Utility of the QSAR modeling system for predicting the toxicity of substances on the European inventory of existing commercial chemicals∗

Many chemicals are in common commercial use for which no information on the environmental fate or toxicity exists. Recent legislation requires that many substances be assessed for their toxicity to aquatic organisms within a very short time and determine which of these chemicals need to be studied in greater detail. It would be impossible to measure the acute and chronic effects of all of these compounds on a single organism, let alone a battery of different types of organisms, communities or ecosystems. Initially, the chemicals on the European Inventory of Existing Commercial Chemical Substances (EINECS) need to be screened and relative hazard to the environment determined. In response to OECD directives, there has been a great deal of activity by government and industry scientists. At the International Workshop on Advances in Environmental Hazard and Risk Assessment it was concluded that quantitative structure activity relationships (QSAR) could and should be used in the hazard assessment process. Papers published in that volume outline the advantages, disadvantages, limitations, advances and research requirements.

The QSAR, structure‐activity based chemical modeling and information system, which was developed by the US‐Environmental Protection Agency was used to predict the acute toxicity of 113 substances from the “Old Substances”; list of the German government to the four commonly used aquatic toxicity test organisms: Daphnia magna (DM), fathead minnow (FHM), rainbow trout (RBT), and blue‐gill sunfish (BG).

Of these compounds the QSAR system predicted the acute toxicity of 87 substances towards fathead minnow. For the other three species examined the QSAR system could be used to predict toxicity for 78 compounds.

The predicted toxicities were compared to observed toxicities of compounds which have been evaluated and stored in the “Aquire”; data base. Observed toxicity values were available for at least one species for 38 compounds. The toxicities of some compounds are well predicted while those of other compounds were not well predicted. Overall, the QSAR system accurately classified the acute toxicity ranges of 50%, 64%, 56% and 56% of the compounds investigated for DM, FHM, RBT and BG, respectively. Of the compounds studied 10 were very poorly predicted, of these the QSAR system overpredicted the toxicity of three, while underpredicting the toxicity of seven. Of these seven compounds, five contained amino groups.     

Absorption and retention of polydimethylsiloxanes (silicones) in fish: Preliminary experiments†

Hydrophobicity and bioaccumulation potential of linear and cyclic polydimethylsiloxane (PDMS) oligomers were estimated by reversed‐phase liquid chromatography and feeding experiments with guppies (Poecilia reticulata). PDMS concentrations in fish were determined by capillary column gas chromatography and gas chromatography‐mass spectrometry. In contrast to polychlorinated biphenyls (PCBs), only very small amounts of PDMS were retained by the fish after six weeks feeding.     

The bacterial metabolism of 4,4′-dichlorobiphenyl,and its suppression by alternative carbon sources

The metabolism of 4,4′-dichlorobiphenyl by mixed cultures of bacteria, isolated from activated sludge, was studied in shake cultures and in soil, both in presence and absence of alternative carbon sources. When 4,4′-dichlorobiphenyl was used as sole carbon source, 4-chlorobenzoic acid and 4,4′-dichloro-2,3-biphenyldiol could be isolated from the culture medium. Polar metabolites however, could not be detected in soil and in media in which alternative carbon sources such as glucose, glycerol, peptone, yeast-extract, humic acid or activated sludge were present. No hydroxylated or carboxylic acid derivatives could be isolated when 2,4′, 5-tri-, 2,2′,5,5′,-textra-, 2,2′,3,4,5′-penta-, 2,2′,3,4,5,5′-hexa- and decachlorobiphenyl were used as the sole carbon sources for incubation with bacteria in shake culture.     

UNITTree A multimedia compartment model to estimate the fate of lipophilic compounds in plants

A dynamic model was developed for the distribution of chemicals in plants, and was applied to describe the distribution in 15 year old spruce (Picea abies). A scenario with underlying air and rain concentrations, and emission onto plant surfaces into soilwater and on air particles is calculated for 2,4-Dichlorophenol, 2,3,7,8-TCDD, simazine, hexachlorobenzene and lindane for a period of 5000 h.