Fumonisin contamination of food: progress in development of biomarkers to better assess human health risks.

Fumonisins, fungal toxins produced by Fusarium moniliforme, contaminate maize based foods and feeds throughout the world. They cause liver and kidney toxicity in animals in addition to leukoencephalomalacia in horses and pulmonary edema in pigs. Fumonisin B(1) is carcinogenic in rats and mice. Ecological studies have linked consumption of fumonisin contaminated maize with oesophageal cancer in human populations in South Africa and China. This review discusses the potential health risks for people exposed to the fumonisins, and describes how mechanistic studies of toxicity in animal models have allowed the development of putative biomarkers of fumonisin exposure at the individual level. The requirements for an applicable biomarker include sample availability as well as a high specificity and sensitivity for the exposure of interest. Most environmental toxic insults involve complex exposures both to other toxins and to infections; these confounding factors need to be considered in assessing both the validity of the biomarker and the exposure-disease associations. Fumonisins can be detected in the urine of animals in feeding studies but the sensitivity of the current methodology means only highly exposed people could be monitored. Mechanistic studies indicate that ceramide synthase, an enzyme involved in sphingolipid synthesis, is one cellular target for fumonisin toxicity and carcinogenicity, and this disruption to sphingolipid metabolism increases the ratio of two sphingoid precursors, sphinganine and sphingosine. The altered ratio has been observed in tissues, serum and urine for a number of animal models suggesting it as a good candidate marker of fumonisin exposure. Despite development of analytical methods to measure this biomarker there have been no studies to date correlating it to fumonisin intake in people. Given the toxic effects of fumonisins in animals and the widespread human exposure, which has been calculated to reach 440 micrograms kg(-1) body weight day(-1) in a population consuming high quantities (460 g day(-1)) of contaminated maize, then the development of biomarkers and their application in epidemiological studies should be a priority for research on these toxins. Copyright 1999 Elsevier Science B.V.