QSAR analyses of the toxicity of aliphatic carboxylic acids and salts to Tetrahymena pyriformis.

Carboxylic acids have been conspicuously absent from the quantitative structure activity relationship (QSAR) literature. This study investigated the aquatic toxicity (log(IGC50(-1)) of selected mono- and di-carboxylic acids and their sodium, or disodium salts, tested in the Tetrahymena population growth assay. The relationship between log(IGC50(-1)) and hydrophobicity as described by the 1-octanol/water partition coefficient (log Kow) revealed a distinct sub-class. The relationship [log(IGC50(-1)) = 0.27(log Kow) - 0.68; n = 16, r2 = 0.943, s = 0.07, F = 233, Pr > F = 0.0001] was derived for mono-carboxylic acids. The QSAR [log(IGC50(-1)) = 0.19(log Kow) - 0.66; n = 9, r2 = 0.951, s = 0.08, F = 135, Pr > F = 0.0001] was generated for the di-carboxylic acids. Regression analysis of data for the monosodium carboxylic acid salts yielded the model, log(IGC50(-1)) = 0.60 (log Kow) + 0.58; n = 4, r2 = 0.932, s = 0.19, F = 41.2, Pr > F = 0.008. Values for the ionization constant (pKa) and the energy of the lowest unoccupied molecular orbital (ELUMO) do not vary within the sub-class for saturated acids. Moreover, pKa and ELUMO did not describe differences in toxicity between sub-classes of saturated aliphatic carboxylic acids and salts. However, these descriptors did vary for unsaturated acids. Inclusion of unsaturated acids afforded the derivation of a global response-surface for all aliphatic carboxylic acids, log(IGC50(-1)) = 0.25(log Kow) - 0.13(ELUMO) - 0.54; n = 34, r2 = 0.850, s = 0.138, F = 87.9, Pr > F = 0.0001. Outliers to the response-surface included small molecules that provided 2-positions in which the molecule could potentially undergo electrophilic attacked and other more large, hydrophobic molecules.