The goal of the Encyclopedia of DNA Elements (ENCODE) project has been to characterize all the functional elements of the human genome. These elements include expressed transcripts and genomic regions bound by transcription factors (TFs), occupied by nucleosomes, occupied by nucleosomes with modified histones, or hypersensitive to DNase I cleavage, etc. Chromatin Immunoprecipitation (ChIP-seq) is an experimental technique for detecting TF binding in living cells, and the genomic regions bound by TFs are called ChIP-seq peaks. ENCODE has performed and compiled results from tens of thousands of experiments, including ChIP-seq, DNase, RNA-seq and Hi-C. These efforts have culminated in two web-based resources from our lab—Factorbook and SCREEN—for the exploration of epigenomic data for both human and mouse. Factorbook is a peak-centric resource presenting data such as motif enrichment and histone modification profiles for transcription factor binding sites computed from ENCODE ChIP-seq data. SCREEN provides an encyclopedia of ~2 million regulatory elements, including promoters and enhancers, identified using ENCODE ChIP-seq and DNase data, with an extensive UI for searching and visualization. While we have successfully utilized the thousands of available ENCODE ChIP-seq experiments to build the Encyclopedia and visualizers, we have also struggled with the practical and theoretical inability to assay every possible experiment on every possible biosample under every conceivable biological scenario. We have used machine learning techniques to predict TF binding sites and enhancers location, and demonstrate machine learning is critical to help decipher functional regions of the genome.

When conspecific individuals are crossed, the ensuing hybridization creates a spectrum of phenotypes in the resulting offspring. Many of hybrid traits will be additive, similar to the parental phenotypes. In some cases however, transgressive phenotypes are formed, outside the range of that of the parental phenotypes. Transgressive phenotypes can either be restricted to the F1 generation or be heritable throughout the hybrid lineage. While the mechanism behind heritable transgressive phenotyped is yet to be determined, transgressive gene expression is thought to be the root cause of their formation. Epigenetics modifications, heritable variation separate to the DNA code, can alter gene expression, persist through generations, and vary between individuals and over time. This makes them ideal candidates to be involved in the formation of transgressive phenotypes. RNA silencing is an epigenetic mechanism of gene regulation relying on 20Q24nt single stranded small RNAs (sRNAs). Small RNAs, due to their ability to set up persistent epigenetic marks at a locus, have the potential to create heritable transgressive gene expression. For example, when genetic variation from one parental genome presents novel targets to the sRNAs of the other parental genome, new epigenetic marks such as DNA methylation or secondary sRNAs can be created at target sites. In order to understand the potential of small RNAs to influence hybrid phenotype, I designed crossing experiments with Chlamydomonas reinhardtii, choosing this unicellular alga due to the genetic tools available and the haploid nature of its vegetative cells. The specific aim of the experiment was to identify transgressively expressed sRNA populations. Crossing two geographically distinct strains of C. reinhardtii, and sequencing both the genomes and sRNAomes of parents and recombinants, I was able catalogue both genetic and epigenetic variation in the parental strains providing unique insight into the inheritance of small RNAs in this alga. In this thesis, I first compare the genomes of the parental strains, identifying polymorphisms and assessing genetic variation in RNA silencing pathway components. I then describe the sRNA profiles of the parental strains, identifying differentially expressed sRNA loci. I then describe my approach to identifying transgressively expressed sRNA loci in the hybrids. While many sRNA loci in the recombinants exhibit additive sRNA expression, I found multiple transgressively expressed sRNA loci. Using the available bioinformatics tools, I identified potential miRNAs and phased secondary sRNAs within the list of transgressively expressed loci. Target analysis of one of the transgressively expressed miRNAs linked it with the transgressive expression of certain phased loci, suggesting a potential for sRNAs to be able to set up heritable epigenetic marks in recombinant C. reinhardtii cells.

Senescent cells interact with the surrounding microenvironment achieving both pro- oncogenic and tumour-suppressive outcomes. In addition to autocrine and paracrine signalling mediated by factors of the senescence-associated secretory phenotype (SASP), we have recently identified that NOTCH1 can drive a unique form of senescence in adjacent cells via juxtacrine signalling. Here, we show that NOTCH1 signalling confers a dramatic impact on chromatin structure during senescence. RAS-induced senescent (RIS) fibroblasts often develop chromatin structures called senescence-associated heterochromatic foci (SAHF). We find that NOTCH1 inhibits SAHF formation at least partially through transcriptional repression of a critical structural component, high-mobility group A (HMGA). Using ATAC-sequencing (assay for transposase accessible chromatin) we demonstrate that nucleosome positioning is substantially altered in RIS and that this re-distribution is also antagonised by NOTCH1, resulting in a distinct chromatin landscape. Importantly, normal or cancer cells that express the NOTCH ligand jagged-1 can drive similar chromatin structural changes in adjacent cells in a cell-cell contact dependent manner. In addition, using a highly optimised chromatin immunoprecipitation (ChIP-seq) protocol and the proximity ligation assay ‘Hi-C’, we demonstrate that HMGA proteins are directly involved in the formation of long-range interactions in RIS cells that may underpin SAHF formation. These ChIP-seq data have also allowed us to identify a unique HMGA1 binding profile, potentially suggesting a novel role for HMGA1 in gene regulation. Together, our data indicate that NOTCH signalling, both cell-autonomously and non-cell-autonomously, can repress HMGA1, a multi-faceted protein that regulates nucleosome positioning (1D structure), SAHF formation (3D structure) and potentially mRNA abundance.

La méthylation de l’ADN est une modification épigénétique sur les cytosines des dinucléotides CpG catalysée par les enzymes DNMT. Les cellules cancéreuses présentent des hyperméthylations aberrantes sur les promoteurs de gènes dits suppresseurs de tumeurs, ce qui contribue à leur répression transcriptionnelle et favorise la progression tumorale. De par sa nature réversible, la méthylation de l’ADN est une cible de choix pour des thérapies épigénétiques ; cependant, les inhibiteurs de DNMT ont une action de déméthylation globale du génome qui conduit à une forte toxicité. Mon travail a consisté à développer des stratégies de déméthylation ciblée sur des régions spécifiques du génome. Premièrement, j’ai validé une stratégie induisant une reprogrammation épigénétique spécifique et durable du gène suppresseur de tumeurs SERPINB5 dans des cellules de cancer du sein. Deuxièmement, j’ai optimisé des stratégies d'édition de l’épigénome comme outil en recherche fondamentale. / DNA methylation takes place on cytosines of CpG dinucleotides in mammals and is catalysed by DNMT enzymes. Cancer cells are characterised by frequent promoter hypermethylation leading to transcriptional repression of tumor suppressor genes and favouring tumor progression. Because of its reversible nature, DNA methylation is a target of choice in epigenetic therapies. However, current DNMT inhibitors act in a global and non-specific manner, leading to side effects and toxicity in normal cells. During my thesis I have developed strategies to perform targeted demethylation in specific regions of the genome without affecting global methylation. First, I have validated a strategy inducing the specific and durable epigenetic reprogramming of the tumor suppressor gene SERPINB5 in a breast cancer cell line, which can pave the way to further biomedical research. Second, I have optimised epigenome editing strategies as a regular tool in basic research.

The human genome is a dynamic structure, recently recognized to present with significant large-scale structural variation. DNA-copy number changes represent one common type of such variation and is found both between individuals and within the somatic cells of the same individual, especially in disease states like cancer.  Apart from DNA-rearrangements, epigenomic changes are increasingly acknowledged as important events in the maintenance of genomic integrity. In this thesis, different array-based methods have been applied for global genomic and epigenomic profiling of both normal and cancer cells. In paper I, a genomic microarray was established and used to determine DNA-copy number variants (CNVs) in a cohort of 76 healthy individuals from three ethnic populations. We identified 315 CNV regions that in total encompassed ~3,5% of the genome. In paper II, the array was utilized to discover CNVs within several differentiated tissues from the same subject. Six variants were identified providing evidence for somatic mosaicism. In paper III and IV we studied pheochromocytomas and paragangliomas, rare endocrine tumours that most often present as benign and sporadic with unclear genetic/epigenetic cause. Genome-wide DNA-copy number analysis of 53 benign and malignant samples in paper III revealed numerous common and novel chromosomal regions of losses and gains. High frequencies of relatively small overlapping regions of deletions were detected on chromosome 1p arm, encompassing several candidate tumour suppressor genes. In paper IV, an epigenomic map for two histone modifications associated with silent (H3K27me3) or active (H3K4me3) gene transcription, was generated for one malignant pheochromocytoma. Integrated analysis of global histone methylation, copy number alterations and gene expression data aided in the identification of candidate tumour genes. In conclusion, the performed studies have contributed to gain knowledge of CNVs in healthy individuals, and identified regions and genes which are likely associated with the development and progression of pheochromocytoma/paraganglioma.

Yes / Infinium Human Methylation BeadChip is an array platform for complex evaluation of DNA methylation at an individual CpG locus in the human genome based on Illumina’s bead technology and is one of the most common techniques used in epigenome-wide association studies. Finding associations between epigenetic variation and phenotype is a significant challenge in biomedical research. The newest version, HumanMethylationEPIC, quantifies the DNA methylation level of 850,000 CpG sites, while the previous versions, HumanMethylation450 and HumanMethylation27, measured >450,000 and 27,000 loci, respectively. Although a number of bioinformatics tools have been developed to analyse this assay, they require some programming skills and experience in order to be usable. Results We have developed a pipeline for the Galaxy platform for those without experience aimed at DNA methylation analysis using the Infinium Human Methylation BeadChip. Our tool is integrated into Galaxy (http://galaxyproject.org), a web-based platform. This allows users to analyse data from the Infinium Human Methylation BeadChip in the easiest possible way. Conclusions The pipeline provides a group of integrated analytical methods wrapped into an easy-to-use interface. Our tool is available from the Galaxy ToolShed, GitHub repository, and also as a Docker image. The aim of this project is to make Infinium Human Methylation BeadChip analysis more flexible and accessible to everyone. / Research Development Fund Publication Prize Award winner, May 2020.

Psychiatric disorders are believed to result from the combination of genetic predisposition and many environmental triggers. While the large number of disease-associated genetic variations have been recognized by previous genome-wide association studies (GWAS), the role of epigenetic mechanisms that mediate the effects of environmental factors on CNS gene activity in the etiology of most mental illnesses is still largely unclear. A growing body of evidence suggested that the abnormalities (changes in gene expression, formation of neural circuits, and behavior) involved in most psychiatric syndromes are preserved by epigenetic modifications identified in several specific brain regions. In this thesis, we developed the second generation of one of our microfluidic technologies (MOWChIP-seq) and used it to profile genome-wide histone modifications in three mental illness-related biological studies: the effect of psychedelics in mice, schizophrenia, and the effect of maternal immune activation in mice offspring. The second generation of MOWChIP-seq was designed to generate histone modification profiles from as few as 100 cells per assay with a throughput as high as eight assays in each run. Then, we applied the new MOWChIP-seq and SMART-seq2 to profile the histone modification H3K27ac and transcriptome, respectively, using NeuN+ neuronal nuclei from the mouse frontal cortex after a single dose of psychedelic administration. The epigenomic and transcriptomic changes induced by 2,5-Dimethoxy-4-iodoamphetamine (DOI), a subtype of psychedelics, in mouse neuronal nuclei at various time points suggest that the long-lasting effects of the psychedelic are more closely related to epigenomic alterations than the changes in transcriptomic patterns. Next, we comprehensively characterized epigenomic and transcriptomic features from the frontal cortex of 29 individuals with schizophrenia and 29 individually matched controls (gender and age). We found that schizophrenia subjects exhibited thousands of neuronal vs. glial epigenetic differences at regions that included several susceptibility genetic loci, such as NRXN1, RGS4 and GRIN3A. Finally, we investigated the epigenetic and transcriptomic alterations induced by the maternal immune activation (MIA) in mice offspring's frontal cortex. Pregnant mice were injected with influenza virus at GD 9.5 and the frontal cortex from mice pups (10 weeks old) were examined later. The results offered us some insights into the contribution of MIA to the etiology of some mental disorders, like schizophrenia and autism. / Doctor of Philosophy / While this field is still in its early stage, the epigenetic studies of mental disorders present promise to expand our understanding about how environmental stimulates, interacting with genetic factors, contribute to the etiology of various psychiatric syndromes, like major depression and schizophrenia. Previous clinical trials suggested that psychedelics may represent a promising long-lasting treatment for patients with depression and other psychiatric conditions. These research presented the therapeutic potential of psychedelic compounds for treating major depression and demonstrated the capability of psychedelics in increasing dendritic density and stimulating synapse formation. However, the molecular mechanism mediating the clinical effectiveness of psychedelics remain largely unexplored. Our study revealed that epigenomic-driven changes in synaptic plasticity sustain psychedelics' long-lasting antidepressant action. Another serious mental illness is schizophrenia, which could affect how an individual feels, thinks, and behaves. Like most other mental disorders, schizophrenia results from a combination of genetic and environmental causes. Epigenetic marks allow a dynamic impact of environmental factors, including antipsychotic medications, on the access to genes and regulatory elements. Despite this, no study so far has profiled cell-type-specific genome-wide histone modifications in postmortem brain samples from schizophrenia subjects or the effect of antipsychotic treatment on such epigenetic marks. Here we show the first comprehensive epigenomic characterization of the frontal cortex of 29 individuals with schizophrenia and 29 matched controls. The process of brain development is surprisingly sensitive to a lot of environmental insults. Epidemiological studies have recognized maternal immune activation as a risk factor that may change the normal developmental trajectory of the fetal brain and increase the odds of developing a range of psychiatric disorders, including schizophrenia and autism, in its lifetime. Given the prevalence of the coronavirus, uncovering the molecular mechanism underlie the phenotypic alterations has become more urgent than before, for both prevention and treatment.

During the process of spermatogenesis, histones are replaced by protamines, basic proteins enabling transmission of DNA to the oocyte during fertilization. In mouse sperm, there is only 1% of remaining histones whose N-terminal tails contain post-translationally modified residues. In this study, I was interested in contribution of paternal histone H4 acetylated on lysine K12 residues (H4K12ac) that is present in mature sperm head in remaining nucleosomes. Physiologically, H4K12ac has an important role in transcription factor accumulation and in regulation of gene expression. The presence and abundance of H4K12ac modification in various pronuclei stages of 1-cell embryo and parthenotes were assessed by imunnoflourescent detection with utilization of anti-H4K12ac antibody. Altogether, the paternal pronucleus exhibits a strong acetylation signal on H4K12 since its formation, while in the maternal one, there is a slow continual increase of H4K12ac getting on the same level as paternal pronucleus till the pronuclei fusion. Simultaneously DNA methylation status in both pronuclei was detected. In paternal pronucleus there is a continual decrease in the DNA methylation detectable as a decrease of 5mC and an increase of 5hmC signal. Meanwhile, the maternal pronucleus stays widely methylated. DNA...

Chez les Mammifères, le succès de la fécondation et d’un développement embryonnaire viable résident principalement dans la qualité des cellules reproductrices : les gamètes mâles (spermatozoïdes) et femelles (ovocytes) apportant chacun la moitié du patrimoine génétique du futur individu. Dès lors que ce patrimoine génétique est endommagé par différents processus tels que les attaques radicalaires communément appelées stress oxydant, cela peut compromettre la qualité des gamètes et par conséquent le succès de la reproduction. En effet, les dommages oxydants à l’ADN spermatique sont une des causes d’infertilité masculine fréquemment associés aux échecs reproductifs dans le cadre de conception naturelle ou artificielle chez l’homme. Afin de mieux appréhender l’impact des dommages oxydants sur la qualité des spermatozoïdes, notre équipe a généré des modèles murins présentant des atteintes oxydantes de l’ADN associées à des échecs reproductifs. Dans un de ces modèles nous avons pu montrer que tout le noyau spermatique n’était pas concerné de la même façon par les altérations oxydantes. De même, à l’échelle des chromosomes murins les investigations de l’équipe suggéraient que tous n’étaient concernés de la même façon. Dans ce travail de thèse, j’ai développé et mis en œuvre des protocoles d’hybridation in situ en fluorescence (FISH) permettant de montrer que la susceptibilité des chromosomes à l’oxydation est déterminée par la position des chromosomes dans le noyau spermatique murin. En parallèle de cette étude, je me suis intéressé à l’épigénome spermatique et plus particulièrement à la méthylation de l’ADN en conditions de stress oxydant. Les résultats préliminaires montrent qu’un environnement post-testiculaire pro-oxydant peut altérer le profil épigénétique spermatique. Dans un développement pré-clinique du sujet, j’ai contribué à montrer qu’une supplémentation orale antioxydante permettait de corriger les dommages oxydants à l’ADN, en modifiant la réponse antioxydante globale du tissu épididymaire des animaux. / In Mammals, the success of fertilization and embryonic development, are associated to the quality of the reproductive cells: male gametes (spermatozoa) and female gametes (oocytes); each bringing half of the genetic heritage of the future individual. This genetic heritage may be compromised by various processes a very common one being radical attacks in cases of oxidative stress, it could diminish the quality of gametes and therefore the success of reproduction. It is well described that sperm DNA oxidative damage (SDOD) is one of the causes of male infertility frequently associated with reproductive failures in the context of natural or artificial conception in humans. To better understand the impact of oxidative damage on the quality of sperm, our team generated mouse models with SDOD. Characterization of these transgenic mice has shown that SDOD alone was associated with increased embryo defects, miscarriages and perinatal mortality. The team also brought forward that SDOD concerns preferentially specific regions of the sperm nucleus. In addition, the team suggested that sperm chromosomes were not identically susceptible to oxidative damage.In this thesis work, I developed and implemented fluorescence in situ hybridization (FISH) protocols to study chromosome positioning in the mouse sperm. I confirmed that susceptibility of chromosomes to oxidation is determined by their position in the murine sperm nucleus. In parallel, I addressed the question whether the sperm epigenome (particularly DNA methylation) was modified in a situation of post-testicular oxidative stress. Preliminary results seem to show that SDOD disturbs the sperm epigenetic profile.In parallel, in a pre-clinical approach, I contributed to show that an oral antioxidant supplementation can correct SDOD by modifying the antioxidant response of the epididymal tissue of the treated animals.

DNA methylation is an important epigenetic mark spanning all of life's kingdoms. In humans, DNA methylation has been associated with a wide range of age-related pathologies, including type II diabetes and cancer. More recently, in humans, changes in DNA methylation at specific positions in the genome have been found to be predictive of chronological age. Interestingly, DNA methylation age is also predictive of health status and time-to-death. A better understanding of what these DNA methylation changes represent and whether they might be causative in the ageing process will be important to ascertain. However, at present there is no animal model system with which this process can be studied at a mechanistic level. Furthermore, it is becoming increasingly apparent that many disease states that increase in prevalence with age are not caused by all cells within the individual, but are often the result of changes to a subset of cells. This underscores the importance of studying these processes at the single cell level. The recent advances in single cell sequencing approaches now mean that we can study multiple layers of biology within the same single cell, such as the epigenome and the transcriptome (scM&T-Seq). Unfortunately, we are still only able to probe these important aspects of single cell biology in a static sense. This is a major limitation in the study of ageing because ageing and age-related disease processes are inherently dynamic. As such, it is incumbent upon us to develop approaches to assay single cell biology in a dynamic manner. 
In this thesis, I describe an epigenetic age predictor in the mouse. This predictor is tissue-independent and can accurately predict age (with an error of 3.33 weeks) and can record deviations in biological age upon interventions including ovariectomy and high fat diet both of which are known to reduce lifespan. Next, I describe the analysis of a homogeneous population of muscle satellite cells (MuSCs) that I have interrogated at the single cell level, using single cell combined transcriptome and methylome sequencing (scM&T-seq). I found that with age there was increased global transcriptional variability and increased feature-specific methylome variability. These findings explain the loss of functionality of these cells with age. Lastly, I describe two imaging approaches to study DNA methylation dynamically in single cells. Using these methods, I demonstrate that it is possible to accurately determine methylation status across a wide spectrum of global methylation levels and that by using such approaches novel information about dynamic methylation processes can be obtained. These methods represent the first to study DNA methylation dynamically.