Discriminating toxicant classes by mode of action: 3. Substructure indicators.

Decision support for selecting suitable QSARs for predictive purposes is suggested by a stepwise procedure: The first tier pre-filters the compounds based on substructure indicators for baseline versus excess toxicity. This step, if sufficiently conservative, discriminates chemicals, whose toxicity can be reliably estimated from their log KOW from those, that require further classification by biological and chemical domain. A test set of 115 chemicals from 9 different MOA classes was used to compare the discriminatory power of several classification schemes based on substructure indicators. Performance, evaluated by contingency table statistics, is varied and no single scheme provides sufficient applicability and reliability for pre-filtering chemical inventories. Major improvements are feasible with combined use of three classification schemes: assignments of baseline toxicants are protective, recognition of excess toxicants is acceptable and applicability range increases favourably.

Predictive QSAR models for estimating biodegradation of aromatic compounds.

The development of valid structure biodegradation relationships (SBRs) is restricted by the lack of reproducible published data and by the considered endpoint of degradation data. A classification scheme is required for comparative evaluation of degradation data obtained by different test methods. SBRs based on substructure indicators are available for application to most compounds, but the reliability is still uncertain. SBRs based on physico-chemical parameters are only available for a few classes of compounds based on specific test methods. A combination of several SBRs covering the various transformation pathways provides a promising tool for predicting biodegradability. Two models describing biodegradation are introduced.

Predictive QSAR models for estimating ecotoxic hazard of plant-protecting agents: target and non-target toxicity.

Quantitative structure-activity relationships (QSARs) are evaluated for predicting pesticides' effects on non-target and target organisms: aquatic biota, mammals, soil micro-organisms, and plants. Satisfactory estimates of pesticides' fish toxicity are obtained from log POW-dependent QSARs derived using chemicals of similar polarity and reactivity. Algal toxicity of herbicidal compounds reflects interactions with the electron transport chain in photosystem II and can be modelled by QSARs describing the Hill reaction inhibition. The ranking in effects on soil micro-organisms is evaluated from models derived using in-vitro bacteria systems. Mammalian toxicity can be estimated by QSARs using the partition coefficient log POW and electronic terms derived by MNDO quantum mechanical calculations. In general, the targets are more susceptible than the non-target species towards phenylurea herbicides. Plants and algae constitute the most sensitive populations, corresponding to the same mode of action. Differences in mode of action towards bacteria, rats, and fish, which are similarly sensitive organisms, are revealed by QSAR analyses.

Potential for secondary poisoning and biomagnification in marine organisms.

For selected priority pollutants, like organochlorine pesticides, PAHs and PCBs, and mercury and cadmium, the transfer along marine food chains was assessed based on monitoring data. Comparison of the acquired body burden for marine fish and the toxicity thresholds for predating marine birds and mammals provides evidence for the relevance of contaminant uptake with the food and the liability for secondary poisoning. As a consequence, contaminant residues in prey organisms (critical body burden) should be used for marine hazard and risk assessments. Evaluations solely from aquatic exposure concentrations are not adequate to account for potential secondary effects in marine ecosystems.

Classification of contaminants by mode of action based on in vitro assays.

A concept for toxicity assessments based on in vitro assays of variant levels (i.e., from whole cells to isolated enzymes) is presented. Due to the complexity of organisms of different species, it is evident that no single in vitro test can represent the entire spectrum of toxic potency of chemicals, rather a carefully designed battery of tests has to be employed to account for the various targets attacked in organisms yielding the different modes of action. Cytotoxicity tests like the Neutral-Red Assay predominantly reflect non-specific toxicity, which can be modelled according to a log Pow dependent baseline QSAR. Specific toxicants (e.g., decouplers, acetylcholinesterase inhibitors or photosystem II inhibitors) may be identified based on according in vitro tests and eventually modelled by the respective mode of action related QSARs employing also, e.g., steric or polarizability descriptors to account for specific interactions. The complementation and partial replacement of in vivo (eco)toxicological testing by in vitro assays depends on two criteria: (a) the sensitivity of the tests to reliably detect environmentally relevant concentrations of toxicants and (b) the specificity of the assays to provide an unambiguous classification of toxicants by modes of action: the pattern of interaction with the various targets allows to recognize those compounds of specific effects that frequently occur as outliers in QSAR analyses.

Discriminating toxicant classes by mode of action: 4. Baseline and excess toxicity.

Functional similarity of chemicals combines toxicological knowledge (which toxicity pathways can happen in which species under which exposure conditions) with chemical expertise (which parts of the chemical structures and physico-chemical properties are involved in which interactions) to discriminate between baseline and excess toxicants. The objective is to identify as many baseline toxicants as possible because their acute fish toxicities can be predicted with sufficient accuracy from their log Kow. Established tools like structural alerts are used to indicate modes of action (MOAs) that are typical causes of excess toxicity. Verhaar classifications are supplemented with additional chemical attributes and physico-chemical property thresholds to cover a larger range of compounds within the baseline toxicity domain. Our approach is precautionary to avoid false negatives with a sensitivity of 96.3%. It classifies 57.1% of the compounds of the EPA Fathead Minnow Acute Toxicity Database (EPAFHM) as baseline toxicants and suggests that more than 50% of acute fish toxicity testing could be replaced by reliable QSAR predictions. Furthermore, functional similarity can support the MOA classification of chemicals in different species. Toxicity profiles with fish, Daphnia and algae reveal specific targets for the compounds and, particularly for chemicals with multiple MOA, identify the most sensitive species.

Screening for low aquatic bioaccumulation (2): physico-chemical constraints.

Physico-chemical properties related to the bioavailability of xenobiotics in aquatic environments have been tested for their ability to identify chemicals with low bioconcentration potential. Cut-offs in lipophilicity (log K(OW) < 3 or > 10), solubility and volatility (log Henry constant <-11 [atm (mol L(-1))(-1)]), degradability (ready biodegradability, hydrolysis) and ionisation (>5% ionisation at pH 7) have been adopted and combined into a decision tree based on 382 industrial chemicals. The five-parameter classification scheme was externally validated with 49 pesticides and successfully confirmed with 83 bioaccumulative compounds. The applicability domain of the model has been described in terms of chemical classes (excluding polybrominated compounds (>4 Br), organometallics, compounds with perfluorinated fragments, substances with an acyclic alkyl moiety (chain length > C7) and thiols) and ranges of physico-chemical properties. The present tool allows to securely de-prioritize more than 50% chemicals of low concern with regard to the B criterion (BCF < 2000). Bioassays with compounds with these physico-chemical constraints may be waived because testing may be technically not possible and does not appear scientifically necessary in persistent, bioaccumulative, toxic (PBT) substances and risk assessments.

Screening for low aquatic bioaccumulation. 1. Lipinski's 'Rule of 5' and molecular size.

Aquatic bioconcentration factors are critical in PBT assessment of industrial chemicals under REACH. Reliable indicators based on physico-chemical properties and molecular attributes of chemicals with low bioconcentration potential have been searched to de-prioritize non-accumulative chemicals in order to avoid unnecessary biotests that do not produce risk-relevant information. Developed to screen drug candidates, Lipinski's 'Rule of 5' identifies chemicals with poor oral absorption based on criteria in partitioning, molecular weight and hydrogen bonding. This parameter ensemble has been supplemented with molecular diameter and tested for its adequacy to filter chemicals with low bioconcentration potential. Perhaps (not) surprisingly, the application of the 'Rule of 5' fails to protectively identify non-accumulative compounds because other processes dominate the uptake in aquatic environments as compared with oral absorption. No robust evidence was found for cut-offs in bioconcentration related to molecular size. However, pragmatic thresholds in molecular weight (>650 g mol(-1)) and lipophilicity (log K(OW) < 3 or > 10) have been verified to securely de-prioritize 30-40% of chemicals of low concern with regard to the B criterion.

Application of bacterial growth kinetics to in vitro toxicity assessment of substituted phenols and anilines.

Bacterial growth kinetics were applied to determine toxicity of substituted phenols and anilines serving as model toxicants. The effects observed on Escherichia coli can be quantified reliably. Additional information is obtained about the onset, the duration, and the time course of the toxic action. For most compounds under study a uniform mode of toxicity was observed, but multihalogen-and/or nitrosubstituted phenols differ in toxicity pattern. Possible reasons for the change in mechanism were examined. A comparison with published quantitative data for fish toxicity of halogenated phenols indicates good agreement with toxic effects observed in E. coli cultures. This may lead to the possibility of replacing the more costly, less stable fish test system by this simple and reliable in vitro test system.

QSAR modelling of the ERL-D fathead minnow acute toxicity database.

1. Regression analysis has been applied to examine the structure-activity relationships regarding the acute fish toxicity (96 h LC50 fathead minnow) of organic chemicals. The log P dependent baseline toxicity model has been confirmed for a data set composed of 618 compounds from 24 chemical classes associated with a putative common mode of action. 2. Covariance analysis of the discrete by class regression functions resulted in the combination of chemicals to subsets associated with their mode of action. Separate models were derived for nonpolar (Class I) and polar (Class II and III) compounds. Chemicals which are more toxic than estimated from the baseline model are identified.

Environmental and health risks of hydroquinone.

Hazard assessment of hydroquinone has been evaluated from bibliographical and original data on the physicochemical properties, the environmental behavior, and the biological effects of this aromatic compound. Hydroquinone, which is produced in large amounts and widely used, must be considered as an environmental contaminant. However, it is not persistent. The ecotoxicity of this molecule, which must be linked to its physicochemical properties, varies from species to species. Its acute and chronic toxicity toward higher terrestrial organisms is moderate. Hydroquinone is estimated to be nonmutagenic by the Ames test but induces chromosome aberrations or karyotypic effects in eucaryotic cells. Carcinogenic and teratogenic potentials have been at present inadequately studied. The study underlines the complementarity of QSAR models and experimental approaches when an attempt is made to obtain ecotoxicological profiles of pollutants.