Virtual screening for environmental pollutants
Öberg, T.
Presentation vid the Seventeenth International Conference on Quantitative Methods for the Environmental Sciences - TIES2006, Kalmar, 18-22 juni 2006.

Sammanfattning
Many chemicals can inflict damage if released into environmental media (air, soil and water). The current risk paradigm calls for individual consideration and evaluation of each separate compound. In this report it is suggested that mixture exposure and additive effects are of major importance in assessing the environmental hazards. Furthermore, there is no distinct demarcation between persistent and non-persistent compounds. Both these observations put the sequential, one-compound-at-a-time, risk assessment approach in question. Virtual screening of chemical libraries using computational methods is now widely used in drug discovery, and the same approach can also be applied in environmental science. Two validated structure-activity relationships (SARs) have been applied to estimate baseline toxicity and atmospheric persistence among 100,000 compounds in such a library [1-3]. The predicted baseline toxicities indicated an acute hazard for one fourth of the investigated compounds, with the 96-h LC₅₀ for fish (Pimephales promelas) below 10 mg/L. Similarly, atmospheric degradation seems to be very slow (> 2 days) for 15 % of the investigated compounds. When these two SARs were applied sequentially, then 50,074 compounds were within the domain of applicability for both models. The ratio between estimated persistence and toxic concentration (LC₅₀) can provide an indication of the potential environmental impact, and those ranking highest in this data set were: DDT metabolites and derivatives, polychlorinated biphenyls (PCB), diphenyl ethers (PCDE) and dibenzofurans (PCDF), chlorinated paraffins, chlorinated benzenes and derivatives, HCFCs and dichlorononylphenol. Chemoinformatics and computational chemistry can thus help identify persistent organic pollutants, but also provide rational tools for design of products that are easily degradable and have low toxicity. The effect of the global mixture exposure to small (< 300-500 amu) anthropogenic chemicals is possible to evaluate only by simultaneously considering the exposure for all these compounds, and SAR models is then the only viable approach.

1. Öberg, T. (2004) A QSAR for baseline toxicity: Validation, domain of application, and prediction. Chemical Research in Toxicology 17, 1630-1637.
2. Öberg, T. (2005) A QSAR for the hydroxyl radical reaction rate constant: Validation, domain of application, and prediction. Atmospheric Environment, 39, 2189-2200.
3. Öberg, T. (2006) Virtual screening for environmental pollutants - Structure-activity relationships applied to a database of industrial chemicals. Environmental Toxicology and Chemistry 25, 1178–1183.