Arsenic is a prevalent human carcinogen whose mutagenicity has not been characterized fully. Exposure to either form of inorganic arsenic, As(III) or As(V), can result in the formation of at least four organic metabolites: monomethylarsonic acid, monomethylarsonous acid (MMA(III)), dimethylarsinic acid, and dimethylarsinous acid (DMA(III)). The methylated trivalent species, as well as some of the other species, have not been evaluated previously for the induction of chromosome aberrations, sister chromatid exchanges (SCE), or toxicity in cultured human peripheral blood lymphocytes; for mutagenicity in L5178Y/Tk(+/-) mouse lymphoma cells or in the Salmonella reversion assay; or for prophage-induction in Escherichia coli. Here we evaluated the arsenicals in these assays and found that MMA(III) and DMA(III) were the most potent clastogens of the six arsenicals in human lymphocytes and the most potent mutagens of the six arsenicals at the Tk(+/-) locus in mouse lymphoma cells. The dimethylated arsenicals were also spindle poisons, suggesting that they may be ultimate forms of arsenic that induce aneuploidy. Although the arsenicals were potent clastogens, none were potent SCE inducers, similar to clastogens that act via reactive oxygen species. None of the six arsenicals were gene mutagens in Salmonella TA98, TA100, or TA104; and neither MMA(III) nor DMA(III) induced prophage. Our results show that both methylated As(V) compounds were less cytotoxic and genotoxic than As(V), whereas both methylated As(III) compounds were more cytotoxic and genotoxic than As(III). Our data support the view that MMA(III) and DMA(III) are candidate ultimate genotoxic forms of arsenic and that they are clastogens and not gene mutagens. We suggest that the clastogenicity of the other arsenicals is due to their metabolism by cells to MMA(III) or DMA(III).

Seafood frequently contains high concentrations of arsenic (approximately 10-100 mg/kg dry weight). In marine algae (seaweed), this arsenic occurs predominantly as ribose derivatives known collectively as arsenosugars. Although it is clear that arsenosugars are not acutely toxic, there is a possibility of arsenosugars having slight chronic toxicity. In general, trivalent arsenicals are more toxic than their pentavalent counterparts, so in this work we examine the hypothesis that trivalent arsenosugars might be significantly more toxic than pentavalent arsenosugars in vitro. We compared the in vitro toxicity of (R)-2,3-dihydroxypropyl-5-deoxy-5-dimethylarsinoyl-beta-D-riboside, a pentavalent arsenosugar, to that of its trivalent counterpart, (R)-2,3-dihydroxypropyl-5-deoxy-5-dimethylarsino-beta-D-riboside. The trivalent arsenosugar nicked plasmid DNA, whereas the pentavalent arsenosugar did not. The trivalent arsenosugar was more cytotoxic (IC50 = 200 microM, 48 h exposure) than its pentavalent counterpart (IC50 > 6000 microM, 48 h exposure) in normal human epidermal keratinocytes in vitro as determined via the neutral red uptake assay. However, both the trivalent and the pentavalent arsenosugars were significantly less toxic than MMA(III), DMA(III), and arsenate. Neither the pentavalent arsenosugar nor the trivalent arsenosugar were mutagenic in Salmonella TA104. The trivalent arsenosugar was readily formed by reaction of the pentavalent arsenosugar with thiol compounds, including, cysteine, glutathione, and dithioerythritol. This work suggests that the reduction of pentavalent arsenosugars to trivalent arsenosugars in biology might have environmental consequences, especially because seaweed consumption is a significant environmental source for human exposure to arsenicals.

No abstract

No abstract