An Agent-Based Model of Cryoprotectant Equilibration in Secondary Stage Preantral Ovarian Follicles

Abstract

Young cancer patients have limited options for fertility treatment when facing gonadotoxic treatment. One promising fertility treatment for young cancer patients is the cryopreservation of immature ovarian follicles followed by maturation and subsequent reimplantation. However, preantral ovarian follicles currently have lower post-thaw success rates compared to mature oocytes and embryos. Previous research suggests that damage to vital intercellular connections, Transzonal Projections (TZPs), occurs during the cryopreservation process and may account for the observed lower post-thaw success rate in this tissue. It is likely that cryoprotective agent (CPA) equilibration is the cryopreservation step during which TZP damage occurs. Constructing a biologically relevant model of CPA equilibration and the associated damage may allow for improved protocols as measured by increased post-thaw success rates. Agent-based models are a promising technique to capture steps in the cryopreservation process, such as CPA equilibration. In this thesis, I conducted a series of experiments with typical CPAs and nonpermeating solutes at different temperatures using preantral ovarian follicles from a non-human primate (Rhesus monkeys) to measure TZP damage. In these experiments, I also estimated relevant permeability parameters within the tissue. I found that the majority of TZP damage was likely the result of mechanical forces that occurred during the cell volume reduction phase of CPA equilibration. Furthermore, through these experiments, I demonstrate that for this tissue type, parameters collected either during monolayer or single-cell experiments can be used to construct full tissue models. Using the derived experimental parameters and available literature values, I constructed and validated a 3-D agent-based model to capture CPA equilibration in preantral ovarian follicles. My agent-based model utilizes parallel computing on an average desktop computer and allows for the rapid design and testing of CPA equilibration protocols. The model I constructed can account for both mechanical and toxic damage. Importantly, my model accurately captures the experimental damage to TZPs in the majority of simulations. Lastly, I propose several theoretically improved cryopreservation protocols for preantral ovarian follicles. The research presented in this thesis demonstrates that agent-based models can be utilized to capture steps in the cryopreservation in silico and represents a non-invasive, less costly means to test and improve CPA equilibration protocols.