Genetic polymorphism of As3MT and delayed urinary DMA excretion after organic arsenic intake from oyster ingestion

Abstract

Organic arsenic intake from seafood is one of the major arsenic exposure routes among the general population. However, organic arsenic metabolism in the human body is not yet clear. The goal of this study was to explore the effects of genetic polymorphisms of human PNP, As3MT and GSTO1 on organic arsenic metabolism among study subjects after oyster ingestion. During the one-week dietary controlled study, fifty study subjects were provided all their daily meals without seafood, except for two designated amounts of oyster given on the fourth day. First morning voided urine samples were provided by the study subjects for 7 consecutive days and analyzed with HPLC-ICP-MS for As³⁺, As⁵⁺, monomethylarsonic acid, and dimethylarsinic acid (DMA). Blood samples were collected later for genetic polymorphisms analysis of PNP, As3MT and GSTO1. Study subjects were categorized into “fast-” (n = 32), “medium-” (n = 13) and “slow-metabolizing” (n = 5) groups based on the number of days after ingestion needed for each subject's urinary DMA level reaching peak. Allele frequencies of single nucleotide polymorphisms (SNP) in intron 6 (G/C, p = 0.024) and in intron 10 (T/C, p = 0.039) of As3MT were significantly associated with the urinary DMA excretion. General estimating equation model analysis indicated that the variants of SNP (G>C) in intron 6 and SNP (T > C) in intron 10 of As3MT were respectively associated with higher or lower urinary DMA level by approximately 9 μg L⁻¹. As3MT was suggested to be one of the major factors affecting the metabolism of dietary organic arsenic in terms of urinary DMA level.

---

Environmental impact

Urinary arsenic metabolites, including dimethylarsinic acid, monomethylarsonic acid, As⁵⁺ and As³⁺, are the most frequently used biological markers for toxic inorganic arsenic exposure. Meanwhile, dietary organic arsenic is another type of common sources for arsenic exposure, and one of its metabolites, dimethylarsinc acid, may confound the biological marker for inorganic arsenic exposure. This study illustrates that the metabolism of dietary organic arsenic is associated with the polymorphism of As3MT, which is also known for inorganic arsenic metabolism in human body. It implies that As3MT may be used to explain the individual variation, or even ethnic variation, in the background urinary arsenic levels, as biological markers, after dietary organic arsenic intake.

Introduction

Arsenic is a natural element widely distributed in the environment and is commercially used as a raw material in many industries, including the manufacture of pesticides, semiconductors and wood preservatives. Therefore, humans have many chances to be exposed to arsenic through environmental or occupational pathways. In fact, there are more than 20 different chemical arsenic species which have been identified in structure and toxicity. According to some previous studies, inorganic arsenic, like arsenate (As⁵⁺) and arsenite (As³⁺), was thought to be more toxic than organic arsenic such as monomethylarsonic acid (MMA), and dimethylarsinic acid (DMA), arsenobetaine, arsenocholine and arsenosugars.^1,2 However, other studies have proved that certain types of organic arsenic, such as the intermediate metabolite, trivalent monomehtylarsonic acid (MMA³⁺), are more toxic than trivalent arsenic (As³⁺).³ Furthermore, even though the tetramethylarsnoium ion is characterized by a tetramethyl group, its toxicity has been demonstrated to be greater than DMA and MMA.⁴ In fact, these organic arsenic species with questionable toxicity are usually found in seafood and these aquatics are the main food resources for people living in islands or coastal areas. The metabolism of dietary arsenic species in the human body is, therefore, of great concern.

On the other hand, one person's susceptibility to arsenic exposure may differ, given that each individual's capacity for dietary arsenic metabolism varies and results in different arsenic metabolites, which may cause different levels of toxicity. Factors affecting an individual's susceptibility to arsenic metabolism include age, gender, race, dose, exposure route, arsenic species, nutrition status, and environment factors.⁵ In addition, different enzymes may, directly or indirectly, regulate the reduction and methylation involved in arsenic metabolism. The activity of these enzymes may also be associated with genetic polymorphisms.⁵ Polymorphisms are therefore considered to be one of the most significant factors affecting arsenic metabolism. At the present time, animal studies have demonstrated that the metabolism of inorganic arsenic is affected by purine nucleoside phosphorylase (PNP), glutathione-S-transferase (GSTO1, GSTP1, GSTM1) and As³⁺ methyltransferase (As3MT).^6–11 In the proposed pathway for inorganic arsenic metabolism in humans, PNP, GSTO, and As3MT have distinct functions in the reduction and oxidative methylation processes.^6,11,12 Meanwhile, PNP has the ability to reduce As⁵⁺ to As³⁺, and GSTO is in charge of the reduction for all pentavalent arsenic species. In addition, the oxidative methylation of trivalent arsenic species is regulated by As3MT. Among these enzymes, genetic polymorphisms of GSTO1 and As3MT were reported to be strongly associated with the distribution of urinary arsenic metabolites following inorganic arsenic exposure.^8,9,13−15

Although it is believed that the toxicity of organic arsenic is less than that of inorganic arsenic, the metabolism of organic arsenic is still of great concern, since dietary organic arsenic is one of the most common types of exposures among the general population.¹⁶ Additionally, a previous study has indicated that urinary arsenic species varied widely during a 3-day period after a meal of seafood, and suggested that human metabolism of organic arsenic should be explored further.¹⁷ Most of the previous studies focused on the effect and variations of the human genes with inorganic arsenic exposure, but few research has highlighted the influence of these genes on arsenic metabolism when humans are exposed to dietary organic arsenic. The goal of the present study was therefore set to explore the effects of polymorphisms of the human genes PNP, As3MT and GSTO1 on variations in arsenic metabolism following organic arsenic exposure through the ingestion of oyster.

Materials and methods

Subjects

A total of 65 subjects were recruited for this dietary arsenic exposure study: 31 men and 34 women. The participants were students and research assistants in the authors' institute who were not occupationally exposed to arsenic-containing compounds. The study protocol was reviewed and approved by the institutional review board of the College of Public Health, National Taiwan University. Informed consents were obtained from the study subjects prior to their participating in the present study. All participants were non-smokers. After exclusion of the subjects whose urinary arsenic speciation data were incomplete (n = 8) or whose blood samples were unavailable (n = 7), the remaining 50 study subjects were involved in the final data analysis. During the one-week dietary controlled study, the study subjects were provided every day with three daily meals without seafood and refrained from consuming any other foods containing seafood, such as snacks or soups. Only on the 4th day, all study subjects were provided with a total of around 60∼80 grams oyster for lunch and dinner.

Oyster sample analysis for arsenic species

For arsenic species determination, oyster samples were first cut into approximately 0.5 × 0.5 cm² size pieces and degreased with acetone, then dryly frozen at −50 °C for 20 h. The dried samples were later ground into powder, prior to being extracted for arsenic species with 1 : 1 methanol–water. Then extractions were ultra-sonicated for 20 min and centrifuged at 4000 rpm for 15 min. Then, the supernatants were filtered in succession with 0.8 μm and 0.45 μm polyethersulfone filters. After this, the filtrates were concentrated under reduced pressure at 35 °C to remove methanol and water. The remaining sample was diluted to 10 ml, and then analyzed for arsenic species by high performance liquid chromatography (HPLC, Agilent 1100 series) coupled with inductive couple plasma – mass spectrometer (ICP-MS, Agilent 7500c). The determined arsenic species included As³⁺, As⁵⁺, MMA, and DMA, arsenobetaine, arsenocholine, trimethylarsine oxide, tetramethylarsonium, and total arsenic with corresponding detection limits of 0.13 μg L⁻¹ for As³⁺, 0.18 μg L⁻¹ for As⁵⁺, 0.25 μg L⁻¹ for MMA, 0.28 μg L⁻¹ for DMA, 0.14 μg L⁻¹ for arsenobetaine, 0.35 μg L⁻¹ for arsenocholine, 0.13 μg L⁻¹ for trimethylarsine oxide, and 0.25 μg L⁻¹ for tetramethylarsonium. In addition, for a level of 2 μg L⁻¹, the recovery rates for urinary As³⁺, As⁵⁺, MMA, arsenocholine, trimethylarsine oxide, tetramethylarsonium were 100.3%, 70.1%, 85.7%, 90.3%, 115.1% and 101.5%, respectively, while those for DMA and arsenobetaine were 98.2% and 108.0%, respectively, for a level of 10 μg L⁻¹.

Urine sample collection and analysis

During the one-week dietary controlled study, daily first morning-voided urine samples were collected from the subjects for seven consecutive days and then stored at 4 °C. Prior to laboratory analysis, stored urine samples were placed on a mixer to return them to room temperature. The urine samples were then centrifuged at 3000 rpm for 3 min and filtered with a Sep-Pak Octadecyl (C₁₈) disposable extraction column (BAKERBOND spe™). The first 5 ml of urine filtrate was discarded; the succeeding 5 ml urine filtrate was used for the analysis of urinary arsenic metabolites, including As³⁺, As⁵⁺, MMA, and DMA. Analysis was conducted with HPLC coupled with ICP-MS as mentioned in previous paragraph.

Single Nucleotide Polymorphisms (SNPs) analysis for PNP, As3MT and GSTO1

Blood samples of about 10 mL were drawn intravenously from the subjects with vacutainers and DNA was extracted. All DNA samples were stored at −20 °C until the SNP analysis, at which time the samples were thawed to 4 °C for sequencing analysis. A total of 12 site points on the genes of PNP, As3MT and GSTO1 were sequenced. Genomic sequences for these study genes were searched from the National Center for Biotechnology Information (NCBI) and the selected SNPs were based on the GenBank of the National Center for Biotechnology Information (NCBI).¹⁸ Some genetic variants in PNP, As3MT, and GSTO1 were found associated with arsenic metabolism in previous studies.^14,19,20 Polymorphic sites of these genetic variants were sequenced and tested in this study, including 5 sites in the PNP gene, 3 sites in the As3MT gene and 4 other sites in the GSTO1 gene. Table 1 presents the selected SNPs and polymerase chain reaction (PCR) primers sequences. The DNA of the tested genes was amplified by PCR using multiblock system thermocycler (ThermoHybaid, Ashford, UK). PCR cycle sequencing reactions were mixed with 2 μL of 50 ng μL⁻¹ genomic DNA, 0.3 μL of 10 pmol forward primer and reverse primer, 1 μL of 2.5 mM dNTPs, 0.5 units AmpliTaq Gold™ enzyme (PE Applied Biosystems, Foster City, CA, USA), 2.5 μL of GeneAmp × 10 buffer II (10 mM Tris-HCl pH = 8.3, 50 mM KCl), 2 μL of 25 mM MgCl₂ and 16.8 μL of double de-ionized water in a final volume of 25 μL. The PCR conditions for diverse genetic SNPs were slightly different and shown in Table 2. After PCR, amplicons were purified by solid-phase extraction (GFXTM PCR DNA and Gel Band Purification Kit, Amersham Biosciences, UK) and the products were sent for sequencing by a well-trained professional using the Applied Biosystems 3100 sequencer.

Table 1 The discovered gene context of polymorphisms and the designed PCR primers in studied genes

Gene	Functional context	polymorphic site	PCR Primers Sequences (5′ → 3′)
a F: Forward primer, R: Reverse primer. b Reference SNP ID.
PNP	intron 1	rs1760940^b	F^a: ATGCTAGGGTCTGGGTAAAG
		rs17884106	R^a: AAAGCAACTGTGATGCCTAA
	exon 2	rs1049562 (His20His)	F: GAAGGTCATTTGTCTGTGAT
		rs1049564 (Gly51Ser)	R: GGTCAAGGAGTAGAAACATT
		rs1130650 (Pro57Pro)
As3MT	intron 3	rs12767543	F: CTATGGGACAGAAACCTTAC
			R: CATAGTGAAACCCTGTCTCT
	intron 6	rs3740393	F: CTTGGCACTTGACTATTGAT
			R: TGTGTGTCCTGATTTCTTCT
	intron 10	rs11191453	F: GTGCTCTTGATTTCCTATCT
			R: AAAGTGAGTCCCAGATGTAG
GSTO1	exon 4	rs4925 (Ala140Asp)	F: CTAGAACACCTTGACACCAG
		rs56204475 (Glu155del)	R: CCTTAAAGTGACTTGGAAAGTGG
	exon 6	rs11509439 (Ala236Val)	F: CTGTGATGTCATCCTAGTTG
		rs11509438 (Glu208Lys)	R: CATGCAACCTGAACCTTGGT

---

Table 2 Parameters of PCR for sequencing PNP, As3MT and GSTO1

Gene	Functional context	Denaturation	Annealing	Extension
PNP	intron 1	1. 95 °C, 10 min.	56 °C, 45 s, 35 cycles	1. 72 °C, 30 s, 35 cycles
	exon 2	2. 94 °C, 30 s, 35 cycles.	52 °C, 45 s, 35 cycles	2. 72 °C, 10min
As3MT	intron 3		52 °C, 45 s, 35 cycles
	intron 6	(All with the same sequencing procedures.)	53 °C, 45 s, 35 cycles	(All with the same sequencing procedures.)
	intron 10		52 °C, 45 s, 35 cycles
GSTO1	exon 4		56 °C, 45 s, 35 cycles
	exon 6		56 °C, 45 s, 35 cycles

---

Statistical analysis

The study subjects were categorized into three groups: fast (5th day of the study week), medium (6th day) and slow (7th day) groups, respectively, based on the day on which the study subject's urinary DMA concentration reached the highest level after the oyster ingestion on the 4th day. In addition, ANOVA was applied to compare the differences for the above-mentioned factors among these three groups while t-test was performed to compare the differences between the Fast group and the combined group of Medium and Slow. Study subjects with SNPs on either one or two strands of DNA were defined as the SNP group whereas the others were set as the wild-type group without SNPs. Chi-square testing was performed to examine the association of urinary DMA excretion group with the types of SNP. With a logistic regression model, the effect of SNP type was examined for the odds for being in the combined Medium and Slow group of urinary DMA excretion. Meanwhile, the generalized estimating equation model (GEE) was also conducted to examine the factors affecting the fluctuation of urinary DMA level, including age, gender, day after oyster ingestion, and the SNP type at intron 6 and intron 10 in As3MT, etc.

Results

Among the final 50 eligible study subjects involved in the one-week controlled study, 21 were males (42%) and 29 were females (58%), with an average age of 24.4 ± 2.1 years and an average body mass index (BMI) of 21.9 ± 2.7 kg m⁻². The average ingested amount of oyster during the two designated meals was 65.2 ± 14.1 g per study subject, i.e., 1.10 ± 0.30 g kg⁻¹ body weight. The total arsenic content in the oyster for the two designated meals was 4.23 μg g⁻¹, including 0.04 μg g⁻¹ As³⁺, 0.10 μg g⁻¹ DMA, 1.28μg g⁻¹ arsenobetaine, 0.05 μg g⁻¹ arsenocholine, 0.02 μg g⁻¹ TetMA, 0.83 μg g⁻¹ speciated arsenosugars, and certain uncharacterized arsenic species, believed to be mostly arsenosugar species.²¹

In general, the daily average urinary DMA and As³⁺ levels of the study subjects reached their highest levels of 28.2 ± 12.0 and 2.20 ± 1.13 μg L⁻¹, respectively, on the 5th day of the study period and declined after this (Fig. 1). The levels of these two urinary arsenic species were also highly correlated (p < 0.0001). Urinary MMA and As⁵⁺ did not present similar fluctuation patterns. With each study subject's respective urinary arsenic species levels on the 4th day used as that individual's specific references, the average urinary DMA levels on the 5th and 6th days were significantly elevated (p < 0.0001) (Fig. 1). 32 subjects' urinary DMA concentrations reached their highest levels one day after oyster ingestion, i.e., the 5th day of the study week, and were categorized as the fast group of urinary DMA excretion. Another 13 study subjects and the other 5 study subjects' urinary DMA levels reached their highest levels two days (6th day) and three days (7th day) after oyster ingestion, respectively, and were categorized as the medium and the slow group accordingly (Table 3).

Fig. 1 Fluctuation of urinary arsenic species before and after oyster ingestion.

Table 3 Study subjects' demographics, ingested oyster amount and urinary DMA levels by group of urinary DMA excretion ^a

Variables	Fast (n = 32)	Medium (n = 13)	Slow (n = 5)	p	Medium and Slow (n = 18)	p
a All data presented as Mean ± SD. b Ingested oyster amount, g, divided by the body weight of study subject, kg. c Individual study subject's highest daily urinary DMA level after ingesting the designated amount of oyster subtracted by the self-reference of the urinary DMA level on the 4th day of each study subject. d Sum of the daily increment of urinary DMA concentrations over the three consecutive days following the ingestion of designated amount of oyster. e p-value of ANOVA test for the comparison among the fast, medium and slow groups. ‘NS’ indicates the p value is greater than 0.100. f p-value of t-test for the comparison of the fast group with the medium and slow combined group. ‘NS’ indicates the p value is greater than 0.100.
Age, years	24.3 ± 1.9	24.3 ± 2.6	25.7 ± 1.5	NS	24.7 ± 2.4	NS
BMI	21.8 ± 3.1	22.0 ± 1.6	22.4 ± 1.5	NS	22.1 ± 1.5	NS
Amount of ingested oyster, g/(kg body weight)^b	1.08 ± 0.33	1.20 ± 0.29	1.02 ± 0.10	NS	1.15 ± 0.26	NS
Urinary DMA Level, μg L^−1
4th day	13.5 ± 4.7	15.3 ± 10.3	10.1 ± 5.0	NS	13.8 ± 9.3	NS
5th day	33.3 ± 11.3	21.1 ± 6.6	14.7 ± 6.7	<0.0001	19.3 ± 7.1	<0.0001
6th day	21.6 ± 7.6	36.0 ± 15.6	14.0 ± 9.0	<0.0001	29.9 ± 17.1	0.0209
7th day	13.5 ± 6.9	17.6 ± 10.7	28.7 ± 12.9	0.0022	20.7 ± 12.1	0.0104
Highest Daily DMA Increment ^c	19.9 ± 11.1	21.1 ± 17.9	18.6 ± 8.9	NS	20.4 ± 15.7	NS
Accumulated DMA Increment ^d	31.0 ± 21.2	32.3 ± 24.9	27.8 ± 14.2	NS	31.0 ± 22.2	NS

---

Among the fast, medium and slow groups, there were no statistically significant differences in age, BMI or amount of ingested oyster per kilogram of body weight (Table 3). For the 12 sequenced site points in the genes of PNP, As3MT and GSTO1 (Table 1), 7 of them presented SNPs (Table 4). The highest frequencies of polymorphism among subjects were found in the intron 3 SNP (G > A) of As3MT (70%), followed by intron 10 SNP (T > C) (60%) and intron 6 SNP (G>C) (48%) of the same gene, while the lowest frequencies of polymorphism of 14%, 12% and 10%, respectively, were present in codon 140 SNP (C > A) of GSTO1, intron 1 SNP (A > C) and codon 20 SNP (C > T) of PNP. The examination of the relationship of the genetic polymorphisms with urinary DMA excretion showed that SNPs in intron 6 and intron 10 of As3MT were significantly associated with urinary DMA excretion after oyster ingestion (p = 0.024 and 0.039, respectively) (Table 5). Furthermore, after controlling for the effects of age, gender, BMI, amount of ingested oyster, and accumulated increment of urinary DMA level, logistic regression analysis indicated that the SNPs in intron 6 and intron 10 of As3MT were still moderately to mildly associated with the group of fast urinary DMA excretion (p = 0.014 and 0.071), respectively (Table 5).

Table 4 PNP, As3MT and GSTO1 polymorphisms among study subjects.^a

Gene	Polymorphism	N (%)	Gene	polymorphism	N (%)
a Note: Only genetic sites with SNPs were presented.
GSTO1	C > A(codon 140)		PNP	A > C (intron 1)
	C/C	43 (86)		A/A	44 (88)
	C/A	7 (14)		A/C	6 (12)
	A/A	0 (0)		C/C	0 (0)
As3MT	G > A(intron 3)		PNP	C > T(codon 20)
	G/G	15 (30)		C/C	45 (90)
	G/A	19 (38)		C/T	5 (10)
	A/A	16 (32)		T/T	0 (0)
As3MT	G > C (intron 6)		PNP	G > A(codon 51)
	G/G	26 (52)		G/G	35 (70)
	G/C	18 (36)		G/A	14 (28)
	C/C	6 (12)		A/A	1 (2)
As3MT	T > C (intron 10)
	T/T	20 (40)
	T/C	24 (48)
	C/C	6 (12)

---

Table 5 Effects of PNP, As3MT, GSTO1 polymorphisms on odds ratio for Combined Medium and Slow Group for urinary DMA excretion after oyster ingestion

Gene	Polymorphism	Fast (n = 32)	Medium (n = 13)	Slow (n = 5)	p	Combined Medium and Slow (n = 18)	p	Crude OR^c	p	Adjusted OR^c	p
a p-value for Chi-square test. 'NS’ indicates p value is greater than 0.100. b Comparison between the Combined Medium & Slow Group and the Fast Group. c OR: odds ratio for Combined Medium and Slow Group for urinary DMA excretion after oyster ingestion. ‘Adjusted OR’ indicates an OR value after adjusting for variables other than the study genetic polymorphism, including age, gender, BMI, ingested oyster amount, and elevated urinary DMA concentration. 95% confidence interval for odds ratio is present in parenthesis. d ‘NS’ indicates p-value is greater than 0.100.
GSTO1	C > A(codon 140)
	C/C	29	11	3	NS^d	14	NS	1	NS^d	1	NS
	C/A	3	2	2		4		2.76 (0.54–14.1)		3.07 (0.47–20.4)
As3MT	G > A(intron 3)
	G/G	9	5	1	NS	6	NS	1	NS	1	NS
	G/A + A/A	23	8	4		12		0.78 (0.22–2.72)		0.57 (0.14–2.34)
As3MT	G > C (intron 6)
	GG	12	10	4	0.024	14	0.0062	1	0.0089	1	0.014
	G/C + C/C	20	3	1		4		0.17 (0.05–0.64)		0.15 (0.03–0.68)
As3MT	T > C (intron 10)
	T/T	10	9	1	0.039	10	0.092	1	0.097	1	0.071
	T/C + C/C	22	4	4		8		0.36 (0.11–1.20)		0.27 (0.07–1.11)
PNP	A > C (intron 1)
	A/A	28	11	5	NS	16	NS	1	NS	1	NS
	A/C	4	2	0		2		0.88 (0.14–5.32)		0.55 (0.08–4.0)
PNP	C > T(codon 20)
	C/C	29	11	5	NS	16	NS	1	NS	1	NS
	C/T	3	2	0		2		1.21 (0.18–8.0)		1.21 (0.16–9.43)
PNP	G > A(codon 51)
	G/G	20	10	5	NS	15	NS	1	NS	1	0.068
	G/A + A/A	12	3	0		3		0.33 (0.08–1.39)		1.29 (0.30–2.35)

---

The SNPs in intron 6 and intron 10 of As3MT were further examined, along with other affecting factors, for their roles in the fluctuation of urinary arsenic metabolites after oyster ingestion. The results of GEE model analysis indicated that, after controlling for the effects of age, gender, number of days after oyster ingestion, and amount of ingested oyster per kilogram of body weight, the variants of both intron 6 and intron 10 of As3MT presented moderately associations with the levels of urinary DMA (p = 0.0045 and 0.0053, respectively) (Table 6). A strong association was also observed for certain variants of these two SNPs with the As⁵⁺ levels, but in the reverse direction (p<0.001).

Table 6 Generalized estimating equation (GEE) model for factors affecting urinary DMA, MMA, As⁵⁺, As³⁺ levels after oyster ingestion.^a

	DMA, μg L^−1	MMA, μg L^−1	As^5+, μg L^−1	As^3+, μg L^−1
a P-value was shown in parenthesis. ‘NS’ indicates the p-value is greater than 0.1000.
Intercept	15.2(NS)	1.70(NS)	1.20 (0.0410)	3.75 (0.0025)
Amount of ingested oyster, g kg^−1	−0.881 (NS)	0.746 (0.0512)	1.01 (<0.0001)	0.106 (NS)
Age, yr	−0.0018 (NS)	−0.0207 (NS)	−0.0735 (0.0004)	−0.0942 (0.0331)
Gender (female as reference)	−1.18 (NS)	0.291 (NS)	0.366 (<0.0001)	0.344 (0.0621)
Day
4th	—	—	—	—
5th	14.9 (<0.0001)	−0.0313 (NS)	−0.110 (NS)	0.468 (0.0268)
6th	11.5 (<0.0001)	−0.518 (0.0456)	−0.060 (NS)	0.196 (NS)
7th	1.21 (NS)	−0.675 (0.0087)	−0.112 (NS)	−0.245 (NS)
Intron 6
G/G	—	—	—	—
G/C	−0.739 (NS)	0.801 (0.0526)	−0.225 (NS)	−0.545 (NS)
C/C	9.59 (0.0045)	−0.520 (NS)	−0.529 (0.0008)	−0.101 (NS)
Intron 10
T/T	—	—	—	—
T/C	−0.0237 (NS)	−0.0742 (NS)	0.386 (0.0166)	0.370 (NS)
C/C	−8.84 (0.0053)	0.695 (0.0674)	0.539 (0.0003)	0.109 (NS)

---

Discussion

As organic arsenic species, arsenosugars are dimethylarsinoylribosides containing two methyl groups and a 5-deoxyriboside group attached to the arsenic atom.²² Anaerobic marine microbial activity can quickly convert algal dimethylarsinoylribosides to dimethylarsinoylethanol, which is the precursor of the two-carbon (carboxymethyl) side-chain in arsenobetaine. Furthermore, the microbial degradation of arsenobetaine to dimethylarsinate via dimethylarsinoylacetate was reported in the marine mussel (Mylitus edulis).²³ However, arsenobetaine was not thought to be transformed in humans or other mammals, i.e., it was believed to be excreted essentially unchanged.^24,25 As to arsenosugar metabolism by humans, the major arsenic compound excreted after the consumption of arsenosugar-rich seaweed and mollusks or the ingestion of synthetic arsenosugars was found to be DMA.^26–30 It is known that seaweed and mollusks, including oyster, do not contain significant amounts of DMA. Furthermore, the specific arsenosugars that were ingested were not found in the urine of seaweed consumers, but trace amounts of uncharacterized arsenicals were detected.^27,30 These findings suggested that some decomposition of arsenosugars occurs during the process of metabolism in the human body. However, little is documented for the metabolism of organic arsenic with respect to the specific decomposing pathways, such as decarboxylation and/or demethylation, which have been found for marine microbial activity.^22,23

Nevertheless, certain studies have indicated that the metabolism of arsenosugars varies from person to person in terms of patterns and quantities of arsenic metabolites found in the urine of study subjects eating the same quantity of an arsenic-rich seaweed.²⁶ These findings imply that arsenosugar metabolism is likely to be influenced by genetic or other unknown factors. Among them, individual variation in the metabolic rate of ingested organic arsenic in seafood was reported in previous studies with significant but various extents of increments in urinary DMA level at different time periods.^26,28,31 Similar findings were also observed in the present study and used as the criteria for study subject groupings. On the other hand, certain other additional factors have been associated with individual variation in inorganic arsenic metabolism as well, including genetic polymorphism.^5,7–9 As to the metabolism of dietary organic arsenic in the human body, the results of the present study indicate that organic arsenic metabolism was also significantly associated with the allele frequency of SNPs in intron 6 (G/C), and probably intron 10 (T/C), too, of As3MT gene (Table 4, Table 5, Table 6). For instance, the variants of SNP (G>C) in intron 6 of As3MT were characterized as having higher urinary DMA level by around 9.6 μg L⁻¹, while the variant of SNP (T > C) in intron 10 of As3MT was characterized as having lower urinary DMA level by around 8.8 μg l⁻¹ (Table 6). Similar trends were also observed for urinary As⁵⁺ and As³⁺ levels, but in a reverse trend along with a limited magnitude of change. These findings suggest a fundamental role of As3MT in organic arsenic metabolism that variation in this gene could account for the variations in efficiency of dietary organic arsenic metabolism in humans.

In previous studies of inorganic arsenic metabolism, As3MT was reported to be a critical factor in the regulation of DMA metabolism, playing a role as catalyst in the process of providing a methyl group from S-adenosyl-methionine (SAM) for As³⁺ and MMA in order to form MMA and DMA;^6,14,32 however, this pathway involves several distinct steps and many details regarding the nature of each constitutive step of this pathway remain uncertain. The As3MT gene was considered to be involved in genetic susceptibility to arsenic and variant of As3MT might increase arsenic methylation, resulting in deceased risk of toxic effects from inorganic arsenic exposure.^7,14,33–35 In contrast, As3MT has rarely been discussed from the aspect of organic arsenic metabolism in the human body, even though the results of the present study suggested that the variant of As3MT was also associated with the metabolic process of organic arsenic, such as that urinary DMA excretion speed was faster among subjects with a high variant of As3MT as compared to others with a low variant. Meanwhile, results of the present study also indicated that the polymorphism of As3MT was associated with the urinary DMA/MMA ratio (data not shown). A similar situation has been reported by Meza et al. for these three SNPs (rs12767543, rs3740393, rs11191453) on As3MT.¹⁴ This suggested that the metabolism of inorganic and organic arsenic in the human body might share certain similar properties. Future studies are warranted to characterize the potential role of As3MT in the human metabolism of organic arsenic, to determine whether it is directly involved in the decomposition of organic arsenic species from ingested seafood by decarboxylation and/or demethylation and results in an elevation of DMA in urine, or if it simply plays a similar role as in inorganic arsenic metabolism, by regulating the methylation of inorganic arsenic species. In addition, the inverse effect of polymorphism of As3MT on urinary As⁵⁺ and As³⁺ levels (Table 6), compared to DMA level, also revealed the influence of As3MT on the arsenic methylation, as indicated by the classical arsenic metabolic pathways.

For monitoring arsenic exposure, the most commonly used biological marker is urinary arsenic level, which is considered to be a composite of DMA, MMA, As⁵⁺ and As³⁺, standing for the major urinary arsenic metabolites after inorganic arsenic exposure. For East Asian islanders, background urinary arsenic levels are generally around 30 μg L⁻¹ and 40∼50 μg L⁻¹, respectively, for Taiwanese and Japanese.^36–38 Meanwhile, the average urinary arsenic levels in western countries are generally around 5∼10 μg L⁻¹.^39–42 This difference is mostly attributed to differences in daily dietary seafood intake because seafood is more common in Asian islanders' regular meals. However, the findings of the present study suggested that, in addition to increased organic arsenic intake from daily meals, the elevated urinary arsenic background levels among the Asian islanders might be partly affected by additional factors, such as genetic susceptibility. Previous studies have reported that the allele frequency of genetic polymorphisms might vary widely among different ethnicities.^14,19 For instance, the allele frequency of the M287T (T/C) polymorphism of As3MT among Asian populations, such as the Korean and Japanese populations (both 99.0% to 1.0%), were relatively higher than the Mongolian (96.0% to 4.0%), African-American (89.2% to 10.8%), Caucasian American (90.0% to 10.0%), and Central European populations (89.1% to 10.9%).^33,35,43 These studies suggested that genetic polymorphisms may explain not only the individual variation, but also ethnic variations in the metabolism of dietary organic arsenic. Since the aforementioned biological marker for arsenic exposure, i.e., urinary arsenic level, is widely utilized for monitoring occupational or environmental exposures to inorganic arsenic, as per the Threshold Limit Value – Biological Exposure Index (TLV-BEI) of the American Conference of Governmental Industrial Hygienists and the World Health Organization recommendation,^44,45 any interpretation of such biological markers for inorganic arsenic exposure should consider not just differences in seafood intake but also the genetic polymorphisms of the relevant genes affecting organic and inorganic arsenic metabolism, such as As3MT, as observed in the present study.

The effect of age is another major concern for affecting urinary DMA excretion and has been reported in previous studies that evaluated an individual's capability for inorganic arsenic metabolism by examining the percentage distributions of various urinary arsenic metabolites.^46,47 It has been reported that, after long-term exposure to high inorganic arsenic drinking water, urinary DMA percentages among adults were significantly higher than for children. It was suggested that the immaturity of a child's metabolic system results in a lower capability for children to metabolize ingested inorganic arsenic, as compared to the adult. On the other hand, for organic arsenic metabolism, a dietary control study has also indicated that the time for urinary DMA to reach the highest concentration after seafood ingestion was positively associated with age.³¹ However, no such effect from age on urinary DMA level or delay of urinary DMA peak was observed in the present study (Table 3, Table 6), probably due to the limited age range among the study subjects.

Conclusions

In this study, organic arsenic metabolism was suggested to be associated with the allele frequency of SNPs in intron 6 (G/C), and probably intron 10 (T/C), too, of As3MT gene. The elevated urinary arsenic background levels in certain populations, like the Asian islanders, might be partly affected by such genetic susceptibility. Future studies are warranted to explore the potential role of As3MT in the human metabolism of organic arsenic.

Acknowledgements

We gratefully acknowledge the financial support of the National Science Council, Taiwan, R.O.C. (NSC 91-2320-B-002-173, NSC 92-2320-B-002-169, NSC93-2320-B-002-084).

References

1. M. M. Moore, K. Harrington-Brock and C. L. Doerr, Mutat. Res., Rev. Mutat. Res., 1997, 386, 279–290.
2. F. Lucas, Posters of the Third International Conference on Arsenic Exposure and Health Effects, San Diego, 1998.
3. W. R. Cullen, B. C. McBride, H. Manji, A. W. Pickett and J. Reglinski, Appl. Organomet. Chem., 1989, 3, 71–78.
4. D. Velez and R. Montoro, J. Food. Protection., 1998, 61(9), 1240–1245.
5. M. Vahter, Toxicol. Lett., 2000, 112–113, 209–217.
6. S. Lin, Q. Shi, F. B. Nix, M. Styblo, M. A. Beck, K. M. Herbin-Davis, L. L. Hall, J. B. Simeonsson and D. J. Thomas, J. Biol. Chem., 2002, 277(13), 10795–10803.
7. A. Hernández, N. Xamena, J. Surrallés, C. Sekaran, H. Tokunaga, D. Quinteros, A. Creus and R. Marcos, Mutat. Res., Fundam. Mol. Mech. Mutagen., 2008, 637(1–2), 80–92.
8. R. Marcos, V. Martínez, A. Hernández, A. Creus, C. Sekaran, H. Tokunaga and D. Quinteros, J. Occup. Environ. Med., 2006, 48(3), 334–341.
9. M. Meza, A. J. Gandolfi and W. T. Klimecki, Toxicol. Appl. Pharmacol., 2007, 222, 381–387.
10. T. R. Radabaugh, A. Sampayo-Reyes, R. A. Zakharyan and H. V. Aposhian, Chem. Res. Toxicol., 2002, 15(5), 692–698.
11. R. A. Zakharyan, A. Sampayo-Reyes, S. M. Healy, G. Tsaprailis, P. G. Board, D. C. Liebler and H. V. Aposhian, Chem. Res. Toxicol., 2001, 14, 1051–1057.
12. B. Nemeti, I. Csanaky and Z. Gregus, Toxicol. Sci., 2003, 74, 22–31.
13. L. L. Marnell, G. G. Garcia-Vargas, U. K. Chowdhury, R. A. Zakharyan, B. Walsh, M. D. Avram, M. J. Kopplin, M. E. Cebrián, E. K. Silbergeld and H. V. Aposhian, Chem. Res. Toxicol., 2003, 16(12), 1507–1513.
14. M. M. Meza, L. Yu, Y. Y. Rodriguez, M. Guild, D. Thompson, A. J. Gandolfi and W. T. Klimecki, Environ. Health. Perspect., 2005, 113(6), 775–781.
15. D. J. Thomas, J. Li, S. B. Waters, W. Xing, B. M. Adair, Z. Drobna, V. Devesa and M. Styblo, Exp. Biol. Med., 2007, 232(1), 3–13.
16. European Food Safety Authority, EFSA. Journal., 2009, 7(10), 1351.
17. V. W. M. Lai, Y. Sun, E. Ting, W. R. Cullen and K. J. Reimer, Toxicol. Appl. Pharmacol., 2004, 198, 297–306.
18. http://www.ncbi.nlm.nih.gov/, accessed December 1, 2007.
19. L. Yu, K. Kalla, E. Guthrie, A. Vidrine and W. T. Klimecki, Environ. Health Perspect., 2003, 111(11), 1421–1427.
20. S. D. Chaudhuri, P. Ghosh, N. Sarma, P. Majumdar, T. J. Sau, S. Basu, S. Roychoudhury, K. Ray and A. K. Giri, Environ. Health. Perspect., 2008, 116, 501–505.
21. A. Chatterjee, Talanta, 2000, 51, 303–314.
22. J. S. Edmonds and K. A. Francesconi, in Organometallic Compounds in the Environment, ed. P. E. Craig, John Wiley and Sons, Chichester, 2nd edn, 2003, ch. 5, pp. 195–222.
23. R. O. Jenkins, A. W. Ritchie, J. S. Edmonds, W. Goessler, N. Molenat, D. Kuehnelt, C. F. Harrington and P. G. Sutton, Arch. Microbiol., 2003, 180, 142–150.
24. R. M. Brown, D. Newton, C. J. Pickford and J. C. Sherlock, Hum. Exp. Toxicol., 1990, 9, 41–46.
25. W. R. Cullen and K. J. Reimer, Chem. Rev., 1989, 89, 713–764.
26. X. C. Le, W. R. Cullen and K. J. Reimer, Clin. Chem., 1994, 40(4), 617–624.
27. X. C. Le, M. Ma and V. W. M. Lai, in Arsenic Exposure and Health Effects III, ed. W. R. Chappell, C. O. Abernathy and R. L. Calderon, Elsevier Science, Amsterdam, 1999, ch. 2, pp. 69–79.
28. M. S. Ma and X. C. Le, Clin. Chem., 1998, 44(3), 539–550.
29. C. Wei, W. H. Li, C. Zhang, M. V. Hulles, R. Cornelis and X. R. Zhang, J. Agric. Food Chem., 2003, 51, 5176–5182.
30. K. A. Francesconi, R. Tanggaard, C. J. McKenzie and W. Goessler, Clin. Chem., 2002, 48(1), 92–101.
31. R. Heinrich-Ramm, S. Mindt-Prüfert and D. Szadkowski, J. Chromatogr., B: Anal. Technol. Biomed. Life Sci., 2002, 778, 263–273.
32. T. Hayakawa, Y. Kobayashi, X. Cui and S. Hirano, Arch. Toxicol., 2005, 79, 183–191.
33. A. L. Lindberg, R. Kumar, W. Goessler, R. Thirumaran, E. Gurzau, K. Koppova, P. Rudnai, G. Leonardi, T. Fletcher and M. Vahter, Environ. Health Perspect., 2007, 115(7), 1081–1086.
34. K. Schlawicke Engstrom, K. Broberg, G. Concha, B. Nermell, M. Warholm and M. Vahter, Environ. Health Perspect., 2007, 115(4), 599–605.
35. T. C. Wood, O. E. Salavagionne, B. Mukherjee, L. Wang, A. F. Klumpp, B. A. Thomae, B. W. Eckloff, D. J. Schaid, E. D. Wieben and R. M. Weinshilboum, J. Biol. Chem., 2006, 281(11), 7364–7373.
36. A. Hata, Y. Endo, Y. Nakajima, M. Ilebe, M. Ogawa, N. Fujitani and G. Endo, J. Occup. Health, 2007, 49, 217–223.
37. Y. M. Hsueh, M. K. Hsu, H. Y. Chiou, M. H. Yang, C. C. Huang and C. J. Chen, Toxicol. Lett., 2002, 133(1), 83–91.
38. H. Yamauchi, K. Takahashi, M. Mashiko and Y. Yamamura, Am. Ind. Hyg. Assoc. J., 1989, 50(11), 606–612.
39. J. P. Buchet, R. Lauwerys and H. Roels, Int. Arch. Occup. Environ. Health, 1981, 48, 111–118.
40. J. G. Farmer and L. R. Johnson, Br. J. Ind. Med., 1990, 47, 342–348.
41. J. Morton and H. Mason, J. Analytical. Toxicol., 2006, 30, 293–301.
42. M. Vahter, L. Friberg, B. Rahnster, A. Nygren and P. Nolinder, Int. Arch. Occup. Environ. Health, 1986, 57, 79–91.
43. J. Fujihara, T. Kunito, T. Agusa, T. Yasuda, R. Iida, Y. Fujii and H. Takeshita, Toxicol. Appl. Pharmacol., 2007, 225, 251–254.
44. American Conference for Governmental Industrial Hygienists, Threshold Limit Values for chemical substances and physical agents and Biological Exposure Indices, American Conference for Governmental Industrial Hygienists, Cincinnati, 2009.
45. World Health Organization, Environmental Health Criteria 224—Arsenic and arsenic compounds, World Health Organization, Geneva, 2nd edn, 2001.
46. G. Concha, G. Vogler, B. Nermell and M. Vahter, Environ. Health Perspect., 1998, 106(6), 355–359.
47. P. Kurttio, H. Komulainen, E. Hakala, H. Kahelin and J. Pekkanen, Arch. Environ. Contam. Toxicol., 1998, 34, 297–305.

---