Quantitative Structure-Activity Relationships of Environmental Pollutants

Abstract

SummaryTechniques for the estimation of rates of environmental fate processes as well as effect concentrations of organic micropollutants are based on comparisons with descriptors related to the structure or to physical chemical properties by quantitative structure activity relationships (QSAR). The hydrophobic character of chemicals, estimated by the octanol water partition coefficient (Kow), influences many processes. Simply due to the influence of hydrophobicity on the absorption in aquatic organisms, Kow explains completely the variance in effect concentrations of relatively unreactive chemicals that act by narcosis [155, 156]; effect concentrations that cover seven orders of magnitude while the internal concentrations are almost constant [57, 83]. An expression of toxicity based on internal effect concentrations, instead of exposure concentrations as in LC50 data, is an interesting new development because it makes it more easier to compare data from different species, exposure regimes and exposure times [83]. QSARs established for these unreactive chemicals predict the minimal or base line effect concentrations of all organic micropollutants.Effect concentrations of other classes of chemicals are usually lower than those from the first group with as examples: aromatic amines, phenols, some nitroaromatics, reactive chemicals and toxicants with specific modes of action. This higher activity can be the result of a specific interaction with a target or to a more or less unselective reactivity to nucleophiles. Because the affinity or reactivity to a target molecule can vary among chemicals from each of these classes, it is not surprising that effect concentrations are also influenced by descriptors other than Kow and which are more related to electronic or steric properties. QSAR equations for aquatic toxicity data of these kind of chemicals are less well understood than the equations for unreactive chemicals because the processes, responsible for their effects, are more complex. Chemical reactivity seems an important feature, but most of the present QSAR equations for reactive chemicals are based on measured rate constants [205–207] and it will be a challenge to explore descriptors for chemical reactivity based on calculated parameters from e.g. molecular orbital calculations.Outliers are usually recognized after a QSAR is established and explained in terms of differences in mode of action or toxicokinetic behaviour (e.g. metabolism). This points to the problem that the structural requirements, related to a particular QSAR, are usually not very well defined. More information on mode of action such as in studies to fish acute toxicity syndromes [249] and the structural requirements related to mode of action, are a necessity.Predictions based on QSAR can be useful for a variety of purposes: regulating activities (evaluating new or existing chemicals); priority setting for existing chemicals; evaluation during the development of new chemicals. When predictions are used for practical purposes it is important to be familiar with the prediction models and their limitations. Besides the possibility to use QSAR for prediction, it is very helpful in the analysis of toxicity data, because it is a tool in the classification of large numbers of chemicals into a limited number of classes and in addition outliers can be recognized and studied in more detail. The distinction of classes of chemicals with similar modes of action is also valuable for the estimation of the effects of mixtures.