Description:
Background: The freezing process of tissue samples is crucial for the preservation of morphological and molecular features. Several biobanking guidelines describe freezing techniques for optimal outcomes. As the Vetbiobank standard freezing protocol does not comply with those recommendations in detail, a process validation was performed to demonstrate that samples are suitable for downstream applications. Here we give a formal example of a process validation in the biobanking setting, as required by the biobanking guideline ISO 20387 (2018).Methods: Three different freezing protocols, freezing in liquid nitrogen, freezing via isopentane precooled on dry ice and freezing via liquid nitrogen vapor, were assessed based on morphological integrity of mouse liver and muscle tissue samples. Samples were either frozen in cryotubes (without Optimal Cutting Temperature compound, OCT) or in cryomolds (with OCT). The protocol providing the best results was validated for reproducibility and robustness in terms of defined acceptance criteria for morphological evaluability, A260/A280 ratio, and RNA integrity number values (RIN). In addition, performance tests were run by gene expression analyzes of selected, tissue specific biomarkers to confirm that processed samples are fit for purpose.Results: From the three applied freezing protocols, freezing in liquid nitrogen generated best results. Reproducibility acceptance criteria were met for both, morphological integrity and RNA quality. The freezing method was robust for the tested tissue types and the application of OCT, with exception of liver tissue, where it led to a significant decrease of the RIN value. Gene expression analyzes showed good comparability of results regardless of the applied freezing method.Conclusion: Freezing of tissue samples in liquid nitrogen provides samples of adequate quality for subsequent RNA investigations. A negative impact of OCT on the RIN value of liver samples was observed, which was independent from the applied freezing protocol and showed no impact on subsequent gene expression analysis.