Survey of the QSAR and in vitro approaches for developing non-animal methods to supersede the in vivo LD50 test

FG-BERT: a generalized and self-supervised functional group-based molecular representation learning framework for properties prediction

Artificial intelligence-based molecular property prediction plays a key role in molecular design such as bioactive molecules and functional materials. In this study, we propose a self-supervised pretraining deep learning (DL) framework, called functional group bidirectional encoder representations from transformers (FG-BERT), pertained based on ~1.45 million unlabeled drug-like molecules, to learn meaningful representation of molecules from function groups. The pretrained FG-BERT framework can be fine-tuned to predict molecular properties. Compared to state-of-the-art (SOTA) machine learning and DL methods, we demonstrate the high performance of FG-BERT in evaluating molecular properties in tasks involving physical chemistry, biophysics and physiology across 44 benchmark datasets. In addition, FG-BERT utilizes attention mechanisms to focus on FG features that are critical to the target properties, thereby providing excellent interpretability for downstream training tasks. Collectively, FG-BERT does not require any artificially crafted features as input and has excellent interpretability, providing an out-of-the-box framework for developing SOTA models for a variety of molecule (especially for drug) discovery tasks.

Discovery of 4-aminoindole carboxamide derivatives to curtail alpha-synuclein and tau isoform 2N4R oligomer formation

Alzheimer's disease (AD) is a multifactorial, chronic neurodegenerative disease characterized by the presence of extracellular β-amyloid (Aβ) plaques, intraneuronal neurofibrillary tangles (NFTs), activated microglial cells, and an inflammatory state (involving reactive oxygen species production) in the brain. NFTs are comprised of misfolded and hyperphosphorylated forms of the microtubule-binding protein tau. Interestingly, the trimeric form of the 2N4R splice isoform of tau has been found to be more toxic than the trimeric 1N4R isoform in neuron precursor cells. Few drug discovery programs have focused on specific tau isoforms. The present drug discovery project is centered on the anti-aggregation effect of a series of seventeen 4- or 5-aminoindole carboxamides on the 2N4R isoform of tau. The selection of the best compounds was performed using α-synuclein (α-syn). The anti-oligomer and -fibril activities of newly synthesized aminoindole carboxamide derivatives were evaluated with biophysical methods, such as thioflavin T fluorescence assays, photo-induced cross-linking of unmodified proteins, and transmission electron microscopy. To evaluate the reduction of inclusions and cytoprotective effects, M17D neuroblastoma cells expressing inclusion-forming α-syn were treated with the best amide representatives. The 4-aminoindole carboxamide derivatives exhibited a better anti-fibrillar activity compared to their 5-aminoindole counterparts. The amide derivatives 2, 8, and 17 exerted anti-oligomer and anti-fibril activities on α-syn and the 2N4R isoform of tau. At a concentration of 40 μM, compound 8 reduced inclusion formation in M17D neuroblastoma cells expressing inclusion-prone αSynuclein3K::YFP. Our results demonstrate the potential of 4-aminoindole carboxamide derivatives with regard to inhibiting the oligomer formation of α-syn and tau (2N4R isoform) for further optimization prior to pre-clinical studies.

MG-BERT: leveraging unsupervised atomic representation learning for molecular property prediction

MOTIVATION

Accurate and efficient prediction of molecular properties is one of the fundamental issues in drug design and discovery pipelines. Traditional feature engineering-based approaches require extensive expertise in the feature design and selection process. With the development of artificial intelligence (AI) technologies, data-driven methods exhibit unparalleled advantages over the feature engineering-based methods in various domains. Nevertheless, when applied to molecular property prediction, AI models usually suffer from the scarcity of labeled data and show poor generalization ability.

RESULTS

In this study, we proposed molecular graph BERT (MG-BERT), which integrates the local message passing mechanism of graph neural networks (GNNs) into the powerful BERT model to facilitate learning from molecular graphs. Furthermore, an effective self-supervised learning strategy named masked atoms prediction was proposed to pretrain the MG-BERT model on a large amount of unlabeled data to mine context information in molecules. We found the MG-BERT model can generate context-sensitive atomic representations after pretraining and transfer the learned knowledge to the prediction of a variety of molecular properties. The experimental results show that the pretrained MG-BERT model with a little extra fine-tuning can consistently outperform the state-of-the-art methods on all 11 ADMET datasets. Moreover, the MG-BERT model leverages attention mechanisms to focus on atomic features essential to the target property, providing excellent interpretability for the trained model. The MG-BERT model does not require any hand-crafted feature as input and is more reliable due to its excellent interpretability, providing a novel framework to develop state-of-the-art models for a wide range of drug discovery tasks.

An In Vitro Cytotoxicity Assay Using Rat Hepatocytes and MTT and Coomassie Blue Dye as Indicators

Isolated hepatocytes from male Sprague-Dawley rats suspended in culture medium supplemented with either 0.2 or 2% bovine serum albumin (BSA) were allowed to attach to collagen coated 96-well dishes. Ten test chemicals from the MEIC list and salicylic acid were added individually to the dishes, and at the end of 24 and 48 hours, cytotoxicity was determined by measuring MTT (tetrazolium salt) reduction (mitochondrial integrity) and total cellular protein using Coomassie blue dye (reflecting cell number). Total cellular lactate dehydrogenase activity was also determined in some experiments, as an indicator of plasma membrane integrity. The relative toxicities of the test chemicals were quantified by the estimation of EC10, EC20 and EC50 values for each parameter. Except for one chemical, digoxin, in the MTT assay, cytotoxic potency increased with incubation time. The hepatocytes tended to be more sensitive to the chemicals in medium containing 0.2% BSA than in medium containing 2% BSA. Simple linear regression analyses of the log transformed data from the MTT assay versus log oral LD50 in rats for the test chemicals gave the best results using EC10 at 24 hours (r2 = 0.86). With protein as the cytotoxic indicator, the best results were obtained with EC values in the medium containing 2% BSA, again at 24 hours (r2 = 0.83). These results suggest that the MTT and Coomassie blue dye assays could be useful indicators for testing the cytotoxic potential of chemicals in rat hepatocyte cultures.

Two metabolites isolated from endophytic fungus Coniochaeta sp. F-8 in Ageratina adenophora exhibit antioxidative activity and cytotoxicity

Ageratina adenophora is an invasive plant of global importance and has a broad distribution in the Western Himalayas. Endophtytic fungus Coniochaeta sp.F-8 was found in Ageratina adenophora. In this paper, we aim to investigate the antioxidative activity and cytotoxicity of the metabolites from Coniochaeta sp.F-8. Consequently, two compounds Phomoxanthone A and Penialidin A were isolated from the endophytic fungus of Ageratina adenophora for the first time. The structures of compounds were identified by IR, NMR and ESI-MS methods. Moreover, both of those compounds showed considerable antioxidative activity in vitro and resulted cytotoxicity in mouse embryo fibroblasts cell line Balb/c3T3. The present study provides a theoretical foundation for the development and utilization of endophytic fungi in Ageratina adenophora as a medicinal substance.

Estimation of reliability of predictions and model applicability domain evaluation in the analysis of acute toxicity (LD50)

This study presents a new type of acute toxicity (LD(50)) prediction that enables automated assessment of the reliability of predictions (which is synonymous with the assessment of the Model Applicability Domain as defined by the Organization for Economic Cooperation and Development). Analysis involved nearly 75,000 compounds from six animal systems (acute rat toxicity after oral and intraperitoneal administration; acute mouse toxicity after oral, intraperitoneal, intravenous, and subcutaneous administration). Fragmental Partial Least Squares (PLS) with 100 bootstraps yielded baseline predictions that were automatically corrected for non-linear effects in local chemical spaces--a combination called Global, Adjusted Locally According to Similarity (GALAS) modelling methodology. Each prediction obtained in this manner is provided with a reliability index value that depends on both compound's similarity to the training set (that accounts for similar trends in LD(50) variations within multiple bootstraps) and consistency of experimental results with regard to the baseline model in the local chemical environment. The actual performance of the Reliability Index (RI) was proven by its good (and uniform) correlations with Root Mean Square Error (RMSE) in all validation sets, thus providing quantitative assessment of the Model Applicability Domain. The obtained models can be used for compound screening in the early stages of drug development and prioritization for experimental in vitro testing or later in vivo animal acute toxicity studies.

An evaluation of the in vitro cytotoxicities of 50 chemicals by using an electrical current exclusion method versus the neutral red uptake and MTT assays

According to the 2001 National Institutes of Health guidance document on using in vitro data to estimate in vivo starting doses for acute toxicity, the performance of the electrical current exclusion method (ECE) was studied for its suitability as an in vitro cytotoxicity test. In a comparative study, two established in vitro assays based on the quantification of metabolic processes necessary for cell proliferation or organelle integrity (the MTT/WST-8 [WST-8] assay and the neutral red uptake [NRU] assay), and two cytoplasm membrane integrity assays (the trypan blue exclusion [TB] and ECE methods), were performed. IC50 values were evaluated for 50 chemicals ranging from low to high toxicity, 46 of which are listed in Halles Registry of Cytotoxicity (RC). A high correlation was found between the IC50 values obtained in this study and the IC50 data published in the RC. The assay sensitivity was highest for the ECE method, and decreased from the WST-8 assay to the NRU assay to the TB assay. The consistent results of the ECE method are based on technical standardisation, high counting rate, and the ability to combine cell viability and cell volume analysis for detection of the first signs of cell necrosis and damage of the cytoplasmic membrane caused by cytotoxic agents.

In vitro cytotoxicity of selected chemicals commonly produced during fire combustion using human cell lines

Fire combustion products contain a broad range of chemicals, which have a multitude of possible toxic interactions in humans. The aim of this study was to evaluate the cytotoxicity of selected substances commonly produced during fire combustion. A range of human cell lines and cultures including: skin fibroblasts, HepG2 (liver derived), and A549 (lung derived cells) were used to represent different human target organs. The colorimetric MTS assay (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) was used to detect the cytotoxic effects of selected substances including: acetic acid, ammonia, formaldehyde, hydrobromic acid, hydrochloric acid, hydrofluoric acid, potassium cyanide, sodium fluoride, sodium nitrite, sodium sulphide, and sulphurous acid. In this study, the NOAEC (No Observable Adverse Effect Concentration), IC(10) (10% inhibitory concentration), IC50 (50% inhibitory concentration), and TLC (Total Lethal Concentration) values were determined. The ratio between in vitro IC50 to in vivo human toxicity data (Lowest Lethal Dose-LDLo and Lowest Lethal Concentration--LCLo) was also established. Results indicated a strong relationship between IC50 values on the cell types used: fibroblast and A549 (R2: 0.92), A549 and HepG2 (R2: 0.72), fibroblast and HepG2 (R2: 0.69). Good correlation was obtained between the IC50 against LDLo and LCLo when an appropriate adjustment factor was implemented. Results of this study indicated that in vitro methods could be a potential technique for assessing the toxicity of fire combustion products.

What is toxicology and how does toxicity occur?

Toxicology has matured since it was defined as the 'science of poisons'. Modern toxicology is no longer anthropocentric but takes on different views at various biological systems, including ecosystems. Each will interact specifically when exposed to defined chemical agents, including drugs. Adverse effects during drug therapy or after (accidental) poisoning are the result of some negative interactions between the agent and the exposed biological system. Toxicity is no longer a specific property of drugs and chemicals but an operative term to describe the adverse outcome of a specific drugs-host interaction. Newer developments in toxicology have focused on the host. Toxicogenetics continues to provide answers to variations of host response to xenobiotics, including drugs. Clinically relevant genetic polymorphisms and gene defects have been detected, and their number is rapidly growing. The key to understanding is in the host proteins that interact with the drug and mediate the cellular response. Hence, the proteom, i.e. the complete set of proteins of a cell, an individual or a species, determines how an exposed biological system may interact with the manifold of different xenobiotics. Structure-activity studies try to find out useful predictive parameters for risk and toxicity assessment.

Estimation of cytotoxicity to HEP-G2 cells of 255 environmental pollutants and water using QSAR (Quantitative Structure-Activity Relationship)

Although bioassays are considered to be a rational method for environmental management, the procedure is generally too complicated to be applied to daily water quality management. In this study, the feasibility of using for application of a conventional QSAR (Quantitative Structure-Activity Relationship) method was examined to estimate the cytotoxicity of various pollutants found in environmental water. logP, pKa, and molecular weight were chosen as the physico/chemical properties of the pollutants, and defined equations for estimating cytotoxicity based on multiple linear regression analysis between these properties and in vitro cytotoxicity data from our previous results. As a result, a method for estimating cytotoxicity of environmental pollutants that had a certain probability (R>0.8) for the 255 chemicals was successfully developed. Considerably high reliability was shown in the leave-one-out prediction of multi-regression analysis. In addition, the cytotoxicity of environmental water samples was estimated based on multi-regression analysis, using as our samples leachates from 25 landfill sites in Japan. The method developed in this study estimated quantitatively the cytotoxicity of the environmental water from chemical analysis data without conducting a cytotoxicity test.

Regression comparisons of Tetrahymena pyriformis and Poecilia reticulata toxicity

The toxicity data of chemicals common to both the Poecilia reticulata mortality assay and the Tetrahymena pyriformis growth impairment assay were evaluated. Two chemicals were not toxic at saturation in the T. pyriformis assay. In addition, due to abiotic transformation, a third chemical was removed from further consideration. Each chemical was a priori assigned a mode of toxic action: neutral non-covalent, polar non-covalent, or electrophilic covalent toxicity. To further investigate comparisons between endpoints, polar and electrophilic chemicals were separated into class-based groups. The polar non-covalent chemicals were separated into phenols and anilines, while the electrophilic chemicals were separated into those reacting via Schiff-base formation (i.e., aldehydes) and those reacting via bimolecular substitution to a nucleophile (i.e., selected nitroaromatics). A comparison of toxic potency as a collective set was statistically described by the relationship; log(LC50(-1)) = 1.05(log(IGC50(-1))) + 0.56, n = 124; r2 = 0.85; s = 0.42; F = 682; Pr > F = 0.0001. The relationship between endpoints was inversely proportional to reactivity associated with the mode of action. While the comparative toxicity for neutral narcotics exhibited an excellent fit (r2 = 0.94), the fits for polar narcotics and electrophiles were poorer, r2 = 0.69 and 0.62, respectively. Investigations into class-based groupings indicated fit of toxic potency data for aldehydes (r2 = 0.85) and phenols (r2 = 0.81) were quite good. However, fits for anilines (r2 = 0.43) and nitroaromatics (r2 = 0.68) revealed that toxicity was not as well related between endpoints for these chemicals.

The pathologist and toxicologist in pharmaceutical product discovery

Significant change is occurring in the drug discovery paradigm; many companies are utilizing dedicated groups from the toxicology/ pathology disciplines to support early stage activities. The goal is to improve the efficiency of the discovery process for selecting a successful clinical candidate. Toxicity can be predicted by leveraging molecular techniques via rapid high-throughput, low-resource in vitro and in vivo test systems. Several important activities help create a platform to support rapid development of a new molecular entity. The proceedings of this symposium provide excellent examples of these applied concepts in pharmaceutical research and development. Leading biopharmaceutical companies recognize that a competitive advantage can be maintained via rapid characterization of animal models, the cellular identification of therapeutic targets, and improved sensitivity of efficacy assessment. The participation of the molecular pathologist in this quest is evolving rapidly, as evidenced by the growing number of pathologists that interact with drug discovery organizations.

Toxicity assessment of 16 inorganic environmental pollutants by six bioassays

The relative toxicity of 16 environmental pollutants, such as inorganic elements (Ba, Cd, Co, Cr, Cu, Fe, Ge, Hg, Mn, Nb, Pb, Sb, Sn, Ti, V, and Zn), is evaluated on the L-929 established cell line of murine fibroblasts, with five bioassays [RNA synthesis rate assay (RNA), MTT reduction assay (MTT), neutral red incorporation assay (NRI), Coomassie blue assay, and cellular growth rate assay], and on the ciliated protozoa Tetrahymena pyriformis GL [doubling time of T. pyriformis GL population assay (DTP)]. For each inorganic substance, the six bioassays allowed the toxicological index IC50 ("inhibitory concentration 50%") to be calculated. The IC50 values are useful to rank the tested elements and to compare the features of the six bioassays. The most sensitive assays were the RNA, MTT, NRI, and DTP assays. Moreover, the in vitro IC50 values correlated with the in vivo LD50 values; these results were close to those obtained with established lines of human, murine, or fish cells. The sensitivity and the complementarity of these bioassays would be in favor of their incorporation in a "battery" of tests used for toxicological screening studies of xenobiotics.

Research to develop a predicting system of mammalian subacute toxicity, (3). Construction of a predictive toxicokinetics model

A new predictive toxicokinetics model was developed to estimate subacute toxicity (target organs, severity, etc.) of non-congeneric industrial chemicals, where the chemical structures and physico-chemical properties are only available. Thus, a physiological pharmacokinetics model, which consists of blood, liver, kidney (these were experimentally found as major toxicological targets), muscle and fat compartments, was established to simulate the chemical concentrations in organs/tissues with pharmacokinetic parameters by means of Runge-Kutta-Gill algorithm. The pharmacokinetic parameters, i.e. absorption rate, absorption ratio, hepatic extraction ratio of metabolism and renal clearance were calculated by using separately established Quantitative Structure-Pharmacokinetics Relationship equations. The developed predictive model was then applied to simulations of 43 non-congeneric industrial chemicals. The chemical concentrations in organs/tissues after single oral administration were simulated, and their maximum concentrations (Cmax's) and area under the concentration-time curves (AUC's) were calculated. Fast Inverse Laplace Transform was newly applied for the purpose of simulation of 28-day repeated dose toxicity. Simulated concentrations of 28 days repeated dose were, however, found to be the same as those of simple repetitions of a single administration per day because of the short half-lives of non-congeneric industrial chemicals. A comparison of subacute toxicity data with Cmax's and AUC's in a single dose scenario suggested that the organs/tissues with relatively high concentrations of tested chemical substances were the most sensitive targets within a chemical. Chemical concentrations in liver, for instance, were correlated with the severity of hepatotoxicity among the chemicals. It was also suggested that to improve and widen the present approach, data of metabolite and reactivity of non-congeneric industrial chemicals to organs/tissues, receptors, etc. should be incorporated into the model.

FRAME Recommendations for the Application of the Three Rs to the Regulatory Toxicity Testing of Food Additives

Direct food additives are tested for genotoxicity, acute and subchronic toxicity, carcinogenicity and teratogenicity. International guidelines differ in the types of tests required, the duration of the tests, the species of animals to be used, the number of animals recommended and the method of housing experimental animals. This lack of harmonisation is wasteful in terms of animal use and creates additional and, perhaps, unnecessary work for the food industry. In addition, unlike other chemicals, food additives pose a special problem for toxicity testing due to repeated low-dose, life-time human exposure, which is difficult to model in animal studies. In an assessment of the extent to which the Three Rs (reduction, refinement and replacement) can be applied to food additive toxicity testing, it was concluded that differences in regulatory requirements and testing protocols can be improved in both the short term and longer term. Suggestions for improvements to existing alternative approaches for food toxicity testing are made.

A research to develop a predicting system of mammalian subacute toxicity. I. Prediction of subacute toxicity using the biological parameters of acute toxicities

Predicting equations of subacute toxicity were developed by analyzing rat acute and subacute toxicity data of 56 chemicals of various structures. Minimum or 10% effect level in acute or subacute toxicity was estimated as a "biological parameter". Good regression equations were established between the geometrical means ("combined parameters") of any two of the parameters of acute and subacute toxicities and introduction of log P to the equations improved the correlations with a statistically significant multiple regression coefficient. The lowest predicted effect level of the subacute toxicity, which is selected from the data calculated by the above several correlations, can predict the upper limit of the no observed effect level. In recent years, for research and development of new chemical substances, it becomes one of the important factors that these have a lower environmental load in nature. Eventually, it becomes essential to evaluate not only their acute effects on human or environments, but also their chronic influences when they are to be exposed for a long period of time, and the cost for such verification is becoming a breaking factor for research and development. Thus, the development of a new technique which estimates the environmental load of a chemical substance including toxic effects on human with lower cost is now being attempted. For example, the development of in vitro new alternative methods using cultured cells, the utilization of a data base or software which relates mammalian and environmental toxicities and so on are internationally carried forwards. As for the latter, quantitative structure-activity relationship (QSAR) techniques have been applied practically to decide appropriate toxicity tests needed for regulatory purposes by EPA/TSCA, ITC/TSCA and FDA/FDAA in the USA. In addition, it was concluded recently in a joint meeting of EU and US-EPA that the approach by QSAR techniques was useful to specify the new chemical substances which are to be required toxicological examinations. In these QSAR techniques, however, while there are some considerations about common mechanisms of toxicity among chemicals with a similar structure (Congeneric chemicals, Congeners) for establishing of correlation formulas, for chemicals with various structures (Non-congeneric chemicals, Non-congeners) there often lack such common considerations. In addition, biological or physiological factors which are basic toxic indices are often ignored. In a previous study, we researched acute and subacute oral toxicities of industrial common chemicals in rats and reported the followings; 1) their subacute toxicological spectrum in target organs/tissues and morphologic changes was very limited and specific, 2) the important targets were liver, kidneys, blood (spleen) and stomach and these are considered the sites of dominant exposure due to the kinetics of chemical substances, 3) the morphological changes were hepatocellular hypertrophy, deposition of substance in renal tubules, extramedullary hematopoiesis in spleen and mucous lesion in stomach, and these are implying adaptation such as induction of drug-metabolizing enzymes, overload to renal function, anemia from erythrocyte destruction, and direct reaction, respectively, 4) there seems to occur a series of direct and adaptive reactions to exclude the "foreign compounds" which do not show any specific biological activities, 5) it is also considered that there is a possibility to establish a correlation between toxicological findings or target organs/tissues of both acute and subacute toxicities by their continuity. Therefore, in the present study, a predicting equation of subacute (28-day repeated dose) toxicity is attempted to develop from acute (single dose) toxicity data by considering both common mechanisms and biological factors for non-congeneric industrial chemical substances.

Biologic testing of leachable aromatic compounds from denture base materials

The aromatic compounds phenyl benzoate (PB), phenyl salicylate (PS), and biphenyl (BP), which have previously been found to leach from poly(methyl methacrylate) denture base materials, were tested for cytotoxicity and biologic effects by L929 cells in culture. The octanol-water partition coefficient (log P(ow), a descriptor for the lipophilicity, was determined for the compounds. Cytotoxicity was evaluated by total cell growth and the plating efficiency test, and biologic effects by the total fatty acid composition of L929 cells. The commonly used tests, total cell growth and plating efficiency, did not show any significant changes of the cells due to the compounds. On the other hand, BP and PS, in particular, induced changes in the total fatty acid composition of L929 cells. The problem of bioavailability of aromatic compounds in cell culture assays and the relation to lipophilicity was addressed.

Modelling of human acute toxicity from physicochemical properties and non-vertebrate acute toxicity of the 38 organic chemicals of the MEIC priority list by PLS regression and neural network

Linear and non-linear modelling of human acute toxicity (as human lethal concentrations; HLCs) of the 38 organic chemicals from the 50 priority compounds of the Multicentre Evaluation of In Vitro Cytotoxicity (MEIC) programme was investigated. The models obtained were derived either from a set of 23 physicochemical properties of the compounds or from their acute toxicities to five aquatic non-vertebrates together with the physicochemical properties. For the linear type, modelling was performed using a partial least square projection to latent structures (PLS) regression method; for the non-linear models, both PLS regression and neural network were utilized. A neural network using a combination of backpropagation and cascade-correlation algorithms was applied in this study. The results generally reveal a slightly better predictive performance of the models obtained from PLS regression than those obtained from neural networks. However, the model composed of physicochemical properties (PC-model) from the trained neural network using a back propagation algorithm with pruning technique proved superior to that trained with a combination of backpropagation and cascade-correlation algorithms after leave-one-out cross-validation. The predictive power of the PC-models, whether linear or non-linear, was comparable with that of the corresponding models consisting of both structural descriptors and the ecotoxicological tests (ECOPC-models), except for the battery (ECOPC-model) from the neural network. The composition of the 'best' PLS and neural network models points to the importance of the combination of physicochemical properties reflecting lipophilicity, size, volume, intermolecular binding forces and electronic properties of the molecule. All the aquatic non-vertebrate tests are shown to be essential in explaining human acute toxicity. However, the degree of contribution differed, with the crustacean (Artemia salina) and the bacterial (Microtox) bioassays being more important to the linear and non-linear PLS models, whereas the crustacean (Artemia salina and Streptocephalus proboscideus) tests, and the rotifer (Brachionus calyciflorus) assay were important to the neural network models. The organochlorine (lindane) and bipyridinium (paraquat) pesticides were common outliers in all the models. Moreover, the latter two compounds and the organophosphate (malathion) pesticide were also common outliers in all ECOPC-models. Other types of pesticides, however, fit the models. The predicted HLCs of a number of non-pesticides, including some chlorinated compounds, also deviated from the observed HLCs by more than one order of magnitude.

Human acute toxicity prediction of the first 50 MEIC chemicals by a battery of ecotoxicological tests and physicochemical properties

Five acute bioassays consisting of three cyst-based tests (with Artemia salina, Streptocephalus proboscideus and Brachionus calyciflorus), the Daphnia magna test and the bacterial luminescence inhibition test (Photobacterium phosphoreum) are used to determine the acute toxicity of the 50 priority chemicals of the Multicentre Evaluation of In Vitro Cytotoxicity (MEIC) programme. These tests and five physiocochemical properties (n-octanol-water partition coefficient, molecular weight, melting point, boiling point and density) are evaluated either singly or in combination to predict human acute toxicity. Acute toxicity in human is expressed both as oral lethal doses (HLD) and as lethal concentrations (HLC) derived from clinical cases. A comparison has also been made between the individual tests and the conventional rodent tests, as well as between rodent tests and the batteries resulting from partial least squares (PLS), with regard to their predictive power for acute toxicity in humans. Results from univariate regression show that the predictive potential of bioassays (both ecotoxicological and rodent tests) is generally superior to that of individual physicochemical properties for HLD. For HLC prediction, however, no consistent trend could be discerned that indicated whether bioassays are better estimators than physicochemical parameters. Generally, the batteries resulting from PLS regression seem to be more predictive than rodent tests or any of the individual tests. Prediction of HLD appears to be dependent on the phylogeny of the test species: cructaceans, for example, appear to be more important components in the test battery than rotifers and bacteria. For HLC prediction, one anostracan and one cladoceran crustacean are considered to be important. When considering both ecotoxicological tests and physicochemical properties, the battery based on the molecular weight and the cladoceran crustacean predicts HLC substantially better than any other combination.

QSAR models for predicting the acute toxicity of selected organic chemicals with diverse structures to aquatic non-vertebrates and humans

The linear and non-linear relationships of acute toxicity (as determined on five aquatic non-vertebrates and humans) to molecular structure have been investigated on 38 structurally-diverse chemicals. The compounds selected are the organic chemicals from the 50 priority chemicals prescribed by the Multicentre Evaluation of In Vitro Cytotoxicity (MEIC) programme. The models used for the evaluations are the best combination of physico-chemical properties that could be obtained so far for each organism, using the partial least squares projection to latent structures (PLS) regression method and backpropagated neural networks (BPN). Non-linear models, whether derived from PLS regression or backpropagated neural networks, appear to be better than linear models for describing the relationship between acute toxicity and molecular structure. BPN models, in turn, outperform non-linear models obtained from PLS regression. The predictive power of BPN models for the crustacean test species are better than the model for humans (based on human lethal concentration). The physico-chemical properties found to be important to predict both human acute toxicity and the toxicity to aquatic non-vertebrates are the n-octanol water partition coefficient (Pow) and heat of formation (HF). Aside from the two former properties, the contribution of parameters that reflect size and electronic properties of the molecule to the model is also high, but the type of physico-chemical properties differs from one model to another. In all of the best BPN models, some of the principal component analysis (PCA) scores of the 13C-NMR spectrum, with electron withdrawing/accepting capacity (LUMO, HOMO and IP) are molecular size/volume (VDW or MS1) parameters are relevant. The chemical deviating from the QSAR models include non-pesticides as well as some of the pesticides tested. The latter type of chemical fits in a number of the QSAR models. Outliers for one species may be different from those of other test organisms.

The role of in vitro experiments in animal welfare

The prospects of replacing animal experiments with other types of toxicological studies are considered, and the use of human data and in vitro experiments are discussed. Ongoing validation studies of in vitro methods for evaluation of acute toxicity, local irritation, target organ toxicity, tumour promotion and teratogenicity are presented.

Comparative lethality of methanol, ethanol and mixtures in female rats

The lethalities of pure methanol and pure ethanol were compared to two mixtures of ethanol/methanol with the following percentages (95/5% and 65/35% v/v). This study was conducted to simulate situations of human exposure to denaturated alcohol (by 5% methanol) or adulterated alcohol (by 35% methanol). Four groups of female adult virgin albino rats were treated with the four mixtures. A fifth group was used as a vehicle control. Graded oral doses were given to eight animals per dose. Lethality over 24 h was used as an end-point. The LD50 was calculated for each of the four treatments on a molar basis. A dose-response function for each mixture was plotted of percentage lethality vs. mmol kg-1 equivalent to the given ml kg-1 dose. Results showed a significantly different LD50 estimates (P less than 0.03) for the four mixtures. The order of lethal toxicity was as follows: 95/5% methanol/ethanol, pure methanol, pure ethanol then 65/35% methanol/ethanol. Slope comparisons indicated two pairs: 65/35% ethanol/methanol and pure ethanol yielding a steep slope, and 95/5% ethanol/methanol and pure methanol yielding a shallow slope. These data indicated that the acute lethality of ethanol/methanol mixtures is a complex unpredictable function. This toxicity presumably depends in a complicated way on the differences in the effective molecular weights of the two alcohols in each of the mixtures.

Experimental design-based quantitative structure-toxicity relationship of some local anaesthetics using the PLS method

A quantitative structure-toxicology (LD50) relationship for some N-alkylsuccinimides is presented. The relationship is based on a small and carefully selected training set using experimental design methodology. The good predictability, an overall r2 value of 0.75, with such an approach is demonstrated. Factors such as large N-alkyl groups, substitution in the aromatic ring and a long side chain between the two nitrogens are favourable for low toxicity.

Comparison of the up-and-down method and the fixed-dose procedure for acute oral toxicity testing

The acute oral toxicity data for 10 compounds, generated by using two alternative methods in rats, the up-and-down method and the fixed-dose procedure, were compared with those obtained from the classical LD50 test. In this evaluation, both alternative methods offered a reduction in animal use, while providing adequate information to rank the compounds according to the EEC classification for acute oral toxicity. In addition to the ranking, these alternative methods also provided useful information on signs of toxicity and gross autopsy findings, although the results varied depending on the method used. Of the three methods, the up-and-down method required the fewest animals. Although the up-and-down method used only females, the LD50 values obtained were in good agreement with those obtained by the classical method, which used both sexes. It is concluded that the up-and-down method and the fixed-dose procedure are acceptable alternative methods to the classical LD50 test, and the choice of method depends on the type of toxicity information required.