Mechanism and Prediction of the Non-Specific Toxicity of Individual Compounds and Mixtures

Abstract

The toxicity of a range of hydrocarbons, phenols and pyridines were assessed. Such chemicals are classed as non-specific toxicants ie. they are neither an acid, base or a salt. Non-specific toxicants exert their toxicity by a mechanism termed narcosis. The toxicity of the test compounds to a mixed marine bacterial culture (containing thirteen different strains) was assessed both individually and in mixtures under static conditions. Thus, the reported toxicity values were actually the average toxicity of a compound or mixture to the thirteen different strains of bacteria. For homologous series of chemicals the toxicity tended to increase with increasing molecular weight until a point was reached beyond which a toxic effect could no longer by exerted. Evidence was presented that this cut off effect may not be due to aqueous solubility limiting accumulation of the toxicant but rather a volume based mechanism. The toxicity of structural isomers were different. These findings are suggestive of a more specific mechanism of action than narcosis. The toxicity of eighteen different mixtures of the test compounds were found to generally be synergistic. A currently accepted model for the toxicity of mixtures states that the toxicity of mixtures composed of toxicants with the same mechanism of action should be additive. Based on the observed mixture toxicity data and the above principle the test compounds exerted their toxicity by three different mechanisms of action. The compounds that exerted their toxicity in the same manner were: firstly polyaromatic hydrocarbons, 1-alkenes, alkyl-substituted benzenes and alkyl-substituted naphthalenes, secondly alkyl-substituted phenols and thirdly alkyl-substituted pyridines. The non-additive toxicity of mixtures composed of structural isomers indicates that structural isomers have different mechanisms of action; which is unusual considering the degree of chemical and structural similarity. High quality relationships were developed to predict the toxicity of individual compounds as well as other environmentally important properties ie. aqueous solubility and octanol-water partition coefficients. Relationships were developed using high performance liquid chromatography (HPLC) capacity factors. HPLC columns containing different stationary phases were combined in series, for the first time, and successful relationships with toxicity developed. This is a significant finding as it indicates there is the potential to specifically design stationary phases to reflect particular properties of the system being modelled. Relationships to describe the toxicity of mixtures were not of as high quality as those for individual compounds, but were satisfactory. A problem with such relationships is that it is difficult to infer a mechanism of action. The exact mechanism of action for non-specific toxicants remains unclear, despite being the subject of extensive research throughout this century. There are currently three major hypotheses for the mechanism of action; the critical concentration, critical volume and the protein binding hypotheses. Whichever, of the hypotheses is correct should be able to explain the toxicity of mixtures of non-specific toxicants as well as that for individual compounds. The ability of the critical concentration and volume hypotheses to describe the toxicity of individual compounds and mixtures was assessed by comparing literature data with predictions based on each of the hypotheses. The critical volume hypothesis, which states that toxicity occurs when the volume of target tissue increases to a certain volume due to the presence of a toxicant, was better able to describe the toxicity of non-specific toxicants. QSARs based upon parameters that measure the change in volume of a solution due to the presence of a solute (partial molar volume) were derived for the toxicity of individual compounds and mixtures. Both relationships lent further support to the critical volume hypothesis. A new hypothesis, called the funnel hypothesis, based upon the principles of the critical volume hypothesis and explaining the toxicity of mixtures in terms of partial molar volume was developed. This hypothesis is mechanistic and led to several predictions about the toxicity of equitoxic and non-equitoxic mixtures. Of these, the predictions for equitoxic mixtures were confirmed based upon comparison of the hypothesis' predictions and available literature toxicity data. Testing the validity of predictions for non-equitoxic mixtures was not possible due to the paucity of non-equitoxic mixture toxicity data. The predictions of the funnel hypothesis are: 1) the probability of a mixture, composed of non-specific toxicants with different mechanisms of action, deviating from toxic additivity will increase as the number of components decreases. The magnitude of the deviation is also likely to increase as the number of mixtures decreases. 2) the probability of a mixture, composed of non-specific toxicants with different mechanisms of action, deviating from toxic additivity will increase as the concentration of the components increases. The magnitude of the deviation is also likely to increase as the concentration increases.