Investigating Human Embryo Implantation – Developing Clinical Applications from in vitro Models

Hunter, Helen Rachel (2024) Investigating Human Embryo Implantation – Developing Clinical Applications from in vitro Models. Doctoral thesis (DClinSci), Manchester Metropolitan University.

Preview

Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (9MB) | Preview

Abstract

Introduction: While assisted conception success rates have increased, factors limiting IVF success include inadequacies in identifying viable embryos, and transfer of embryos into uteri with an unknown state of receptivity. Aims and experimental approaches: The aims of this project are to determine the possibility of using non-invasive techniques to reveal differences between preimplantation human embryos which successfully form a pregnancy and those that fail to implant. The experimental approaches are: 1 Sampling of conditioned media and co-culture with a 3D in vitro model of mid-secretory phase normal human endometrium, followed by transcriptomic analysis of these endometrial cells; 2 Development of a time lapse annotation system to improve selection of PN stage frozen embryos cultured to blastocyst and replaced in FET cycles. Methods: Endometrial epithelial and stromal cells in an in vitro model of mid-secretory phase human endometrium were exposed to conditioned media samples from 10 human embryos cultured singly to the blastocyst stage, with known pregnancy outcomes. These cells were subjected to RNA sequencing and transcriptomic analysis. Time lapse recordings of these embryos were taken through an experimental AI model (eM-Life). Retrospective analysis and annotation of time lapse videos of embryo development of 193 PN stage frozen embryos thawed and cultured to the blastocyst stage for replacement in an FET cycle was performed. Results: Endometrial epithelial cells showed changes in gene expression in response to media from successful embryos, while stromal cells responded to a lesser extent to media from unsuccessful embryos. The deep learning model ranked embryos on morphology but did not correlate with endometrial response in this project. From the analysis of 193 PN stage frozen embryos, statistically significant differences in several morphokinetic parameters between implanting and non-implanting embryos were found and morphological differences not previously studied in frozen thawed embryos relating to embryo viability were identified. Conclusions: Both experimental approaches revealed differences between embryos which implant successfully and those which fail, not detected by standard morphological grading. Further work is needed to identify upstream factors in conditioned media which cause gene expression changes in the in vitro endometrial model, and to test the morphokinetic model developed for frozen embryos in culture.