Alternative methods to safety studies in experimental animals: role in the risk assessment of chemicals under the new European Chemicals Legislation (REACH)

Machine learning of chemical reactivity from databases of organic reactions

Databases of chemical reactions contain knowledge about the reactivity of specific reagents. Although information is in general only explicitly available for compounds reported to react, it is possible to derive information about substructures that do not react in the reported reactions. Both types of information (positive and negative) can be used to train machine learning techniques to predict if a compound reacts or not with a specific reagent. The whole process was implemented with two databases of reactions, one involving BuNH2 as the reagent, and the other NaCNBH3. Negative information was derived using MOLMAP molecular descriptors, and classification models were developed with Random Forests also based on MOLMAP descriptors. MOLMAP descriptors were based exclusively on calculated physicochemical features of molecules. Correct predictions were achieved for approximately 90% of independent test sets. While NaCNBH3 is a selective reducing reagent widely used in organic synthesis, BuNH2 is a nucleophile that mimics the reactivity of the lysine side chain (involved in an initiating step of the mechanism leading to skin sensitization).

Implementation of the Three Rs in Biomedical Research —Has the Turn of the Century Turned the Tide?

There has been increasing pressure from the public against animal experimentation for testing and research purposes. The Three Rs ( replacement, reduction, and refinement) principle is thought to be a key foundation concept in optimising the welfare of animals used in experiments. This retrospective study attempts to investigate the transition of the Three Rs in biomedical research through a review of articles published in Nature Medicine. We categorised all of the articles published in Nature Medicine from 1998 to 2003, on the basis of the pain and distress of the animals used in the experiments featured in the analysed article. We found there were no large fluctuations in the distribution of these categories over this time period. We also examined each article for the presence of a statement relating to the humane use of laboratory animals, and found that the number of articles which included such a statement dramatically increased in 2002. Over the years studied, there was a decreasing trend in the total number of animal types used for the experiments in the articles. Our results suggest that: a) more encouragement by journal editors might improve the attitude of scientists in terms of animal welfare; and b) the progress of replacement appears to be a more long-term effort in the field of biomedical research.

Is the fish embryo toxicity test (FET) with the zebrafish (Danio rerio) a potential alternative for the fish acute toxicity test?

The fish acute toxicity test is a mandatory component in the base set of data requirements for ecotoxicity testing. The fish acute toxicity test is not compatible with most current animal welfare legislation because mortality is the primary endpoint and it is often hypothesized that fish suffer distress and perhaps pain. Animal alternative considerations have also been incorporated into new European REACH regulations through strong advocacy for the reduction of testing with live animals. One of the most promising alternative approaches to classical acute fish toxicity testing with live fish is the fish embryo toxicity (FET) test. The FET has been a mandatory component in routine whole effluent testing in Germany since 2005 and has already been standardized at the international level. In order to analyze the applicability of the FET also in chemical testing, a comparative re-evaluation of both fish and fish embryo toxicity data was carried out for a total of 143 substances, and statistical approaches were developed to evaluate the correlation between fish and fish embryo toxicity data. Results confirm that fish embryo tests are neither better nor worse than acute fish toxicity tests and provide strong scientific support for the FET as a surrogate for the acute fish toxicity test.

The transgenic mouse assay as an alternative test method for regulatory carcinogenicity studies--implications for REACH

REACH, an EU regulation that requires the submission of safety data in support of the protection of human and environmental health, mandates that registration should be achieved with the minimum amount of animal testing possible. Under REACH, a two-year carcinogenicity assay may be required for certain chemicals produced at >1000 metric tonnes per year. In addition, some chemicals that are found to be genotoxic will also require testing. Alternative methods have been explored in an attempt to improve the predictivity of this bioassay as well as to reduce the number of animals used for such testing. This research has focused on the use of transgenic/knockout mouse models. Study results from selected models indicate that they are useful in hazard identification, even if they are not entirely suitable for risk assessment on their own. Carcinogenic hazard assessment can be greatly enhanced and animal use reduced if the traditional two-year rat bioassay is combined with a well conducted transgenic mouse assay. Importantly, the use of transgenic animals to supplement a traditional two-year carcinogenicity study may help reduce the number of false negatives, one of the unstated goals of REACH via the precautionary principle.

Modelling tissues in 3D: the next future of pharmaco-toxicology and food research?

The development and validation of reliable in vitro methods alternative to conventional in vivo studies in experimental animals is a well-recognised priority in the fields of pharmaco-toxicology and food research. Conventional studies based on two-dimensional (2-D) cell monolayers have demonstrated their significant limitations: the chemically and spatially defined three-dimensional (3-D) network of extracellular matrix components, cell-to-cell and cell-to-matrix interactions that governs differentiation, proliferation and function of cells in vivo is, in fact, lost under the simplified 2-D condition. Being able to reproduce specific tissue-like structures and to mimic functions and responses of real tissues in a way that is more physiologically relevant than what can be achieved through traditional 2-D cell monolayers, 3-D cell culture represents a potential bridge to cover the gap between animal models and human studies. This article addresses the significance and the potential of 3-D in vitro systems to improve the predictive value of cell-based assays for safety and risk assessment studies and for new drugs development and testing. The crucial role of tissue engineering and of the new microscale technologies for improving and optimising these models, as well as the necessity of developing new protocols and analytical methods for their full exploitation, will be also discussed.

Primary rat hepatocytes as in vitro system for gene expression studies: comparison of sandwich, Matrigel and 2D cultures

Recent studies have presented evidence that in vivo obtained gene expression data can be used for carcinogen classification, for instance to differentiate between genotoxic and non-genotoxic carcinogens. However, although primary rat hepatocytes represent a well-established in vitro system for drug metabolism and enzyme induction, they have not yet been systematically optimized for toxicogenomic studies. The latter may be confounded by the fact that cultured hepatocytes show strong spontaneous alterations in gene expression patterns. Therefore, we addressed the following questions: (1) which culture system is optimal, comparing sandwich, Matrigel and 2D cultures, (2) how critical is the impact of culture period on substance-induced alterations in gene expression and (3) do these substance-induced alterations in cultured hepatocytes occur already at in vivo relevant concentrations? For this purpose we analyzed the expression of four genes, namely Abat, Gsk3beta, Myd116 and Sult1a1 that recently have been reported to be influenced by the antihistamine and non-genotoxic carcinogen methapyrilene (MPy). The most reproducible effects of MPy were observed in sandwich cultures. Induction factors of Gsk3beta and Myd116 at 100 microM MPy were 2 and 4 (medians), respectively, whereas expression of Abat and Sult1a1 were inhibited by factors of 7 and 5, respectively. Similar results were observed in hepatocytes maintained for 24 h or 3 weeks in sandwich culture with respect to the influence of MPy on the expression of Abat, Gsk3beta, Myd116 and Sult1a1. To determine whether MPy influences gene expression at in vivo relevant concentrations, 3.5 mg/kg MPy were administered to male Wistar rats intraperitoneally, resulting in plasma concentrations ranging between 1.72 and 0.32 microM 5 and 80 min after injection. Inhibition of Abat and Sult1a1 expression in vitro already occurred at in vivo relevant concentrations of 0.39 microM MPy. Induction of Myd116 was observed at 6.25 microM which is higher but in the same order of magnitude as in vivo relevant concentrations. In conclusion, the presented data strongly suggest that sandwich cultures are most adequate for detection of MPy-induced gene expression alterations and the effect of MPy was detected at in vivo relevant concentrations.

Current in vitro methods in nanoparticle risk assessment: limitations and challenges

Nanoparticles are an emerging class of functional materials defined by size-dependent properties. Application fields range from medical imaging, new drug delivery technologies to various industrial products. Due to the expanding use of nanoparticles, the risk of human exposure rapidly increases and reliable toxicity test systems are urgently needed. Currently, nanoparticle cytotoxicity testing is based on in vitro methods established for hazard characterization of chemicals. However, evidence is accumulating that nanoparticles differ largely from these materials and may interfere with commonly used test systems. Here, we present an overview of current in vitro toxicity test methods for nanoparticle risk assessment and focus on their limitations resulting from specific nanoparticle properties. Nanoparticle features such as high adsorption capacity, hydrophobicity, surface charge, optical and magnetic properties, or catalytic activity may interfere with assay components or detection systems, which has to be considered in nanoparticle toxicity studies by characterization of specific particle properties and a careful test system validation. Future studies require well-characterized materials, the use of available reference materials and an extensive characterization of the applicability of the test methods employed. The resulting challenge for nanoparticle toxicity testing is the development of new standardized in vitro methods that cannot be affected by nanoparticle properties.