The enigma of arsenic carcinogenesis: role of metabolism

Inorganic arsenic is considered a high-priority hazard, particularlybecause of its potential to be a human carcinogen. In exposed humanpopulations, arsenic is associated with tumors of the lung, skin, bladder,and liver. While it is known to be a human carcinogen, carcinogenesis inlaboratory animals by this metalloid has never been convincinglydemonstrated. Therefore, no animal models exist for studying molecularmechanisms of arsenic carcinogenesis. The apparent human sensitivity,combined with our incomplete understanding about mechanisms of carcinogenicaction, create important public health concerns and challenges in riskassessment, which could be met by understanding the role of metabolism inarsenic toxicity and carcinogenesis. This symposium summary covers threecritical major areas involving arsenic metabolism: its biodiversity, therole of arsenic metabolism in molecular mechanisms of carcinogenesis, andthe impact of arsenic metabolism on human risk assessment. In mammals,arsenic is metabolized to mono- and dimethylated species bymethyltransferase enzymes in reactions that require S-adenosyl- methionine(SAM) as the methyl donating cofactor. A remarkable species diversity inarsenic methyltransferase activity may account for the wide variability insensitivity of humans and animals to arsenic toxicity. Arsenic interfereswith DNA methyltransferases, resulting in inactivation of tumor suppressorgenes through DNA hypermethylation. Other studies suggest thatarsenic-induced malignant transformation is linked to DNA hypomethylationsubsequent to depletion of SAM, which results in aberrant gene activation,including oncogenes. Urinary profiles of arsenic metabolites may be avaluable tool for assessing human susceptibility to arsenic carcinogenesis.While controversial, the idea that unique arsenic metabolic properties mayexplain the apparent non-linear threshold response for arseniccarcinogenesis in humans. In order to address these outstanding issues,further efforts are required to identify an appropriate animal model toelucidate carcinogenic mechanisms of action, and to define dose-responserelationships.