Correlation of Tetrahymena and Pimephales toxicity: Evaluation of 100 additional compounds

Citation
Gd. Sinks et Tw. Schultz, Correlation of Tetrahymena and Pimephales toxicity: Evaluation of 100 additional compounds, ENV TOX CH, 20(4), 2001, pp. 917-921
Citations number
20
Categorie Soggetti
Environment/Ecology
Journal title
ENVIRONMENTAL TOXICOLOGY AND CHEMISTRY
ISSN journal
07307268 → ACNP
Volume
20
Issue
4
Year of publication
2001
Pages
917 - 921
Database
ISI
SICI code
0730-7268(200104)20:4<917:COTAPT>2.0.ZU;2-O
Abstract
In the summary/recommendations for the Ecotoxicology Session of TestSmart-A Humane and Efficient Approach to Screening Information Data Sets (SIDS) Da ta Workshop, it was recommended that more population growth impairment data using Tetrahymena be generated and compared with available lethality data for the fathead minnow. To comply with this recommendation, 100 additional chemicals were tested in the ciliate assay. Toxicity values for the 96-h Pi mephales promelas mortality assay (log[LC50(-1)]) and the 2-d Tetrahymena p yriformis growth assay (log[IGC50(-1)]) were compared. Each chemical was a priori assigned a mode of action. The majority of compounds were classified as either narcotics (n = 46) or direct-acting electro(nucleo)philes (n = 4 3), while 11 chemicals were listed as carboxylic acids, diesters, proelectr ophiles, or weak acid respiratory uncouplers. Toxicities for narcotics show ed an excellent relationship between endpoints with the coefficient of dete rmination (r(2)) being 0.93. A weaker relationship, r(2) = 0.78, was observ ed for the electro(nucleo)philes. The poorer fit for the covalent-reactin. electro(nucleo)philes is attributed to differences in protocol, in particul ar, to test-medium composition and exposure scheme. Those chemicals whose p otency is mediated by metabolism in fish (diesters and proelectrophiles) as well as the acids exhibited poor correlation between endpoints, with toxic ity in the fish assay being greater than that predicted from the ciliate da ta. The regression analysis between endpoints, regardless of mode or mechan ism of toxic action, yielded the model log(LC50(-1)) = 1.12(log[IGC50(-1)]) + 0.46, with n= 92, r(2) = 0.82, s (root of the mean square error) = 0.87, F = 399, and p > F = 0.0001. A result for the present investigation suppor ts earlier findings that, with noted exceptions, there is a strong relation ship between toxicity potency as quantified by P. promelas mortality and T. pyriformis growth impairment.