Clonal cell populations are known to exhibit marked phenotypic heterogeneity at the single cell level, a phenomenon usually masked by conventional population-wide analyses. Two models of yeast cell heterogeneity were investigated during this study. First, expression of the Saccharomyces cerevisiae Rad6 protein, known to facilitate DNA damage tolerance, was investigated. However, apparent bimodal Rad6 expression proved to be an artefact of a mixed-genotype culture source. The other model was adhesin expression heterogeneity in the opportunistic yeast pathogen Candida glabrata. Adherence to the host cell is an important step in the establishment of C. glabrata infection, mediated by adhesin proteins. The subtelomeric EPA family of adhesin genes encodes a large class of GPI-anchored cell wall proteins in C. glabrata, among which Epa1 is the best studied. Epa1 expression is highly heterogeneous between individual C. glabrata cells, a factor that can dictate adherence capacity and may have important implications for infection. Such cell-to-cell variability was dependent upon strain background. Variation in cell surface Epa1 level was correlated with variation in EPA1 mRNA, consistent with transcriptional regulation of heterogeneity. Indeed Sir-dependent silencing was found to be a major driver of heterogeneous Epa1 expression in a strain demonstrating high cell-to-cell variability but not in an alternative genetic background demonstrating lower heterogeneity. Inefficient silencing in the latter strain was overcome by ectopic SIR3 expression, and was not due to differences in EPA1 sequence or distance from the chromosome end compared with the heterogeneous strain. Moreover, strain-to-strain variation in the silencing-dependence of EPA1 expression was observed across a range of clinical isolates and was found to correlate with the extent of Epa1 heterogeneity. Thus, marked variation in adhesin expression exists between cells and between strains of C. glabrata. In addition, the data presented shed light on the regulation of such heterogeneity in particular the role of Sir-dependent transcriptional silencing.

Embryonic stem (ES) cells are a potential source of neural derivatives that can be used in stem cell-based therapies. To generate specific cell types in a predictable manner, a detailed understanding of cell fate specification is required. To address this, this study employed mouse ES cells as a model to explore how neural identity was acquired and how regional identities were specified in ES cell-derived neural progenitor cells (NPCs). Chemical inhibitors are more stable than recombinant proteins, hence they give more reproducible biological responses. This is crucial when generating specific cell types from ES cells for large-scale analysis. This study explored the effect of two newly discovered small molecule inhibitors of BMP signalling, dorsomorphin (DM) and LDN193189 (LDN) on the neural induction of ES cells, and compared their capabilities with those of recombinant noggin. Both DM and LDN treatments increased the expression of neural markers to levels that were comparable to that achieved by noggin treatment, suggesting that both LDN and DM can potentially substitute recombinant noggin in the generation of NPCs in vitro. Retinoic acid (RA) is an important regulator of regional specification in vivo, but the underlying mechanisms for giving region-specific response are unclear. Here, early and late NPCs were exposed to RA at specific periods and were analysed for neural and positional markers. Results showed that region-specific responses were produced by RA at specific periods, indicating that NPCs display temporal changes in patterning responsiveness to morphogenic cues. The ability of morphogens to impose positional response is lost in late cultures. Since Sox1 expression is associated with a naïve neural progenitor state and neural commitment, the expression of Sox1 was examined over time to test whether it was associated with the responsiveness to patterning cues. Both caudal and ventral markers were induced by RA and, because it plays an essential role in ventral patterning, Hedgehog (Hh) agonist purmorphamine respectively in early NPCs but not in late NPCs in both Sox1 positive and negative populations. This suggested that the expression of positional markers was not dependent on a temporally-defined Sox1 progenitor state.

To control cellular processes synthetic biologists and biotechnologists often use regulation of gene expression; by regulating transcription it is assumed protein levels will follow. However the use of a common pool of gene expression resources results in the emergence of hidden interactions, couplings, between genes which are not immediately apparent from circuit topologies. This can result in a breakdown in the relationship between transcriptional regulation (input) and protein levels (output). Current evidence suggests that it is the number of free ribosomes which limits protein synthesis capacity and therefore creates these non-regulatory linkages. In this thesis, the feasibility of dividing the cell's translational resources to reverse the breakdown in input-output responses and decouple co-expressed genes is demonstrated. A model of microbial growth which captures key aspects of host-circuit interactions is developed and demonstrates the feasibility of using orthogonal (circuit specific) ribosomes for circuit gene expression; showing that by allocating circuit genes to both host and orthogonal translational pools these genes can be decoupled and the flux through a model biotechnological pathway can be improved. The design of negative feedback controllers to allocate translational resources is investigated. These act to increase supply of orthogonal ribosomes as the demand for translational resources by the circuit increases. The stability of a number of controller architectures is investigated. An experimental prototype of the best performing controller architecture is able to reduce gene coupling by 50%. The best controller architecture is carried forward for further analysis, and a detailed mechanistic model which can be used as a design tool is developed. A hard trade of between gene expression and decoupling activity is identified and designs which manage this trade-o suggested. The controller is shown to ameliorate resource mediated failures of modularity in a range of synthetic biology circuits. Finally, a discussion on ways to produce a second generation of translational resource allocation controllers is provided.

Sp 1 is a ubiquitously expressed transcription factor and regulates a range of genes including housekeeping and tissue-specific genes. Studies using a DNA binding domain (DBD)-deficient Spl have shown that Spl is required for haematopoietic specification. Here, we generated a Spl-DBD deficient ESC line to recapitulate the previous model, as well as a novel Spl null ESC line. Spl knockout cells demonstrated a complete absence ofhaematopoietic differentiation, indicating a crucial role for Sp 1 at the early stages of blood cell specification. In contrast, Sp 1 DBD-deficient cells were able to differentiate to haematopoietic progenitors, but failed to terminally differentiate, suggesting a different mechanism of Sp !-mediated transcriptional regulation in early and later stages. Gene expression analysis in Spl knockout cells indicated a novel role for Spl in ESC differentiation potential and mesoderm formation, while chromatin accessibility profiling revealed changes in chromatin structure in the absence ofSpl. We found Sp3, a close family member ofSpl , is able to compensate for loss ofSpl at most sites, but not at some important genes encoding developmental regulators. This work provides novel insights into the interplay between Sp 1 and Sp3 and furthers our understanding of the function of one of the earliest discovered TFs.

Genome duplication is an essential cellular process and its appropriate execution is required to maintain genome integrity. Ubiquitin and SUMO belong to the protein family of ubiquitin-like modifiers, and they are required to maintain genome integrity by acting as key regulators in the response to DNA damage. However, their role in the regulation of unperturbed DNA replication is less understood. My PhD work has focused on studying the roles of ubiquitin and SUMO during unperturbed DNA replication in Xenopus laevis egg extract. Members of our lab showed that polyubiquitylation of Mcm7 – a subunit of the CMG replicative helicase, promoted CMG disassembly during DNA replication termination. Importantly, my work identified the ubiquitin ligase cullin2LRR1 as the enzyme required for Mcm7 polyubiquitylation and CMG disassembly. I also showed that these events were restricted to termination by the regulated recruitment of cullin2LRR1 to the terminated replisome. Moreover, I showed that the p97 protein remodeller was recruited to the polyubiquitylated CMG and drove chromatin extraction of the terminated replisome. My work has also provided the first insights into how the chromatin associated proteome changes in response to impaired sumoylation during S phase. Finally, I have identified several potential SUMO2 substrates associated with replicating chromatin.

Huntington disease (HD) is an extremely variable inherited neurodegenerative disorder caused by expansion of an unstable CAG trinucleotide repeat in the huntingtin gene (HTT). Somatic instability in HD exhibits an age-dependent, expansion-biased and tissue-specific pattern, and the highest level of somatic instability is found in tissues that are most susceptible to the disease pathology. Therefore, the aim of this project was to test the hypothesis that somatic instability of the HD CAG repeat plays a major role in disease pathology by quantifying somatic instability in the number of CAG repeats by next generation sequencing (NGS) technology in buccal cell DNA. We developed a method to sequence and genotype HTT alleles from blood and buccal swab DNA of the Scottish and Venezuelan populations respectively. A total of 210 individuals from the Scottish general population and 742 HD patients and unaffected individuals from the Venezuelan HD cohort were sequenced on the MiSeq platform. We established that it was possible to sequence and genotype the CAG repeats, the polymorphic CCG repeat and the flanking sequences. Our data highlight the utility of NGS technology as an approach to genotype HTT alleles, detect sequence variants and quantify somatic instability of the CAG repeat. Our data emphasise that the somatic instability in HD is age-dependent and expansion-biased, also could be a major factor in disease progression, and could be a potential therapeutic target in HD. We also investigated the possibility that there are trans-acting modifier factors involved in determining the degree of somatic instability in HD patients. We genotyped polymorphisms in candidate mismatch repair (MMR) genes and examined their effect, if any, on the residual variation of somatic instability. Individuals carrying the minor allele of rs3512 in FAN1 have a higher level of somatic instability than average, suggesting that some of the variations in HD somatic instability could be accounted for by genetic variation in the DNA mismatch repair pathway. The search for modifier genes might have consequences in understanding the pathological process in HD, and may therefore provide therapeutic targets for future investigations.

Central to this thesis is the investigation of how several factors (e.g. morphology, ecology, and social conditions) co-vary with consistent individual behavioural variation. Additionally, conceptual work explores the reach of consistent inter-individual behavioural variation in several novel contexts. In doing so, the aim is to contribute novel findings to the quickly growing compendium of knowledge concerning 'animal personality' in various settings. In summary, during this PhD the following research has been performed: • Through the use of lab-reared juvenile three-spined stickleback (Gasterosteus aculeatus) whose parents came from different habitats, I investigated the relative contribution of plasticity and (epi)genetic mechanisms on three behaviours (boldness, activity, and sociality). While these behaviours did not differ between juveniles from both populations, correlations of body length with specific behaviours did. (Chapter 2) • During fieldwork on the Scottish island of North Uist, seven populations of wild three-spined stickleback were surveyed for boldness, exploratory behaviour, and activity. Furthermore, the defensive morphology of these fish was quantified following an armour-staining protocol. Bold fish were found to be less armoured, directly contradicting the phenotypic compensation hypothesis, while more active fish were more armoured, which contradicts the idea that armour is energetically costly. (Chapter 3) • Spurred on by the findings on morphology-behaviour correlations, the ecology of the studied populations was assessed with a focus on parasitism by three common macro-parasites of three-spined stickleback, predation by trout, competition with the nine-spined stickleback, and the availability of resources. I found that all these factors, to some extent, influence individual behavioural variation. Importantly, this study provides the first observational evidence that less-explored factors, such as resource availability and interspecific competition, can be highly relevant influences on individual behaviour. (Chapter 4) • Next, a study on the effect of group behavioural composition on group foraging behaviour was done. With the help of a research visit to Aarhus University, similar experiments were performed on stickleback and social spiders (Stegodyphus dumicola). Group behavioural composition had a strong effect in both study systems, but experience was a more relevant influence on group foraging behaviour in stickleback. By using two different study organisms from quite disparate animal taxa, this allows me to develop hypotheses concerning the evolutionary origin and maintenance of individual behavioural variation and its tentative link with sociality. (Chapter 5) • Lastly, the phenomenon of consistent individual variation in behaviour is explored in novel contexts through conceptual essays (Chapter 6). o Do bacteria offer a suitable system to study ‘personality’? o How can animal personality be leveraged to improve animal welfare? o Are non-antagonistic interspecific interactions an overlooked yet important factor in the study of animal personality?

The human amylase gene family on chromosome 1p21.1 is highly copy number variable. The salivary (AMY1) and pancreatic (AMY2A and AMY2B) amylase genes encode the starch-digesting enzymes expressed in the salivary gland and pancreas, respectively. Different molecular approaches have been used to investigate copy number variation (CNV) in this region, more specifically the salivary AMY1 copy number (CN). High AMY1 CN has been shown to be correlated with adaptation to human dietary starch intake, and low AMY1 CN reported to be a predisposition factor to obesity. These findings have not been replicated independently, and reliable measurement methods and accurate structural characterisation of the region are important to address such findings. The large dynamic copy number range of AMY1 genes and the extent of sequence identity in this region demands accurate, reliable and high-throughput CNV measurement methods. In this study, high-resolution paralogue ratio test (PRT) measurement methods have been developed to define the full scope of variations in human amylase genes. The data demonstrated independent allelic series of amylase CN variants (CNVs) in sub-Saharan Africans, in which the region has undergone homologous and non-homologous rearrangements to create new haplotypes, some of which contain five copies each of the AMY2A and AMY2B genes. These expansions have taken AMY2B/AMY2A /AMY1 repeat unit as a starting point to create new allelic series represented by triplication, quadruplication and quintuplication. With a better understanding of amylase region variation and the high-resolution measurement methods developed in this study, the effect of amylase CNVs on diseases was assessed in eight common diseases. The results of case-control association studies showed no evidence of significant association between amylase CNVs and the eight diseases being tested. Furthermore, the significant association between amylase CN and body mass index (BMI) has been reassessed in three different cohorts (in a total of 4237 samples). The data demonstrated no evidence for an association between amylase CNVs and BMI. The lack of evidence of significant associations suggest that amylase CNVs does not influence BMI or obesity.

Transposable elements are an almost ubiquitous feature in all three domains of life, and are a significant driver of genomic evolution. These discrete genetic elements are capable of mobilisation and amplification, which allows them to multiply. A subset of these elements, the cut-and-paste transposons, excise themselves entirely from the surrounding DNA and integrate themselves elsewhere in the genome. Hsmar1, a cut-and-paste element of the mariner transposon family, was recently investigated to examine the kinetics of the complex formation pathway. This transposase forms a dimer in solution before binding to first one transposon end, then recruiting the other. This pathway, Synapsis by Naked End Capture (S-NEC), is found in both prokaryotes and eukaryotes. Tn5, of the IS4 family, is an example of the alternative pathway, Synapsis by Protein Dimerisation (S-PD). In this pathway, the transposase is monomeric in solution. Each monomer binds to a transposon end independently, before combining to form a synaptic complex. This pathway has been seen widely in prokaryotes, but no eukaryotic elements tested to date employ it. To understand what limitations the complex formation pathway might place upon an S-PD transposase in a eukaryotic genome, the Tn5 transposase was manipulated. Forming the transposase into a single-chain dimer allowed it to mimic S-NEC, an alteration that made it hyperactive in prokaryotes and eukaryotes. As the eukaryotic genome contains chromatin, adapting the Tn5 transposase to this might overcome some of the low activity seen. The fusion of a chromatin binding domain to the transposase was not able to enhance the frequency of transposition, which may suggest that the presence of chromatin is not causing significant inhibition. The results presented in this work suggest two significant conclusions. The first is that the S-NEC reaction dynamics are significantly preferable to S-PD within eukaryotes. The second is that S-PD itself may be a system of regulation of transposition, to prevent activity levels that could be detrimental to the host.

Humans have highly variable number of alpha defensin genes, with between 3-16 diploid copies. Alpha-defensin genes have important roles in human innate immunity and diseases. Recently, GWAS studies reported this locus associated with IgA nephropathy and periodontitis. However, the underlying mechanism of association is not clear. In this Ph.D. thesis, human alpha defensin CNV flanking haplotype diversity in global populations was studied and the association between diseases and haplotype classes was discussed. Then a novo method to detect variants from inside the DEFA1A3 CNV was developed and a list of potential disease-related mutations for further functional studies was generated. The association between CNV internal variants and flanking haplotype classes was studied. Non-allelic homologous recombination was found to be the major mechanism of CNV formation of alpha defensin CNV. Analysis results were verified by PCR and Sanger sequencing-based methods. Additional to that, the haplotype diversity analysis highlighted an unusual haplotype 5T/7C which is only found in European populations but highly diverged from other human haplotypes. Further evidence was provided to suggest that this is an introgressed haplotype from Neanderthals. Furthermore, we used Oxford Nanopore to reconstruct haplotype structure in DEFA1A3 CNV and discussed its advantages and limitations by our analysis results. In brief, this Ph.D. research greatly improved our understanding of DEFA1A3 global diversity, evolutionary history, diseases and haplotype association.