Pre-validation of the PCLS Ex Vivo Model for the Prediction of Respiratory Toxicology
For acute inhalation toxicity studies, animals inhale chemicals at certain doses. In particular,at the beginning, it is difficult to estimate non-toxic doses for in vivo inhalation studies. In thecontext of REACH and the principles of Three Rs, there is an increasing public demand foralternative methods. The goal of this BMBF-funded project is the standardisation andprevalidation of precision-cut lung slices (PCLS) as a suitable ex vivo alternative approach toreplace pre-studies for inhalation toxicology. The project is conducted in three independentlaboratories (Fraunhofer ITEM, BASF SE, RWTH Aachen), together with BfR providingsupport in validation and biostatistics. In all participating laboratories, PCLS were freshlyprepared and exposed to five increasing concentrations of industrial chemicals in serum-freeDMEM under standard submerged cell culture conditions for 1 hour. After 23 hours of postincubationwith serum-free DMEM, chemical-induced toxicity was assessed by the release oflactate dehydrogenase by using the LDH assay, and by determination of metabolic enzymeactivity by using the WST-1 assay. In addition, PCLS protein content and pro-inflammatorycytokine IL-á were measured with the BCA assay and ELISA, respectively. For all endpoints,a sigmoid dose-response model was fitted to the data and EC50 values were calculated. Inaddition, based on the variability obtained in positive and negative controls and test samples,test acceptance criteria were established for each endpoint. This study shows the results ofthe first six tested substances out of 20. In all laboratories, concentration-dependent toxicitycould be shown for aniline, glutaraldehyde, Triton X-100 and paracetamol, but not for lactoseand methyl methacrylate with the WST-1 and LDH assays. EC50 values obtained for theWST-1, LDH and BCA data were very similar in all participating laboratories. No increase inIL-1á level was observed for these chemicals. We conclude that local respiratory toxicology,including irritation and inflammation induced by chemicals, could be tested with comparableresults in the PCLS model without in vivo experiments in three independent laboratories. Thestandardisation of the PCLS method was successful and the reproducibility of the results isvery promising after the testing of the first six substances.