CELLULAR METHODS FOR IDENTIFICATION OF NEUROTOXIC CHEMICALS AND ESTIMATION OF NEUROTOXICOLOGICAL RISK

This review critically addresses key aspects of neurotoxicity that can be assessed by using in vitro test procedures and batteries. Such tes t schemes must most probably be hierarchical and multi-optional in ord er to be able to cope with the large number of possible mechanisms of neurotoxicity. Although the regenerative capacity of the nervous syste m is low, lesions can be compensated for by a number of cellular dynam ic functions (e.g. intracellular calcium sequestration, membrane-bound ion transport systems and increases in the rates of energy metabolism and protein synthesis). Therefore, cellular tests included in primary screens of a multiple system should be based on determinations of cel l physiological parameters rather than on measurements of single bioch emical reactions. A general neurotoxicity test system for the determin ation of critical neurotoxic concentrations is suggested to include, a s a first step, the assessment of basal cytotoxicity in a human neurob lastoma cell line. In a second step, differential cytotoxicity is assa yed in highly developed primary cultures of neuronal and non-neuronal cells. In order to find out whether the compound is likely to produce axonopathy, a test procedure in mouse neuroblastoma cells is carried o ut. Toxicokinetic information is obtained from hepatocyte/target cell and endothelial cell/astrocyte co-cultures. To disclose alterations in cell physiology, studies of cell respiration, protein synthesis, memb rane permeability and calcium homoeostasis are suggested. When informa tion from these test steps is evaluated together with available data o n in vivo toxicity, toxicokinetics and physical/chemical parameters, i t may be necessary to proceed to mom neuronal specific determinations or mechanistically oriented studies. If a consistent pattern of effect s or non-critical and critical concentrations is found, the toxicokine tic distribution over the blood-brain barrier must be considered in re lation to actual in vivo blood concentrations if estimates of neurotox ic risk am to be made.