Virtual Sub-zero Life : an agent based approach to modelling cryopreservation of biological tissues

Abstract

The long-term storage of organs and tissue at low temperatures has become a popular area of study for biomedical science and biophysics over the past century. The applications of cryobiology extend to multiple fields including organ transplantation, pharmaceutical research, in vitro fertilisation, food preservation and others. Currently, however, cryopreservation of cells and tissue above the 1-3mm length scale is challenging. This is due to different cell and tissue types requiring specific combinations of cooling rates, cryoprotective agent concentrations and thawing trajectories. The wide variation in the responses of cells to cryopreservation therefore requires extensive trial and error wet lab work, a highly wasteful approach to the optimisation of cryopreservation protocols. The aim of this PhD project is to investigate the use of a computational framework that accounts for key physical differences between cell types during cryopreservation with suitable extensibility for future improvements to include predictions for cryopreserved tissue. The ultimate goal of this work being the optimisation of experimental cryopreservation through computational methods from first principles. Our models take into account multiple factors of the cryopreservation process. These include cellular membrane and osmotic properties, statistics for intracellular ice formation, chemical diffusion and heat transfer in 3D space and finally intercellular forces. The response of cells to these factors is then simulated and our models can make post-thaw survival predictions for cells based on their individual responses to local environmental variables. Based on our computational approaches, we have successfully made post-thaw survival predictions for three cell types: (1) Jurkat T lymphocyte cells, (2) HeLa cells and (3) Human induced pluripotent stem cells. Our modelling shows excellent agreement with experimental literature for the post-thaw survival, cooling rate and cryoprotectant concentrations for all three cell types. In addition, we also have also utilised our model to investigate the effect of nonstandard cooling profiles upon the post-thaw survival of cells in suspension. Our results show that cell death post-thaw can be reduced by more than one order of magnitude via the utilisation of multi-step cooling profiles during the freezing process. This outcome was first predicted through our computational software and we later validated these findings from our modelling process experimentally. Our research has demonstrated that the use of in silico models in combination with experimental work represents a powerful method for reducing the wastage of cells for optimising cryopreservation procedures. The ability of our computational models to accurately predict the post-thaw survival of multiple cell types in 3D space, such as Jurkat cells, means we can better quantify and optimise the post-thaw viability of cells in suspension. In addition, the findings from our work that utilisation of a multi-step cooling profile can significantly improve cellular post-thaw viability represents an additional strength of computational assistance. This is a key example where physical optimisation would not be practical or feasible. The final design of our work constitutes a foundational framework for the optimisation of cells undergoing cryopreservation with high extensibility for future work investigating tissues and organs