ABSTRACT
BACKGROUND: Mercury is an immunotoxic metal that induces autoimmune disease in rodents. Highly susceptible mouse strains such as SJL/N, A.SW, B10.S (H-2s) develop multiple autoimmune manifestations after exposure to inorganic mercury, including lymphoproliferation, elevated levels of autoantibodies, overproduction of IgG and IgE, and circulating immune complexes in kidney and vasculature. A few studies have examined relationships between mercury exposures and adverse immunological reactions in humans, but there is little evidence of mercury-associated autoimmunity in humans. METHODS: To test the immunotoxic effects of mercury in humans, we studied communities in Amazonian Brazil with well-characterized exposures to mercury. Information was collected on diet, mercury exposures, demographic data, and medical history. Antinuclear and antinucleolar autoantibodies (ANA and ANoA) were measured by indirect immunofluorescence. Anti-fibrillarin autoantibodies (AFA) were measured by immunoblotting. RESULTS: In a gold mining site, there was a high prevalence of ANA and ANoA: 40.8% with detectable ANoA at > or =1:10 serum dilution, and 54.1% with detectable ANA (of which 15% had also detectable ANoA). In a riverine town, where the population is exposed to methylmercury by fish consumption, both prevalence and levels of autoantibodies were lower: 18% with detectable ANoA and 10.7% with detectable ANA. In a reference site with lower mercury exposures, both prevalence and levels of autoantibodies were much lower: only 2.0% detectable ANoA, and only 7.1% with detectable ANA. In the gold mining population, we also examined serum for AFA in those subjects with detectable ANoA (> or =1:10). There was no evidence for mercury induction of this autoantibody. CONCLUSIONS: This is the first study to report immunologic changes, indicative of autoimmune dysfunction in persons exposed to mercury, which may also reflect interactions with infectious disease and other factors.