Current limitations of biodegradation screening tests and prediction of biodegradability: A focus on fragrance substances

Abstract Key parameters of OECD biodegradation screening tests were analyzed for impact on the assessment of biodegradation potential of selected fragrance compounds. Comparison of official Japanese OECD 301C test results for 48 predominantly readily biodegradable chemicals with recent OECD 301F screening data demonstrated a significantly higher occurrence of false negatives potentially attributable to inoculum limitations for the former, while for the latter, restriction to 28 days was frequently limiting. Influence of test concentration was studied for two groups of quaternary carbon-containing compounds under high (OECD 301F) and low (OECD 301D) concentrations and extended incubations. Ionones were generally ultimately biodegradable within 28 days or readily biodegradable, whereas damascones reached ultimate biodegradation in OECD 301D tests only and required extended incubations. Predictions from Biowin and Catalogic models were analyzed for ionones, damascones and other selected structural groups of fragrance compounds. Comparison with previously published experimental data revealed significant differences in results between OECD screening tests, thus illustrating the shortcomings of currently available data collections and the potential risk of training prediction models with false negative results, particularly if used for models that are primarily based on data from one specific test. In addition, the analysis revealed specific limitations for prediction models that cannot take into account positioning of fragments within a structure. As a result, it appears that the presence of more than one unfavorable fragment usually does not allow reaching currently accepted threshold scores that would identify a given compound as biodegradable. The outcome of the analyses of this study underlines the need to take into account results from OECD screening tests that are best adapted to physico-chemical properties of fragrances (hydrophobicity, volatility and inhibition potential at high test concentration) and conducted at reduced stringency compared to OECD guidelines (extended incubation owing to reduced bioavailability). Negative MITI database results should be critically reviewed, particularly for quaternary carbon-containing compounds. For future efforts aimed at designing more environmentally benign chemicals, the apparent underestimations will have to be overcome by providing more realistic experimental data that should subsequently be used to improve current biodegradation prediction models for better guidance.