Environmental arsenic exposure in humans : toxicity and adaptation in the Bolivian Andes

Abstract

Arsenic, a potent toxicant and carcinogen, is naturally present in soil and leaches into groundwater. More than 140 million people worldwide are exposed to arsenic through drinking water. How well humans metabolize arsenic is a susceptibility factor for arsenic toxicity: individuals with a less efficient arsenic metabolism are at higher risk of arsenic-related health effects, such as cancer or cardiovascular disease. In fact, positive selection of a more efficient arsenic metabolism phenotype has been described in the Argentinean Andes, where indigenous populations have been presumably exposed to arsenic for centuries. However, whether this genetic adaptation to arsenic has occurred elsewhere in the Andes was not clear. The overall aim of this thesis was to assess the exposure, toxicity, and potential genetic adaptation to arsenic in the Bolivian Andes. For this, we recruited indigenous women from the Bolivian Andes living in 10 villages around Lake Poopó.

In Paper I, we described that these Bolivian women were exposed to arsenic with varying arsenic concentrations in urine (range 12–407 μg/L, median 65 μg/L). The women had on average an efficient arsenic metabolism compared to other populations across the world. Ethnicity, body weight, fish consumption, and tobacco smoking were identified as influencing their capacity to metabolize arsenic. We then showed that these Bolivian women had molecular signs of arsenic toxicity by measuring four toxicity biomarkers in Paper II. Using multivariable-adjusted linear regression models, arsenic exposure was associated with longer telomeres and more copies of mitochondrial DNA in blood, two biomarkers for cancer risk, particularly in women with a less efficient arsenic metabolism. Urinary 4-hydroxy nonenal mercapturic acid, a metabolite of lipid peroxidation, was associated with increasing arsenic exposure. Urinary 8-oxo-2'-deoxyguanosine, a biomarker of oxidative DNA damage, showed discrepant results depending on the biomarker of exposure used: a positive association with urinary arsenic vs. a negative association with blood arsenic. In Paper III, we found four putative cancer-related proteins in urine that were associated with arsenic exposure in blood: tumor necrosis factor ligand superfamily member 6, FASLG; seizure 6-like protein, SEZ6L; Ly6/PLAUR domain-containing protein 3, LYPD3; and tissue factor pathway inhibitor 2, TFPI2. Other factors influencing the variation of these proteins in urine were identified, including urinary osmolality, leukocytes in urine, and age. By combining genotype-phenotype association analyses and genome-wide selection scans for positive selection in Paper IV, we identified genetic signatures of positive selection near AS3MT, the main arsenic methylating enzyme, in the indigenous Bolivian study group. These Bolivian communities have the highest frequency of protective AS3MT alleles associated with a more efficient arsenic metabolism described in the literature.

In conclusion, indigenous communities living in the Bolivian Andes were exposed to environmental arsenic and were genetically adapted to methylate arsenic more efficiently at a population level. Despite their efficient arsenic metabolism on average, arsenic exposure was still associated with toxicity biomarkers and changes in urinary cancer-related proteins, stressing the need to tackle the public health concern of arsenic in the Andes.