The effect of divalent nickel (Ni2+) on in vitro DNA replication by DNA polymerase alpha.

The effects of the carcinogenic metal nickel on DNA polymerase alpha (pol alpha) activity and fidelity have been analyzed. In the absence of Mg2+, the presence of Ni2+ ions at concentrations below 0.25 mM gave rise to a dose-dependent activation of pol alpha as monitored by [3H]dTMP incorporation into an activated DNA template. The apparent Km for Ni(2+)-dependent pol alpha incorporation of dTTP was estimated to be 25 microM, which was about 10 times higher than the Km for Mg2+ (2.3 microM). Above 0.25 mM, Ni2+ caused a dose-dependent inhibition of pol alpha activity and the Ki was calculated to be 1.5 mM. Scatchard analyses showed that Ni2+ binds to affinity-purified pol alpha and associated proteins at two tight binding sites with a Kd of approximately 50 microM and at eight weak binding sites with a Kd of approximately 4 mM. In the presence of 2 mM Mg2+, the addition of Ni2+ to the reactions caused an inhibition of polymerase activity. The inhibition patterns tended to switch from competitive to mixed-type to noncompetitive as a function of Ni2+ concentration. Lastly, Ni2+ increased the incorporation of the modified nucleotide dideoxy-CMP in reactions using varying ratios of dideoxy-CTP/dCTP.

Metal carcinogenesis: mechanistic implications.

Cancer epidemiology has identified several metal compounds as human carcinogens. Recent evidence suggests that carcinogenic metals induce genotoxicity in a multiplicity of ways, either alone or by enhancing the effects of other agents. This review summarizes current information on the genotoxicity of arsenic, chromium, nickel, beryllium and cadmium compounds and their possible roles in carcinogenesis. Each of these metals is distinct in its primary modes of action; yet there are several mechanisms induced by more than one metal, including: the induction of cellular immunity and oxidative stress, the inhibition of DNA metabolism and repair and the formation of DNA- and/or protein-crosslinks.

Effects of chromium on DNA replication in vitro.

Chromium is an environmentally significant human carcinogen with complicated metabolism and an unknown mechanism of mutagenesis. Chromium(VI) is taken up by cells as the chromate anion and is reduced intracellularly via reactive intermediates to stable Cr(III) species. Chromium(III) forms tight complexes with biological ligands, such as DNA and proteins, which are slow to exchange. In vitro, CrCl3.6H2O primarily interacts with DNA to form outer shell charge complexes with the DNA phosphates. However, at micromolar concentrations, the Cr(III) binds to a low number of saturable tight binding sites on single-stranded M13 DNA. Additional chromium interacts in a nonspecific manner with the DNA and can form intermolecular DNA cross-links. Although high concentrations of Cr(III) inhibit DNA replication, micromolar concentrations of Cr(III) can substitute for Mg2+, weakly activate the Klenow fragment of E. coli DNA polymerase I (Pol I-KF), and act as an enhancer of nucleotide incorporation. Alterations in enzyme kinetics induced by Cr(III) increase DNA polymerase processivity and the rate of polymerase bypass of DNA lesions. This results in an increased rate of spontaneous mutagenesis during DNA replication both in vitro and in vivo. Our results indicate that chromium(III) may contribute to chromate-induced mutagenesis and may be a factor in the initiation of chromium carcinogenesis.

A possible role for chromium(III) in genotoxicity.

Chromium is found in the environment in two major forms: reduced CrIII and CrVI, or chromate. Chromate, the most biologically active species, is readily taken up by living cells and reduced intracellularly, via reactive intermediates, to stable CrIII species. CrIII, the most abundant form of chromium in the environment, does not readily cross cell membranes and is relatively inactive in vivo. However, intracellular CrIII can react slowly with both nucleic acids and proteins and can be genotoxic. We have investigated the genotoxicity of CrIII in vitro using a DNA replication assay and in vivo by CaCl2-mediated transfection of chromium-treated DNA into Escherichia coli. When DNA replication was measured on a CrIII-treated template using purified DNA polymerases (either bacterial or mammalian), both the rate of DNA replication and the amount of incorporation per polymerase binding event (processivity) were greatly increased relative to controls. When transfected into E. coli, CrIII-treated M13mp2 bacteriophage DNA showed a dose-dependent increase in mutation frequency. These results suggest that CrIII alters the interaction between the DNA template and the polymerase such that the binding strength of the DNA polymerase is increased and the fidelity of DNA replication is decreased. These interactions may contribute to the mutagenicity of chromium ions in vivo and suggest that CrIII can contribute to chromium-mediated carcinogenesis.

Metal mutagenesis in transgenic Chinese hamster cell lines.

Metals are toxic agents for which genotoxic effects are often difficult to demonstrate. To study metal mutagenesis, we have used two stable hprt/gpt+ transgenic cell lines that were derived from Chinese hamster V79 cells. Both the G12 and G10 cell lines are known to be very sensitive to clastogens such as X-rays and bleomycin, with the mutagenic response of the integrated xanthine guanine phosphoribosyl transferase (gpt) gene in G10 usually exceeding that of the same gene in the transgenic G12 cells. In studies with carcinogenic insoluble nickel compounds, a high level of mutagenesis was found at the gpt locus of G12 cells but not at the endogenous hypoxanthine phosphoribosyl transferase (hprt) locus of V79 cells. We have since demonstrated the similar recovery of a high frequency of viable G12 mutants with other insoluble nickel salts including nickel oxides (black and green). The relative mutant yield for the insoluble nickel compounds (G12 > G10) is the opposite of that obtained with nonmetal clastogens (G10 > G12). In the G12 cells, nickel mutagenesis may be related to the integration of the gpt sequence into a heterochromatic region of the genome. For some of the insoluble nickel compounds, significant inhibition of both cytotoxicity and mutant yield resulted when the G12 cells were pretreated with vitamin E. In comparison with the nickel studies, the mutagenic responses to chromium compounds in these cell lines were not as dramatic. Mutagenesis of the gpt target could not be demonstrated with other metals such as mercury or vanadium.

Propylene oxide mutagenesis at template cytosine residues.

Propylene oxide (PO) is a widely used industrial reagent which is mutagenic and carcinogenic. We have recently shown that a variety of aliphatic epoxides, including propylene oxide, can react with DNA to form hydroxyalkyl adducts at N-3 of cytosine which rapidly undergo hydrolytic deamination to produce uracil adducts. These 3-hydroxyalkyl uracil adducts are stable in DNA and are postulated to be an important class of potentially mutagenic lesions. Mutagenesis at cytosine residues due to PO modification of single-stranded M13mp2/C141 DNA was studied by transfection of modified DNA into SOS and non-SOS induced E. coli host cells. Mutations of the proline (CCC) codon at C141 which result in reversion of the lacZ phenotype (blue plaques) were scored. It was found that PO treatment of single-stranded DNA results in dose-dependent mutagenesis that is highly SOS dependent. The spectrum of base-substitution mutations found at this site differed when PO-modified DNA was transfected into E. coli with different DNA repair backgrounds. These results indicate that propylene oxide induced DNA adducts at template cytosine residues are mutagenic in E. coli and that this mutagenesis is greatly increased by SOS processing. They also show that these lesions may be repaired by one or more mechanisms.

Characterization of gpt deletion mutations in transgenic Chinese hamster cell lines.

The transgenic cell lines G12 and G10, each with a bacterial gpt gene stably integrated at a single but different position in the Chinese hamster genome, were evaluated for deletion of the gpt transgene following exposures to several clastogens. More than 150 independently cloned G12 and G10 6-thioguanine-resistant mutants have been characterized by polymerase chain reaction (PCR) amplification and Southern blots in this study. Despite differences in the integration sites for the gpt gene in the G12 and G10 cells, PCR amplification of the gpt gene from both cell lines can be performed using the same single set of primers. By PCR deletion screening, about 20% of recovered spontaneous 6-thioguanine resistant (6TG) gpt G12 mutants had deleted the transgene, whereas the deletion mutant frequency was increased to about 50% of the X-ray- and bleomycin-induced G12 mutants. In contrast, both spontaneous and induced deletion frequencies are considerably higher for the G10 cell line. Among spontaneous G10 mutants, up to 50% have deleted the gpt transgene, whereas almost all of the X-ray- and bleomycin-induced G10 mutants have lost the integrated gene sequence. These results are discussed in the context of the transgene integration sites and the influences of the surrounding genome that may render certain genetic regions prone to deletion.

Effects of nickel ions on polymerase activity and fidelity during DNA replication in vitro.

Nickel is a genotoxic carcinogen. However, the mechanisms of nickel-induced genotoxicity are not well understood. We have investigated the effects of Ni2+ ions on DNA polymerase activity and the fidelity of DNA replication in vitro. The effect of Ni2+ on different DNA polymerases is quite variable. The amount of enzyme inhibition and degree of alteration in replication fidelity induced by Ni2+ are dependent both on the polymerase and its associated 3'-5' exonuclease activity. Some polymerases, such as E. coli DNA polymerase I, AMV reverse transcriptase and human DNA polymerase alpha, can utilize Ni2+ as a weak substitute for Mg2+ during DNA replication. Other polymerases are very sensitive to inhibition by Ni2+ and the IC50 can vary by an order of magnitude. T4 polymerase is relatively insensitive to inhibition by Ni2+, although the sensitivity is enhanced in the absence of added Mg2+, and Ni preferentially inhibits the 3'-5' exonuclease function of T7 DNA polymerase. The fidelity and processivity of DNA polymerases may be either increased or decreased by Ni ions in a polymerase dependent manner. The inhibition DNA polymerase activity and altered replication fidelity may contribute significantly to Ni-induced mutagenesis and genotoxicity in vivo.

The role of DNA repair in development.

Several pathways of DNA repair are essential for maintaining genomic integrity in mammalian cells. Mismatch repair is the final line of defense against polymerase errors during normal cellular replication. Base excision repair removes endogenous DNA damage resulting from normal cellular metabolism. Nucleotide excision repair removes bulky, transcription blocking, lesions resulting from endogenous and environmental insults to the DNA. The role of DNA repair in mammalian development is not well understood. Nevertheless, clues to the essential nature of these processes are evident in the human DNA repair syndromes, in the nature of the interactions between DNA repair and other proteins, and in the phenotypes of genetically engineered, knockout mice lacking functional repair genes. Questions remain: what is the relative importance of endogenous vs. environmental DNA damage and is repair itself critical for normal development or are transcription-repair interactions more crucial?

Role of carcinogen-modified deoxynucleotide precursors in mutagenesis.

Agents which damage or modify cellular DNA will generally also modify the nucleotide precursor pools, sometimes preferentially (Topal and Baker, 1982). There are at least two different ways that incorporation of modified (possibly promutagenic) nucleotides could, theoretically, make a significant contribution to the mutations induced by these agents. Modified bases may exhibit ambiguous base pairing and produce mutations during normal replication or they may induce secondary mutations as a result of processing subsequent to incorporation. There are important precedents for such possibilities. Classical studies on mutagenesis with prototype mutagens like 2-aminopurine (2-AP) and 5-bromouracil clearly show that mutations can occur by incorporation of deoxynucleotides of tautomeric or ionized (Sowers et al., 1987) bases into newly synthesized DNA (Ronen, 1979; Lasken and Goodman, 1984, Coulondre and Miller, 1977). 5-Hydroxymethyl-2'-deoxyuridine (HMdU), a product of oxidative DNA damage, can also be (re)incorporated into cellular DNA with both toxic and mutagenic consequences (Kaufman, 1987; Shirname-More et al., 1987). Furthermore, modified nucleotides may alter the pool sizes of the normal nucleotides and indirectly produce toxic and mutagenic effects. However, these effects are generally seen at high, nonphysiological, concentrations of the modified precursors and may not be relevant under physiological conditions. The relative importance of modified deoxynucleotide precursors in the production of mutations by alkylating and oxidative DNA-damaging agents is discussed.

Arsenic toxicity is enzyme specific and its affects on ligation are not caused by the direct inhibition of DNA repair enzymes.

The molecular mechanism of arsenic toxicity is believed to be due to the ability of arsenite [As(III)] to bind protein thiols. Numerous studies have shown that arsenic is cytotoxic at micromolar concentrations. Micromolar As can also induce chromosomal damage and inhibit DNA repair. The mechanism of arsenic-induced genotoxicity is very important because arsenic is a human carcinogen, but not a mutagen, and there is a need to establish recommendations for safe levels of As in the environment. We have measured the dose-response for arsenic inhibition of several purified human DNA repair enzymes, including DNA polymerase beta, DNA ligase I and DNA ligase III and have found that most enzymes, even those with critical SH groups, are very insensitive to As. Many repair enzymes are activated by millimolar concentrations of As(III) and/or As(V). Only pyruvate dehydrogenase, one of eight purified enzymes examined so far, is inhibited by micromolar arsenic. In contrast to the purified enzymes, treatment of human cells in culture with micromolar arsenic produces a significant dose-dependent decrease in DNA ligase activity in nuclear extracts from the treated cells. However, the ligase activity in extracts from untreated cells is no more sensitive to arsenic than the purified enzymes. Our results show that direct enzyme inhibition is not a common toxic effect of As and that only a few sensitive enzymes are responsible for arsenic-induced cellular toxicity. Thus, arsenic-induced co-mutagenesis and inhibition of DNA repair is probably not the result of direct enzyme inhibition, but may be an indirect effect caused by As-induced changes in cellular redox levels or alterations in signal transduction pathways and consequent changes in gene expression.

An Escherichia coli plasmid-based, mutational system in which supF mutants are selectable: insertion elements dominate the spontaneous spectra.

A new system is described to determine the mutational spectra of mutagens and carcinogens in Escherichia coli; data on a limited number (142) of spontaneous mutants is presented. The mutational assay employs a method to select (rather than screen) for mutations in a supF target gene carried on a plasmid. The E. coli host cells (ES87) are lacI- (am26), and carry the lacZ delta M15 marker for alpha-complementation in beta-galactosidase. When these cells also carry a plasmid, such as pUB3, which contains a wild-type copy of supF and lacZ-alpha, the lactose operon is repressed (off). Furthermore, supF suppression of lacIam26 results in a lactose repressor that has an uninducible, lacIS genotype, which makes the cells unable to grow on lactose minimal plates. In contrast, spontaneous or mutagen-induced supF- mutations in pUB3 prevent suppression of lacIam26 and result in constitutive expression of the lactose operon, which permits growth on lactose minimal plates. The spontaneous mutation frequency in the supF gene is approximately 0.7 and approximately 1.0 x 10(-6) without and with SOS induction, respectively. Spontaneous mutations are dominated by large insertions (67% in SOS-uninduced and 56% in SOS-induced cells), and their frequency of appearance is largely unaffected by SOS induction. These are identified by DNA sequencing to be Insertion Elements; IS1 dominates, but IS4, IS5, gamma-delta and IS10 are also obtained. Large deletions also contribute significantly (19% and 15% for -SOS and +SOS, respectively), where a specific deletion between a 10 base pair direct repeat dominates; the frequency of appearance of these mutations also appears to be unaffected by SOS induction. In contrast, SOS induction increases base pairing mutations (13% and 27% for -SOS and +SOS, respectively). The ES87/pUB3 system has many advantages for determining mutational spectra, including the fact that mutant isolation is fast and simple, and the determination of mutational changes is rapid because of the small size of supF.

Base substitution mutagenesis by terminal transferase: its role in somatic mutagenesis.

We have addressed the possibility of terminal transferase involvement in somatic mutagenesis and the creation of N-region diversity, by measuring the ability of TdT to enhance single-base substitution mutagenesis during in vitro DNA synthesis. Using 3 independent assays we find that terminal transferase produces only a small increase in base-substitution mutagenesis when assayed in the presence of DNA polymerase-beta. In the presence of either polymerase-alpha or E. coli polymerase-I, however, no detectable increase in TdT-induced mutagenesis is seen. Furthermore, in an assay capable of detecting a variety of mutational events, terminal transferase primarily produces complex addition/deletion mutations, as well as a few multiple, tightly-clustered, single-base mutations. We conclude that the majority of the scattered single-base changes that occur during antibody gene differentiation are not catalyzed by terminal transferase, but instead result from another error-prone DNA synthetic process (possibly utilizing DNA polymerase-beta).

Transgenic gpt+ V79 cell lines differ in their mutagenic response to clastogens.

Several gpt+ transgenic cell lines were derived from hprt V79 cells to study mutagenesis mechanisms in mammalian cells. The G12 cell line was previously shown to be hypermutable by X-rays and UV at the gpt locus compared to the endogenous hprt gene of the parental V79 cells (Klein and Rossman, 1990), and is now shown to be highly mutable by the clastogenic anti-tumor agent bleomycin sulfate. A second transgenic cell line G10, which has a different gpt insertion site, was studied in comparison with G12. Both G12 and G10 cell lines carry the stable gpt locus at a single integration site in the Chinese hamster genome, and neither spontaneously deletes the integrated gpt sequence at a high frequency. Although spontaneous mutation to 6-thioguanine resistance in G10 cells is 3-4 times higher than in G12 cells, the cell lines differ to a much greater extent when mutated by clastogens. In comparison to G12 cells, the gpt locus in G10 cells is up to 13 times more sensitive to bleomycin mutagenesis and 5 times more responsive to X-ray mutagenesis. In contrast, there is much less difference in UV-induced mutagenesis and no differences in mutagenesis induced by alkylating agents such as N-methyl-N'-nitro-N-nitrosoguanidine (MNNG). The dose-dependent decrease in survival of the transgenic cells is the same for all mutagens tested, and does not differ from that of V79 cells. Neither transgenic cell line is generally hypermutable, since mutagenesis at an endogenous gene, Na+K+/ATPase, is similar to that of the parental V79 cell line. Although both cell lines can be induced to delete the transgene following clastogen exposure, deletions are not the only recovered mutations, and the cell lines can also be used to study mutations within the PCR recoverable gpt gene. The utility of these transgenic cells to investigate genome position effects related to mammalian mutagenesis mechanisms is discussed.

Arsenic and heavy metal contamination of vegetables grown in Samta village, Bangladesh.

Drinking of arsenic (As) contaminated well water has become a serious threat to the health of many millions in Bangladesh. However, the implications of contamination of agricultural soils from long-term irrigation with As-contaminated groundwater for phyto-accumulation in food crops, and thence dietary exposure to As, and other metals, has not been assessed previously in Bangladesh. Various vegetables were sampled in Samta village in the Jessore district of Bangladesh, and screened for As, Cd, Pb, Cu and Zn by inductively coupled plasma emission spectrometry (ICP-AES) and inductively coupled plasma mass spectrometry (ICP-MS). These local food products are the basis of human nutrition in this region and of great relevance to human health. The results revealed that the individual vegetables containing the highest mean As concentrations microg x g(-1)) are snake gourd (0.489), ghotkol (0.446), taro (0.440), green papaya (0.389), elephant foot (0.338) and Bottle ground leaf (0.306), respectively. The As concentration in fleshy vegetable material is low. In general, the data show the potential for some vegetables to accumulate heavy metals with concentrations of Pb greater than Cd. Some vegetables such as bottle ground leaf, ghotkol, taro, eddoe and elephant foot had much higher concentrations of Pb. Other leafy and root vegetables, contained higher concentrations of Zn and Cu. Bioconcentration factors (BCF) values, based on dry weight, were below 1 for all metals. In most cases, BCF values decreased with increasing metal concentrations in the soil. From the heavily As-contaminated village in Samta, BCF values for As in ladies finger, potato, ash gourd, brinjal, green papaya, ghotkol and snake gourd were 0.001, 0.006, 0.006, 0.014, 0.030, 0.034 and 0.038, respectively. Considering the average daily intake of fresh vegetables per person per day is only 130 g, all the vegetables grown at Samta had Pb concentrations that would be a health hazard for human consumption. Although the total As in the vegetables was less than the recommended maximum intake of As, it still provides a significant additional source of As in the diet.

Effects of chromium(III) on DNA replication in vitro.

A number of metal compounds are important environmental carcinogens; however, the molecular mechanisms of metal-induced genotoxicity are not yet understood. Chromium, for example, is substantially mutagenic in vivo and has been shown to decrease the DNA replication fidelity in vitro. But the mechanism of chromium-induced mutagenesis is unknown and the role of replication fidelity in chromium-induced carcinogenesis is unclear. We have used in vitro DNA replication assays to investigate the effects of chromium ions on DNA polymerase activity preliminary to studying their role in chromium-induced mutagenesis. Biologically active M13mp2 DNA was replicated with purified DNA polymerases in the presence of micromolar amounts of chromium with or without the normal divalent cation, magnesium. Nucleotide incorporation kinetics were determined and sequence specific pausing was analyzed by primer-extension. Our results have demonstrated an unexpected polymerase activation by low (0.5-5.0 microns) concentrations of chromium (III), although higher concentrations of chromium are increasingly inhibitory. The increased incorporation seem at low chromium(III) concentrations is the result of increased enzyme processivity and is not polymerase specific. The possible relationship between processivity and metal-ion mutagenesis is discussed.

Chromium(III) bound to DNA templates promotes increased polymerase processivity and decreased fidelity during replication in vitro.

Carcinogenic chromium [Cr(VI)] compounds are reduced intracellularly to DNA- and protein-reactive chromium(III) species. However, the role of Cr(III) ions in chromium-induced genotoxicity remains unclear. We have investigated the effects of chromium(III) binding on DNA replication and polymerase processivity in vitro. Chromium ions bind slowly and in a dose-dependent manner to DNA. Micromolar concentrations of free chromium inhibit DNA replication, but if the unbound chromium is removed by gel filtration, the rate of DNA replication by polymerase I (Klenow fragment) on the chromium-bound template is increased greater than 6-fold relative to the control. This increase is paralleled by as much as a 4-fold increase in processivity and a 2-fold decrease in replication fidelity. These effects are optimum when very low concentrations of chromium ions are bound to the DNA [3-4 Cr(III) ions per 1000 nucleotide phosphates]. Increased concentrations of chromium lead to the production of DNA-DNA cross-links and inhibition of polymerase activity. These results suggest that low levels of DNA-bound chromium(III) ions may contribute to chromium mutagenesis and carcinogenesis by altering the kinetics and fidelity of DNA replication.

Chromium(III) decreases the fidelity of human DNA polymerase beta.

Certain particulate compounds of hexavalent chromium are well-known occupational and environmental human carcinogens. Hexavalent chromium primarily enters the cells and undergoes metabolic reduction; however, the ultimate trivalent oxidation state of chromium, Cr(III), predominates within the cell. DNA-bound Cr(III) has been previously shown to decrease the fidelity of replication in the M13 phage mutation assay. This study was done to understand how Cr(III), in the presence of physiological concentrations of magnesium, affects the kinetic parameters of steady-state DNA synthesis in vitro across site-specific O6-methylguanine (m6dG) residues by DNA polymerase beta (pol beta). Cr(III) binds to the short oligomer templates in a dose-dependent manner and stimulates the activity of pol beta. Cr(III) stimulates the mutagenic incorporation of dTTP opposite m6dG more than the nonmutagenic incorporation of dCTP, and thereby Cr(III) further decreases the fidelity of DNA synthesis across m6dG by pol beta. In contrast, Cr(III) does not affect the fidelity of DNA synthesis across the normal template base, dG. Both the enhanced activity and the mutagenic lesion bypass in the presence of Cr(III) may be associated with Cr(III)-dependent stimulation of pol beta binding to DNA as reported here. This study shows some of the mechanisms by which mutagenic chromium affects DNA synthesis.

In vitro effect of arsenical compounds on glutathione-related enzymes.

The mechanism of arsenic toxicity is believed to be due to the ability of arsenite (As(III)) to bind protein thiols. Glutathione (GSH) is the most abundant cellular thiol, and both GSH and GSH-related enzymes are important antioxidants that play an important role in the detoxification of arsenic and other carcinogens. The effect of arsenic on the activity of a variety of enzymes that use GSH has been determined using purified preparations of glutathione reductase (GR) from yeast and bovine glutathione peroxidase (GPx) and equine glutathione S-transferase (GST). The effect on enzyme activity of increasing concentrations (from 1 microM to 100 mM) of commercial sodium arsenite (As(III)) and sodium arsenate (As(V)) and a prepared arsenic(III)-glutathione complex [As(III)(GS)(3)] and methylarsenous diiodide (CH(3)As(III)) has been examined. GR, GPx, and GST are not sensitive to As(V) (IC(50) > 50 mM), and none of the enzymes are inhibited or activated by physiologically relevant concentrations of As(III), As(III)(GS)(3), or CH(3)As(III), although CH(3)As(III) is the most potent inhibitor (0.3 mM < IC(50) < 1.5 mM). GPx is the most sensitive to arsenic treatment and GST the least. Our results do not implicate a direct interaction of As with the glutathione-related enzymes, GR, GPx, and GST, in the mechanism of arsenic toxicity. CH(3)As(III) is the most effective inhibitor, but it is unclear whether this product of arsenic metabolism is produced at a sufficiently high concentration in critical target tissues to play a major role in either arsenic toxicity or carcinogenesis.