|
(15-07) Nitrophenols in Milan area: analysis, toxicology and QSAR modelling . Bolzacchini, Ezio*,1, Gramatica, Paola2, Pozzoli, Luca2, Santagostino, Angela1, Rindone, Bruno1, 1 Dipartimento di Scienze dell'Ambiente e del Territorio, Università di Milano-Bicocca, P.zza della Scienza 1, Milano, Italy2 Department of Structural and Functional Biology, QSAR Research Unit, University of Insubria, Via J.H.Dunant 3, Varese, Italy ABSTRACT- Aromatics with polar substituents such as phenol, benzoic acid, nitrophenols, nitrocresols and others have been found in appreciable concentrations in many areas. Liquid chromatography coupled to a photodiode array detector and atmospheric-pressure chemical ionization ion sources were used for the determination of a variety of nitrophenols in air and rain samples after solid phase extraction using silica gel and XAD-7 resin-adsorbents (rainwater). Nitration of phenols in the troposphere appears to be acting both in gas- and in aqueous phase, leading to dinitrophenols. The phytotoxicity of nitrophenols and nitrocresols was studied by determining the reduction in germination of soy seeds. The results obtained from this study indicate that phenol and cresols have no effect at tested doses, while the nitration products always show some effect. In particular the activity of the dinitroderivates is higher than the corresponding mononitroderivates. In order to predict the toxicity of these relevant compounds on animals we determined in vitro both the uncoupling activity of the oxidative phosphorylation and the inhibitory effect of oxygen consumption in mammalian mitochondria for 18 available phenol derivatives. IC20 on ADP/O ratio and IC20 on Respiratory Control Index (RCI) are calculated for all the available nitrophenol and nitrocresol derivatives. The experimental data have been modelled in a QSAR approach using theoretical molecular descriptors, selected by Genetic Algorithm procedure: OLS models with satisfactory predictive power have been obtained for all the responses. The models have been validated by leave-one-out and leave-more-out procedure (Q2LOO=85-93% Q2LMO=82-93%). By these models is possible to predict good IC20 values, also for not experimentally tested compounds. The predicted values for all mononitro-, dinitro-phenols and -cresols have been checked for their reliability by the leverage approach. Key words: nitrophenols, troposphere, toxicity, QSAR |
