Analyzing Toxicity Through Electrophilicity

The toxicological structure-activity relationships are investigated using conceptual DFT based descriptors like global and local electrophilicities. In the present work the usefulness of electrophilicity in predicting toxicity of several polyaromatic hydrocarbons (PAH) is assessed. The toxicity is expressed through biological activity data (pIC50) defined as molar concentration of those chemicals necessary to displace 50% of radiolabeled tetrachlorodibenzo-p-dioxin (TCDD) from the arylhydrocarbon (Ah) receptor. The experimental toxicity values (pIC50) for the electron acceptor toxin like polychlorinated dibenzofurans (PCDF) are taken as dependent variables and the DFT based global descriptor electrophilicity index (ω) is taken as independent variable in the training set. The same model is then tested on a test set of polychlorinated biphenyls (PCB). A good correlation is obtained which vindicates the importance of these descriptors in the QSAR studies on toxins. These toxins act as electron acceptors in the presence of biomolecules whereas aliphatic amines behave as electron donors some of which are also taken into account for the present work. The toxicity values of the aliphatic amines in terms of the 50% inhibitory growth concentration (IGC50) towards ciliate fresh-water protozoa Tetrahymena pyriformis are considered. Since there is no global nucleophilicity we apply local nucleophilicity (ωmax+) as the descriptor in this case of training set. The same regression model is then applied to a test set of amino alcohols. Although the correlation is very good the statistical analysis reflects some cross validation problem. As a further check the amines and amino alcohols are used together to form both the training and the test sets to provide good correlation. It is demonstrated that the toxicity of several toxins (both electron donors and acceptors) in the gas and solution phases can be adequately explained in terms of global and local electrophilicities. Amount of charge transfer between the toxin and the biosystem, simulated as nucleic acid bases and DNA base pairs, indicates the importance of charge transfer in the observed toxicity. The major strength of the present analysis vis-à-vis the existing ones rests on the fact that it requires only one descriptor having a direct relationship with toxicity to provide a better correlation. Importance of using the information from both the toxin and the biosystem is also analyzed.