Search

In Vitro Fertilisation (IVF) is quickly becoming an extremely important medical intervention as the prevalence of infertility increases. Therefore, it is vital to ensure that IVF procedures are as safe and successful as possible. There are still many challenges to be addressed, including the embryo assessment process used to select an embryo for transfer. This is a difficult task, as knowledge of pre-implantation development is incomplete and there is still a high degree of subjectivity involved in assessing embryos for viability. In this thesis, we work towards more accurate, objective and versatile embryo selection by employing machine learning (ML) techniques and investigating a range of potential morphological quality markers, including currently neglected sub-cellular features. We first present CNN models trained to predict live birth from a variety of developmental stages. These include the first DL models predicting live birth using solely pre-blastocyst stages, which could allow for earlier embryo transfer, mitigating the harmful effects of prolonged culture. We also showed that information from earlier stages can assist selection at blastocyst stage, allowing for the previously unachievable ranking of high-quality blastocysts. In developing these models, we explored the time period of pre-implantation development to identify the best developmental moments for predicting live birth, therefore providing crucial information for embryo assessment procedures. In order to achieve the best possible assessment of embryo quality at pre-blastocyst stage, we next identified morphological features correlated with transfer outcome and combined these with the CNN model outputs to get an improved overall prediction of live birth. We experimented with different supervised learning techniques and found that linear regression gave the best performance. We also investigated the structure of our dataset via dimensionality reduction techniques and unsupervised clustering, gaining a deeper insight into the challenge of embryo selection. Finally, we investigated changes in nuclear size and appearance during preimplantation development, a sub-cellular feature not currently used to assess embryo viability past the first two embryonic cycles. For this we used the mouse embryo, which mirrors key developmental stages in human embryos. We discovered trends in size and shape both over development, and across different lineages at the same developmental stage. This research could pave the way for better understanding of standard nuclei appearance during pre-implantation development, allowing for the existing embryo selection criteria to be extended. In this thesis we have demonstrated the potential of ML techniques to increase knowledge of the pre-implantation development period, and ultimately lead to improved embryo selection procedures in IVF.

The first few days within mammalian development are arguably some of the most important moments in an organism's lifetime. Between the time of fertilisation of the egg by the sperm and the implantation of the embryo into the uterus wall, a series of morphological and biochemical changes take place that are crucial to the advancement of the pregnancy. This period is known as the preimplantation period and during this time three distinct cell lineages are specified and organised. The trophectoderm (TE), primitive endoderm (PrE) and epiblast (Epi) cell lineages are crucial for the formation of extraembryonic tissues that support the development of the embryo, as well as the formation of the embryo proper itself. The morphological changes within the murine embryo mirror key stages within human development and so the murine embryo offers a promising avenue for understanding our own development and improving assisted reproductive therapies. In this thesis, we work towards the better understanding of the Epi versus PrE specification event through quantitative assessments of embryo architecture and cell neighbourhood in both fixed and live embryos. We first present a novel, user-friendly neighbourhood assessment tool, IVEN, and apply it to murine preimplantation embryos at various fixed stages of development. Through the application of IVEN, we show for the first time a quantitative description of the morphology and cell microenvironment for all cells within the embryo at key stages within the murine preimplantation period. In order to investigate the spatiotemporal dynamics of cells within the embryo, and the corresponding levels of Epi-associated proteins, we present a novel Bayesian landmark matching approach dependent only on the spatial coordinates of the points to be registered. We apply our method to the registration of cells between real time imaging studies with corresponding immunostained images. Within this work, we also present a novel approach to 'Bayesian data selection' which facilitates the inference of cells that can be registered via the chosen model, thus removing the requirement for manual data curation prior to registration. More generally, this data selection approach could be applied to a multitude of statistical problems where subsections of data may be subject to higher noise or corrupted. We therefore envisage that this method could change data handling in the presence of noisy observations. Finally, we introduce sequential neighbourhood analyses of real time imaging data and present preliminary tracking and quantification of neighbourhood dynamics within developing embryos. Through cell registration between real time imaging and corresponding immunostaining images, we then compare the neighbourhood dynamics of Epi and PrE precursors via the comparison of SOX2 positive and negative cells. Within this thesis we demonstrate the value of quantitative approaches within developmental biology and the resulting wealth of information that can be used to progress our understanding of mammalian preimplantation development.

The mammalian preimplantation embryo comprises three distinct lineages; trophectoderm, primitive endoderm and epiblast. The epiblast lineage is the founder line of all the embryonic tissues, while the other lineages give rise to extra-embryonic tissues. The mechanisms behind formation and specification of the first lineages are not fully understood. Pluripotency is a unique characteristic of epiblast lineage. SOX2 and KLF4 transcription factors are a part of the core pluripotency networks, and are known to play a role in somatic cells reprogramming. The role of both factors has been extensively studied in vitro; however, their role in the establishment of pluripotency in vivo is not well understood. This thesis investigated the role of SOX2 and KLF4 in the formation and specification of epiblast lineage in preimplantation mammalian embryo. For many technical reasons, mouse embryo was used as a model in this study. Examination of SOX2 and KLF4 localisation during preimplantation development by immunochemistry revealed that although both proteins were previously proposed to be important components of pluripotency circuit, only SOX2 localisation was restricted to the pluripotent lineages, whereas KLF4 expression was broader and did not mark the pluripotent line throughout the whole preimplantation period. SOX2 levels and localisation were controlled by FGF4/Erk pathway, whereas KLF4 was not. To analyse the function of SOX2 and KLF4, eight-cell stage embryos were injected with Sox2 and Klf4 mRNAs and double-stranded RNAs, and were cultured for 48-72 hours to follow up the progeny of the injected cells. Functional analysis revealed that SOX2 and KLF4 together are crucial to the formation and specification of epiblast lineage. In conclusion, for successful specification of pluripotent line in vivo two mechanisms are needed, first the up-regulation of the pluripotency factors (mediated by SOX2 and KLF4), second simultaneous down-regulation of key endoderm-differentiation factor GATA6 (mediated by KLF4).

The preimplantation stages of mammalian development are dedicated to the differentiation of two extraembryonic epithelia, the trophectoderm (TE) and the primitive endoderm (PrE), and their segregation from the pluripotent embryonic lineage, the epiblast. The TE and PrE are responsible for implantation into the uterus and for producing the tissues that will support and pattern the epiblast as it develops into the foetus. PrE and epiblast are formed in a two step process that involves random cell fate specification, mediated by fibroblast growth factor (FGF) signalling, and cell sorting through several mechanisms. In the present work I have addressed aspects of both steps of this process. Chimaera assays showed that epiblast precursors transplanted onto a recipient embryo rarely differentiate into PrE, while PrE precursors are able to switch their identity and become epiblast. Transient stimulation or inhibition of the FGF4-ERK pathway in the chimaeras can modify the behaviour of these cells and restore the plasticity of epiblast precursors. This work shows that epiblast precursors are refractory to differentiation signals, thus ensuring the preservation of the embryonic lineage. I have also found that atypical Protein Kinase C (aPKC) is a marker of PrE cells and that pharmacological inhibition of aPKC impairs the segregation of PrE and epiblast precursors. Furthermore, it affects the survival of PrE cells and can alter the subcellular localisation of the PrE transcription factor GATA4. These data indicate aPKC plays a central role for the sorting of the PrE and epiblast populations and links cell position within the embryo to PrE maturation and survival. Lastly, I have found that aPKC can directly phosphorylate GATA4 in vitro. Knockdown of GATA4 affects cell position within the embryo, whereas aPKC knockdown reduces the number of GATA4-positive cells. These results indicate GATA4 plays an important role in cell sorting during preimplantation development and suggest phosphorylation by aPKC could determine its presence in the nuclei of PrE cells. My work, in the light of the current knowledge, supports a model where the earliest cell fate decisions during mammalian development depend on cellular interactions and not on inherited cell fate determinants. This robust mode of development underlies the plasticity of the preimplantation embryo and ensures the formation of the first mammalian cell lineages, critical for any further progression in mammalian development.