry wt which may be a level that can induce skin lesion (Arnold et al., 1990), suggesting potential high risk for some residents in this area

Concentration and composition of As compounds in human urine

It is generally known that ingested inorganic As is methylated to MMA, followed by DMA and then they are excreted into the urine in humans (Styblo et al., 2002) Therefore, As speciation is important for assessing exposure and metabolic capacity of As in humans Concentra-tions of As species in human urine from the HH and LT are shown inTable

3 Concentration ranges of SAs and IMDAs were 38.6–397 μg/g creatinine and 33.0–320 μg/g creatinine, respectively There was no significant difference in concentrations of urinary SAs between HH (GM, 92.6μg/g creatinine) and LT (GM, 97.9μg/g creatinine), which was inconsistent with the results of groundwater and human hair The reason is discussed

in the next section Similar composition of urinary As compounds was found in the residents of both HH and LT, showing DMA as the predominant form and inorganic arsenicals as minor species Arseno-betaine, which may be mainly derived from consumption of seafood, was also detected in the urine as the second abundant As species In the present study, AC was not detected in any of the urine samples

Table 3

Concentrations (geometric mean and range) of total As and As compounds in water (μg/l),

human hair (μg/g dry wt) and urine (μg/g creatinine) from Hoa Hau and Liem Thuan in

Vietnam

Location Hoa Hau Liem Thuan

Groundwater

TAs ⁎ 368 (163–502, and 2120 (an outlier)) 1.4 (0.7–6.8)

Filtered water

TAs ⁎ 18.9 (3.2–143) 2.0 (1.0–4.9)

Drinking water

TAs ⁎ 50.1 (3.2–486) 1.7 (0.9–4.9)

Human hair

TAs ⁎ 0.351 (0.028–2.94) 0.232 (0.068–0.690)

Human urine

AB 17.7 (2.5–72.6) 14.5 (2.1–232)

DMA 50.5 (22.5–268) 56.4 (20.2–132)

MMA 9.3 (3.5–23.9) 9.3 (3.8–23.1)

As III 6.9 (b1.0–32.2) 7.1 (b1.0–26.6)

As V 1.5 (b1.0–12.7) 1.7 (b1.0–19.1)

SAs 92.6 (45.2–365) 97.9 (38.6–397)

IAs 9.3 (3.1–38.2) 10.2 (4.0–28.6)

IMDAs 70.5 (33.6–320) 77.3 (33.0–176)

TAs; total As.

SAs; sum of all As compounds.

IAs; sum of As III

+ As V

.

IMDAs; sum of IAs + MMA V

+ DMA V

Drinking water; In a house equipped with sand filter, filtered water instead of raw

groundwater is assumed to be consumed.

⁎ Significant difference (pb0.001) between locations.

Fig 1 Concentrations of TAs in raw and sand-filtrated groundwater from Hoa Hau and Liem Thuan in Vietnam.

Relationships among concentration of As in drinking water, hair

and urine

To understand whether subjects in these areas are exposed to As

mainly through the consumption of groundwater, relationships

between As concentrations in water, and in hair and urine were examined Since it is expected that residents of the house equipped

exposure status for these subjects A significant positive correlation was found between TAs concentrations of drinking water and hair in residents from HH and LT (pb0.001;Fig 2a) On the contrary, there were no positive correlations between concentrations of TAs in water and any of the As compounds in human urine of all the residents (the result of IMDAs is only shown inFig 2b) Hence, it seems that uptake of As from other source(s) such as food might affect the urinary concentration especially for the LT residents However, since food habit was almost similar between locations but TAs levels in human hair were positively correlated with those in drinking water, peoples in HH may be not recently exposed to As from the water

Hair As levels were positively correlated with concentrations of urinary IMDAs (pb0.001; Fig 2c) and SAs (p = 0.002) Similar relationship was obtained in previous investigations in the

(Kurttio et al., 1998) Since hair can be a good indicator of past As exposure status, while As level in urine can represent recent exposure of As (Mandal and Suzuki, 2002), these results indicate that the residents in HH and LT may have been chronically exposed

to As from groundwater

Potential effects of sex, age, and BMI on As concentration and metabolism

Influences of age, sex, BMI, occupation, residential years, and alcohol and smoking habits on As levels and metabolites in the residents were

Fig 2 Relationships between (a) concentrations of TAs in drinking water and human

hair, (b) concentrations of TAs in drinking water and IMDAs in human urine, and

(c) concentrations of TAs in human hair and IMDAs in human urine from Hoa Hau

and Liem Thuan in Vietnam Solid line in each figure represents the regression line

0.123 2 = 0.155, pb0.001; (c) y=0.014x 0.691 2 b0.001).

Fig 3 Comparison of (a) %IAs and (b) DMA/MMA ratio in urine between females and males from Hoa Hau and Liem Thuan in Vietnam Bar and plots indicate arithmetic mean and individual values, respectively.

investigated Occupation, residential period, and alcohol and smoking

habits were not significantly correlated with As concentrations and

profiles in the residents Concentrations of MMA (p=0.034), AsIII

(p = 0.033), and IAs (p = 0.004) in urine and TAs (p = 0.012) in hair of

males were significantly higher than those of females Also, males

showed higher %MMA (p = 0.001), %AsIII(p = 0.003), %AsV(p = 0.043),

and %IAs (pb0.001,Fig 3a) in urine compared to females Concentration

ratio of urinary DMA/MMA, which can be a useful indicator as 2nd

methylation step of IAs, was significantly higher in females (p=0.008;

Fig 3b) These results suggest that males might be highly exposed to IAs

and/or might have a lower 2nd methylation capacity of As compared to

females In our previous study, sexual difference in hair As concentration

was not significant for residents in Gia Lam and Thanh Tri in Hanoi,

Vietnam (Agusa et al., 2006) Interestingly, urinary As concentrations in

females from As-contaminated sites were higher than those in males,

while opposite trend was observed in its reference site (Chiou et al.,

1997; Chowdhury et al., 2003; Loffredo et al., 2003) Although, the

reason still remains unclear, our result is consistent with the study in

non-As contaminated sites.Watanabe et al (2001)reported that from

an As contaminated area of Bangladesh, concentrations of As in urine of

females were high but severe cases of skin pathologies were remarkably

found in males, and suggested that females may prevent toxic effects

(ex skin disorder) from chronic As exposure by immediately excreting

of As in their body compared with males However, further studies on

the mechanism of sexual difference in As excretion are necessary

It is known that creatinine level in urine of males is generally

higher than that of females because its excretion into urine is related

to mass of muscle Therefore, it is expected that sexual difference in

urinary creatinine level might affect the above-mentioned results However, no significant difference was observed for creatinine between males and females in this study Also, even when we used data of urinary As levels and compositions without correction by creatinine, similar results were obtained

Concentration ratio of urinary MMA/IAs, an index of 1st methylation step of IAs, significantly increased with age in all subjects (p=0.005) Greater age-dependence was observed for females (pb0.001,Fig 4a) Furthermore, a significant negative correlation was observed between age and %IAs in urine of females (p = 0.013,Fig 4b) Therefore, older persons may have a higher methylation capacity from IAs to monomethylated As than young people and also this tendency is possibly pronounced in female Several previous studies reported that children may have a higher 2nd methylation capacity compared to adults (Agusa et al., 2009; Chowdhury et al., 2003; Chung et al., 2002), but no significant correlation between age and DMA/MMA in urine was found in the present study This difference may be due to the small sample size of children No age-dependent variation in human hair As concentration in this study was consistent to our previous study conducted for Vietnamese (Agusa et al., 2006)

Significant negative correlations between BMI as an indicator of nutritional status (Bailey and Ferro-Luzzi, 1995) and concentrations of urinary DMA (p = 0.003), MMA (p = 0.044), AsV (p = 0.008), IAs (p = 0.006), SAs (p = 0.005), and IMDAs (p = 0.002), and hair TAs (p = 0.027) were found in all the residents These correlations except IAs in urine and TAs in hair were also observed only in males Representative results of DMA and IMDAs are shown inFigs 5a and b,

Fig 4 Relationships between age and (a) MMA/IAs, and (b) %IAs in human urine from

Hoa Hau and Liem Thuan in Vietnam Solid lines in eachfigure represent the regression

lines for females ((a) [MMA/IAs] = 0.018 × [age] + 0.49, R 2

= 0.303, pb0.001, (b) % IAs =−0.075×[age]+12, R 2

Fig 5 Relationships between BMI and (a) concentration of DMA, and (b) IMDAs in human urine from Hoa Hau and Liem Thuan in Vietnam Solid line in each figure represents the regression line for males ((a) [Log DMA] =−0.032 × [BMI]+2.33,

R 2 = 0.278, pb0.001, (b) [Log IMDAs]=−0.029×[BMI]+2.43, R 2 = 0.254, pb0.001).

respectively These results might suggest that nutrition status was

exacerbated by As exposure in the residents, especially in males

Alternatively, people with poorer nutritional status might accumulate

more As in their body

Genetic polymorphisms in AS3MT

Genetic factors could also be one of critical factors for As

metabolism (Vahter, 2000) The present study investigated potential

influence of nine SNPs (4602, 4740, 7395, 12390, 12590, 14215, 14458,

35587, and 35991) in AS3MT that may be involved in As metabolism

(Meza et al., 2005; Wood et al., 2006; Schläwicke Engström et al.,

2007; Lindberg et al., 2007; Hernandez et al., 2008) Also, four SNPs

(5913, 6144, 8979, and 37853) with relatively high frequency of alleles

in AS3MT (Meza et al., 2005) were identified by PCR-RFLP in the present study Numbers of subjects for each genotype are shown in

Table 4 CC homo type of 5913 was found only in one subject and there was no homo type of 14458CC in this population

Distinct LD groups were found for SNPs in AS3MT (Fig 6); 4602–

35991 (R2= 0.71), 4602–37853 (R2= 0.53), 4740–12590 (R2= 0.61),

6144–12390 (R2= 0.82), 6144–14215 (R2= 0.72), 6144–35587 (R2= 0.62), 12390–14215 (R2

= 0.84), 12390–35587 (R2

= 0.74),

14215–35587 (R2= 0.78), and 35991–37853 (R2= 0.55) In this analysis, three clusters were obtained: SNPs 4602, 35991 and 37853

as cluster 1, 4740 and 12590 as cluster 2, and 6144, 12390, 14215 and

35587 as cluster 3 Among these three LD clusters, some haplotypes were identified: 8 for LD cluster 1, 4 in LD cluster 2, and 9 in LD cluster

3 (Table 5) Haplotype 1 represented the most frequent sequence in each LD cluster of AS3MT and the frequencies in LD cluster 1, 2, and 3 were 0.457, 0.629, and 0.690, respectively The three groups of LD obtained in this study were different from previousfindings reported for the populations of Mexico (Meza et al., 2005) and Argentine (Schläwicke Engström et al., 2007) (Table 6) Mexicans had strong LD among SNPs 7395, 12390, and 35587 (Meza et al., 2005).Schläwicke Engström et al (2007)showed a LD cluster composed of SNPs 12390,

14215, and 35991 in Argentina Remaining genotypes such as SNPs

5913, 7395, 8979, and 14458 were independent from other SNPs in this Vietnam population

Potential effects of genetic polymorphisms in AS3MT on As concentration and metabolism

Results of ANOVA followed by Tukey–Kramer Test showed statisti-cally significant associations between 10 SNPs (4602, 4740, 5913, 8979,

12390, 12590, 14458, 35587, 35991, and 37853), and As concentration and metabolite pattern in urine of Vietnamese (Table 4) On the contrary, SNPs 6144, 7395, and 14215 had no relation to any of the indicators of As exposure and metabolic capacity Although urinary %AB, which is unlikely to be involved in As methylation, had some associations with SNPs 4602, 4640, 8979, and 37853 (pb0.05), these associations might result from relations between these SNPs and %DMA

as shown by a strong negative correlation between %DMA and %AB (r =

−0.865, pb0.001) Concentration of TAs in human hair had no dependence on genotype in AS3MT

Homo types in 4602GG, 35991AA, and 37853AA, which had strong

LD with each other (cluster 1), showed significantly lower %DMA in urine compared to other genotypes in each corresponding SNP (pb0.05) (Table 4) For SNP 37853, %MMA in AA homozygote was also lower than those in GG type (pb0.05,Table 4) The results did not agree with previous studies In Mexicans, SNP 4602 was not correlated with urinary As composition (Meza et al., 2005) (Table 6).Schläwicke Engström et al (2007)revealed that AA variant homozygosis in SNP

35991 was associated with a decrease in %MMA and an increase in % DMA in urine (Table 6), suggesting consequently a higher ratio for the 2nd methylation step To understand relationships between As metabolic capacity and haplotype, subjects were initially divided into several groups, comprising homozygote and heterozygote for AS3MT haplotype in each LD cluster (Table 7) Then, the association was assessed by excluding haplotype groups with less than four subjects to obtain sufficient statistical power Urinary %DMA in G1-2 (4602GG/35991AA/37853AA) of haplotype group in LD cluster 1 was significantly lower than those in G1-1 (4602AG/35991GA/37853GA) and G1-3 (4602AA/35991GG/37853GG) (pb0.05) (Fig 7) Further-more, this G1-2 showed lower %MMA compared with G1-4 (4602GG/

Higher %DMA in urine for TT homozygote in SNP 4740 was observed in the present study (pb0.05) (Table 4), while no similar observation was reported in Mexicans (Meza et al., 2005) (Table 6) Although concentration of IAs in urine for 4740TT was significantly

Table 4

Composition of As compounds (arithmetic mean) and concentration ratios of MMA/IAs

and DMA/MMA (arithmetic mean) in urine for SNPs of AS3MT in residents from Hoa

Hau and Liem Thuan in Vietnam

SNP ID a n %AB %DMA %MMA %IAs DMA/MMA MMA/IAs

4602

AA 40 16.2 y 61 x 10.8 11.9 5.9 1.0

AG 22 19.6 xy 59.8 x 10.0 10.6 6.5 1.1

GG 38 25.8 x 52.7 y 10.3 11.2 5.9 1.0

4740

TT 52 17.1 y 61.2 y 10.3 11.5 6.5 1.0

TC 40 24.9 x 54.9 x 10.1 10.1 6.0 1.2

CC 8 29.4 x 45.6 x 11.3 13.7 4.6 0.9

5913

TT 85 22.1 57.0 9.9 y 11.0 6.4 1.0

TC 14 15.1 60.6 12.7 x 11.6 5.1 1.3

CC 1 31.2 43.6 13.7 xy 11.5 3.2 1.2

6144

AA 57 20.3 57.6 10.7 11.4 5.8 1.1

TA 36 21.4 58.1 9.6 10.8 6.7 1.0

TT 7 27.5 52.5 10.0 9.9 6.1 1.0

7395

GG 38 20.4 57.4 10.8 11.3 5.7 1.1

GA 57 20.8 58.2 10.0 11.0 6.5 1.0

AA 5 31.9 48.0 10.2 10.0 5.1 1.1

8979

TT 36 18.9 y 59.7 x 10.3 11.0 6.3 1.1

TA 46 19.0 y 58.6 x 10.9 11.5 5.9 1.1

AA 18 31.4 x 49.6 y 8.7 10.3 6.5 0.9

12390

GG 59 20.5 56.9 11.0 x 11.6 5.6 y 1.1

GC 37 21.8 58.2 9.5 xy 10.6 6.8 x 1.0

CC 4 25.6 58.5 7.1 y 8.9 8.2 xy 0.8

12590

TT 43 18.2 60.9 x 9.8 11.1 6.7 x 1.0

CT 42 22.5 56.8 xy 10.3 10.4 6.2 xy 1.1

CC 15 26.3 48.9 y 11.7 13.0 4.6 y 1.0

14215

CC 57 20.9 57.0 10.8 11.3 5.7 1.1

CT 37 21.3 57.8 9.8 11.1 6.6 1.0

TT 6 23.3 59.2 8.3 9.2 7.4 0.9

14458

TT 96 21.1 57.5 10.2 11.1 6.2 1.0 y

TC 4 23.0 54.7 11.9 10.4 4.8 1.6 x

35587

TT 59 21.0 56.6 10.9 11.5 5.6 y 1.1

TC 28 18.4 60.1 9.9 11.5 6.9 x 0.9

CC 13 28.1 55.1 8.3 8.5 6.9 xy 1.0

35991

GG 19 17.2 60.2 xy 10.9 11.7 5.8 1.0

GA 45 19.6 60.0 x 10.2 10.3 6.4 1.1

AA 36 25.3 52.7 y 10.2 11.8 6.0 0.9

37853

GG 26 17.5 y 60.5 x 11.0 x 11.1 5.7 1.2

GA 49 18.2 y 59.6 x 10.7 xy 11.6 6.2 1.0

AA 25 31.0 x 50.0 y 8.8 y 10.2 6.5 0.9

Values with same letters are not significantly different at pb0.05.

a

SNP ID indicates the SNP identification number relative to the location in the

consensus sequence, with the first base of the consensus numbered 1.

higher than that for 4740TC (p = 0.002), %IA was not related to the

genetic polymorphism For SNP 12590 belonging to the cluster 2 as

does SNP 4740, the TT type had higher concentrations of DMA and

IMDAs, %DMA, and DMA/MMA in urine than other genotypes

(pb0.05), suggesting the prompted 2nd methylation capacity in this

carrier In the LD cluster 2, concentrations of DMA, IAs, and IMD in

urine of G2-1 (4740TT/12590TT) haplotype group were greater than

those in G2-2 (4740TC/12590TC) (pb0.05) (Table 7) Furthermore,

significant higher %DMA in urine was observed in G2-3 (4740TT/

12590TC) and G2-1 than in G2-4 (4740CC/12590CC) (pb0.05) (Fig 7

and Table 7) Higher hair As concentration was found in G2-3

compared to G2-1, indicating the difference in allele between T and C

in AS3MT 12590

Although SNPs 6144, 12390, 14215, and 35587 showed strong LD and were grouped as cluster 3 in the present study, their associations with urinary As profile were classified into two different patterns; genotypes in SNPs 6144 and 14215 showed no significant correlations with As in hair and urine, while 12390GG and 35587CC had higher urinary %MMA and thus lower DMA/MMA (Table 4) Similarly,Meza

et al (2005) and Schläwicke Engström et al (2007)have reported lower DMA/MMA in urine of 12390GG in Mexican children and Argentina women (Table 6) SNP 14215 in females of Argentine had

Fig 6 Linkage disequilibrium of SNPs in AS3MT of humans from Hoa Hau and Liem Thuan in Vietnam The value shown in each diamond indicates pair wise R 2

value for each SNP pair.

Table 5

Distribution of haplotype group in linkage disequilibrium (LD) cluster og AS3MT in residents from Hoa Hau and Liem Thuan in Vietnam

Haplotype SNP IDs a

Total frequency Cumulative frequency

LD cluster 1

LD cluster 2

LD cluster 3

Haplotype 1 6144A/12390G/14215C/35587T 0.690 0.690

Haplotype 2 6144T/12390C/14215T/35587C 0.215 0.904

Haplotype 3 6144A/12390G/14215C/35587C 0.035 0.939

Haplotype 4 6144T/12390G/14215C/35587T 0.026 0.965

Haplotype 5 6144A/12390G/14215T/35587C 0.010 0.975

Haplotype 6 6144A/12390G/14215T/35587T 0.010 0.985

Haplotype 7 6144A/12390C/14215T/35587C 0.005 0.990

Haplotype 8 6144T/12390C/14215C/35587T 0.005 0.995

Haplotype 9 6144T/12390G/14215T/35587C 0.005 1.000

a SNP ID indicates the SNP identification number relative to the location in the consensus sequence, with the first base of the consensus numbered 1.

strong LD with SNP 12390, and the association with estimated

metabolic capacity of As was also similar between SNPs 14215 and

12390 (Schläwicke Engström et al., 2007) As for Vietnamese, no such

a result of SNP 14215 was observed Among the haplotype groups in

this LD cluster 3, AB concentration in urine of G3-3 (6144AT/

12390GC/14215CT/35587CC) was significantly higher than that in

although the reason was not clear

than that for 5913TC (Table 4) TT genotype in 8979 was associated

with higher %DMA (pb0.05) To our knowledge, this is the first finding

on the association with SNPs in addition to SNP 37853

14458TC hetero type had significantly higher MMA/IAs (p=0.002) than TT homo type (Table 4), although the sample size was small (n = 4) SNP 14458 located at exon 8 in AS3MT corresponds to 287 at amino acid base, in which amino acid substitution occurs from Met to Thr A previous in vitro expression study using COS-1 cell, variant type with 287MetNThr showed significantly higher levels of enzymatic activity and immunoreactive protein than the Met/Met homo type (Wood et al.,

2006) Higher %MMA in urine was reported for 287MetNThr hetero-zygous carriers than the Met/Met homozygosis in general populations

in Hungary, Romania, and Slovakia, (Lindberg et al., 2007), and in male workers of copper smelting plant in Chile (Hernandez et al., 2008) Although there was no significant difference, %MMA in urine for the TC

Table 6

Comparison of significant differences in composition of As compounds and concentration ratios of MMA/IAs and DMA/MMA in urine and linkage disequilibrium for SNPs a of AS3MT among study groups in Vietnam, Mexico, and Argentine

Reference This study Meza et al., 2005 Schläwicke Engström et al., 2007

Study group Vietnam Mexico (children) Argentine (females)

SNP 4602 %DMA; AANGG, AANGG NS

SNP 12390 %MMA; GGbCC DMA/MMA; GG bGC+CC %DMA; GG bGC, GGbCC

DMA/MMA; GG bGC, GGbCC

DMA/MMA; TTNCC

%MMA; CC NCT, CCNTT DMA/MMA; CC bCT, CCbTT SNP 14458 MMA/IA; TTbTC

SNP 35587 DMA/MMA; TTbTC DMA/MMA; TTbCT+CC

MMA/As III ; TTNCT+CC

%MMA; GG NGA, GGNAA DMA/MMA; GG bGA, GGbAA SNP 37853 DMA%; GGNAA, GANAA

MMA%; GGNAA Pair of linkage disequilibrium SNPs 4602, 35991, and 37853 SNPs 7395, 12390, and 35587 SNPs 12390, 14215, and 35991

SNPs 4740 and 12590 SNPs 6144, 12390, 14215, and 35587 NS; not significant.

a SNP ID indicates the SNP identification number relative to the location in the consensus sequence, with the first base of the consensus numbered 1.

Table 7

Composition of As compounds (arithmetic mean) and concentration ratios of MMA/IAs and DMA/MMA (arithmetic mean) in urine for haplotype group of each linkage disequilibrium (LD) cluster of AS3MT in residents from Hoa Hau and Liem Thuan in Vietnam

Group n Haplotype group a SNP IDs b %AB %DMA %MMA %IAs DMA/MMA MMA/IAs

LD cluster 1

G1-1 33 Haplotype 1 × 2, 3 × 8, 4 × 7, 5 × 6 4602AG/35991GA/37853GA 18.9 y 60.2 x 10.0 xy 10.8 6.6 1.0 G1-2 21 Haplotype 1 × 1 4602GG/35991AA/37853AA 30.5 x 50.2 y 9.0 y 10.3 6.4 0.9 G1-3 17 Haplotype 2 × 2 4602AA/35991GG/37853GG 16.4 y 60.3 x 11.2 xy 12.1 5.5 1.1 G1-4 10 Haplotype 1 × 3 4602GG/35991AA/37853GA 17.5 y 55.3 xy 12.6 x 14.5 5.0 1.0

LD cluster 2

G2-1 41 Haplotype 1 × 1 4740TT/12590TT 18.0 60.9 x 9.9 11.2 6.6 1.0 G2-2 31 Haplotype 1 × 2, 3 × 4 4740TC/12590TC 25.6 55.0 xy 9.9 9.6 6.1 1.2 G2-3 11 Haplotype 1 × 3 4740TT/12590TC 13.8 62.0 x 11.7 12.5 6.3 1.0 G2-4 8 Haplotype 2 × 2 4740CC/12590CC 29.4 45.6 y 11.3 13.7 4.6 0.9 G2-5 7 Haplotype 2 × 3 4740TC/12590CC 22.7 52.8 xy 12.2 12.3 4.6 1.2

LD cluster 3

G3-1 54 Haplotype 1 × 1 6144AA/12390GG/14215CC/35587TT 20.4 57.6 10.7 11.3 5.8 1.1 G3-2 25 Haplotype 1 × 2, 4 × 7, 5 × 8 6144AT/12390GC/14215CT/35587TC 18.6 60.0 9.6 11.8 7.0 0.9 G3-3 7 Haplotype 2 × 3 6144AT/12390GC/14215CT/35587CC 32.2 51.6 8.4 7.9 6.5 1.2 G3-4 4 Haplotype 2 × 2 6144TT/12390CC/14215TT/35587CC 25.6 58.5 7.1 8.9 8.2 0.8 Values with same letters are not significantly different at pb0.05.

a

Each haplotype detail is shown in Table 5

b SNP ID indicates the SNP identification number relative to the location in the consensus sequence, with the first base of the consensus numbered 1.

heterozygote (mean,11.9%) showed a tendency to be higher than that for

the TT homozygote (mean, 10.2%) in the present study From these

results, hetero type in 14458TC (287MetNThr at amino acid base) may

have higher methylation ability from IA to MMA

Since sex, age, and BMI were also significantly related to As

accumulation and methylation capacity as shown above, the results of

genetic differences may be submerged by the effects of such factors

Thus, we adjusted concentrations and compositions of As using

multiple regression analysis to remove effects of these co-factors and

then reconsidered the genetic association with As concentration and

methylation %DMA, AB concentration and %AB in urine of local

residents were not corrected by the analysis, because no significant

regression equations were obtained for these variables The

adjust-ment by the multiple regression analysis also provided similar results

to the analysis without the correction as described before However,

IMDAs level and SNP 12590, and between DMA, IAs, and IMDAs

concentrations and haplotype groups in LD cluster 2 disappeared after

the correction On the contrary, following new associations were

obtained; higher MMA/IAs for 5913TC than that for 5913TT was

observed (p = 0.007), suggesting that the SNP may be related to the

1st methylation process of As In addition, concentration of MMA for

TT homozygosis in SNP 4740 was higher than TC heterozygosis

(p = 0.013) and TT homozygote in SNP 5913 was associated with

increased SAs level in urine (p = 0.014) Adjusted DMA and IMDAs

concentrations in urine of G1-1 (4602AG/35991GA/37853GA) were

the highest among LD1 haplotype groups (pb0.05) In haplotype

groups of the LD cluster 3, significant higher DMA/MMA was found in

G3-2 (6144AT/12390GC/14215CT/35587TC) than in G3-1 (6144AA/

12390GG/14215CC/35587TT) This indicates that G3-2 haplotype

group may have higher 2nd methylation capacity

The present study identified 10 SNPs in AS3MT that may affect As methylation process in residents of the Red River Delta, Vietnam Especially, we found that SNP 12390 in AS3MT was greatly associated with DMA/MMA ratios in human urine, which was consistent with the results in previous studies (Meza et al., 2005; Schläwicke Engström

et al., 2007) (Table 6) Therefore, SNP 12390 may be a universal genotype which affects 2nd step methylation process of As Further-more, to our knowledge, significant relationships between SNPs 5913,

8979 and 37853, and urinary As profile were observed for the first time Interestingly, among the 10 SNPs, only SNP 14458 is located at the exon, and other SNPs are at the intron, upstream or downstream region Further investigation is necessary to link non-exonic polymorphisms with the function of AS3MT

Acknowledgments

We are grateful to Dr A Subramanian, CMES, Ehime University for critical reading of the manuscript The authors wish to thank the staff of the CETASD, Hanoi University of Science and Dr H Sakai, CMES, Ehime University for their help in sample collection We also acknowledge Ms H Touma and Ms N Tsunehiro, staff of the es-BANK, CMES, Ehime University for their support in sample management This study was mainly 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 (to M.I.) and by a Grant from Research Revolution 2002 (RR2002) Project for Sustainable Coexistence of Human, Nature and the Earth (FY2002; to H.I.) Financial assistance were also provided by Grants-in-Aid for Scientific Research (S) (No 20221003; to S.T.) and (A) (No 19209025;

to H.T.) from JSPS, and 21st Century and Global COE Programs from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT),

Fig 7 Comparison of %DMA in urine among haplotype groups in linkage distribution (LD) cluster 1 and 2 from Hoa Hau and Liem Thuan in Vietnam Bar and plots indicate arithmetic mean and individual values, respectively.