Search

Early mammalian embryos undergo remarkable changes in their metabolism, with a global transition from oxidative phosphorylation to glycolysis as they transit out of pluripotency and begin to make the first cell fate decisions in preparation for gastrulation. In humans this transition can be modelled with pluripotent cells in culture which exist in a naive (more pluripotent) and a primed (epiblast like) state and show the same differences in metabolic regulation. These two cell states also show major differences in their epigenetic landscape including in DNA methylation, repressive histone modifications, and spatial organisation of chromatin. I am interested in studying this metabolic regulation and its potential interface with epigenetic modifiers. In order to study metabolic regulation and heterogeneity within populations of naive and primed cells, I have adapted an existing method of constraint-based modelling to using constraints based on single cell RNA-seq data. The implementation of single cell models allows us to directly investigate the effect of transcriptional variability on the metabolism of the cells. Analysis of these modelling results shows that PCA of all reaction fluxes reveals clear differences between naive and primed cells, but interestingly also reveals subgroups within each cell type which exhibit different metabolic landscapes, which has not been previously observed. Notably, we have identified two genes which clearly identify these subgroups: SLC15A1 and SLC15A2. These genes encode transporters which are involved in uptake of dietary peptides, and there is no documentation of their function in pluripotency. Here we use several experimental techniques, including single molecule RNA fluorescence in situ hybridisation, to identify if these subpopulations present as our model suggests; future investigation may lead to novel insights into the regulation of these genes and their previously unnoticed role in stem cell metabolism and pluripotency.

Growth in land plants is based on stem cell activity. In seed plants, most of the biomass is produced in the sporophytic generation by stem cells that are embedded in complex multicellular meristems. The CLAVATA pathway is composed of a set of diffusible peptides and their receptors, and mediates intercellular communication and stem cell regulation in multicellular meristems. In non-seed plants, growth is based on the activity of one or a few apical initials. Moss development is based on several single celled meristems, each one constituted by a different stem cell type, which generate filaments, shoots, phyllids and reproductive organs in the gametophytic generation. In this thesis I explore the role of the CLAVATA pathway in the regulation of different types of moss stem cells, using the model moss Physcomitrium patens. By using expression pattern and mutant phenotype analyses, I contribute to uncover the conserved role of CLAVATA in repressing stem cell identity and/or proliferation in P. patens shoots. I then focus on the roles of CLAVATA in the regulation of plant shape and filament development. Here I show that CLAVATA represses the transition from photosynthetic (chloronema) to foraging (caulonemata) filaments by acting on apical stem cell identity. Because this identity shift is controlled by auxin and cytokinin, I then examine the links between CLAVATA and these two hormones, showing that CLAVATA upregulates auxin synthesis and transport in filaments. Finally, I examine CLAVATA roles in the regulation of bud and phyllid development, and reveal a possible role in reproductive development. Overall, in this thesis I show how CLAVATA pathway regulates different stem cell types in P. patens, and identify possible links with auxin and cytokinin homeostasis. This study will help to elucidate how single celled meristems function and how different evolutionary trajectories mold the function of gene families.

Cell polarity is a critical feature of cells and refers to the spatial differences in their shape, structure, and function. Most cells are polarised, with structurally and functionally distinct regions. This polarisation facilitates processes such as asymmetric cell division, differentiation and morphogenesis, directional cell migration, nerve impulse transmission, and transport of molecules across tissues. In fission yeast, the stress-activated protein kinase Sty1, homologous of human p38, is a regulator of the major polarity regulator Cdc42. The activation of Sty1 affects the dynamics of the Cdc42 polarity module, by inducing its dispersal from the sites of growth. However, the link between cell stress and cell polarity is still poorly understood. To identify the targets of Sty1 that are part of the novel polarity regulation pathway, I used a phosphoproteomics approach. I optimised a new triple SILAC protocol to analyse the changes in the whole phosphoproteome of the cells upon Sty1 activation. The implemented protocol allows the identification of up to 50,000 total phosphopeptides per experiment. Interestingly, many of the identified phosphosites that show a significant Sty1-dependent change in phosphorylation belong to proteins that are important for polarity establishment and maintenance. However, the role of Sty1 in controlling cell polarity seems not to be restricted to the regulation of Cdc42: my results suggest that the Sty1 might be a regulator of cell polarity at multiple levels, by controlling the dynamics of proteins that are involved in diverse polarity establishment and maintenance pathways, including the MOR pathway component Sog2, and the eisosome proteins Eis1 and Pil1.

Centromeres are epigenetically defined by the histone H3 variant, CENPA, which is crucial for faithful chromosome segregation and genome stability. CENP-A is distributed equally to sister chromatids during DNA replication in the S phase necessitating CENP-A replenishment in each round of the cell cycle. CENP-A deposition is initiated by the CENP-C-mediated targeting of the Mis18 complex (Mis18a, Mis18b, Mis18BP1) to the centromeres at mitotic exit. Subsequently, the Mis18 complex recruits the CENP-A assembly factor, HJURP, to deposit new CENP-A at the centromeres in early G1. HJURP interacts mainly with the Mis18ab C-terminal domains which are essential but not sufficient for HJURP binding. The Mis18 complex is targeted to the centromeres by direct interaction between Mis18BP1 and Mis18b with the C-terminal region of CENP-C to establish a robust centromeric association of the Mis18 complex. Furthermore, CDK and PLK1-mediated negative and positive regulation, respectively, restrict CENP-A deposition to early G1 in humans. CDK phosphorylation prevents the Mis18 complex assembly and centromeric targeting, while PLK1 directly binds and phosphorylates the Mis18 complex that is important for the centromeric association of the Mis18 complex and CENP-A deposition in early G1. Until today, several important aspects of the CENP-A maintenance pathway were discovered, however, mechanistic details of how the CENP-A deposition machinery is targeted to the centromeres are not fully understood. By using biochemical and mammalian cell-based techniques, this study demonstrates the molecular basis for HJURP/Mis18ab interaction and the implications for CENP-A deposition. Here, we demonstrate that Mis18ab MeDiY domains also contribute to HJURP binding and this implicates CENP-A deposition in vivo. Moreover, phosphorylation of the N-terminal region of Mis18a, potentially by PLK1, positively modulates HJURP/Mis18 complex interaction. Importantly, this study identified new interacting residues between HJURP and Mis18ab MeDiY and C-terminal domains. Finally, biochemical characterisation of CENP-C/Mis18b interaction in vitro showed a robust 2:1 complex formation. Mapping the minimum interaction region of the complex revealed a novel role for the highly conserved CENP-C motif, previously implicated in CENP-A binding, in mediating Mis18b interaction. Collectively, this study provides further insights into novel molecular interactions and regulatory mechanisms of the CENP-A deposition machinery that is essential for centromere maintenance.

In vertebrate embryos, a process called somitogenesis lays the foundations of the adult spine. This process involves elongation and segmentation of the paraxial mesoderm to form somites. Although the segmentation aspect of this has been widely studied, the elongation aspect is not well understood. Posterior growth is widely assumed to be the main driver, but there is very little evidence for this - particularly in fast-developing species like zebrafish. In this thesis, I present the first long term, multi-scale, 3D characterisation of the zebrafish paraxial mesoderm, and show that this tissue elongates through some form of convergent extension, not through growth. In fact, the tissue is compressed over time, and so decreases in volume. I suggest that these processes may be functionally linked, and thus propose a novel mechanism of "compression-extension". Cell tracking, agent-based modelling, and perturbations show that this form of convergent extension does not involve PCP-dependent directional intercalation but, instead, involves convergent flows of cells towards the midline and non-directional intercalation. The cause of compression is not clear, but perturbation experiments suggest that extrinsic forces from the neural tube and TGFβ signalling may be involved. Comparative work in cichlids, chickens, and catsharks suggests that tissue convergence is not unique to zebrafish, and instead is a conserved feature of paraxial mesoderm elongation - even in species that undergo high levels of growth during somitogenesis. This suggests that the relative contributions of growth and tissue convergence to the process of paraxial mesoderm elongation have evolved differently across vertebrate lineages, resulting in a spectrum of elongation strategies.

Macroautophagy (hereafter autophagy) is the process through which portions of the cytosol are sequestered within double-membrane bound autophagosomes. Captured cytosolic contents are subsequently delivered to lysosomes for catabolic degradation. Under optimal growth conditions, autophagy functions to maintain homeostasis through removal of protein aggregates, degradation of longlived proteins, and recycling of excess organelles, to name a few. Under sub-optimal conditions, autophagy can be rapidly induced and functions as an adaptive mechanism to promote cell viability. The best documented stimulator of autophagy is nutrient deprivation, which acts via the canonical signalling pathway involving mechanistic Target of Rapamycin (mTOR) and AMPK. Additional, autophagy modulating, upstream signalling pathways have been described, which include Akt/PKB, MAPK/ERK1 and p53. A diverse array of environmental stressors have also been reported to affect autophagy, many of which utilise these established canonical pathways. Compressive force has been shown to induce autophagy in both the soil dwelling amoeba, Dictyostelium discoideum, and mammalian cells. In both instances, the autophagic response was proportional to the force magnitude and greater than nutrient starvation induction. However, involvement of established autophagy signalling proteins was ruled out, indicating signalling occurred via an unknown non-canonical mechanism. Currently, the underpinning signalling pathways linking mechanical stimulation and autophagy remain elusive. Using D. discoideum, we present novel findings that sustained compression can stimulate production of the secondary messenger cyclic GMP (cGMP). Furthermore, the membrane-permeant cGMP analogue (8Br-cGMP) potently induced autophagy in a dose-responsive manner similar to compression. As hyper-osmotic stress (hyper-OS) drives cGMP production, this stimulus was tested and shown to elicit a strong, dose-dependent, autophagic response similar to compression. For both stimuli, however, cGMP signalling was dispensable as autophagy was unaffected by ablation of guanylyl cyclases or the cGMPregulated kinase, GbpC. While the signalling pathway(s) linking mechanical compression with autophagy induction remain elusive, our findings do, however, highlight a previously unreported mechanism. Subsequent phosphoproteomic analysis has provided candidate proteins which could facilitate this cGMPinduced autophagy. Clearly, the complex web of signalling pathways affecting autophagy is not yet fully resolved. Our findings add to the growing repertoire of stressors which modulate autophagy, and provide promising avenues for future research.

Biology, fundamental to understanding life, remains a vitally important area of research. There is still much left for humankind to understand even after decades of research. This is clear now more than ever; as I write this, research has been disrupted for about 12 months due to various COVID-19 restrictions. Biology, therefore, is ripe for the fresh, new advances that result from interdisciplinary collaboration. Recent years have seen the exciting development of new few-layered materials, providing new possibilities in biology, as well as other areas. To this end, the work presented herein considers the interactions between different cell lines and various synthesised nanomaterial substrates. To distinguish effects due to the inherent biology of the cell and effects due to the nanomaterial substrate, well-defined substrates are crucial to interdisciplinary research. Thin films created by the Langmuir-Schaefer (L-S) deposition technique are a good candidate. This technique provides an easily controllable method of producing single-layer substrates. Here, a method resulting in improved understanding of the physical and chemical influences on L-S film formation is described. Surface pressure-surface coverage data can be normalised to nanosheet size to account for edge effects. This new approach allows the L-S film density to be determined from standard dispersion properties alone. In addition, this work produced the first demonstration of the production of single layer hexagonal boron nitride films using this method. To test nanomaterial-cellular interactions, various cell lines were seeded onto MoS2 L-S substrates. To the best of our knowledge, this study resulted in the first demonstration of the internalisation of MoS2 through mechanotransduction. The material showed localisation to the endoplasmic reticulum, which combined with the innate fluorescence or Raman signal of the MoS2 nanosheet, could lead to a new theranostic tool. This study was expanded to consider cell interactions with other transition metal dichalcogenide materials, WS2 and MoSe2, to investigate the difference between structure and chemistry seen by the cell. This work provides a step change to studying nanomaterial-cellular interactions, opening the door to new therapies and diagnostics.

Centrosomes form the poles of the mitotic spindle in animal cells. Their separation in late G2/early prophase requires precise spatial coordination. Failure in this process can cause an increase in sister chromatid attachment and segregation errors that may contribute to aneuploidy and tumorigenesis. As cells enter mitosis, centrosome positioning is determined by a balance of forces that support and antagonise centrosome separation. Although the major effector of centrosome separation is kinesin Eg5, the antagonising forces that regulate Eg5 are not as well characterised. In this Thesis, we analyse the balance of forces that counteract Eg5 in centrosome separation by the use of a chemical-genetical assay in mammalian cell lines. By manipulating candidate components we found that microtubule polymerisation against a stiff actin cortex acts as the major Eg5 antagonising positioning mechanism. Surprisingly, the system directs centrosomes towards the centre of the nucleus rather than the cell centre. Accordingly, we identified perinuclear actin structures that form in late G2/early prophase that directly interact with microtubules emanating from the centrosomes. These structures cause microtubule bending at the nuclear boundary further supporting the notion of perinuclear actin/microtubule interactions in prophase. Data fitting of microtubule images against an Euler-Bernoulli rod model allowed us to estimate this interaction force in the range of tens of pN. Interestingly, the microtubule (MT)-network branches at points of high curvature, suggesting a load regulation mechanism. Disrupting these structures by breaking the interactions of the LINC complex with perinuclear actin filaments abrogates the nuclear centring mechanism of the centrosomes and causes a dramatic increase in subsequent chromosome segregation errors. Overall, our results provide a systematic analysis of the major components that coordinate centrosome position and separation before nuclear envelope breakdown to ensure accurate chromosome segregation.

Primordial germ cells (PGCs) are the founder cells of germline which follow a specific programme of development characterized by a unique transcriptional network and the resetting of epigenome. Studies in human PGCs (hPGCs) are limited to late migratory and gonadal stages as earlier embryos are almost not accessible. Previous study showed that pig PGC (pPGC) share similar developmental programme to human germ cells in vitro. Therefore, pig may serve as an alternative model to provide insight on transcriptional network and epigenetic reprogramming applicable to hPGC. Using single-cell RNA-seq and whole-genome bisulfite-seq with post-bisulfite adaptor tagging (PBAT-seq), here I show that extensive epigenetic reprogramming, including active DNA demethylation, depletion of macroH2A1, global enrichment of H3K27me3 and X chromosome reactivation (XCR), starts in pre- and early migratory pPGC. Concomitantly, there is dampening of glycolytic gene expression and re-expression of some pluripotency genes like those in preimplantation embryos. Evolutionarily young transposable elements (TEs) and gene coding regions resistant to DNA demethylation, i.e. escapees, have also been identified in extremely hypomethylated gonadal pPGCs. Some of those aforementioned events show conservation with hPGCs but not mPGCs, suggesting that the pig is a relevant model for studies on non-rodent PGC specification and transgenerational epigenetic inheritance. Detailed insights into the pig germline will likely contribute to advances in human germline biology, including in vitro gametogenesis.

Collective cell migration is involved in a plethora of developmental and physiological processes. One well-studied example is the collective migration of the cephalic neural crest (NC); an embryonic stem cell population that gives rise to a diverse cell lineage in vertebrate embryos. NC migration requires the formation of integrin-based cell matrix adhesions. However, the exact molecular composition of these adhesions and how they regulate cell polarity during NC migration remains unclear. Furthermore, it was recently demonstrated that stiffening of the migratory substrate is required to trigger the onset of NC migration. This finding raises the question of how the neural crest transduces mechanical cues from its substrate to trigger migration. Whilst integrin-based adhesions serve as the mechanosensory unit of adherent cells, their functional relevance in mechanotransduction during NC migration remains unknown. Here, we investigate the role of Crk Associated Substrate (CAS) protein, X-chef-1, within the cephalic neural crest in Xenopus laevis. This scaffolding protein within the integrin signaling pathway is expressed in the neural crest prior to the onset of migration, however its function is unknown. Through loss of function experiments, we investigated the requirement of X-chef-1 during NC migration. Knock down of X-chef-1 inhibited migration in vivo and cell dispersion and motility in vitro. Through the targeted expression of X-chef-1 dominant negative constructs, we observed that the migratory deficit was primarily attributed to the loss of tyrosine phosphorylation of the X-chef-1 substrate domain. Taken together, we propose that tyrosine phosphorylation of the X-chef-1 substrate domain promotes migration through activating the cell polarity effector Rac-1 at the leading edge of the neural crest. Furthermore, preliminary experiments suggested that constitutive phosphorylation of the X-chef-1 substrate domain may rescue NC migration on mechanically non-permissive substrates in vivo. Hence, our results set up the framework for further investigation into the role of X-chef-1 in the response to mechanical cues during NC migration.

Unlike humans, some animals can regenerate their lost appendages, such as limbs and tails. To identify cellular principles driving this ability, I investigated Xenopus laevis tadpoles. Tadpoles can regrow their lost tails and limbs only at specific developmental stages, allowing comparative studies to determine principles of regeneration. I first compared tail amputation responses of regeneration-competent and regeneration- incompetent tadpoles by using single-cell transcriptomics. This approach allowed me to identify a previously uncharacterised cell population that we termed Regeneration-Organising- Cells (ROCs). Upon tail amputation, ROCs are relocated from the body to the amputation plane only in regeneration-competent tadpoles. ROCs secrete multiple growth factors for the underlying stem/progenitor cell types to proliferate and grow a new tail. Meanwhile, ROCs in regeneration-incompetent tadpoles are not able to relocalise on the amputation plane, and regeneration is stalled. Eliminating ROCs block regeneration, and grafting ROC-containing tissues enable ectopic-outgrowths. Identification of ROCs suggests that signalling centres and stem/progenitor cells in their proximity are critical requirements to build a lost appendage. Next, I found that the immune system of regeneration-competent tadpoles use reparative myeloid lineage cells that have an impact on the environmental changes, allowing ROCs to relocalise to the amputation plane. Meanwhile, in the regeneration-incompetent tadpoles, inflammatory myeloid lineage cells prevent ROCs' relocalisation. These findings suggest a regeneration-permissive environment is also required to regenerate an appendage. Finally, I focused on Xenopus limb regeneration by comparing limb amputation responses of various regeneration-competency stages by single-cell transcriptomics. Limb regeneration ability is correlated with the re-deployment of a developmental signalling centre cell type, apical-ectodermal-ridge (AER) cells. By using ex vivo regenerating limb cultures, I then demonstrated that, during tadpole development, there is an enrichment of mature chondrogenic lineage cells which express inhibitory secreted factors, including NOGGIN, blocking AER cell formation. Inhibitory factors can be overridden by the application of FGF10 which operates upstream of the effect of NOGGIN, prevents chondrogenesis, and restores regeneration. As mammals have AER in their embryonic development, revealing their connection to chondrogenesis opens new avenues for research aiming to regenerate mammalian limbs. Comparing cellular views of two appendage regeneration scenarios suggests appendage regeneration is context-dependent at the cellular level. Nonetheless, Xenopus appendage regeneration requires the presence of three main cellular principles: (1) Signalling centres (e.g. ROCs, AER cells) to instruct the growth of the surrounding cells, (2) Stem and progenitor cell types which will be the building blocks of a new appendage (e.g. existing stem/progenitor cells, injury-induced plastic cells), (3) A regeneration-permissive-environment (e.g. immune system, secreted factors).

The "maternal age effect" describes the striking increase in risk of miscarriage and chromosomally abnormal embryos and children from women older than 35. Studies in mice have shown that the protein complex Cohesin is reduced in an age-dependent manner. This protein complex ensures accurate segregation during both rounds of meiosis, by holding the chromosomes together and providing a counteracting force to spindle microtubules. Despite a wealth of knowledge generated from human oocytes, there are few live-cell studies, in part due to the paucity of material. This thesis uses human oocytes specifically donated for research to assess the effect of age on alignment of chromosomes at metaphase I and -II, which is a predictor of missegregation. Using high-resolution live-cell microscopy, it is clear that increased age is associated with chromosomes that are misaligned in metaphase I and -II. At metaphase II, eggs arrest until they are fertilised by sperm. The regulation of how chromosomes separate at this point is poorly understood. While the bulk of Cohesin is removed in anaphase I, a small amount is "protected" by Shugoshin 2 and remains between the centromeres to allow for faithful segregation in meiosis II. Currently, there is poor experimental work to support the hypotheses proposed to explain how the mechanisms that protect Cohesin in meiosis I are removed in meiosis II. One of these hypotheses is the requirement of spindle tension to separate the protector, Shugoshin 2, from Cohesin in meiosis II. Here, I show that spindle tension is not required for deprotection and that alternative models should be considered, such as one which suggests that higher-order regulation around meiosis II resumption is orchestrated by the Anaphase Promoting Complex and its coactivator Cdc20, APC/CCdc20, a protein complex that is active at anaphase onset.

The endosomal sorting complexes required for transport (ESCRT) machinery regulate a broad range of biological processes which require similar membrane topological transformations, including intraluminal vesicle formation, viral budding, nuclear envelope resealing, cytokinetic abscission and membrane repair. These processes rely on the assembly of the main ESCRT-III membrane remodelling complex, which cooperates with AAA ATPase Vps4, to drive membrane curvature and membrane fission. Although 11 mammalian members of the ESCRT-III family have been identified, the exact mechanism by which these components are regulated to induce membrane curvature is poorly understood. Coiled-coil C2 domain containing 1A (CC2D1A) and CC2D1B are evolutionarily conserved mammalian members of the Drosophila Lethal giant disc (Lgd) family. CC2D1A and CC2D1B have been reported to directly regulate ESCRT-III-membrane deforming activity during viral budding and postmitotic nuclear envelope reformation, respectively. In addition, the function of CC2D1A has been associated with cell division but its role here is under debate. In this study, the cellular roles of both CC2D1 proteins were investigated during two biological processes: 1) the regulation of ESCRT-III dependent-EGFR trafficking at the endosome, and 2) the centrosome duplication and maturation cycle during cell cycle progression, using the simultaneous siRNA-mediated knockdown of CC2D1A/B. We report that CC2D1A localises at the early endosome and the centrosome whilst CC2D1B localises at the centrosome. Depletion of CC2D1A/B resulted in the clustering of endosomes at the nucleus but did not alter the localisation of several ESCRT-III components nor impact the rate of EGFR degradation in HeLa cells. CC2D1A/B depletion led to a variety of mitotic defects such as binucleated cells, mitotic spindles and lagging chromosomes. A novel approach was utilised to synchronise RPE-1 cells in G1 using a CDK4/6 inhibitor, palbociclib. Following palbociclib release, CC2D1A/B depletion induced G1 cell cycle arrest and reduced EdU incorporation. Synchronised CC2D1A/B-depleted cells also formed micronuclei, indicating that cells had developed chromosome segregation and stability defects. In addition, the recruitment of several centrosomal components important for centriole assembly was reduced and centriole duplication was prevented. Taken together, we reveal that CC2D1A/B is important for the correct spatial distribution of endosomes, regulating G1/S transition during cell cycle progression and mediating the localisation of several centrosomal components essential for centrosome duplication.

Primordial germ cells (PGCs) are germline competent cells which form the functional gametes of the animal. The potential usage of avian PGCs in producing genetically modified birds has driven research in the derivation, culturing, and genetic manipulation of PGCs. In chicken blastoderm, approximately 50 PGCs are present which proliferate in both male and female embryos until stage HH34 (day 8) and subsequently follow different differentiation pathways in male and female gonads. I investigated the hypothesis that chicken migratory stage PGCs are not initially determined to an oocyte or spermatogonial fate. To understand the differences in genetic mechanisms between male and female chicken PGCs, I studied the RNA transcriptome of PGCs from chicken. Analysis of RNA-Seq data of chicken PGCs reveals transcriptome divergence between the male and female cells and identified 150 differentially expressed genes (DEGs). The cultured female PGCs showed higher expression of cell adhesion genes like NCAM2 and PCDH9, and SMAD7B than male PGCs and also showed that dosage compensation is not maintained throughout the Z sex chromosome. To identify novel germ cell and stem cell factors in avian PGCs, I compared the transcriptome of chicken PGCs with immortalized chicken cell lines. As a result, a set of genes were identified which are specific to germ cells including DAZL, DDX4, DDX43, PNLDC1, DMRT1, DMRTB1, and FKBP6. This analysis also helped to identify a suite of pluripotency genes expressed in PGCs: NANOG, OCT4, LIN28, SOX3, GNOT1, TGIF2, PRDM14 and many others. Furthermore, a cross-species transcriptome comparison between in vitro cultured chicken and goose PGC transcriptomes revealed that the expression of these sets of germ cell-specific genes and pluripotent genes expression is conserved in PGCs from these two avian species. This study also revealed the contrasting gene regulatory networks involved in the selfrenewal are active in chicken and goose PGCs. Chicken PGCs exhibit expression of both Activin and BMP signalling pathway genes whereas BMP signalling pathway genes are active in goose PGCs. PRDM14 belongs to the family of the transcription factors containing a conserved N-terminal SET regulatory domain. In mouse, Prdm14 gene expression is limited to the pluripotent cells and essential for the development of the germ cell lineage. In chicken, the PRDM14 knockout embryos do not form a primitive streak. I characterized germ cell development in PRDM14 null chicken embryos and found that PRDM14 has a crucial role in the survival and maintenance of germ cells. Extending my transcriptome analysis to wild-type and PRDM14 null embryos identified DEGs and regulatory pathways possibly responsible for the gastrulation phenotype in the null embryos.

Cells fate choices are governed by a multitude of signalling pathways during development. Environmental and internal cues converge onto important developmental genes that control transcriptional networks and in turn dictate cell fate outcomes. In budding yeast Saccharomyces cerevisiae, nutrient and mating type signals regulate the expression of a developmental gene called IME1. Upon nutrient starvation, IME1 is expressed and diploid cells are induced to enter a conserved cell differentiation programme known as sporulation or gametogenesis. During sporulation, meiotic cell divisions take place to generate four haploid spores. How nutrient signals modulate the expression of IME1 was unknown. In this thesis, I demonstrate that nutrient signals regulate repression of IME1 transcription. Glucose and rich nutrients including nitrogen compounds control the binding of an array of at least nine sequence-specific transcription factors to the IME1 promoter. The transcription factors in turn recruit Tup1-Cyc8, a conserved transcriptional repressor complex to inhibit IME1 transcription. Importantly, Tup1-Cyc8 is heavily involved in the repression of IME1 transcription, given that IME1 is fully de-repressed when Tup1-Cyc8 is depleted. I find that three transcription factors, Yap6, Sok2, and Phd1, play significant roles in mediating Tup1-Cyc8 repression of the IME1 promoter. Remarkably, IME1 transcription is highly responsive to changes in nutrient availability. I show that the IME1 promoter is poised for activation under Tup1-Cyc8 repression. Furthermore, IME1 transcription is rapidly activated when Tup1-Cyc8 is lost. My findings present a model of how complex signals regulate developmental genes at the transcriptional level to control cell fate decisions in eukaryotes.

Genes indispensable for proper embryonic development show intricate patterns of expression throughout the time, space and magnitude of their activity. This diversity is enabled by elaborate regulatory mechanisms that guide their expression. They also possess a distinct type of core promoters that enable the integration of all regulatory inputs. However, it is still not clear how is coordination of regulation achieved. The first step towards understanding this process is to characterise dynamics of expression, and core promoter features that process the regulation. In this thesis, I explored the diversity of spatio-temporal gene expression during zebrafish development. I defined a novel measure of anatomical specificity that defines how precisely an anatomical structure is defined in the Anatomical Ontology system. Using anatomical specificity measure, I quantified gene expression dynamics from mRNA in situ hybridisation data. Gene expression divergence from in situs was used to predict expression levels from RNA-seq expression data. This analysis allowed me to propose a measure of gene expression complexity which showed that genes with the highest complexity score are developmental genes, whereas genes with low complexity score are involved in housekeeping functions. Next, I developed a method that reports significantly enriched core promoter elements in a group of genes. Using this method, I compared differences in core promoter composition in active genes expressed in different developmental periods. In addition, this method found groups of genes with a specific core promoter structure that are specified for a biological process. Finally, I used scRNA-seq data from zebrafish development to identify patterns of gene co-expression across different cell clusters. Co-expression suggests that a gene pair possesses a common regulatory programme. I show that genes with the most divergent co-expression patterns across development are developmental genes and that housekeeping genes have least diverse co-expression patterns. I went further to create co-expression networks which allowed me to analyse co-expression patterns into more details.

Cells need to accurately decode and integrate information from signalling molecules to regulate cellular processes during development and adult homeostasis. The highly conserved Bone Morphogenetic Protein (BMP) signalling pathway is essential during development. In the Drosophila embryo, a gradient of BMP signalling patterns the dorsal ectoderm through the differential regulation of gene expression. It is well established that spatial patterns of gene expression underpin the specification of different cell types. However, the temporal dynamics with which cells respond to cell signals at the transcriptional level is poorly understood. This project aims to investigate how the BMP signalling gradient is decoded at single-cell resolution to generate transcriptional responses. This is accomplished through the use of quantitative live and fixed imaging complemented with computational modelling of burst kinetics. Static images reveal a graded transcriptional response to the BMP gradient, depending on a cell's position within the expression domain. Highest mRNA numbers are found in cells residing in regions exposed to peak BMP signalling levels. In contrast, some cells at the expression domain border are unable to maintain active transcription from both alleles resulting in reduced transcript numbers per cell. Moreover, evidence is provided that expression of a BMP target gene from both alleles in response to peak signalling requires the full complement of early embryonic enhancers. Correspondingly, investigation of transcriptional bursting parameters, based on live imaging of endogenous BMP target gene loci, showed that cells receiving low BMP signalling levels have poor transcriptional burst kinetics that generate only short, low frequency bursts. The burst profiles of two BMP target genes, u-shaped and hindsight, differ significantly in their profiles but both decode BMP signalling levels by modulating promoter occupancy and burst amplitude, suggesting a common mechanism for BMP gradient interpretation. Furthermore, BMP signalling influences promoter occupancy even in the presence of a heterologous promoter, suggesting that the signal is interpreted by the enhancer, which in turn regulates the rate at which the target promoter switches on. In terms of the promoter's contribution to transcriptional bursting, data is presented that the promoter sequence regulates the transcriptional response primarily by altering burst amplitude. Based on these findings a mRNA threshold model is proposed in which a minimum number of mRNA molecules needs to be produced to ensure robustness of cell fate decisions. Moreover, the results presented here provide a platform for understanding how signals are decoded by individual cells in other contexts during both development and adult homeostasis.

Notch signalling is a highly conserved signalling pathway; vital for both embryonic development and adult tissue homeostasis. Aberrant Notch signalling can result in both developmental abnormalities and disease pathogenesis. Such a core signalling pathway is highly regulated. This regulation influences which trafficking route the plasma membrane presented receptor subsequently takes. A key focus of the Baron lab is to better characterise the endocytic routes of Notch activation in Drosophila melanogaster, with an aim to visualise Notch signalling in real-time in living systems. This would greatly benefit the Notch field by shedding light on both the temporal regulation and localisation of the Notch receptor in response to different stimuli. To this end, the main aim of the project presented here is to develop and characterise the tools required to allow visualisation of both the extracellular and intracellular domains of the same Notch molecule, in real time. The Notch receptor is comprised of two main domains; the extracellular domain which functions as a receptor, and the intracellular domain which functions as a transcription factor. To date, the published fluorescently tagged Notch constructs are limited to labelling only one of these two domains. As the cleavage of the extracellular domain from the intracellular domain is a key step in Notch activation, the ability to observe both domains and how they traffic with respect to each other has the potential of providing better understanding of the fundamental basics of Notch signalling. The existing fluorescent Notch tools predominantly tag the transcriptionally active intracellular domain of Notch, with very few examples of tools which fluorescently tag the extracellular domain. This limitation is addressed here by the production and characterisation of the NBFP-GFP construct, which tags both the extracellular and intracellular domains of the same Notch molecule. NBFP-GFP behaves like wild-type Notch in terms of production of full-length protein, localisation to the membrane and transcriptional activity. The use of this construct is applied to live microscopy, and the relative localisation of both extracellular and intracellular domains to distinct endosomal domains is assessed in vitro. To better understand the roles of individual EGF-like repeats of the extensive extracellular EGF-domain of Notch, two cysteine mutant receptors (Nnd3 and NAxM1) are characterised in vitro. Finally, an example of how dual-tagged Notch tools can be applied to biological questions is provided. Inserting the notchoid 3 (nd3) mutation into NBFP-GFP showed that this mutation results in the apparent loss of the extracellular domain. This provides insight into how the nd3 mutation affects normal Notch trafficking as well as better understanding the role of the individual EGF-like repeats of the extracellular domain of Notch.

Asymmetric cell division (ACD) is the process whereby one dividing cell gives rise to two daughter cells of differing fates. ACD is known to be controlled, at least in part, by protein complexes that reside asymmetrically at the cortex of dividing cells. These asymmetric domains are involved in orienting the division plane, localising proteins important for cell fate, as well as for generating polarity in cytosolic structures important for the dispatch of cell fate determinants. One protein complex of particular interest is the Par complex, which is known to asymmetrically localise to the plasma membrane in many systems. It is known that an asymmetric Par complex is required (in many systems), for orienting the division plane, and for generating asymmetry in the density of microtubules in the anaphase-B central spindle (at least in Drosophila). However, whether the Par complex is sufficient to generate these features, or whether they rely on other polarity pathways, is unknown. In this work, I sought to address this question, by developing two techniques for reimparting cortical polarity back into unpolarised cells (cells that have lost polarity). The first of these techniques relies on protein micropatterning, which is the process by which proteins can be ‘printed’ onto a surface. For successful establishment of this assay, a new system of micropatterning, based on ‘anchoring’ proteins to Fibrinogen (a large glycoprotein complex), was developed. Fibrinogen-anchoring remarkably improves the micropatterning of previously difficult-to-pattern proteins, and gives a universal patterning platform upon which complex, multi-protein patterning experiments can be performed. For the first time, protein micropatterns were used to control the subcellular localisation of cell receptors. Further, with this technology, intracellular plasma membrane asymmetries, akin to those that occur in multicellular organisms, could be reconstituted. The second technology for reconstituting plasma membrane asymmetry was based on novel de novo designed proteins, which, upon mixing, spontaneously assemble into large, two-dimensional lattices. Using quantitative fluorescence microscopy and Atomic Force Microscopy, we demonstrated that these lattices assemble in a near-crystalline fashion on cells. With these proteins, robust asymmetric ‘caps’ of plasma membrane complexes could be generated. Using this technique, I showed that an asymmetric Par complex is indeed sufficient to orient division, and to generate asymmetry in the central spindle. Finally, I showed that these two pathways, while both downstream of an asymmetric Par complex, can be untangled. This represents the most complete reconstitution of asymmetric cell division in symmetrically dividing, unpolarised cells to date. This is the first demonstration that an asymmetric Par complex is sufficient to reinstate polarised signalling and important aspects of ACD in unpolarised cells, and is also the first example of central spindle asymmetry in mammalian cells, revealing this to be a conserved feature of ACD. Further, the technology developed in this study will be of broad use to the field, and will no doubt be used to shed light onto more of the mechanisms controlling cortical symmetry breaking and polarised signalling.

The gap genes encode transcription factors that play a central role in the process of segment patterning in Drosophila. Specifically, they interact to form the well-characterised ‘gap gene network’, which directs the formation of segment boundaries (through regulation of pair-rule genes), and the subsequent diversification of segments (through regulation of Hox genes). Although homologues of the gap genes play important roles in segment patterning in many insects, there is as yet no clear understanding of the network as a whole outside of Drosophila. In particular, the existence of a ‘gap gene gap’ (a region of the axis that expresses no gap genes) in embryos of the beetle Tribolium castaneum may indicate the existence of additional, as yet unidentified gap genes in this species. In this work, I investigate the hypothesis that the neuroblast timer genes nubbin (nub) and castor (cas) may act as components of the gap gene network in Tribolium. I first utilise Hybridisation Chain Reaction in situ hybridisation to produce a comprehensive description of the dynamics of gap gene expression in Tribolium, concluding that, although the non-canonical gap genes Tc-mille-pattes and Tc-shavenbaby are expressed in the gap gene gap, their role may be distinct from that of the canonical gap genes, and that there are likely to be other, unknown gap genes expressed alongside them. I show that the four genes of the neuroblast timer series (hunchback, Krüppel, nub and cas) are expressed sequentially in the segment addition zone, with the result that nub and cas are expressed in the gap gene gap. Knocking down the expression of nub, but not cas, using RNAi results in weak homeotic transformations of the first abdominal segment towards a thoracic fate. Finally, I show that this phenotype is dramatically increased in severity and penetrance when Tc-nub is knocked down in addition to the trunk gap genes Tc-gt and/or Tc-kni. In triple knockdowns, all abdominal segments are transformed into thoracic segments due to a posterior expansion of the thoracic gap gene Tc- Kr and subsequent loss of expression of the abdominal Hox genes Tc-abdominal A and Tc- Ultrabithorax. These data indicate that Tc-nub, Tc-gt and Tc-kni act redundantly to repress the expression of Tc-Kr in the abdomen, and that both Tc-nub and Tc-kni should therefore be considered as components of the gap gene network in Tribolium. My work strengthens the hypothesis that the gap gene network may have ancient evolutionary ties to the neuroblast timer network, and promotes a more ‘modular’ view of the gap gene network. I hope that this will inform future studies that aim to unravel the developmental role of the gap genes in sequentially segmenting arthropods, and the evolutionary origins of the gap gene network.

The yeast centrosome or Spindle Pole Body (SPB) is an organelle situated in the nuclear membrane, where it nucleates spindle microtubules and acts as a signalling hub. The SPB is reproduced conservatively, meaning an existing SPB is required to form a new SPB. Additionally, the Mitotic Exit Network (MEN), a group of proteins controlling exit from mitosis localize to the SPB, in an essential step in MEN activation. These properties are conserved in the MicroTubule Organising Centre’s (MTOC) of higher eukaryotes, which also duplicate conservatively and act as signalling hubs. Various studies have explored the effects of forcing individual proteins to interact with the yeast SPB, however no systematic study has been performed. Furthermore, while the MEN has been studied intensively, a unified understanding of how localization and protein activity function together as a system is lacking. I have used Synthetic Physical Interaction (SPI) screening to detect proteins that inhibit growth when forced to associate with the SPB. I found that the SPB is especially sensitive to relocalization, necessitating a novel data analysis approach for the SPI data. In addition, I found a set of forced associations that perturb SPB duplication, resulting in elevated SPB number and, in some cases, multi-polar spindles. In order to study spatial aspects of MEN regulation in parallel to control of enzymatic activity by post-translational modifications I developed a compartmental, logical model of the MEN that is capable of representing both aspects of regulation. I found that my model is capable of correctly predicting the phenotype of 80% of mutants we tested, including mutants representing mislocalizing proteins. I use this model to uncover new aspects of MEN regulation, study the role of the FEAR (Cdc Fourteen Early Anaphase Release) network in determining anaphase length and investigate stochasticity in the Spindle Position Checkpoint (SPoC). Altogether this project presents a systematic view of which proteins are regulated through interaction with the SPB and how this functions as a mechanism of protein regulation.

Pollination networks have traditionally been seen as static, specialised systems. This view has supported the concept of pollination syndromes; the idea that plants evolve particular traits that attract specific pollinator types. However, this view has been challenged in the last few decades as a community level approach has revealed plants are visited by a wide range of taxa. Conversely, xeric environments are rich in oligolectic bee species, suggesting specialism is dominant in this habitat type. This thesis used the Hajar Mountains in Oman as a study site to investigate how visitation networks change temporally and spatially. Specialism appears to be the governing process across all types of flower visitor, not just bee species. However, when a temporal approach is taken, species replacement was extremely high, indicating a constant fluctuation in the composition of the networks in this mountain range, which like in other studies challenges the notions regarding floral syndromes. Species distribution models show a degree of homogenisation in bee communities across the Arabian Peninsula. This contrasts at more local level as seen in the Hajar Mountains. Further studies are now needed to DNA barcode pollen loads from bee specimens collected during the study and region as a whole. This would allow a clearer understanding of network dynamics and help clarify whether the observed specialism seen in this study reflects evolutionary specialism or simply floral constancy.

Plant protection products (PPPs) is a term that has a very similar definition to pesticides, however pesticides can be used for wider applications such as biocides and other chemicals harmful to fauna. PPPs specifically refer to chemicals used to guard and increase the yield of profitable or desirable crops. This term originated from the Plant Protection Products directive (91/414/EEC) which pertains to the registration of products on the market (Jess et al., 2014). Regulator interest in the safety of pesticides began with the development of synthetic organic chemicals and the creation of DDT. The publication of Silent Spring by Rachel Carson in 1962 brought the mal effects of synthetic pesticides to the forefront of public thinking and proposed that Springtime will be devoid of birdsong if they continue to be used as they are (Zadoks and Waibel, 2000; Beard, 2006). This preceded the setup of the Environment Protection Agency (EPA) in 1970. Within Europe the environmental policy has undergone many iterations since this, with the current strategy for regulation of pesticides within Europe now governed by the European Union (EU) and the individual states within a tiered setup. The first tier is that the producer must provide a dossier to the EU specifying that the product is safe for use within their framework. This is then assessed by the 'rapporteur member state (Regulation No. 1107/2009). As a result of the increased regulatory interest and environmental impact, agrochemical companies began to produce their products to not only have deleterious effects on the target organism but also to have less impact on the environment. This means that pesticides are often designed to reduce the risk to nontarget organisms. Methods of doing so include: limiting the transport of the compound from the site of application; reducing the persistence of the compound within the soil profile; and ensuring that the mode of action and toxicity is as organism specific as possible. Despite this, the assessment of the producer needs to be verified by the rapporteur member state and a full risk assessment must be conducted to identify potential hazards, risk of exposure, and determine the likelihood of risk to the environment and non-target organisms. The second tier is then for the product to be approved for use within the individual member states. Following the approval by the member states, EFSA are then responsible for conducting a peer-review of the risk assessment (EFSA, 2014). Due to increasing demand and agricultural intensification, production of naturally occurring and synthetic plant protection products (PPPs) has increased, causing a rise in regulatory interest of the use and potential application of these compounds to the soil. To protect human and environmental health, it is imperative to be able to predict how the compound will behave in the soil profile prior to releasing it into the environment. Environmental fate is the study of the behaviour of compounds once they enter the soil profile and analyses the likelihood that they will be exposed to living organisms. One such pathway is through leaching down the soil profile into potable water supplies (Arias-Estévez et al., 2008). Plant uptake has been proposed as a sink pathway that could help reduce the concentration of PPPs being transported (Lamshoeft et al., 2018). As it cannot be known how much a newly developed product would leach from the soil, it cannot be released to the environment prior to assessment. To conduct these studies, environmental fate models have been developed to simulate example locations and indicate how likely it is that the underlying groundwater will be polluted. A plant uptake value was inputted into these models based on a statistical relationship derived from laboratory experiments. Previously, within these models the statistical relationship was used to determine how much of the compound was taken up by the plants. However, a lack of agreement and reproducibility of results within the scientific community has led to this approach being abandoned, with research calling for a new method to be established. This thesis aims to test the main assumptions used within the current laboratory framework and improve the realism of the test system. This work will then be linked with modelling experiments to determine the effects of plant uptake on the leaching output on simulated soil profiles.

Entry into mitosis is mediated by the switch-like activation of CDK/Cyclin B. This needs to be tightly regulated to ensure that mitosis occurs at the right time in the cell cycle after completion of DNA replication. Mathematical models of this feedback system predict that mitotic entry is bistable, meaning that cells can only exist in either an Interphase or mitotic state, but not arrest in intermediate stages. A positive feedback in the Cdk1 activation loop that results in the removal of inhibitory phosphorylation sites on Cdk1, T14 and Y15, has been proposed to generate different thresholds for transitions between Interphase and M phase. Recent work on mitotic control also suggests that the major Cdk1 counteracting phosphatase, PP2A-B55, needs to be inhibited as cells progress into mitosis. This requires Gwl kinase generating an additional feedback loop. In vitro experiments suggest that the regulation of PP2A-B55 By Gwl also constitutes a bistable switch. How Cdk1 activation and PP2A inactivation are interlinked in the G2/M switch and how either pathway contributes to bistability remains to be explored in vivo. We have addressed these questions by combining quantitative cell biology assays and mathematical modelling that includes both kinase and phosphatase regulation using quantitative assays in human cell lines to estimate numerical parameters for the model. Our results show that the mitotic switch displays hysteresis and bistability and that phosphatase inhibition and kinase activation are each suficient to maintain it; only abrogation of both mechanisms eliminates a stable barrier between interphase and mitosis. Loss of bistability has as consequences for mitotic progression and cell proliferation, as they fail to stabilize the metaphase state and are also unable to initiate progression through mitosis toward cell division and G1. Our model predicted a third stable steady state that represents a prophase arrest in between mitosis and interphase. We verified experimentally this prediction of the model which supports the predictive value of it. Further work explored the contribution of other phosphatases to the G2/M switch system and a synergistic effect between Gwl kinase and inhibition of Cdk4. Overall, this work presents a comprehensive quantitative model for the mammalian mitotic entry switch and sheds new light on its regulation.

Zinc finger protein 516, or ZNF516, is a largely understudied protein. Germline homozygous loss of the murine ortholog, ZFP516, has been previously reported to lead to embryonic lethality in mice. However, the cause of this lethality has not been investigated. With the aid of newly developed Zfp516 knock out mouse line in our laboratory I characterised the phenotypes of Zfp516 knock out animals and concluded that severe and 100% penetrant congenital heart defects were the most likely cause of embryonic lethality in Zfp516-/-. I reported that heart defects were not due to selective Zfp516 deletion in cardiomyocytes, and that only some defects were recapitulated when Zfp516 was deleted in Isl1-expressing cells. I further described spatiotemporal expression of Zfp516 during development and showed that Zfp516 ablation at E9.5 leads to upregulation of genes involved in neuronal migration, growth and synapsogenesis. Moreover, I found genetic synergy between Zfp516 and Sox10 during development. Next, I characterized phenotypes associated with deletions of and polymorphisms in human ZNF516 and adult heterozygous Zfp516 knock out mice. I discovered that manifestation of phenotypes in Zfp516+/- mice was largely gender dependent, and that our mouse model phenocopied some of the phenotypes seen in humans. Lastly, I reported that targeting ZNF516 in HCT116 with CRIPSR/Cas9 can result in chromosome arm truncations. I therefore suggested a screening method for future studies that could preselect clones with correct genotype and without arm truncations before further cell expansion for subsequent downstream experiments.

Cellular senescence is a cell fate characterised by the permanent arrest from the cell cycle with an altered, mainly proinflammatory, secretory phenotype. Although, a positive process in development and regeneration, the accumulation of senescent cells, as seen in tissues of aged humans, can cause many negative effects. Within skeletal muscle the resident stem cells, Pax7+ve and CD56+ve satellite cells/muscle precursor cells (MPCs), are essential for muscle maintenance and repair following damage. Impairments either in their inherent behaviour, such as by being senesced, or by being negatively influenced by their environment, such as by the senescence associated secretory phenotype (SASP) of local senescent cells, may impair muscle regeneration in later life and contribute to the age-related loss of skeletal muscle mass, sarcopenia. Furthermore, regeneration is also dependent on the interaction of MPCs with other key cell types in muscle, principally fibroblasts, therefore understanding the senescent characteristics of both cell types is essential to gain further insight into skeletal muscle ageing. Currently, little is known about the senescent phenotype of human skeletal muscle derived cells. The aim of the work in this thesis was thus to investigate the senescent phenotype of human primary skeletal muscle derived MPCs and fibroblasts in a primary cell culture system. Once characterised this phenotype would then be used to investigate senescence within skeletal muscle cells obtained from very old and frail hip fracture patients. MPCs were sorted using CD56+ve magnetic beads after extraction and expansion from muscle biopsy samples taken from young healthy male volunteers (n=6, 22±1 years). Fibroblasts were derived from the negative fraction. In the first series of experiments cumulative population doublings (CPD) across multiple passages was used to induce cellular replicative senescence (RS). In the second, cells were treated with Doxorubicin (DOX; 0.2 μM) for 24h to induce a senescent phenotype and then investigated at a number of timepoints up to 35 days post treatment. In both approaches, senescence was determined by the presence of senescence-associated β-galactosidase (SA β-gal) and the protein expression of p16, together with the absence of proliferation marker, Ki67. With expression determined using custom built cell-by-cell image analysis programs. The mRNA levels of established SASP factors (e.g. PAI-1, TGF-β1, IL-8 and IGFBP-3, amongst others) were also measured as was the differentiation capacity of MPCs for myogenic fusion. In the RS study only two of the six MPC populations remained highly myogenic through to RS. Senescence of these two populations of MPCs was shown by increased number of SA β-gal-positive and p16 expressing cells, and decreased Ki67 expression. The senescent, late passage MPCs showed a markedly reduced myogenic fusion index, compared to proliferative MPCs. The fibroblast experiments were terminated at the same time as the equivalent MPC cultures reached senescence. At this point, none of the fibroblast populations had reached RS, as evidenced by no changes in SA β-gal or p16, although there was a trend for slower population doubling times when compared to early passage rates and significantly decreased Ki67. When both MPCs and fibroblasts were exposed to DOX for 24h, senescence was demonstrated shortly after, with 99% of both cell types being SA β-gal-positive four days after DOX treatment, Ki67 expression markedly reduced and MPCs being less able to fuse and form myotubes. SASP factor (i.e. IGFBP7, PAI-1, IGFBP3) mRNA expression from this 4-day time point until 35 days post-DOX treatment varied over time and magnitude in both MPCs and fibroblasts. MPCs isolated from skeletal muscle of old (82±9 yrs) hip fracture patients showed significant differences in senescence marker and SASP factor expression, compared to young donor cells after a similar time in culture. In conclusion, RS is hard to achieve in human primary muscle cultures because even highly purified MPCs populations can eventually be overrun by initially small populations of fibroblasts, whilst the sorted fibroblast cultures, from the same biopsy samples, appear to take a much longer time to reach RS than the MPCs. DOX treatment allows a synchronised and rapid induction of senescence in both cell types eliminating these problems. Here the results showed both varying temporal and quantitative responses of the different SASP factors between the two cell types. The data highlight the complexity and dynamism of the senescent state within different cell types derived from human skeletal muscle. When myoblasts are taken from a very elderly frail human population undergoing hip fracture surgery, markers of senescence are elevated compared to young myoblasts after similar exposure to cell culture. These findings suggest that cellular senescence may start to be an issue within skeletal muscle cells very late in life. Therefore, this work could contribute to supporting research into healthy human ageing by investigating strategies which prevent the accumulation, or remove, these senescent cells.

Meiosis is a specialised cell division producing the haploid gametes required for sexual reproduction. A key function of this process is to generate genetic diversity through recombination and independent assortment of homologous chromosomes. Delineation of gene expression during this process has been challenging due to the heterogeneity of testis tissue and the lack of faithful in vitro models. Here we jointly analyse a single-cell resolution transcriptomic dataset of over 20,000 cells from both wild type and mutant mice. We use dimensionality reduction methods to infer latent components of variation that represent transcriptional programmes, mutant specific pathological processes, and technical effects. This approach simultaneously soft clusters cells, infers corresponding groups of co-expressed marker genes, and imputes sparse, noisy gene expression. We were also able to infer, de novo, transcription factor binding motifs for each component, revealing a general switch at the meiotic divisions. We facilitate access to this resource by providing an interactive website testisatlas.ml. The high-resolution delineation of gene expression during the spermatogenic programme provides high-resolution clues to the function of understudied genes. One such gene, Zcwpw1, is highly co-expressed with Prdm9, and has domains capable of recognising the unique combination of histone marks that PRDM9 deposits (H3K4me3 and H3K36me3). By using a human in vitro system of Zcwpw1 co-transfection with either human or chimp Prdm9, we show that PRDM9 causes the recruitment of ZCWPW1 to its binding sites. This recruitment is stronger than for sites with H3K4me3 alone and is CpG dependent. Male Zcwpw1-/- mice have completely normal double strand break positioning, but severe repair and synapsis defects leading to complete testicular azoospermia. Although PRDM9’s effect of DSB positioning remains intact, PRDM9’s effect of aiding synapsis appears to be abolished, with persistent DMC1 signal - most dramatically at the most strongly PRDM9 bound hotspots.

Collective migration in epithelial tissues is required for many developmental processes, wound healing, and cancer, and is conserved across vertebrates. It is a process that has been difficult to study for 3D epithelia in vivo as the large timelapse datasets required present significant computational challenges for quantitative analysis. An important example of collective migration in an epithelial context is the migration of Anterior Visceral Endoderm (AVE) cells in the mouse embryo. From E5.5, this subgroup of cells in the visceral endoderm monolayer migrate from the distal tip, specifying the anterior-posterior axis by patterning the underlying epiblast which will later give rise to the embryonic lineage. AVE induction and migration have yet to be fully characterised and understood, and provide an attractive in vivo system to study the biological function of collective migration due to the relatively small cell numbers at this stage and the experimental accessibility of the visceral endoderm tissue. In this thesis, a new computational framework is developed to analyse and extract quantitative information from light sheet fluorescence microscopy (LSFM) 3D timelapse datasets of E5.5 mouse embryos during AVE migration. The proposed framework initially isolates the migratory cell movement, and secondly extracts the apical surface data to simplify the 3D volume to a 2D characterisation problem. Applying this computational framework on LSFM datasets from 3 fluorescent reporter mouse lines, I demonstrate the first comprehensive tracking of individual cells and cell divisions in the entire visceral endoderm, the first study of cell division angles, and the movement of cells in the underlying epiblast. Altogether these results reveal a model of AVE migration as an active process that drives large-scale tissue deformation and changes in morphology independent of cell proliferation in the VE, and presents evidence that suggests the A-P axis could be pre-defined.

Computational modelling has become a vital component of biological research, offering the ability to predict biological responses to external stimuli and influencing experimental setup. Modelling has been successfully used to investigate cell growth of both cancer and healthy cells. The modelling of cancer development and treatment response have delivered patient benefit with respect to radiation treatment scheduling and dose fraction size. The modelling approach employed typically varies depending on the desired result with continuous, Boolean and hybrid models all used to investigate cellular processes. This thesis presents a computational model of cell cycle progression controlled by the cyclin proteins, and ultimately the genes that regulate their expression. This interaction network is modelled using a Boolean network with the concentration of the cyclin proteins calculated by ordinary differential equations, thus creating a hybrid model. This cell cycle network was also expanded to include interactions between cell cycle regulators and the p53 network, responsible for cell cycle arrest and the triggering of DNA damage repair. DNA damage included in this model was induced via radiation exposure, and thus can be incorporated at any point in the cell cycle, determining the specific repair mechanism response. This hybrid model was then converted to calculate the absolute number of protein molecules as cells progress through the cell cycle. This quantisation was achieved through the creation of a standard curve relating western blot band intensity, to the weight of the protein in the sample. After synchronising a population of Human Umbilical Vein Endothelial Cells (HUVEC), a time series was collected allowing for a direct comparison between the concentration of the cyclin proteins obtained from the samples, and the calculated concentrations from the model. These results showed strong agreement between the computational data and the biological results allowing for the computational model to be converted to absolute protein concentration, with the maximum protein concentration obtained by cyclin B, on the order of 8E6 molecules per cell. The expansion of the model to include the radiosensitisation due to the treatment with citrate capped gold nanoparticles was achieved by collecting in vitro data from gold nanoparticle treated HUVEC cells before irradiation, HUVEC’s were used as they contain fully intact cell cycle damage checkpoints and a functioning p53 gene. Both the short and long term effects where investigated, showing a clear increase in damage yields at all radiation doses tested and a decrease in the long term survivability. This result was incorporated into the model providing a complete cell cycle progression model that includes DNA damage, repair, cell cycle arrest and radiosensitisation, giving a more complete understanding of the processes in a single model.

Morphogens are secreted signalling molecules that disperse from their region of production to form concentration gradients in the adjacent tissue. By inducing cellular responses in a concentration-dependent manner, morphogens organise pattern formation in various tissues. How extracellular protein gradients form has remained controversial, especially in epithelia. In this work, I used a forward-engineering approach to assess the minimal requirements for diffusion-based gradient formation by transforming the otherwise inert green fluorescent protein (GFP) into an in vivo morphogen. I found that, on its own, GFP cannot form a detectable gradient in the Drosophila wing disc, the epithelial pouch that gives rise to the adult wing. However, GFP does form a gradient in the presence of binding partners engineered from anti-GFP nanobodies. The gradient can be formally described by a diffusion-degradation model that considers GFP leakage into the hemolymph, the larval blood, and GFP reentering into the tissue. To test whether GFP gradients can faithfully provide positional information, I substituted GFP for Dpp, a bona fide morphogen, in the presence of Dpp receptors engineered to be responsive to GFP. GFP was able to partially rescue growth and patterning of the wing disc. However, GFP reentering the tissue from the hemolymph seemed to impair the patterning performance of the gradient. While increasing receptor number improved GFP retention, it also reduced the signalling range of the gradient. I found that the coexpression of low-affinity binding partners mimicking extracellular matrix components was able to both increase the gradient range and provide sufficient ligand retention. My results suggest that extracellular protein gradients can form by diffusion in epithelial tissues and that in a system where signalling receptors are the only binding partners, ligand leakage and signalling range are opposing constraints that can, however, be overcome by the coexpression of low-affinity binding partners.

Axial development in amniotes occurs from a caudal growth zone. Previous work in chick suggested that Hensen’s node contains resident, self-renewing stem cells that generate axial and paraxial tissues along the entire axis. However, it is not clear whether the node represents a stem cell niche that can induce and maintain stem cell behaviour (‘instructive’ stem cell niche). Through a series of heterotopic grafts we show that the node can specify resident cell behaviour in epiblast cells that do not normally enter the node. Single cell transplantation reveals that these can self-renew and contribute to the axis, suggesting that the node can induce and maintain stem cells and therefore acts as an instructive stem cell niche. However, the node is not able to induce any epiblast in this way. When made to enter the node, epiblast from an older donor, while able to contribute to axial and paraxial structures and to cells with resident behaviour, does not always express genes appropriate to its new location, as revealed by in situ hybridisation. Single-cell RNA-seq of identified, individually hand-picked resident cells from control and experimental epiblast grafts (epiblast not normally fated to enter the node) was used to find genes whose expression correlates with cells that are either more or less competent to respond to the node. The RNA-seq also reveals that irrespective of epiblast origin, a cell’s molecular signature appears to reflect its anterior-posterior position within the regressing node. More posterior cells express genes associated with late cell-cycle markers, suggesting they are in the process of dividing. Furthermore, fate maps of regressing node reveal that only posterior cells continue to contribute to cells with resident behaviour as axial development continues, thus raising the possibility that ‘longer-term’ dividing resident stem cells reside in the posterior part of the regressing node.

Cell division is a tightly coordinated process that involves doubling of the cellular content and its separation into two genetically identical daughter cells. In G1 phase of the cell cycle activation of the G1/S transcriptional wave initiates entry into S phase, during which the genomic DNA is replicated, and thereby activation of G1/S transcription commits cells to a new round of division. In the budding yeast, Saccharomyces cerevisiae, this transcriptional wave is controlled by two transcription factors, MBF and SBF. Previous work suggested that changes to the chromatin state through histone acetylation has an important role in regulating G1/S transcription. In this thesis I show that the HAT Gcn5 and HDAC Rpd3, which control histone acetylation, have a limited contribution to G1/S transcriptional regulation. The G1/S transcription factors MBF and SBF are highly similar, yet MBF regulates G1/S transcription through repression of its targets, and SBF through activation. My findings suggest that the local chromatin environment is unlikely to explain their opposing mechanisms. During S phase the process of DNA replication causes an imbalance in gene copy number between early- and late-replicating genes, which could alter their transcription levels. It was previously reported that S. cerevisiae exhibits gene expression homeostasis, defined as the buffering of transcription levels against DNA copy number changes during S phase. My work confirms a previously indicated requirement of the protein Tos4 in this process, and suggests this is dependent upon its binding to HDACs. I also demonstrate that Tos4, and therefore gene expression homeostasis, confers a fitness advantage. Loss of Tos4-dependent gene expression homeostasis may increase dependence upon components of the gene expression and protein production pathways. Overall this work provides new insights into transcriptional regulation at distinct cell cycle stages: in the contribution of the chromatin state to G1/S transcriptional regulation, and the control of transcription during DNA replication.

PADI6 is a maternal effect gene (MEG) that is expressed throughout oocyte maturation and appears to be critical for early embryo development and embryonic genome activation (EGA). It is described as the fifth member of the oocyte-specific subcortical maternal complex (SCMC), which is involved in transcription, epigenetic regulation and cytoskeletal organisation. PADI6 is also essential for the maintenance of critical structures in the oocyte called cytoplasmic lattices that are believed to function as storage sites for maternal transcripts and proteins. It is hypothesised that PADI6 is involved in translational regulation in the oocyte in preparation for EGA. Experiments were conducted on bovine oocytes as a physiologically relevant model for functional investigations of the role of PADI6 and its interaction with the SCMC during human oocyte maturation and embryo development. Initial studies characterising the expression of PADI family genes in different bovine somatic tissues showed that PADI6 expression was restricted to the oocyte and early embryo. PADI6 expression patterns were mapped across bovine oocyte meiotic maturation following the in vitro maturation (IVM) of oocytes and during preimplantation embryogenesis in embryos derived following in vitro fertilisation and embryo culture to the blastocyst stage. Quantitative PCR analysis of Smart-seq2 cDNA libraries from individual oocytes and embryos indicated that PADI6 was highly expressed in the germinal vesicle (GV) and metaphase II (MII) oocytes and embryos prior to EGA (8-16 cell stage in bovine) and thereafter was significantly reduced by the blastocyst stages. This expression pattern was in marked contrast to the tissue distribution of other members of the PADI family PADI1-4 which showed no expression in oocytes and early embryo development in the bovine model. The expression patterns of the other members of the SCMC – namely KHDC3L, NLRP5, OOEP and TLE6 were also mapped across oocyte maturation and preimplantation embryo development. Functional analysis of the role of PADI6 during bovine oocyte maturation was conducted using a validated system for the microinjection of double-stranded short-interfering RNAs (dsiRNAs) into cumulus-enclosed GV oocytes (n=17) relative to duplex buffer injected (n=22) and scrambled siRNA (n=19) controls. Gene knockdown (KD) was assessed following 24 hr of IVM and was found to reduce the expression of PADI6 by an average of 74±3.6% (p<0.05). KD of PADI6 did not affect oocyte meiotic progression to metaphase II or cumulus expansion after IVM. However, targeted real-time PCR highlighted 7 transcripts (DNMT3A, DPPA3, PLAGL1, PRDX1, TRIM28, ZP1 and ZFP57) associated with oocyte and embryo developmental competence and epigenetic regulation that were significantly dysregulated in PADI6KD oocytes. Full transcriptome analysis by RNA sequencing was conducted on 6 individual PADI6KD MII oocytes and 12 controls. Bioinformatic analysis identified 452 differentially expressed genes (DEGs): 165 genes were downregulated and 287 genes were upregulated following PADI6 KD. Network analysis revealed that 61 DEGs were directly or indirectly linked to PADI6. A diverse number of transcripts were involved in RNA processing and translation. Genes from 7 downstream networks and 10 downstream networks were associated with infertility and organisation of the cytoskeleton, respectively. Finally, KD of PADI6 appeared to impact oocyte metabolism as the regulation of 5 amino acids was altered in PADI6KD oocytes compared to controls. Overall, the results of this thesis suggest a role for PADI6 in translational control of maternal transcripts which may impact on oocyte developmental competence in the bovine.

Background: The alteration of Epidermal Growth Factor Receptor (EGFR) plays a major role in the development of Head and Neck cancer. However, the only FDA approved EGFR target for this neoplasm is the drug Cetuximab. Cetuximab can only be used where there is an over-expression of EGFR. Cetuximab works by competing with the ligands Epidermal Growth Factor (EGF) and Transforming Growth Factor alpha (TGFα) to bind the extracellular domain of EGFR. Inhibition of this ligand binding blocks the activation of subsequent signalling pathways. It is known that not all Head and Neck cancers over-expressing EGFR respond to Cetuximab. It has been postulated that some of these tumours have additional genetic lesions. Resistance to Cetuximab has been associated with the genetic lesions to EGFR mutation and Ras mutations. These mutations are more common in Asian than Caucasian patients. More recently the Tyrosine kinase inhibitors (Gefitinib and Erlotinib) have been investigated as drugs to target tumours with EGFR mutation. These inhibitors bind to the cytoplasmic domain of the EGFR and compete with ATP and inhibit phosphorylation of the receptor. Another inhibitor of interest is PD98509 which is a selective ERK1/2 inhibitor and may be used in targeting tumours with a Ras mutation. None of these inhibitors has received FDA approval for Head and Neck cancer treatment. Aim: The main aims of the project were to investigate the effect of the growth factors EGF and TGFα and the inhibitors; Gefitinib, Erlotinib and PD98059 on cancer cell lines obtained from Asian head and neck cancer patients. The first area of study was to investigate the effect of the growth factors EGF and TGFα on cell proliferation, morphological changes, EMT marker expression, single cell migration and collective cell migration. Once the initial data had been collected a second part was to determine whether the inhibitors; Gefitinib, Erlotinib and PD98059 could block the stimulatory effect of EGF and TGFα in these assays. Methods: Three tumour cell lines of Asian origin were used in the study and a normal skin keratinocyte as a control. Cell proliferation was measured using an automated cell counting system. Migration was studied using a cell scatter and cell scratch assay. The movement of the cells was documented by using photomicrographs. Epithelial to Mesenchymal Transition (EMT) and protein expression were determined using immunofluorescence and SDS-PAGE and Western blotting. Results: Cell proliferation data indicated that the growth factors did not stimulate the proliferation of the cells investigated. Addition of the growth factors for 48 hours induced an EMT like morphological change, individual cell migration in the scatter assay and collective cell migration in the scratch assay. However, the specific markers of EMT (E cadherin and vimentin) as measured by Western blotting and immunofluorescence showed no change in intensity following the addition of the growth factors for 48 hours. Pre-treatment of the cells for one hour with the tyrosine kinase inhibitors (Gefitinib, Erlotinib) and the MAPK inhibitor (PD98059) inhibited these morphological changes, and individual cell migration. However, only the Tyrosine kinase inhibitors inhibited collective cell migration. PD98059 partially inhibited collective cell migration of the TYS and HaCaT cell lines and fully inhibited cell migration of the HSG and AZA1 cell lines. Analysis of the MAPK pathway using the expression of phosphorylated MAPK was investigated by Western Blot and Immunofluorescence. Western blot data indicated an increased expression of phosphorylated MAPK202/204 was observed in cells treated with both EGF and TGFα for 24 hours. At the 24hour time point, tyrosine kinase inhibitors (Gefitinib, Erlotinib) fully inhibited the MAPK202/204 phosphorylation. In contrast, in ERK1/2 inhibitor (PD98059) treated dishes, phosphorylated MAPK 202/204 was still expressed which was likely to be the same level of expression as serum free treated dishes. Conclusion and Possible Clinical Impact: The data indicates that the growth factors EGF and TGFα both stimulated single and collective cell migration. The data lead us an unexpected answer as to whether single cell migration and collective cell migration utilised the same signalling pathways. The data indicated that these are separate pathways; the MAPK pathway appears to be the only pathway responsible for morphological change and single cell migration in cell lines used in this project. The role of MAPK signalling pathway in collective cell migration appears to be cell line dependent and MAPK alone might not be responsible for collective cell migration. The ability of the inhibitors may in the future lead to a more personalised therapy for patients in Asian countries and more clinical trials across a diverse group of patients is needed.

Global nitrogen (N) deposition has risen steeply since the mid-19th Century and is forecasted to rise further. This increased flux of N to ecosystems is increasingly considered as a worldwide driver of environmental change. Impacts to botanical communities have been shown across a range of ecosystems and regions, but evidence on consequential impacts to further trophic levels is currently limited. Our understanding of how N affects plant-pollinator interactions is relatively poor. Declining trends in pollinating insects have been widely reported during recent decades and a key factor in this is losses of floral resources. The research chapters presented in this thesis seek to address the wide knowledge gap of how N affects pollinators by investigating how soil N affects 1) the composition of floral functional traits in a plant community, 2) intra-specific variation in key floral functional traits: phenology and nectar, and 3) the potential impacts to plant-pollinator interaction networks. This research was undertaken using the Park Grass Long-term Experiment (PGE) at Rothamsted Research, UK, a long-term nutrient-enrichment experiment. The study found extensive impacts of N on the provision of floral resources. Floral resources were lost with N-enrichment, with morphologically specialised floral units particularly threatened. Flowering phenology of plant species was influenced by N-enrichment with a notable contrast between early-season species, which had an extended flowering duration, and peak-season species that had a reduced flowering duration. Responses of nectar traits varied across species but were shown to respond to N applications. The observed pollinator visitation networks were impacted by the soil treatments; N additions typically led to Diptera-dominant communities and networks that were potentially less robust to stochastic events. The research project provided evidence that soil N can act as an underlying driver of plant-pollinator networks and should be considered as a factor in pollinator ecology and trends.

Bumblebees are important pollinating insects for many crops and wild flowering plants. Due to multiple factors that include agricultural intensification, many populations have experienced severe declines and several species are now listed in the UK Biodiversity Action Plan. Bumblebees rely exclusively on nutrients derived from pollen and nectar, and nutrition is crucial for the development and activation of the immune system. Four microbial pathogens are known to infect bumblebees, and previous work has shown a variety of interactive effects between host nutritional status and pathogen epidemiology. Of the studies that investigated nutritional immunology in bumblebees, almost all have been carried on a limited range of common species because others, notably long-tonged species, have been difficult to rear in captivity. As a result, we lack knowledge of the nutritional needs of many declining species. Here, I test methods for rearing two long-tongued species in captivity and investigate the effect of diet on incipient colony development. A new technique to encourage oviposition and brood care was trialled successfully, and I observed interspecific differences between bumblebee queens on each diet. I also find evidence to support the idea that the nutritional content of pollen, not only plant species diversity, determines bumblebee health. In the wild, floral resources play an important role in regulating bumblebee populations and that of their pathogens. The abundance, diversity and composition of floral resources vary dramatically across the landscape, but studies investigating the effects of floristic composition on host-pathogen dynamics in bumblebees remain scarce. I compare floral resource availability, bee health and pathogen prevalence across three important UK habitats: farmland, gardens and nature reserves. I found that gardens contained the greatest species richness of flowers and had the largest, healthiest bees, despite increased parasitism. Farmland consistently provided the least floral resources, but habitats were complementary to each other in resource provision. I observed interspecific differences in bee health across habitats and report on the prevalence of bumblebee pathogens C. bombi, N. bombi and N. ceranae. Floral resources have a substantial effect on bumblebee health and pathogen dynamics, but these effects appear to vary between species. To support taxonomically diverse bumblebee communities, it is essential that the nutritional needs of a wider range of bumblebee species are considered.

SAG21/AtLEA5 is an Arabidopsis thaliana gene belonging to the late embryogenesis associated (LEA) protein family. Prior research identified tissue specific, spatial and temporal expression of the SAG21 promoter. A 1685 bp upstream region of SAG21 contains cis elements that act as binding sites for several transcription factor families. The aim here was to investigate the transcription factor network regulating SAG21 in response to senescence, development and stress responses. To investigate if transcription factors identified in a yeast- 1-hybrid screen regulate SAG21, expression of SAG21 was analysed in transcription factor mutants of WRKY 15, 63 and 67 under abiotic stresses known to elicit SAG21 expression. WRKY15 was not essential for induction of the SAG21 expression. WRKY63 functioned as a negative regulator of the SAG21 expression under drought stress. WRKY67 may act as a positive or negative regulator of SAG21 under salt stress. SAG21 promoter deletion-GUS reporter constructs were used to understand the role of the cis-elements in regulating senescence and wounding responses. SAG21 showed different expression patterns in floral organs of the deletion constructs. The role of kinetin in regulating senescence and wounding responses was studied in young and older cotyledons and in four-week old wounded leaves. Kinetin inhibited senescence-related and wound-induced SAG21 expression. Induction of SAG21 expression by wounding was dependent on the age of wounded rosette. Overexpression of SAG21 under its own promoter in optimal conditions produced similar primary root length, more lateral roots, and greater lateral root length compared to overexpression of SAG21 from the 35S constitutive promoter. Under oxidative stress SAG21 transgenic over expressors lines produced increased primary root length under all concentrations of H2O2 whereas the number of lateral roots and length of lateral roots was greater on lower concentrations of H2O2.Together, these data suggest effects of specific SAG21 promoter regions on development and stress responses.

Pluripotency is defined as the capacity to give rise to all cell types of the embryo proper. It arises in the early mammalian embryo but is lost after a short period of time as cells differentiate and become committed to different lineages. Prior to implantation, mouse epiblast cells enter the pluripotent naïve state, which can be captured in vitro in the form of embryonic stem cells. These cells are characterized by a capacity for indefinite self-renewal, and the ability to re-enter normal development upon being returned to the naïve epiblast. A complex transcription factor network promotes this state. Overexpression of one of many of these factors leads to stabilisation of the naïve state, with enhanced self-renewal and reduced spontaneous differentiation. However, the transcription factor Oct4 must be maintained within a tight window of expression; depletion results in extraembryonic differentiation, while overexpression also results in exit from pluripotency. Oct4 was identified as a protein expressed in the early embryo and in germ cells, and was subsequently discovered to be essential for the establishment of the naïve epiblast. In vitro studies determined that loss of Oct4 in ESCs induced trophoblast differentiation. Meanwhile, overexpression of Oct4 led to differentiation of ESCs, and constitutive expression of Oct4 was not sufficient to replace any of the extrinsic factors required for ESC self-renewal. Despite these dramatic phenotypes, the essential role of Oct4 remains unclear, further complicated by the finding that a reduced level of Oct4 promotes self-renewal at the expense of differentiation capacity. In this work, I generated a novel Oct4 fusion protein capable of rapid inducible degradation in order to study the immediate responses to removal of Oct4. This system utilizes the auxin responsive degradation domain of the Arabidopsis thaliana IAA17 protein to recruit a transgenic F-box protein Tir1 on addition of the small molecule auxin to the culture medium. Subsequent ubiquitination by the endogenous SCF complex leads to rapid proteolytic degradation of the Oct4 fusion protein, resulting in loss of detectable protein in as little as two hours. This system allows the study of immediate responses to loss of Oct4 in contrast to conventional depletion systems in which Oct4 levels decay over a protracted period making it difficult to disentangle direct and indirect effects. I established that several pluripotency-associated genes require Oct4 for their transcription. RNA levels of these factors decrease rapidly on depletion of Oct4, prior to significant changes in the expression of other key pluripotency factors such as Nanog and before protein levels of other factors can change dramatically. Furthermore, I established that the presence of Oct4 antagonises chromatin binding by a naïve transcription factor, revealing a possible mechanism by which increased Oct4 levels are detrimental to the naïve state. Together, these findings may be sufficient to explain the simultaneous requirement for, and antagonistic activity of, Oct4 in naïve pluripotent cells. I also found that an ESC level of Oct4 facilitates cell identity transitions. Cells constitutively expressing Oct4 and differentiated in vivo could be reverted to naïve pluripotency in vitro through the application of defined naïve pluripotency growth conditions. Additionally, in the course of these experiments I examined the phenotypic abnormalities that occur in mouse embryonic development under continuous expression of Oct4. In keeping with previous work, we observed abnormalities in limb development and in the skin. We also observed exencephaly in a number of embryos. In conventional exit from pluripotency, female cells must inactive an X chromosome in order to balance X-linked gene expression between males and females. This is achieved via expression of Xist from a single X chromosome, where it orchestrates chromosome-wide silencing. I show that Oct4 plays an important role in the regulation of Xist during the exit from pluripotency. Normally, Oct4 expression persists after the downregulation of most naïve transcription factors during early differentiation. I propose that this allows Oct4 to antagonise expression of the Xist, and thus ensure proper control over X chromosome inactivation. Together this work focuses on the dual roles of Oct4 in regulating both pluripotency and differentiation. I address the contradictory phenotypes relating to altered expression of Oct4, and establish a unifying theory to explain them. I put forward evidence that Oct4 promotes cell fate transitions by regulating naïve transcription factors. I propose that the environment plays a key role in determining cell identity, while Oct4 acts to maintain plasticity by preventing cells from being trapped within otherwise stable states.

Meiosis is a conserved eukaryotic cell division that increases genetic diversity in progeny. During meiosis, homologous chromosomes pair and undergo reciprocal exchange, called crossover. Meiotic recombination initiates from DNA double strand breaks, which are repaired using either sister or homologous chromatids as templates. When meiosis occurs in heterozygous (hybrid) organisms, interactions between homologous chromosomes have the potential to generate mismatched DNA structures. Several proteins with roles in DNA mismatch repair (MMR) are known to influence meiotic recombination in several eukaryotes. In Arabidopsis thaliana this includes three MSH heterodimers (MSH2-MSH3, MSH2-MSH6 and MSH2-MSH7) that recognise mismatched nucleotides and have demonstrated roles in repressing meiotic crossovers in hybrid plants. To further investigate the meiotic function of MMR genome-wide, I generated a series of msh2 mutant introgression lines in three different genetic backgrounds – Ct-1, Ler-0 and CLC – which have varying patterns and levels of polymorphism. Consistent with the function of MSH2 as a hybrid-specific anti-recombinase, I observed significant crossover increases in the chromosome arms in all three hybrid msh2 mutants. However, I also found evidence for accession and region specific effects of msh2 on crossover frequency. For example, crossovers appeared to decrease over centromere proximal regions in msh2 compared to wild type. A genotyping-by-sequencing experiment was performed to generate genome-wide crossover maps in two Arabidopsis hybrids, with and without MSH2 function. This revealed that total crossover number remained unchanged in the MMR-deficient hybrids, whilst crossovers redistributed into regions of reduced polymorphism density. This relationship was counter to my expectation that MMR would repress crossovers most strongly in divergent regions. However, this relationship was observed across varying physical scales, from 1–100 kilobases. This confirms a positive relationship between polymorphism and meiotic crossovers in Arabidopsis hybrids, and reveals a novel role for MSH2 in mediating this effect. In addition to the investigation of MSH2 in meiotic recombination, I present an analysis of the genome-wide distribution of MutS homolog 4 (MSH4). MSH4, and its binding partner MutS homolog 5, evolved from their ancestral role in MMR and now function exclusively to promote meiotic crossovers in the ZMM pathway. I present a genome-wide chromatin-immunoprecipitation-sequencing analysis of the binding profile of MSH4, and analyse its distribution at varying scales. This has revealed novel relationships between MSH4, the meiotic cohesin subunit REC8, and features of the chromatin and recombination landscapes. Together, these results advance our understanding of meiotic recombination in plants, and raise further questions about the regulation of crossovers in eukaryotes more broadly.

The establishment of the body axes has been well-studied in Drosophila melanogaster, and is known to occur prior to fertilisation in the developing egg chamber as a result of crosstalk between the germline and the surrounding somatic epithelium. Whilst it is known that the formation of the anterior-posterior and dorsal-ventral axes occurs due to the establishment of oocyte cell polarity and repositioning of the oocyte nucleus, respectively, the nature of the signals sent from the soma to trigger these two events, and the method through which these signals are transduced, remain unclear. In an attempt to elucidate these underlying mechanisms, studies of both the follicle epithelium and germline were undertaken. Posterior follicle cells are known to be responsible for the formation of both body axes through signalling, but very few factors have been identified which perturb this process. The transcriptomic method Targeted DamID was used to quantify gene expression in these cells, compared to their lateral follicle cells counterparts. This resulted in the identification of 140 genes which were either uniquely expressed or upregulated in posterior follicle cells. Specific subsets of these genes were knocked down via RNAi, and the impact of this knockdown on axis formation determined. However, an attempted validation of these observed phenotypes through CRISPR-based mutagenesis or analysis of extant alleles failed to recapitulate many of the expected defects. Subsequent experiments have attempted to further narrow this list of differentially-expressed genes to provide a more specific profile of the posterior follicle cell transcriptome. In parallel, a forward genetic screen was undertaken to identify novel factors required for axis formation in the germline. Whole genome sequencing of a mutant which fails to establish cell polarity identified the causative mutation as a SNP in the enzyme CDP-diacylglycerol synthase (CdsA), required for the production of a number of phospholipid species. An initial characterisation of this mutant phenotype suggested the activity of CdsA is required to recruit the polarity kinase Par-1 to the posterior of the oocyte. A further analysis established that the production of phosphatidylglycerol-based species is dispensable for oocyte polarity establishment, suggesting that the phenotype of cdsa alleles is produced due to its impact on phosphatidylinositol synthesis. Additionally, an analysis of the localisation of the PIP2 derivative diacylglycerol using a fluorescent biosensor revealed its localisation was not uniform within the oocyte. To better understand potential downstream functions of oocyte cell polarity in the context of axis formation, several previously-identified Par-1 phosphorylation target sites were examined. The six sites of interest were located in the eIF4E-binding protein Cup, which is required to repress translation of osk mRNA until it is posteriorly-localised. To analyse the function of these phosphorylation sites, several Cup-GFP transgenes expressed under the endogenous Cup promoter were created, in which these phosphorylation sites were perturbed. Expression of these constructs in a cup mutant background rescued their premature egg chamber degradation phenotype. However, due to technical difficulties in the creation of these phosphorylation mutant transgenes, it remains unclear to what extent Par- 1 phosphorylation affects the function of Cup in the context of osk mRNA translation.

During early mouse embryonic development, a single cell, the fertilised egg, will give rise to a wide range of cell types that become specialised at a functional and molecular level. Gastrulation and early organogenesis are two of the most critical events during these early stages, when pluripotent cells, able to generate any cell in the embryo, proliferate and become lineage-restricted into the progenitors of the major organs. The vast amount of cell fate decisions taking place in this 48-hour window makes these stages a suitable paradigm to study cell type diversification. Nevertheless, the low cell numbers in early mouse embryos and the limited strategies to isolate homogenous cell populations have restricted the study of the transcriptional programs and regulatory processes that underlie these processes in vivo. With the advent of high-throughput single-cell genome-wide technologies, it is now possible to obtain the molecular profiles of hundreds of individual cells at once, thus opening a new window for the study of early embryogenesis. To delineate the molecular events underlying gastrulation and early organogenesis, we have therefore generated a comprehensive single-cell transcriptomic atlas of these stages. In Chapter 3, I introduce this atlas and give a general overview of the lineages that have been captured. Due to the importance of the haemato-endothelial lineages to establish the circulatory system in the embryo for appropriate oxygenation, in Chapter 4, I characterise their emergence using the atlas. My analyses uncover a rapid formation of primitive erythrocytes that do not transition through mature endothelium. Furthermore, I report the transcriptomes of megakaryocytic and myeloid progenitors as well as show that endothelial cells from different embryonic locations present distinctive transcriptional signatures. Getting a better characterisation of embryogenesis gives us a solid baseline to understand the consequences of genetic mutations. In Chapter 5, I explore the effects of disrupting the blood regulator Tal1 using mouse embryonic chimaeras and reveal that endothelial cells are transcriptionally aberrant at early organogenesis and express genes characteristic of other mesodermal lineages. Although single-cell transcriptomics unveils the molecular programs defining each cell type, studying gene expression is not enough if we want to highlight the regulatory events behind cell type diversification. Therefore, in Chapter 5, I examine the use of single-cell transcriptomics to detect RNAs at enhancers, which may represent a surrogate for enhancer activity. Due to the limitations encountered, in Chapter 7, I perform single-nucleus ATAC-seq in cells at early organogenesis, a time-point by which all major progenitors are established. Analysing the resulting chromatin accessibility maps together with subsequent in vivo validation experiments have allowed the discovery of two novel endothelial enhancers as well as a previously unrecognised role for the ETS transcription factor FEV in the establishment of haemato-endothelial lineages. In conclusion, single-cell genome-wide technologies have permitted the comprehensive characterisation of the molecular programs and regulatory events underlying gastrulation and the start of organogenesis in the early mouse embryo. Having acquired this information has not only contributed to our understanding of embryonic development, but it will also help the optimisation of in vitro differentiation protocols in the future.

Investigation of mTOR-independent regulation of macroautophagy Richard Ian Odle Macroautophagy is a critical catabolic response to cellular stress, enabling lysosomal mediated breakdown of cytosolic cargo. The nutrient-responsive mTORC1 kinase complex has been described as a master regulator of cellular metabolism. Indeed, mTORC1 inhibits autophagy via repressive phosphorylation of the key autophagy regulators ATG13, ULK1, ATG14 and TFEB. Consequently, mTORC1 has become a candidate therapeutic target in neurodegeneration and cancer; however, its essential role in other cellular programs has prompted the investigation of mTORC1-independent regulation of autophagy. This thesis explores the role of CMGC kinase family members ERK1/2 and CCNB1-CDK1 in the regulation of autophagy. The ERK1/2 signalling cascade is activated in a high proportion of cancers. ERK2 has been proposed as a regulatory kinase of TFEB; however, we found little evidence to suggest that ERK1/2 was a direct kinase responsible for TFEB phosphorylation, including at the putative site S142. Furthermore, whilst we observed that hyperactivation of the ERK1/2 pathway did lead to increases in total TFEB protein levels in HEK293, this appeared to be a cell line specific finding. We therefore concluded ERK1/2 was not likely to be a critical regulator of TFEB. It has been proposed that autophagy must be repressed during mitosis, otherwise nuclear envelope breakdown will expose the genome to the cytosolic autophagy machinery. Here we show that autophagy initiation, as measured by markers of the omegasome, is indeed repressed throughout mitosis. Furthermore, autophagy regulators undergo mitotic hyperphosphorylation, including at known repressive sites, in a manner dependent on CDK1 but not mTORC1. Indeed, we find mTORC1 is likely inactive as a result of CDK1-dependent hyperphosphorylation of RAPTOR. Thus, we conclude that mTORC1 is substituted by CDK1, as the master repressor of autophagy during mitosis. These results suggest that autophagy regulation is uncoupled from nutrient status during nuclear envelope breakdown as a mechanism to prevent genomic instability, a hallmark of cancer.

Primordial germ cells (PGCs), the precursors of gametes, are specified in early post-implantation human embryos, which are inaccessible for direct investigation. This process can however be modelled using human embryonic stem cells (hESCs) for in vitro derivation of human PGC-like cells (hPGCLCs). hESCs in conventional culture conditions have very low potential for germ cell fate; this increases significantly if the cells are first cultured in PGC competence-promoting media. To address how the PGC-competent state might be gained, I used a new computational tool - Branch-Recombinant Gaussian Processes (B-RGPs) - to identify potentially relevant pathways based on single-cell RNA sequencing of early human embryos, as well as gonadal PGCs and soma. I then tested these predictions experimentally by modulating the candidate signalling pathways in competent hESCs, followed by hPGCLC induction. I found that inhibition of FGF receptor (FGFR) or its downstream effectors (MEK and PI3K) enhanced competency for PGC fate. By contrast, premature MEK/ERK activation in hESCs resulted in aberrant expression of endoderm markers and loss of germ cell potential. Altogether the study showed that FGFR pathway is a negative regulator of cellular competence for human PGCLC fate. PGCLC specification from competent cells occurs in response to BMP signalling and is orchestrated by specific transcription factors (TFs). Emerging evidence points to many differences between TF networks in mouse and human PGCs. For example, SOX17 is essential for human germ cell fate, but is dispensable for mouse PGCs. On the other hand, a key pluripotency factor PRDM14 is crucial for mouse PGC specification, but its role in human PGC development is unclear. I designed genetic approaches to address PRDM14 involvement in human pluripotency versus PGC specification. To achieve inducible loss of function, I deleted endogenous PRDM14 in the background of doxycycline-repressible PRDM14 transgene expression. This approach, however, did not allow fast and homogenous control over protein levels. To circumvent this, I used CRISPR/Cas9 in conjunction with conditional degrons (auxin- and jasmonate-inducible degrons) for rapid and reversible degradation of the endogenous protein. Notably, PRDM14 depletion during differentiation significantly reduced the number of specified hPGCLCs, revealing its role in human PGC fate. Furthermore, RNA sequencing of mutant cells showed aberrant gene expression, with derepression of WNT signalling and somatic genes. Importantly, ectopic PRDM14 expression in mutant cells rescued hPGCLC specification and transcriptional changes, thus proving the specificity of the observed phenotype. Interestingly, loss of PRDM14 in hESCs had a distinct transcriptional effect, indicating a context-dependent role of PRDM14 in PGCs and pluripotent cells. ChIP-sequencing in hESCs and hPGCLCs identified PRDM14 binding to regulatory elements of key pluripotency and germ cell genes and motif analysis suggested its integration in the core TF circuitry in both cell types. Interestingly, the overlap between PRDM14 targets in mouse and human was limited. Altogether, this reveals a key role of PRDM14 in human germ cell fate and exemplifies the importance of complete and fast protein depletion in loss-of-function studies. Overall my research has furthered our understanding of signalling and transcriptional regulators controlling the segregation of human germ cells from soma. It also expanded the toolkit for protein studies in dynamic processes and uncovered hitherto unknown roles of PRDM14 in early human development.

Meiosis is a specialised form of cell division, which halves the chromosome content of a cell. A single round of DNA replication is followed by two rounds of division. Meiosis utilises the key machinery of mitosis but with specific adaptations, particularly in the first division. In the first meiotic division, unlike mitosis or meiosis II, sister chromosomes segregate together. In budding yeast, the fusion of sister kinetochores allows them to attach to a single microtubule from one cell pole. The four-protein complex responsible for this fusion and therefore sister co-segregation in meiosis I is called monopolin. However, the mechanism by which monopolin acts is unclear. What is well understood about monopolin is its structure, it is a four-protein complex of two homodimers of Csm1 and two copies of the proteins Lrs4, Mam1 and the casein kinase, Hrr25. The complex forms a V-shaped structure, with a kinetochore binding region at each point of the V. This structure has led to the proposed mechanism of monopolin forming a mechanical bridge between the two kinetochores directly causing their fusion. There is, however, no direct evidence for this mechanism and many open questions remain, particularly as to how the bridging interaction would be regulated and the role of proteins, Mam1 and Hrr25, which do not form the structure of the V-shape. This project had several aims, firstly to dissect whether all components of the monopolin complex are required for monoorientation or if some only serve to target the complex to the kinetochore. By artificial tethering of monopolin to the kinetochore, I was able to demonstrate the requirement for a complete monopolin complex for the mono-orientation of sister chromatids. Secondly, I aimed to show the molecular basis for the interaction between monopolin and the kinetochore. This was done by following insight gained form the crystal structure of monopolin showing the interaction between Csm1 and its kinetochore receptor Dsn1, I generated mutants in Dsn1 that disrupt this interaction. Through predominantly assessment of cell viability, imaging of chromosome division in meiosis and chromatin immunoprecipitation of monopolin components I demonstrated that these sites within Dsn1 are essential for the localistion of monopolin to the kinetochore, sister chromatid mono-orientation and therefore meiosis. In addition, through the use of mass spectrometry I showed that two serines within this region of Dsn1 are phosphorylated in meiosis, and that they are good candidates for being a point of regulation for the interaction between monopolin and the kinetochore. Finally, although it is known that Hrr25 must be enzymatically active for mono-orientation of sister chromatids in meiosis, no in vivo targets of this phosphorylation have been described. I, therefore, utilized a mutant version of Hrr25 which is not recruited to the monopolin complex and mass spectrometry to identify possible targets of phosphorylation by Hrr25 at the meiotic kinetochore.

Polo kinase is a major regulator of the cell cycle. Polo is present at many different subcellular structures, including centrosomes. Centrosomes are formed when mother centrioles recruit pericentriolar material (PCM) around themselves. The PCM expands dramatically in preparation for mitosis, in a process that relies on Polo phosphorylation of the PCM protein Cnn. Polo is also known to regulate centriole disengagement, and Polo binding to the centriolar protein Sas-4 is crucial for Asl recruitment to daughter centrioles during their conversion into fully functional mothers. The primary aim of my thesis was to fully characterise how Polo is recruited to centrosomes to perform its different functions. As a model, I use syncytial fly embryos. In addition to the previously known Polo binding site in Sas-4, I found that the PCM protein Spd-2 directly interacts with the PBD domain of Polo. While Spd-2 is not required for Polo recruitment to the centriole, it is crucial to recruit Polo to the PCM. I propose that this interaction is part of a Spd-2/Polo/Cnn positive feedback loop that drives mitotic PCM assembly. Furthermore, I found that a third protein (Ana1) is required for overall Polo recruitment to the centrosome. Ana1 has been previously linked to Asl recruitment and/or maintenance and centriole conversion. Losing Ana1-mediated Polo recruitment did not affect these processes, but it did severely hinder mitotic PCM expansion. I propose that, independently of its Asl-related role, Ana1 acts as a crucial reservoir of centriolar Polo. Thus, centriolar Polo recruited via Ana1 can phosphorylate Spd-2 to allow it to bind the PBD, in an example of Polo “self-priming”.