ORGAN TOXICITY AND MECHANISMS

Seishiro Hirano Æ Yayoi Kobayashi Æ Toru Hayakawa

Xing Cui Æ Megumi Yamamoto Æ Sanae Kanno

Amjad Shraim

Accumulation and toxicity of monophenyl arsenicalsin rat endothelial cells

Received: 8 April 2004 / Accepted: 30 June 2004 / Published online: 11 September 2004� Springer-Verlag 2004

Abstract Clark 1 (diphenylarsine chloride) and Clark 2(diphenylarsine cyanide) were used as chemical weaponagents (CWA), and the soil contamination by theseCWA and their degraded products, diphenyl and phenylarsenicals, has been one of the most serious environ-mental issues. In a series of comparisons in toxicity be-tween trivalent and pentavalent arsenicals weinvestigated differences in the accumulation and toxicityof phenylarsine oxide (PAO3+) and phenylarsonic acid(PAA5+) in rat heart microvascular endothelial cells.Both the cellular association and toxicity of PAO3+

were much higher than those of PAA5+, and LC50 val-ues of PAO3+ and PAA5+ were calculated to be0.295 lM and 1.93 mM, respectively. Buthionine sul-foximine, a glutathione depleter, enhanced the cytotox-icity of both PAO3+ and PAA5+. N-Acetyl-L-cysteine(NAC) reduced the cytotoxicity and induction of hemeoxygenase-1 (HO-1) mRNA in PAO3+-exposed cells,while NAC affected neither the cytotoxicity nor the HO-1 mRNA level in PAA5+-exposed cells. The effect ofNAC may be due to a strong affinity of PAO3+ to thiolgroups because both NAC and GSH inhibited the cel-lular accumulation of PAO3+, but PAA3+ increasedtyrosine phosphorylation levels of cellular proteins.These results indicate that the inhibition of protein

phosphatases as well as the high affinity to cellularcomponents may confer PAO3+ the high toxicity.

Keywords Phenylarsine oxide Æ Phenylarsonic acid ÆCytotoxicity Æ Endothelial cell Æ Heme oxygenase-1 ÆN-Acetyl-L-cysteine Æ Glutathione Æ Buthioninesulfoximine Æ Inductively coupled plasma massspectrometry

Introduction

Soil contamination with chemical warfare agents (CWA)has gained toxicological research interest because theyare highly toxic in nature, and many persons who useground water are at risk of intake of these chemicals.Organic arsenic sternutators such as Clark 1 (dipheny-larsine chloride) and Clark 2 (diphenylarsine cyanide)and a blister agent Lewisite (2-chloro-ethenyl dichlo-roarsine) were buried during wars and are contaminat-ing the ground water. Recently diphenylarsinic acid ofhigh concentration was found in ground water in Ka-misu, Ibaraki Prefecture, Japan (Okazaki et al. 2003). Itis speculated that diphenylarsinic acid was formed in soilby oxidation of buried Clark 1 or 2 and leaked into theground water. Clark 1 has been reported to degrade inthe soil and generate bis(diphenylarsine)oxide andtriphenylarsine (Pitten et al. 1999). Autochthonic bac-teria may play an important role in converting organ-oarsenic CWA into soluble inorganic arsenicals and areimplicated in bioremediation of military waste sites(Kohler et al. 2001). Autochthonic fungi have also beenshown to convert soil-contaminating hydrophobic or-ganoarsenic CWA-related compounds to soluble arsenicforms. They oxidize triphenylarsine to triphenylarsineoxide and phenylarsine oxide to phenylarsonic acid(Hofmann et al. 2001). CWA arsenicals are highly toxiccompared to inorganic pentavalent arsenic (As2O5), andconcentrations of Lewisite, Clark 2, Clark 1 and As2O5

that inhibited human leukocyte proliferation have been

S. Hirano (&) Æ Y. Kobayashi Æ T. Hayakawa Æ X. CuiM. Yamamoto Æ S. KannoEnvironmental Health Sciences Division,National Institute for Environmental Studies,16-2 Onogawa, Tsukuba, 305-8506 Ibaraki, JapanE-mail: seishiro@nies.go.jpTel.: +81-29-8502512Fax: +81-29-8502512

T. HayakawaFaculty of Pharmaceutical Sciences, Chiba University,Yayoi, Inage, 263-8522 Chiba, Japan

A. ShraimNational Research Centre for Environmental Toxicology,University of Queensland, 39 Kessels Road, Coopers Plains,4108 Brisbane, Queensland, Australia

Arch Toxicol (2005) 79: 54–61DOI 10.1007/s00204-004-0598-4

reported to be 0.3, 0.75, and 15 lg/ml and 1.7 mg/ml,respectively (Henriksson et al. 1996).

Drinking water in chronic arsenism endemic areas iscontaminated with arsenate (iAs5+) and less amount ofarsenite (iAs3+). However, recent reports show thatiAs3+ is occasionally more abundant than iAs5+, andlow concentrations of organic arsenicals exist in drink-ing water in West Bengal State of India and in Bangla-desh (Harvey et al. 2002; Shraim et al. 2002). It has beenshown that iAs3+ is more toxic than iAs5+ and penta-valent organic arsenicals are less toxic than inorganicarsenicals both in vitro and in vivo (Maitani et al. 1987;Styblo et al. 2000; Vega et al. 2001). Monomethylar-sonous acid (MMA3+) and dimethylarsinous acid(DMA3+) have been reported to be more toxic thaninorganic arsenicals (Petrick et al. 2001; Sakurai et al.2002). Trimethyarsine oxide (TMAO) is essentially non-toxic (Hirano et al. 2004). Over all the toxicity order ofinorganic arsenicals and their methylated metabolitesseems to be as follows: DMA3+, MMA3+ >iAs3+

>iAs5+ >DMA5+, MMA5 +>TMAO.However, the reason that toxicity of trivalent arsen-

icals is higher than pentavalent arsenicals has not beenwell documented. Higher toxicity of trivalent arsenicalscompared to the corresponding pentavalent forms canbe explained by higher affinity of trivalent arsenicals tothiol compounds (Shiobara et al. 2001) and the follow-ing generation of arsenic peroxides (Kato et al. 2003;Yamanaka et al. 2003).

It has been shown that chronic exposure to inorganicarsenicals causes arteriosclerosis, hypertension and othervascular diseases, in addition to cancers (Engel et al.1994; Lewis et al. 1999). Oxidative stress is one of themost probable mechanisms whereby arsenic causes in-jury to vascular endothelial cells (Barchowsky et al.1999; Hirano et al. 2003; Hirano et al. 2004). In thepresent study we exposed rat vascular endothelial cells tophenylarsine oxide (PAO3+) and phenylarsonic acid(PAA5+), CWA-related arsenicals, to investigate thedifference in toxicity and accumulation of arsenic be-tween trivalent and pentavalent arsenicals. We reportthat the toxicity and cellular association of PAO3+ werehigher than those of any other arsenicals ever studied,and that the cytotoxicity level of PAA5+ was ca. 6500-fold less than that of PAO3+ in vascular endothelialcells.

Materials and methods

Preparation of PAA5+

PAO3+ with purity higher than 98% was purchasedfrom Sigma (St. Louis, Mo., USA). PAA5+ was syn-thesized by oxidizing PAO3+ with H2O2. Briefly, 50 mgPAO3+ was dissolved in 25 ml 10% H2O2 and thesolution was left at 25�C for 1 h and then boiled for5 min. After cooling to the room temperature, thesolution was saturated with sodium chloride and

PAA5+ was extracted twice with 10 ml ethyl acetate.The organic extract was dried over sodium sulfate an-hydrate. Ethyl acetate was evaporated to dryness, andthe remaining white reaction product was recrystalizedin ethyl acetate with trace amount of ethanol. The pre-cipitate, PAA5+, was kept in a vacuum desiccator. Theproduct identity was confirmed by fast atom bombard-ment mass spectrometry (FAB-MS, Mstation, Jeol,Tokyo) and high performance liquid chromatography–inductively coupled plasma mass spectrometry (HPLC-ICPMS; HP4500 plus, Yokogawa Analytical Systems,Tokyo). In FAB-MS analysis [M+1] was observed.HPLC-ICPMS analysis showed that PAA5+ was elutedas a single peak (Fig. 1B). The commercially availablePAO3+ was also examined by HPLC-ICPMS and elec-tron impact mass spectroscopy (EI-MS). PAO3+ waseluted later than PAA5+ in an HPLC chromatogram(Fig. 1A) and there seems to be a trimer of PAO3+ [3M,m/z 504] in the sample as examined by EI-MS. Theelution condition of HPLC was the same as that forspeciation of methylated arsenicals (Shraim et al. 2002).The stock solutions of PAO3+ and PAA5+ were pre-pared in dimethyl sulfoxide and basal culture medium,respectively. PAO3+ was added to the culture system sothat the final concentration of dimethyl sulfoxide be-came 0.1%.

Fig. 1 HPLC-ICPMS analyses of phenylarsine oxide (PAO3+; A)and phenylarsonic acid (PAA5+; B) on a reverse-phase column.The arsenic concentration of the sample was adjusted at 10 ppb,and the integral mass current was measured with the time analysismode of ICPMS. The measurement condition was as follows:Column: ODS-3, 150·4.6 mm, particle size 3 mm (GL Science,Tokyo); Column temperature: 50�C. Mobile phase: 3 mM malonicacid solution (pH 5.6) containing 5 mM tetrabutylammoniumhydroxide and 5% methanol. Flow rate: 1.0 ml/min. Injectionvolume: 20 ll

55

Cells

Rat heart microvessel endothelial (RHMVE) cells werepurchased from VEC Tech (fewer than six passages;Rensselaer, N.Y., USA). The cells were subcultured inrat endothelial cell growth medium (Cell Applications,San Diego, Calif., USA) on fibronectin-coated (CellApplications, bovine plasma, 10 lg/ml) culture dishesand exposed to PAO3+ or PAA5+ in growth supple-ment-free endothelial basal medium at the 6th–10thpassages after arrival.

Cytotoxicity assay

The cells were detached by trypsin/EDTA, resuspendedin the growth medium, and cultured in a fibronectin-coated 96-well culture dish (Costar, Cambridge, Mass.,USA). After 3 days of culture the conditioned mediumwas replaced with fresh basal medium containing vari-ous concentrations of PAO3+ and PAA5+, and the cellswere further cultured for another 24 h. The cell viabilitywas measured using a modified 3-(4,5-dimethylthiazole-2-yl)-2,5-diphenyl tetrazolium bromide assay kitaccording to the manufacturer’s instruction (WST-8,Dojindo, Kumamoto, Japan). Briefly, the cell monolayerof quadruplicated wells was washed three times withphenol red-free minimum essential medium, WST-8solution was added to each well, and the dish wasincubated at 37�C for 1 h. The reaction was stopped byadding hydrochloric acid to a final concentration of0.01 mol/l and the OD at 450 nm was measured with areference of 650 nm using a microtitre plate reader(CS9300, Shimadzu, Kyoto, Japan).

Measurement of arsenic concentrations

The cells were grown to monolayer in 12-well culturedishes and exposed to PAO3+ or PAA5+ in the presenceor absence of 10 mM NAC or 5 mM GSH. At pre-determined time points the cell monolayer was washedthree times with phosphate-buffered saline. Nitric acid(0.6 ml) and H2O2 (0.2 ml) were added to each well andthe dishes were incubated overnight at room tempera-ture. The samples were transferred to acid-washed testtubes and digested at 130�C for 2 days in an aluminumblock bath. The samples were diluted with deionizedwater and total arsenic concentrations were measured byICPMS.

Northern analyses of heme oygenase-1 (HO-1) mRNA

The cell monolayer was exposed to 0.05, 0.1, and 0.2 lMPAO5+ or 0.2, 0.4, and 0.8 mM PAA5+ for 6 h in thepresence or absence of 10 mM NAC. Total RNA wasextracted from control and arsenic-exposed cells usingTRIZOL (Gibco BRL, Rockville, Md., USA). TheRNA was electrophoresed on a formaldehyde-denaturedagarose gel (1%), transferred onto a nylon membrane(Hybond-N+, Amersham Pharmacia Biotech, LittleChalfont, Buckinghamshire, UK) and fixed to themembrane by UV irradiation. The probes were labelledwith [a-32P]deoxycytidine triphosphate using RediprimeDNA labelling system (Amersham Pharmacia Biotech).The blot was prehybridized in ExpressHyb (Clontech,Palo Alto, Calif., USA) at 65�C for 90 min, andhybridized with the 32P-labelled cDNA probe for thedetection of HO-1 mRNA at 65�C for 2h. The radio-activity on the membrane was analysed and quantifiedusing a bioimage analyser (BAS2000, Fuji, Tokyo).After the detection and quantification of HO-1 mRNAthe probe was stripped off the membrane in 0.5% so-dium dodecyl sulfate solution at 90�C for 10 min and theblot was rehybridized with 32P-labelled b-actin cDNAprobe to normalize mRNA level of HO-1 against that ofthe housekeeping gene. The method for preparation ofcDNA probes was reported previously (Kitajima et al.1999).

Fig. 2 Dose-dependent viability changes in PAO3+- or PAA5+-exposed rat heart microvessel endothelial (RHMVE) cells. The cellmonolayer was exposed to PAO3+ at concentrations of 0 (control),0.01, 0.03, 0.1, 0.2, 0.3, 0.4, 1, 3, and 10 lM (A) or PAA5+ atconcentrations of 0 (control), 0.2, 0.4, 0.6, 0.8, 1, 2, 3, 5, and10 mM for 24 h (B). Data are shown as percentages of the controlvalue. Each value represents the mean ±SE of quadruplicate wells.LC50 values for PAO3+ and PAA5+ were calculated to be0.295 lM and 1.93 mM, respectively

56

Western blotting for the detection of tyrosine-phos-phorylated proteins

The cells were cultured to confluence in a 6-well culturedish and were exposed to 0.1–0.8 lM PAO3+ or 0.2–2 mM PAA5+ for 1 h. The cells were lysed usingphosphatase inhibitors-containing lysis buffer; 150 mMsodium chloride, 1% Nonidet P-40, 0.5% sodium de-oxycholate, 50 mM N-(2-hydroxyethyl)piperazine-N-(4-butanesulfonic acid), 1 mM phenylmethylsulfonyl fluo-ride, 1 mM sodium orthovanadate, 50 mM sodiumfluoride, 1 mM p-nitrophenyl phosphate, 10 lg/ml ap-rotinin and 5 mM benzamidine. The lysate was centri-fuged at 1,000 g, for 10 min at 4�C, and the proteinconcentration of the supernatant was adjusted at1.5 mg/ml. The soluble proteins were resolved on 4-20%sodium dodecyl sulfate polyacrylamide gel electropho-resis and transferred electrophoretically to a nitrocellu-lose membrane. The membrane was blocked in BlockAce (Dainippon, Osaka, Japan) and probed with anti-

phosphotyrosine (HRP-conjugated 4G10, Upstate, LakePlacid, N.Y., USA). Tyrosine-phosphorylated proteinson the membrane were visualized using electrochemilu-minescence and a Hyperfilm (Amersham PharmaciaBiotech).

Data analysis

The viability values were shown as the mean ±SE offour determinations, and LC50 values were calculated byfitting a sigmoid curve using Graph Pad PRIZM (GraphPad Software, San Diego, Calif., USA). The measure-ment of arsenic concentrations was performed for trip-licate wells and each value represents the mean ±SE.

Results

Figure 2 shows changes in the viability of RHMVE cellsafter 24-h exposure to PAO3+ (Fig. 1A) and PAA5+

(Fig. 1B). The cytotoxicity of PAO3+ was much higherthan that of PAA5+ and LC50 values for PAO3+ andPAA5+ were calculated to be 0.295 lM and 1.93 mM,respectively. The cytotoxicity of PAO3+ and PAA5+

was enhanced by treatment with 0.5 mM BSO (Fig. 3),suggesting that intracellular GSH plays an importantrole in detoxification of those arsenicals.

Figure 4a shows that NAC at a concentration of 10mM reduced cytotoxicity of PAO3+, whereas NAC wasnot effective for detoxification of PAA5+ (Fig. 4B).Sublethal concentrations of PAO3+ (less than 0.2 lM)

Fig. 3 Effects of BSO on cytotoxicity of PAO3+ (A) and PAA5+

(B) in RHMVEC. The cells were exposed to 0.01, 0.1, and 1 lMPAO3+ or 0.2, 0.4, and 0.8 mM PAA5+ for 24h in the presence orabsence of 0.5 mM BSO. Open column without BSO; hatchedcolumn with 0.5 mM BSO

Fig. 4 Effects of NAC on cytotoxicity of PAO3+ (A) and PAA5+

(B). The cells were exposed to 0.01, 0.1, and 1 lM PAO3+ or 0.2,0.4, and 0.8 mM PAA5+ for 24 h in the presence or absence of 5 or10 mM NAC. Open column without NAC; hatched column with5 mM NAC; closed column with 10 mM NAC

57

and PAA5+ (0.8 mM) increased the HO-1 mRNA levelafter 6-h exposure. The HO-1 mRNA level in PAO3+-exposed cells was decreased by 10 mM NAC, while thetranscription of HO-1 was not reduced by NAC inPAA5+-exposed cells (Fig. 5).

Figure 6 indicates that the accumulation of arsenic inPAO3+-exposed cells appeared to be distinct from thatin PAA5+-exposed cells. The large difference in cyto-toxicity between PAO3+ and PAA5+ did not enable usto compare the cellular arsenic accumulation at the samearsenic concentration. However, it is clear that the cellstook up about a half amount of PAO3+ in the culturemedium in 30 min, while less than 0.01% of arsenic inthe culture medium was taken up in PAA5+-exposedcells even after 6 h of culture. It is of interest to note thatboth NAC and GSH reduced the cellular association ofPAO3+, while neither of these thiol compounds changedthe cellular association of PAA5+ in RHMVE cells(Fig. 7).

We exposed the cells to various concentrations ofPAO3+ and PAA5+ for 1 h to investigate tyrosinephosphorylation levels of cellular proteins by westernblotting. Figure 8 shows that the tyrosine phosphoryla-tion level of 74-kDa protein was increased, and theelectromobility of 62-kDa protein was retarded inPAO3+-exposed cells at concentrations higher than 0.2–

Fig. 5 Northern analyses of dose-related changes in HO-1 mRNAlevels in PAO3+- and PAA5+-exposed RHMVE cells. Thebioimage photographs (A) and the quantitative results of northernblots (B) are shown. The cells were cultured in the presence of 0.05-0.2 lM PAO3+ or 0.2-0.8 mM PAA5+ with (closed column) orwithout 10 mM NAC (hatched column) for 6h; open column controlvalue. Data are presented as the mean values of two differentexperiments

Fig. 6 Time-course changes in the accumulation of arsenic inPAO3+- and PAA5+-exposed RHMVE cells. The cell monolayerscultured in a 12-well culture dish were exposed to 0.1 lM PAO3+

(A) or 0.2 mM PAA5+ (B) for up to 6h. The cell monolayers werewashed three times with phosphate-buffered saline and digested inHNO3-H2O2 as described in the text. The ratio of cellular to total(cells plus medium) arsenic is shown. Data are the presented as themeans ±SE of three different wells

58

0.4 lM. However, PAA5+ changed neither the tyrosinephosphorylation nor the electromobility even at the le-thal concentrations, suggesting that PAO3+ may exertits cytotoxic effects by inhibiting protein phosphatases,whereas PAA5+ injured the cell without changing thetyrosine phosphorylation levels.

Discussion

Inorganic arsenicals are methylated in the metabolicprocess and excreted as MMA5+ and DMA5+ mainly inurine in mammals, although some animal species lackarsenic methyltransferase and excrete inorganic arseni-cals (Aposhian 1997; Wildfang et al. 2001). Recently,methylated trivalent arsenicals, such as MMA3+ andDMA3+, have been reported to be more cytotoxic andgenotoxic than inorganic arsenicals (Mass et al. 2001;

Nesnow et al. 2002; Petrick et al. 2001; Sakurai et al.2002). However, it has not clearly shown why trivalentarsenicals are more toxic than pentavalent ones. Inaddition, it is not clear why methylated trivalent arsen-icals are more toxic, whereas methylated pentavalentarsenicals are less toxic than inorganic arsenicals. Wehave further confirmed that the large difference in tox-icity between trivalent and pentavalent organic arseni-cals using phenyl arsenicals (Fig. 2). It should be notedthat the cell viability decreased sharply from 0.2 lM (nochange) to 0.5 lM (total viability loss) in PAO3+-ex-posed cells. The LC50 values of arsenic compounds inRHMVE cells are summarized in Table 1. PAO3+ is themost cytotoxic of all those arsenic compounds investi-gated in our laboratory. The high toxicity of PAO3+ isprobably due to the rapid association of this compoundwith the cells as shown in Fig. 6 and Table 1, and astrong affinity to thiol groups as described below.

The intracellular GSH level is very important forsurvival in phenyl arsenical-exposed cells, because thetreatment of the cells with BSO, a GSH depleter, effec-tively enhanced the toxicity of both PAO3+ and PAA5+

(Fig. 3). Extracellularly added NAC was very effectiveto increase the survival of PAO3+-exposed cells, while

Fig. 7 Effects of NAC and GSH on the accumulation of arsenic inPAO3+- and PAA5+-exposed RHMVE cells. The cell monolayerwas exposed to 0.1 lM PAO3+ (A) or 0.2 mM PAA5+ (B) in thepresence or absence of 10 mM NAC or 5 mM GSH for 1 h. Theratio of cellular to total (cells plus medium) arsenic is shown. Dataare presented as the means ±SE of three different wells

Fig. 8 Changes in proteintyrosine phosphorylation levelsin RHMVE cells. The cells wereexposed to 0.1–0.8 lM PAO3+

or 0.2–2 mM PAA5+ for 1 hand were lysed usingphosphatase inhibitors-containing lysis buffer. Thesoluble proteins were resolvedon 4–20% sodium dodecylsulfate polyacrylamide gelelectrophoresis and transferredelectrophoretically to anitrocellulose membrane. Themembrane was probed withHRP-conjugated anti-phosphotyrosine (4G10)

59

NAC did not increase the viability of PAA5+-exposedcells as shown in Fig. 4. The similar observation wasfound in the induction of HO-1 mRNA as shown inFig. 5. The induction of HO-1 mRNA was diminished inPAO3+-exposed cells, whereas the induction of HO-1mRNA was not changed in PAA3+-exposed cells byNAC. NAC has been used to increase cellular GSH le-vel. However, we have recently reported that the addi-tion of NAC to the culture medium did not increase, ifany, the intracellular GSH concentration in RHMVEcells (Hirano et al. 2004). Changes in intracellular GSHlevel by extracellularly added NAC or GSH appear todepend on the cell types. For example, A549 humanepithelial cells take up a significant amount of neitherNAC nor GSH, while mouse glial cells are able to takeup extracellular GSH from the medium (Riganti et al.2003). It seems that the protective effect of NAC is dueto the inhibition of cellular association of PAO3+ ratherthan antioxidative effect of NAC in the present study,because the cellular arsenic content was greatly reducedby both NAC and GSH (Fig. 7). We also reported thatalthough NAC did not affect the cellular association ofiAs3+, it effectively reduced the cytotoxicity of iAs3+ inRHMVE cells (Hirano et al. 2003). Thus we do notexclude the possibility that NAC reduced extracellularlygenerated oxidative effects caused by arsenic com-pounds.

It has also shown that PAO3+ has a high bindingaffinity to enzymes such as galectin I, thioredoxin per-oxidase II, and GST-P (Chang et al. 2003) and inhibitstyrosine phosphatase in endothelial cells (Young et al.2003). It is plausible that PAO3+ binds to thiol-richenzymes rapidly, inhibits their activity and exerts itscytotoxic effects. In the present study PAO3+ changedtyrosine phosphorylation levels, while PAA5+ did notchange the tyrosine phosphorylation patterns even at thecytotoxic concentrations (Fig. 8). The inhibition ofphosphatase activities is one of the mechanisms wherebyPAO3+ exerts its cytotoxic effects at lower concentra-tions than the other arsenicals.

In summary, the cytotoxicity of PAO3+ was muchhigher than that of PAA5+. The difference in the cyto-toxicity between PAO3+ and PAA5+ was most likelydue to the difference in the cellular association betweenthese phenyl arsenicals. The accumulation and toxicityof PAO3+ in RHMVE cells were reduced while those ofPAA5+ were not changed by NAC. The toxicity of bothPAO3+ and PAA5+ was enhanced by the depletion of

intracellular GSH. The higher toxicity of trivalent ar-senicals than pentavalent ones can be explained by fasterassociation of trivalent arsenicals with the cells.

Acknowledgements The authors thank Ms. Kyoko Takata and Ms.Kimiyo Nagano of NIES for ICPMS- and FAB- and EI-MSmeasurements. This study was partially supported by Grant-in-Aidfor Scientific Research (14390058) from the Japan Society of Pro-motion of Science.

References

Aposhian HV (1997) Enzymatic methylation of arsenic species andother new approaches to arsenic toxicity. Annu Rev PharmacolToxicol 37397–37419

Barchowsky A, Roussel RR, Klei LR, James PE, Ganju N, SmithKR, Dudek EJ (1999) Low levels of arsenic trioxide stimulateproliferative signals in primary vascular cells without activatingstress effector pathways. Toxicol Appl Pharmacol 159:65–75

Chang KN, Lee TC, Tam MF, Chen YC, Lee LW, Lee SY, Lin PJ,Huang RN (2003) Identification of galectin I and thioredoxinperoxidase II as two arsenic-binding proteins in Chinese ham-ster ovary cells. Biochem J 371:495–503

Engel RR, Hopenhayn Rich C, Receveur O, Smith AH (1994)Vascular effects of chronic arsenic exposure: a review. Epi-demiol Rev 16:184–209

Harvey CF, Swartz CH, Badruzzaman ABM, Keon-Blute N, YuW, Ali MA, Jay J, Beckie R, Niedan V, Brabander D, OatesPM, Ashfaque KN, Islam S, Hemond HF, Ahmed MF (2002)Arsenic mobility and groundwater extraction in Bangladesh.Science 298:1602–1606

Henriksson J, Johannisson A, Bergqvist PA, Norrgren L (1996)The toxicity of organoarsenic-based warfare agents: In vitroand in vivo studies. Arch Environ Contam Toxicol 30:213–219

Hirano S, Cui X, Li S, Kanno S, Hayakawa T, Shraim A (2003)Difference in uptake and toxicity of trivalent and pentavalentInorganic arsenics in rat microvessel endothelial cells. ArchToxicol 77:305–312

Hirano S, Kobayashi Y, Cui X, Kanno S, Hayakawa T, Shraim A(2004) The accumulation and toxicity of methylated arsenicalsin endothelial cells: important role of thiol compounds. ToxicolAppl Pharmacol 198:458–467

Hofmann K, Hammer E, Kohler M, Bruser V (2001) Oxidation oftriphenylarsine to triphenylarsineoxide by Trichoderma har-zianum and other fungi. Chemosphere 44:697–700

Kato K, Yamanaka K, Hasegawa A, Okada S (2003) Active ar-senic species produced by GSH-dependent reduction of dime-thylarsinic acid cause micronuclei formation in peripheralreticulocytes of mice. Mutat Res 539:55–63

Kitajima H, Hirano S, Suzuki KT (1999) Upregulation of hemeoxygenase gene expression in rat lung epithelial cells followingexposure to cadmium. Arch Toxicol 73:410–412

Kohler M, Hofmann K, Volsgen F, Thurow K, Koch A (2001)Bacterial release of arsenic ions and organoarsenic compoundsfrom soil contaminated by chemical warfare agents. Chemo-sphere 42:425–429

Table 1 Summary of cellular

association and cytotoxicity ofarsenicals in RHMVE cells

Conc. of As inculture medium (lM)

Cellular As/totalAs at 1 h (%)

LC50 at 24 h(lM)

Reference

iAs3+ 10 0.129±0.0024 36 Hirano et al. 2003iAs5+ 50 0.017±0.00094 220 Hirano et al. 2003MMA5+ 500 0.0048±0.00021 36,600 Hirano et al. 2004DMA5+ 500 0.0022±0.00015 2,540 Hirano et al. 2004MMA3+(GS)2 0.92 1.91±0.065 4.1 Hirano et al. 2004PAO3+ 0.1 46±2.1 0.295 This reportPAA5+ 200 0.0061±0.0011 1,930 This report

60

Lewis DR, Southwick JW, Ouellet-Hellstrom R, Rench J,Calderon RL (1999) Drinking water arsenic in Utah: A co-hort mortality study. Environ Health Perspect 107:359–365

Maitani T, Saito N, Abe M, Uchiyama S, Saito Y (1987) Chemicalform-dependent induction of hepatic zinc-thionein by arsenicadministration and effect of co-administered selenium in mice.Toxicol Lett 39:63–70

Mass MJ, Tennant A, Roop BC, Cullen WR, Styblo M, ThomasDJ, Kligerman AD (2001) Methylated trivalent arsenic spe-cies are genotoxic. Chem Res Toxicol 14:355–361

Nesnow S, Roop BC, Lambert G, Kadiiska M, Mason RP,Cullen WR, Mass MJ (2002) DNA damage induced bymethylated trivalent arsenicals is mediated by reactive oxygenspecies. Chem Res Toxicol 15:1627–1634

OkazakiM, SakaueM,Kunimoto S,MoritaM,KunimotoM (2003)Assessment of potential neurotoxic actions of organoarseniccompounds using human neuroblastoma NB-1 cells and ratcerebellar neurons in primary culture. J Health Sci 49:410–415

Petrick JS, Jagadish B, Mash EA, Aposhian HV (2001) Mo-nomethylarsonous acid (MMA (III)) and arsenite: LD (50) inhamsters and in vitro inhibition of pyruvate dehydrogenase.Chem Res Toxicol 14:651–656

Pitten FA, Muller G, Konig P, Schmidt D, Thurow K, KramerA (1999) Risk assessment of a former military base con-taminated with organoarsenic-based warfare agents: uptakeof arsenic by terrestrial plants. Sci Total Environ 226:237–245

Riganti C, Aldieri E, Bergandi L, Tomatis M, Fenoglio I,Costamagna C, Fubini B, Bosia A, Ghigo D (2003) Longand short fiber amosite asbestos alters at a different extentthe redox metabolism in human lung epithelial cells. ToxicolAppl Pharmacol 193:106–115

Sakurai T, Qu W, Sakurai MH, Waalkes MP (2002) A majorhuman arsenic metabolite, dimethylarsinic acid, requires re-

duced glutathione to induce apoptosis. Chem Res Toxicol15:629–637

Shiobara Y, Ogra Y, Suzuki KT (2001) Animal species difference inthe uptake of dimethylarsinous acid (DMA (III)) by red bloodcells. Chem Res Toxicol 14:1446–1452

Shraim A, Sekaran NC, Anuradha CD, Hirano S (2002) Speciationof arsenic in tube-well water samples collected from WestBengal, India, by high-performance liquid chromatography-inductively coupled plasma mass spectrometry. Appl Organo-met Chem 16:202–209

Styblo M, Del Razo LM, Vega L, Germolec DR, LeCluyse EL,Hamilton GA, Reed W, Wang C, Cullen WR, Thomas DJ(2000) Comparative toxicity of trivalent and pentavalent inor-ganic and methylated arsenicals in rat and human cells. ArchToxicol 74:289–299

Vega L, Styblo M, Patterson R, Cullen W, Wang C, Germolec D(2001) Differential effects of trivalent and pentavalent arsenicalson cell proliferation and cytokine secretion in normal humanepidermal keratinocytes. Toxicol Appl Pharmacol 172:225–232

Wildfang E, Radabaugh TR, Aposhian HV (2001) Enzymaticmethylation of arsenic compounds. IX. Liver arsenite methyl-transferase and arsenate reductase activities in primates. Toxi-cology 168:213–221

Yamanaka K, Mizoi M, Tachikawa M, Hasegawa A, Hoshino M,Okada S (2003) Oxidative DNA damage following exposure todimethylarsinous iodide: the formation of cis-thymine glycol.Toxicol Lett 143:145–153

Young BA, Sui X, Kiser TC, Hyun SW, Wang P, Sakarya S,Angelini DJ, Schaphorst KL, Hasday JD, Cross AS, RomerLH, Passaniti A, Goldblum SE (2003) Protein tyrosine phos-phatase activity regulates endothelial cell-cell interactions, theparacellular pathway, and capillary tube stability. Am J PhysiolLung Cell Mol Physiol 285:L63–L75

61