Environmental Toxicology and Pharmacology 19 (2005) 1–7

Effects of dietary folate intake and folate binding protein-2 (Folbp2)on urinary speciation of sodium arsenate in mice

Ofer Spiegelsteina,∗, Xiufen Lub, X. Chris Leb, Aron Troenc, Jacob Selhubc,Stepan Melnykd, S. Jill Jamesd, Richard H. Finnella

a Center for Environmental and Genetic Medicine, Institute of Biosciences and Technology, Texas A&MUniversity System Health Science Center, Houston, TX 77030, USA

b Department of Public Health Sciences, University of Alberta, Edmonton, Alta., Canada T6G 2G3c Vitamin Metabolism and Neurocognitive Laboratories, Jean Mayer USDA Human Nutrition Research Center on Aging

at Tufts University, Boston, MA 02111,USAd Department of Pediatrics, University of Arkansas for Medical Sciences, Little Rock, AR 72205, USA

Received 14 August 2003; accepted 7 January 2004Available online 11 September 2004

A

arsenic.A istry anda oulds sformation.F served ina in Folbp2m omeostasisi©

K

1

glastos

PT

Ds;-

lmostnot

sure.not

alu-

wither ofple-

, andailyhas70%

1d

bstract

Folate binding protein-2 (Folbp2−/−) knockout mice have been previously shown to be highly susceptible to the teratogenic effects ofrsenic biotransformation is achieved primarily by biomethylation. Given the potential close relationship between folate biochemrsenic biotransformation, the aims of our study were to: (1) test whether Folbp2−/− mice have altered arsenic biotransformation which wuggest a potential mechanism for their enhanced susceptibility; (2) examine whether dietary folate deficiency alters arsenic biotranolbp2−/− mice were found to have slightly lower plasma folate levels than wildtype mice. No genotype-specific effects were obrsenic speciation thereby negating altered biotransformation of arsenic as the mechanism of the enhanced teratogenicity seen−/−

ice. Reduction in excretion of organic arsenicals was observed during folate deficiency, suggesting an important role for folic acid hn arsenic biotransformation.

2004 Elsevier B.V. All rights reserved.

eywords:Folic acid; Arsenic; Excretion; Urinary speciation

. Introduction

Inorganic arsenic has long been suspected of being terato-enic in humans; however, the existing published literature is

imited to a few case reports and epidemiological studies. Asresult, the association between in utero human fetal expo-

ure to arsenic and adverse pregnancy outcome remains con-roversial (Shalat et al., 1996; DeSesso, 2001). In a numberf laboratory animal species, arsenic has consistently beenhown to be teratogenic, primarily inducing exencephaly, the

∗ Corresponding author. Present address: Global Innovative R&D, Tevaharmaceutical Industries, P.O. Box 8077, Netanya 42504, Israel.el.: +972 9 863 9713; fax: +972 9 863 1460.

E-mail address:ofer.spiegelstein@teva.co.il (O. Spiegelstein).

major form of teratogen-induced neural tube defects (NTShalat et al., 1996; Golub et al., 1998). However, arsenicinduced teratogenicity in animals has been achieved aexclusively by acute injections of high doses, which doesrepresent the typical conditions of human arsenic expoThus, extrapolating from animal studies to humans isstraightforward, and warrants additional scientifically vable human epidemiological studies.

Folate deficiency has been shown to be associatedthe occurrence of birth defects, and a considerable numbstudies have shown the protective role of folic acid supmentation in preventing neural tube, conotruncal heartcraniofacial malformations in humans. For example, dsupplementation with folic acid in the range of 0.4–5 mgbeen shown to reduce the incidence of NTDs by up to

382-6689/$ – see front matter © 2004 Elsevier B.V. All rights reserved.oi:10.1016/j.etap.2004.01.007

2 O. Spiegelstein et al. / Environmental Toxicology and Pharmacology 19 (2005) 1–7

(Mulinare et al., 1988; Czeizel and Dudas, 1992). However,despite the accumulating evidence regarding the embryo-protective effect of folic acid, the mechanism by which thisbeneficial effect is achieved remains unknown.

Folates enter cells by way of the folate receptors (FRs),also known as folate binding proteins (Folbps) in mice, andthe reduced folate carrier. While the reduced folate carrieris a ubiquitously expressed membrane transporter, most FRsare externally bound to the plasma membrane by glycosyl-phosphatidylinositol anchors, and have tissue-specific andcell-specific expression patterns (Antony, 1992, 1996; Rosset al., 1994). The human FR-� is variably expressed in mosthuman tissues and binds various folate forms with relativelyhigh affinity (Ratnam et al., 1989; Ross et al., 1994). Usinghomologous recombination, we have previously inactivated(knocked out) the Folbp2 gene, the mouse homologue ofFR-�, and generated nullizygous Folbp2 (Folbp2−/−) mice(Piedrahita et al., 1999). Folbp2−/− mice develop normallyand have no apparent pathologies or abnormalities associatedwith the mutation. However, when pregnant Folbp2−/− damswere treated with sodium arsenate during gestational days 7.5and 8.5, the surviving fetuses had an elevated incidence ofNTDs compared to control mice (Wlodarczyk et al., 2001).Folate deficiency further exacerbated this phenomenon in themutant Folbp2 mice but not in controls, and thus it was con-

FA5

cluded that Folbp2−/− mice have an increased susceptibilityto in utero arsenic exposure.

In most mammalian species, the major route of arsenicdetoxification occurs via biomethylation reactions (Fig. 1).For the first biomethylation reaction to occur, arsenate(As(V)) must be activated by reduction to its trivalent form,arsenite (As(III)). Arsenite serves as a substrate for a methyl-transferase that generates the organic arsenical monomethy-larsonate (MMA(V)) (Vahter and Concha, 2001; Lin et al.,2002). MMA(V) is then reduced to monomethylarsonousacid (MMA(III)) which is biomethylated to dimethylarsonicacid (DMA(V)), the major arsenic metabolite excreted in theurine (Vahter and Marafante, 1988). Both biomethylation re-actions are catalyzed by methyltransferases that requireS-adenosylmethionine (SAM) as the methyl donor (Vahter andEnval, 1983; Aposhian, 1997; Lin et al., 2002). The metabolicprocess that generates SAM occurs primarily via the trans-methylation pathway (Fig. 1), where the conversion of ho-mocysteine (Hcy) to methionine requires the transfer of amethyl group from 5-methyl-tetrahydrofolate (5M-THF), theprimary bioactive folate form. Methionine is then convertedto SAM by methionine adenosyltransferase and ATP. SAM,after donating its methyl group to acceptor molecules, isconverted toS-adenosylhomocysteine (SAH), which is hy-drolyzed to Hcy and adenosine.

ig. 1. Schematic representation of arsenic biotransformation pathway, thes(V): arsenate; DMA(V): dimethylarsonic acid; Hcy: homocysteine; MMA(-methyl tetrahydrofolate; THF: tetrahydrofolate; SAH:S-adenosylhomocystein

homocysteine transmethylation and folate metabolic recycling. As(III): arsenite;III): monomethylarsinic acid; MMA(V): monomethylarsonic acid; 5M-THF:

e; SAM:S-adenosylmethionine.

O. Spiegelstein et al. / Environmental Toxicology and Pharmacology 19 (2005) 1–7 3

As folate transport and metabolism seem to be importantfor arsenic biotransformation, in the current study we hypoth-esize that Folbp2−/− mice have diminished ability to metab-olize inorganic arsenic when compared to wildtype controlmice, which would provide a plausible explanation to theirincreased susceptibility. We tested this hypothesis by inject-ing arsenate to Folbp2−/− and wildtype control mice andperforming 24-h urinary speciation of inorganic and organicarsenic species. In addition, we examined the dependency ofarsenic biotransformation on folate homeostasis by assess-ing urinary speciation during folate deficiency. Our hypoth-esis was that folate deficiency would result in altered arsenicbiotransformation, thus pointing toward the importance ofadequate nutritional folate intake for arsenic biotransforma-tion.

2. Methods

2.1. Animal studies

In the first phase of the study, five wildtype and fiveFolbp2−/− adult male mice, 3–5 months of age, were placedon an amino acid defined diet containing 2.7 mg folate/kg diet(folate replete control diet; Dyets #517839, Bethlehem, PA.C diet)f le tor vail-a t,a alsa .I ed as se-n fori of1 cedi ster,N pro-v xi-m MO)a nty-f d thec ofw to-t wasf

t thec thes on ana (re-d r 27d den-t erf plew tion

was performed in a similar manner as in the first phase of thestudy.

Mice were allowed free access to water and food andwere housed in clear polycarbonate cages with Harlan Sani-chips® bedding (Harlan Teklad, Madison, WI) to preventvitamin intake from bedding consumption. The mice weremaintained on a 12-h light/dark cycle at the vivarium of theInstitute of Biosciences and Technology, Houston, TX. Theinstitutional animal care and use committee approved thisstudy.

2.2. Arsenic urinary speciation

During the first phase of the study, once the samples werereceived in the analyzing lab which was generally less than48-h after the completion of urine collection, they were im-mediately processed and prepared for analysis, in order toincrease the chances of detecting unstable trivalent arseni-cals, such as MMA(III). However, since no trivalent methy-lated arsenicals were detected in the urine of freshly deliveredsamples, all subsequent urine samples were stored at−80◦Cand shipped together on dry ice.

Arsenic analysis was performed by ion pair chromato-graphic separation and anion exchange chromatographicseparation with hydride generation atomic fluorescenced idc ton,W orts ,a tionw S-3,1 x,T ing5 id,a fort ed at5 ana3 ands obilepm ner-a 003,P ofa

im-iMM .

2

theL e-s

holine content was 2.0 g/kg diet, methionine 8.2 mg/kgor 20 days. The diet included 1% succinyl sulphathiazoeduce gastrointestinal bacterial flora that produce bioable folates (Bills et al., 1992). On the day of the experimenblood sample (∼300�L) was collected via the retro-orbitinus, which was subsequently spun for 5 min at 5000× gnd the plasma separated and stored at−80◦C until analyzed

mmediately after blood collection, each mouse receivingle 1 mg/kg intraperitoneal (ip) injection of sodium arate (Sigma, St. Louis, MO) dissolved in sterile water

njection (Abbott Laboratories, Chicago, IL), at a volume0�L/g body weight. Each mouse was individually pla

n a metabolic cage (Nalge Nunc International, RocheY) with powdered diet and a 10% sucrose solutionided ad libitum. Into each urinary collection tube, approately 100 mg of sodium ascorbate (Sigma, St. Louis,nd 7.5 mg of sodium azide (Acros, NJ) were added. Twe

our hours post-injection, the animals were removed anage interior thoroughly rinsed with approximately 10 mLater to ensure collection of all the dried-up urine. The

al volume of diluted urine was recorded and an aliquotrozen at−80◦C until analyzed.

As this study was performed in a crossover design, aompletion of the first phase of the study (control diet),ame mice were returned to regular cages and placedmino-acid defined diet containing 0.3 mg folate/kg dietuced folate diet; Dyets #517796, Bethlehem, PA) foays. The composition of the reduced folate diet was i

ical to the control diet with the exception of having lowolate content. At the day of the experiment, a blood samas taken prior to arsenic injection and 24-h urine collec

etection (Le et al., 2000). The high performance liquhromatography system consisted of a Gilson (MiddleI) pump (Model 307), a Rheodyne (Cotati, CA) 6-p

ample injector (Model 7725i) with a 20-�L sample loopnd a column. The ion pair chromatographic separaas performed on a reversed-phase C18 column (OD50 mm × 4.6 mm, 3-�m particle size, Phenomeneorrance, CA). A mobile phase solution (pH 5.9) containmM tetrabutylammonium hydroxide, 3 mM malonic acnd 5% methanol (flow rate was 1.2 mL/min) was used

he separation. The column temperature was maintain0◦C. For the determination of trimethylarsine oxide,nion exchange column (PRP-X100, 150 mm× 4.1 mm,-�m particle size, Hamilton, Reno, NV) was used,eparation was performed at room temperature. The mhase solution (pH 8.2) contained 5 mM Na2HPO4 and 5%ethanol, and its flow rate was 0.8 mL/min. A hydride getion atomic fluorescence detector (Model Excalibur 10..S. Analytical, Kent, UK) was used for the detectionrsenic.

Using the described analytic system, the following lts of detection were obtained: 0.5�g/L for As(III) and

MA(V); 1 �g/L for DMA(V) and As(V); 2�g/L forMA(III) and DMA(III); 7 �g/L for trimethylarsine oxide

.3. Folate quantification in plasma

Total folate levels in plasma were determined usingactobacillus Caseimicrobiological assay as recently dcribed (Spiegelstein et al., 2003b).

4 O. Spiegelstein et al. / Environmental Toxicology and Pharmacology 19 (2005) 1–7

2.4. SAM, SAH and adenosine quantification in plasma

Plasma concentrations of SAM, SAH and adenosine weredetermined by high-pressure liquid chromatography withelectrochemical detection as previously described (Melnyket al., 2000).

2.5. Statistical analysis

Genotype-specific analysis for statistical significance interms of plasma folate, SAM, SAH and adenosine concen-trations, the SAM/SAH ratios, and the urinary excretion ofarsenicals was performed using the individual animal data ofeach group and using a student’st-test (P value set at 0.05,two-sided). Since all dietary-specific effects were comparedin the same animals, statistical significance was analyzed us-ing a pairedt-test (P value set at 0.05, two-sided).

3. Results

3.1. Urinary arsenic speciation

The 24-h urinary speciation data is presented inTable 1.Overall, DMA(V) was the arsenic species excreted to theg f thea und,a xtento % oft e oft ted int senceo td

cesic lated xcre-t thec nics n the

TT

D As(III)

C 10.8 ± 38.1 ± 3

F 6.3 ± 05.7 ± 1

V the ad nic acid:M

t: folat

pe andthe co

wildtype mice as well; however, it was short of reaching sta-tistical significance (P> 0.05). Similar results were obtainedin the Folbp2−/− mice; however, none reached statistical sig-nificance (P > 0.05).

3.2. Plasma folate concentrations

Total folate concentrations in plasma are presented inTable 2. Wildtype mice fed the control diet had signifi-cantly higher (42.6%) plasma folate levels than the mutantFolbp2−/− mice (P = 0.002). The genotype related differ-ences of plasma folate levels were maintained during folatedeficiency, i.e. wildtype mice had 36.0% higher folate levelsthan the knockouts (P= 0.020). Folate levels dropped signif-icantly by approximately sevenfold during the second phasein both strains (P< 0.0001 in the wildtype mice;P= 0.004 inthe Folbp2−/− mice), indicating that the reduction of dietaryfolate content from 2.7 to 0.3 mg/kg diet was sufficient forinducing folate deficiency.

3.3. Plasma SAM and SAH concentrations, andSAM/SAH ratios

The plasma concentrations of SAM and SAH, as well asthe SAM/SAH ratio, often used as a measure of the over-at ere5( theS ani sti-c

theFl em sedt hant

ereo -t the

reatest extent in the urine, accounting for 43–53% odministered dose. Excretion of As(V), the parent compoccounted for 14–26% of the administered dose. The ef As(III) excretion was 6–11%, whereas no more than 1

he dose was excreted as MMA(V) during the first phashe study. The average total amount of arsenicals excrehe urine was 65–88% of the administered dose. The pref MMA(III), DMA(III) and trimethylarsine oxide was noetected.

No statistically significant genotype-specific differenn excretion of arsenicals between Folbp2−/− and wildtypeontrol mice were observed, either on the control or the foeficient diets. Folate deficiency reduced total arsenic e

ion by 24.6% in the wildtype mice when compared toontrol diet (P= 0.036). Reduction in excretion of all arsepecies as a result of folate deficiency was observed i

able 1wenty-four hour urinary speciation of arsenicals

ieta Genotype As(V) %

ontrolb Wildtype 23.6± 5.2Folbp2−/− 26.0± 6.8

olate deficient Wildtype 14.3± 4.9Folbp2−/− 18.9± 10.1

alues are expressed as the mean± standard deviation percentage ofMA(V); dimethylarsonic acid: DMA(V).a Control diet: folate content was 2.7 mg/kg diet; folate deficient dieb Data reproduced fromSpiegelstein et al. (2003).c Not detected: MMA(V) was not detected in the urine of 5/5 wildtyd Dietary effect: statistically significant compared to the wildtype on

% MMA(V) % DMA(V) % Total %

.6 0.86± 0.28 53.0± 14.3 88.3± 15.6

.7 0.58± 0.45 52.6± 13.3 87.3± 12.1

.8 NDc 44.4± 6.2 65.0± 6.3d

.6 NDc 42.8± 7.2 67.4± 12.6

ministered dose. Arsenate: As(V); arsenite: As(III); monomethylarso

e content was 0.3 mg/kg diet.

4/5 Folbp2−/− mice.ntrol diet.P < 0.05, pairedt-test, two-tailed.

ll methylation capacity, are presented inTable 2. Wild-ype mice fed the control diet had SAM levels that w2.2% higher than levels obtained in the Folbp2−/− miceP = 0.005), whereas SAH levels were similar. AlthoughAM/SAH ratio was 27.6% higher in the wildtype mice th

n the Folbp2−/− mice, the differences did not reach statial significance (P > 0.05).

During diet-induced folate deficiency, SAH levels inolbp2−/− mice were 18.7% higher (P = 0.034) and SAM

evels were 27.2% lower (P = 0.026) than in the wildtypice. The opposing trends in SAM and SAH levels cau

he SAM/SAH ratio to be 37.7% lower in the knockouts the wildtype mice (P < 0.0001).

Interestingly, statistically significant dietary-effects wbserved only for the Folbp2−/− mice, but not in the wild

ype controls. Folate deficiency increased SAH levels in

O. Spiegelstein et al. / Environmental Toxicology and Pharmacology 19 (2005) 1–7 5

Table 2Plasma concentrations of folate, SAM and SAH

Dieta Strain Folate concentration (ng/mL) SAH (pmol/mL) SAM (pmol/mL) SAM/SAH

Control Wildtype 32.8 ± 1.3 52.0± 4.5 140.8 ± 27.3 2.75± 0.73Folbp2−/− 18.5 ± 6.8b 47.9± 10.9 92.5 ± 8.1b 1.99± 0.37

Folate deficient Wildtype 4.4 ± 1.0c 55.1± 8.1 131.1 ± 27.3 2.36± 0.18Folbp2−/− 2.8 ± 0.8b,c 65.4± 4.0b,c 95.5 ± 10.1b 1.47± 0.18b,c

Values are expressed as the mean± standard deviation.a Control diet: folate content was 2.7 mg/kg diet; folate deficient diet: folate content was 0.3 mg/kg diet.b Genotype effect: statistically significant compared to the wildtype mice on the same diet. Student’st-test,P < 0.05, two-tailed.c Dietary effect: statistically significant compared to the same genotype on the control diet. Pairedt-test,P < 0.05, two-tailed.

mutant mice by 36.5% (P = 0.033), but did not affect SAM.As a result of the increase in SAH levels, the SAM/SAH ratiowas decreased by 26.1% (P = 0.049).

4. Discussion

We have previously shown that Folbp2−/− mice weremore susceptible to the teratogenic effects of arsenic than thewildtype controls (Wlodarczyk et al., 2001). To test whetherthe increased susceptibility of the mutant mice was due toaltered arsenic biotransformation and elimination, we havepreviously performed a urinary speciation study using a ‘ter-atogenic relevant’ 30 mg/kg dose; however, our previous find-ings did not support the hypothesis as both genotypes dis-played similar excretion of arsenicals. It is important to note,however, that the ‘metabolic hypothesis’ could not be firmlyruled out due to the relatively high dose that was found toskew excretion values of DMA(V) and As(V). In the presentstudy we used a lower 1 mg/kg dose of arsenic, and excretionof arsenicals was consistent with previously published val-ues (Hood et al., 1987; Hughes et al., 1999). Nevertheless,the current results indicate that there were no differences inarsenic biotransformation and excretion between Folbp2−/−and wildtype mice, regardless of the dietary folate intake reg-i senicb rato-g

olbp2r a fo-l wasmm . Ina erer sed.T on inf ver,d orma-t s ine taryr

otala utedl ofM et,

but was completely undetectable during folate deficiency.Similar results were also recently seen in Folbp1−/− mice(Spiegelstein et al., 2003a), and thus it appears that folatedeficiency impairs methylation capacity that is evident in thediminished excretion of both organic arsenic species, andthe total amount excreted. The mechanism by which thisphenomenon occurs is likely due to the ability of folate de-ficiency to cause elevation of plasma Hcy, resulting in el-evated SAH levels with concomitant inhibition of methyl-transferase enzymes (Lin et al., 1989; Hoffman et al., 1979).In a likely similar mechanism, mice and rabbits fed dietslow or devoid of methyl donor amino acids choline and me-thionine, had increased bodily retention and decreased uri-nary excretion of arsenic, predominantly due to diminishedmetabolism to DMA(V) (Vahter and Marafante, 1987; Ticeet al., 1997). Similarly, inhibition of methyltransferases inrabbits resulted in a significant decrease in the productionof DMA(V) ( Marafante and Vahter, 1984), and in an in vitromodel SAM maximized DMA(V) production, whereas SAH,decreased its production (Styblo et al., 1996).

Our present data provides supportive evidence that arsenicmetabolism and excretion is likely to be impaired during fo-late deficiency in humans, a condition that is not uncom-mon in certain sub-populations (Wright et al., 1998; Marshallet al., 2001). Impaired arsenic biotransformation and excre-t hu-m ly toi op-u

upss ncen-t en-h latel ctedt ter-a dra-m rato-g gedw xen-c 2m pec-t

sus-c nice ation

men. Hence, we can conclude that altered maternal ariotransformation is unlikely to cause the increased teenicity in Folbp2−/− mice.

In this study, we demonstrated that the absence of a Feceptor results not only in a modest reduction of plasmate levels, but also in altered SAM and SAH levels. This

ostly apparent during folate deficiency, when Folbp2−/−ice had significantly lower folate and higher SAH levelsddition, the SAM levels of the Folbp2 nullizygous mice weduced and the SAM/SAH ratio significantly decreahese changes are consistent with a modest alterati

olate homeostasis and methylation capacity. Howeespite these biochemical changes, arsenic biotransf

ion remained unaltered as evident by the similaritiexcretion of methylated arsenicals under the two dieegimens.

Overall, folate deficiency caused reduction in the tmount of arsenic excreted in the urine, which is attrib

argely to the reduction in DMA(V) excretion. ExcretionMA(V) was very small to begin with on the control di

ion results in increased body retention and exposure inans. In turn, an increase in arsenic exposure is like

ncrease the risk of toxicity, especially in certain at-risk plations due to genetic predisposition (Vahter, 2000).

Our findings of plasma folate levels in the two mice grouggest that the moderately reduced folate plasma corations in the Folbp2−/− mice are also not the cause ofanced susceptibility in the mutant mice. If reduced fo

evels were to explain this phenomenon, it would be expehat the wildtype mice would have a significant increase intogenicity under folate deficiency. However, despite aatic reduction in folate levels in both genotypes, the teenic outcome in the wildtype controls remained unchanhen compared to the control diet (25.7 and 24.0% eephaly, respectively), whereas exencephaly in Folbp−/−ice was significantly elevated (40.6 and 64.0%, res

ively) under the same conditions.In the current study we ruled out that the enhanced

eptibility of Folbp2 nullizygous mice to in utero arsexposure is due to alterations in arsenic biotransform

6 O. Spiegelstein et al. / Environmental Toxicology and Pharmacology 19 (2005) 1–7

or intrinsic folate deficiency per se. Future studies will beaddressing several possible hypotheses. One will be focusedon testing whether alterations of gene expression occurredin the embryo secondary to alterations in DNA methylation,which is a major epigenetic mechanism. In addition, we willbe testing whether differences in distribution and accumu-lation of arsenicals into embryonic tissues may account fordifferences in developmental susceptibilities. An additionalhypothesis being considered is that FRs, including Folbp2have biological role(s) other than just transporting folatesacross cell membranes, and that their absence renders em-bryos more susceptible to teratogenic insults. This latter hy-pothesis is especially intriguing and is supported by severalstudies regarding the cellular localization of FRs and theirpotential interactions with signaling molecules (Okamotoet al., 1998; Miotti et al., 2000).

Acknowledgments

This project was supported in part by grants ES 04917, ES09106 and ES 11775 from the National Institute of Environ-mental Health to RHF; grants from the Natural Sciences andEngineering Research Council of Canada and Canadian Wa-ter Networks NCE to XCL. The paper’s contents are solelyt epre-s uldl ie,a andG nol-o ncei DAH ni-v ; Dr.L andG nol-o itht

R

A lood

A 1.A other

xicol.

B ive

C ural-gl. J.

D gy 63,

G andman

Hoffman, D.R., Cornatzer, W.E., Duerre, J.A., 1979. Relationship be-tween tissue levels ofS-adenosylmethionine,S-adenylhomocysteine,and transmethylation reactions. Can. J. Biochem. 57, 56–65.

Hood, R.D., Vedel-Macrander, G.C., Zaworotko, M.J., Tatum, F.M.,Meeks, R.G., 1987. Distribution, metabolism, and fetal uptake ofpentavalent arsenic in pregnant mice following oral or intraperitonealadministration. Teratology 35, 19–25.

Hughes, M.F., Kenyon, E.M., Edwards, B.C., Mitchell, C.T., Thomas,D.J., 1999. Strain-dependent disposition of inorganic arsenic in themouse. Toxicology 137, 95–108.

Le, X.C., Lu, X., Ma, M., Cullen, W.R., Aposhian, V., Zheng, B., 2000.Speciation of key arsenic metabolic intermediates in human urine.Anal. Chem. 72, 5172–5177.

Lin, J.Y., Kang, S.S., Zhou, J.M., Wong, P.W., 1989. Homocysteinemiain rats induced by folic acid deficiency. Life Sci. 44, 319–325.

Lin, S., Shi, Q., Nix, F.B., Styblo, M., Beck, M.A., Herbin-Davis, K.M.,Hall, L.L., Simeonsson, J.B., Thomas, D.J., 2002. A novelS-adenosyl-l-methionine:arsenic(III) methyltransferase from rat liver cytosol. J.Biol. Chem. 277, 10795–10803.

Marafante, E., Vahter, M., 1984. The effect of methyltransferase inhibitionon the metabolism of [74As] arsenite in mice and rabbits. Chem. Biol.Interact. 50, 49–57.

Marshall, T.A., Stumbo, P.J., Warren, J.J., Xie, X.J., 2001. Inadequatenutrient intakes are common and are associated with low diet varietyin rural, community-dwelling elderly. J. Nutr. 131, 2192–2196.

Melnyk, S., Pogribna, M., Pogribny, I.P., Yi, P., James, S.J., 2000.Measurement of plasma and intracellularS-adenosylmethionine andS-adenosylhomocysteine utilizing coulometric electrochemical detec-tion: alterations with plasma homocysteine and pyridoxal 5′-phosphateconcentrations. Clin. Chem. 46, 265–272.

M S.,andarci-

M ncep-fects.

O olins,nal-419–

P B.A.,999.

in

R mol-oning

R n ofand

ncer

S as as. J.

S S.,and

ium

S 03b.tein-74–

S tionpell,and

he responsibility of the authors and do not necessarily rent the official views of the NIEHS, NIH. The authors woike to thank Ms. Michelle Merriweather, Ms. Marlene Tsnd Mr. Joe Wicker from the Center for Environmentalenetic Medicine at the Institute of Biosciences and Techgy, Texas A&M University System HSC for their assista

n this study; Ms. Marie Nadeau from the Jean Mayer USuman Nutrition Research Center on Aging at Tufts Uersity for assisting with the folate determination assaysaura E. Mitchell from the Department of Environmentalenetic Medicine at the Institute of Biosciences and Techgy, Texas A&M University System HSC for assisting w

he statistical analysis.

eferences

ntony, A.C., 1992. The biological chemistry of folate receptors. B79, 2807–2820.

ntony, A.C., 1996. Folate receptors. Annu. Rev. Nutr. 16, 501–52poshian, H.V., 1997. Enzymatic methylation of arsenic species and

new approaches to arsenic toxicity. Annu. Rev. Pharmacol. To37, 397–419.

ills, N.D., Koury, M.J., Clifford, A.J., Dessypris, E.N., 1992. Ineffecthematopoiesis in folate-deficient mice. Blood 79, 2273–2280.

zeizel, A.E., Dudas, I., 1992. Prevention of the first occurrence of netube defects by periconceptional vitamin supplementation. N. EnMed. 327, 1832–1835.

eSesso, J.M., 2001. Teratogen update: inorganic arsenic. Teratolo170–173.

olub, M.S., Macintosh, M.S., Baumrind, N., 1998. Developmentalreproductive toxicity of inorganic arsenic: animal studies and huconcerns. J. Toxicol. Environ. Health B: Crit. Rev. 1, 199–241.

iotti, S., Bagnoli, M., Tomassetti, A., Colnaghi, M.I., Canevari,2000. Interaction of folate receptor with signaling molecules lynG(alpha)(i-3) in detergent-resistant complexes from the ovary cnoma cell line IGROV1. J. Cell Sci. 113, 349–357.

ulinare, J., Cordero, J.F., Erickson, J.D., Berry, R.J., 1988. Pericotional use of multivitamins and the occurrence of neural tube deJ. Am. Med. Assoc. 260, 3141–3145.

kamoto, T., Schlegel, A., Scherer, P.E., Lisanti, M.P., 1998. Cavea family of scaffolding proteins for organizing “preassembled siging complexes” at the plasma membrane. J. Biol. Chem. 273, 55422.

iedrahita, J.A., Oetama, B., Bennett, G.D., van Waes, J., Kamen,Richardson, J., Lacey, S.W., Anderson, R.G., Finnell, R.H., 1Mice lacking the folic acid-binding protein Folbp1 are defectiveearly embryonic development. Nat. Genet. 23, 228–232.

atnam, M., Marquardt, H., Duhring, J.L., Freisheim, J.H., 1989. Hoogous membrane folate binding proteins in human placenta: cland sequence of a cDNA. Biochemistry. 28, 8249–8254.

oss, J.F., Chaudhuri, P.K., Ratnam, M., 1994. Differential regulatiofolate receptor isoforms in normal and malignant tissues in vivoin established cell lines. Physiologic and clinical implications. Ca73, 2432–2443.

halat, S.L., Walker, D.B., Finnell, R.H., 1996. Role of arsenicreproductive toxin with particular attention to neural tube defectToxicol. Environ. Health 48, 253–272.

piegelstein, O., Lu, X., Le, X.C., Troen, A., Selhub, J., Melnyk,James, S.J., Finnell, R.H., 2003a. Effects of dietary folate intakefolate binding protein-1 (Folbp1) on urinary speciation of sodarsenate in mice. Tox. Lett. 145, 167–174.

piegelstein, O., Merriweather, M.Y., Wicker, N.J., Finnell, R.H., 20Valproate-induced neural tube defects in folate binding pro2 (Folbp2) knockout mice. Birth Defects Res., Part A 67, 9978.

piegelstein, O., Lu, X., Le, X.C., Finnell, R.H., 2003. Urinary speciaof sodium arsenate in folate receptor knockout mice. In: ChapW.R., Abernathy, C.O., Calderon, R.L. (Eds.), Arsenic ExposureHealth Effects, vol. V. Elsevier, pp. 337–344.

O. Spiegelstein et al. / Environmental Toxicology and Pharmacology 19 (2005) 1–7 7

Styblo, M., Delnomdedieu, M., Thomas, D.J., 1996. Mono- and dimethy-lation of arsenic in rat liver cytosol in vitro. Chem. Biol. Interact. 99,147–164.

Tice, R.R., Yager, J.W., Andrews, P., Crecelius, E., 1997. Effect of hepaticmethyl donor status on urinary excretion and DNA damage in B6C3F1mice treated with sodium arsenite. Mutat. Res. 386, 315–334.

Vahter, M., 2000. Genetic polymorphisms in the biotransformation ofinorganic arsenic and its role in toxicity. Toxicol. Lett. 112–113,209–217.

Vahter, M., Concha, G., 2001. Role of metabolism in arsenic toxicity.Pharmacol. Toxicol. 89, 1–5.

Vahter, M., Enval, J., 1983. In vivo reduction of arsenate in mice andrabbits. Environ. Res. 32, 14–24.

Vahter, M., Marafante, E., 1988. In vivo methylation and detoxification ofarsenic. In: Craig, P.J., Glockling, F. (Eds.), The Biological Alkylationof Heavy Elements. Royal Society of Chemistry, London, UK, p. 105.

Vahter, M., Marafante, E., 1987. Effects of low dietary intake of methio-nine, choline or proteins on the biotransformation of arsenite in therabbit. Toxicol. Lett. 37, 41–46.

Wlodarczyk, B., Spiegelstein, O., Gelineau van-Waes, J., Vorce, R.L., Lu,X., Le, X.C., Finnell, R.H., 2001. Arsenic-induced congenital malfor-mations in genetically susceptible folate binding protein-2 knockoutmice. Toxicol. Appl. Pharmacol. 177, 238–246.

Wright, J.D., Bialostosky, K., Gunter, E.W., Carroll, M.D., Najjar, M.F.,Bowman, B.A., Johnson, C.L., 1998. Blood folate and vitamin B12:United States, 1988–1994. Vital Health Stat. 243, 1–78.