The requirement for a second species in chronic and developmental toxicitystudies will have a major impact on the use of rabbits and dogs in Europe, if no alternatives can be applied. A second, non-rodent species like dogs or rabbits is used not just for chemicals, but in many areas of toxicology. This happens where rodents do not sufficiently-well predict the human toxicity, and therefore a safety gap would exist if only rats and mice were used. The recent possibility to examine toxicity directly on undifferentiated, differentiating and differentiated human cells in various model systems may bridge this safety gap, and should make any large animal testing, especially on companion animals, obsolete.

Methods/approach: The large 2-generation reproduction studies and the developmental toxicity animal experiments cannot be substituted by a single alternative assay. Therefore, e.g. ECVAM follows presently the strategy to assemble an array of different tests covering all aspects. The use of embryonic stem cells (ESC) and their differentiation to mature cells is one of the important elements of such a test battery. In close interaction with ECVAM and other laboratories, we want to establish an ESC differentiation model to test for compounds that disturb early embryonic development, impair neurodevelopment or act toxic on neural cells. As part of a test battery coordinated by ECVAM, such systems should be validated as alternative system for the replacement of large and companion animals in all developmental toxicity tests, and should considerably reduce the number of rodents required in first tier testing.

Study plan: We will aim for two major goals. The first goal is the development of an advanced, and importantly, human version of the already-validated murine embryonic stem cell test (EST). The second goal is the development of an assay that will detect neurotoxicity relevant to human exposure conditions. To reach these goals, we will follow three major strategies throughout our work: Firstly, we will work with murine and human ESC in parallel. It has been shown in the past that methods, once established, could often be directly transferred from murine to human systems. To get to an established method, the murine system promises to be faster as many more tools and methods are already available. Secondly, we will work with reporter constructs that will stain the cells in different colors once they have reached a certain developmental stage. This approach allows for more exact and quantitative data in a more time-effective way than the endpoints currently used. Thirdly, we will make use of co-cultures to provide for robust and realistic assay systems. Three methods provide novelty to our approach in relation to the classical EST and neurotoxicity tests: 1. In parallel to the classical approach via embryonic bodies, we will differentiate ESC with help of stromal bone marrow feeder cells towards the neuronal linage.
2. We will introduce defined mechanistic endpoints that sensitively detect sublethal neurotoxicity (ROS formation, JNK activation, heat shock response, disturbed neurotransmitter metabolism, subtle cytoskeletal changes, and reporter activation). 3. We will work with cells that are prelabeled and can therefore be detected easily also in co-cultures or after toxic changes of shape, and quantified by fluorescence imaging methods.

Relevance to 3R: The problem addressed here is a core area of legally required toxicity testing with the highest use of experimental animals (millions). Embryotoxicity and neurotoxicity are domains that still resort to the use of dogs in an attempt to obtain data closer to humans. With our assays we plan to reduce the number of animals required, and eliminate the use of any companion animals. Our assays are developed in constant interaction with ECVAM and should be incorporated into a broad test strategy for developmental and chronic toxicity.