archives@tulane.edu / 0 / Anna Smith

Le séquençage de nombreux génomes eucaryotes indique que l’augmentation de la «biocomplexité» au cours de l’évolution n’est pas directement corrélée à l’accroissement du nombre de gènes. En d’autres termes, nous sommes plus que la somme de nos gènes et l’ère post-génomique actuelle promet de cerner de façon plus précise les bases moléculaires de notre identité. A cet égard, il semble de plus en plus clair que l’épigénétique est riche d’une information qui se superpose à celle du code génétique. L’information épigénétique est principalement véhiculée par des modifications de l’ADN et des histones. La modification majeure de l’ADN est la méthylation de la cytosine qui est la marque d’une chromatine transcriptionnellement silencieuse. Quant aux histones, différentes modifications posttraductionnelles ont été décrites, comme l’acétylation, la phosphorylation, la méthylation et l’ubiquitinylation. L’ensemble de ces modifications constituerait un «code histone», dont le décryptage n’en est qu’à ses prémices, permettant d’associer à chaque combinaison de modifications un état particulier de la chromatine, et ainsi de l’expression génique. La méthylation des histones a longtemps été considérée comme irréversible mais l'identification récente de déméthylases des histones spécifiques de certains sites a révélé que cette modification est régulée de façon dynamique et réversible. La découverte de ces enzymes a ouvert de nouveaux axes de recherche dans le domaine de l'épigénétique (Klose and Zhang, 2007). Au cours de ma thèse de doctorat, nous nous sommes intéressés à la déméthylase PADI4 (peptidylarginine déiminase 4) qui convertit des résidus arginines des histones H3 et H4, associés à l'activation des gènes, en résidus citrullines, ce qui a pour conséquence d'entraîner une répression transcriptionnelle. Cette réaction porte le nom plus particulier de déimination/citrullination des histones. A l’heure actuelle, il est primordial de cerner comment la déméthylation des histones, et plus précisément la peptidylarginine deiminase 4 (PADI4), réprime la transcription. Dans un premier temps, nous avons mis en évidence un lien mécanistique entre la deméthylation et la désacétylation des histones. Nous avons montré que PADI4 interagit avec l’histone désacétylase HDAC1. Cette enzyme est responsable du décrochage des groupements acétyls des histones, conduisant à la fermeture de la chromatine. Des expériences d’immunoprécipitation de la chromatine indiquent que PADI4 et HDAC1 s’associent au promoteur du gène de réponse aux oetrogènes pS2 simultanément et de manière cyclique. L’utilisation d’une construction shRNA dirigée contre la protéine endogène HDAC1 indique que la liaison de PADI4 au promoteur du gène pS2 est dépendante de la présence de HDAC1. Dans la deuxième partie de notre travail, un lien mécanistique entre la déméthylation des histones par PADI4 et la méthylation de l’ADN a été mis en évidence. La méthylation de l’ADN est catalysée par des enzymes, appelées méthyltransférases de l’ADN (DNMTs), qui transfèrent des résidus méthyls sur les cytosines. Nous avons montré que la protéine DNMT3A interagit avec PADI4. Nous avons également démontré que l’enzyme PADI4 était capable de citrulliner/déiminer (convertir des résidus arginines en résidus citrullines) la méthyltransférase de l’ADN DNMT3A in vitro et que cette citrullination de la protéine DNMT3A par PADI4 stabiliserait DNMT3A in vivo. Enfin, nos récents travaux révèlent une relation mécanistique entre la protéine MeCP2, interprète des signaux de méthylation de l’ADN, et la protéine polycomb EZH2. Celle-ci possède une activité méthyltransférase d’histone sur les lysines 27 de l’histone H3. Nos données montrent que MeCP2 interagit avec EZH2 et que ces protéines fixent des régions promotrices communes. De plus, la déplétion en MeCP2 affecte la présence de EZH2 au niveau de ces régions communes. En conclusion, ce travail de thèse devrait permettre une meilleure compréhension des mécanismes moléculaires de l’épigénétique. Plus particulièrement, il devrait aider à mieux cerner comment la première histone déméthylase décrite, la peptidylarginine déiminase 4 ou PADI4, verrouille l’expression génique.

The Wnt/β-catenin signaling pathway is often deregulated in cancer. Here we investigate Wnt/β-catenin signaling, aberrant accumulation of β-catenin, and epigenetic deregulation in breast cancer and parathyroid tumours. An aberrantly spliced Wnt coreceptor LRP5 (LRP5Δ) is important for accumulation of nonphosphorylated active β-catenin and tumour growth in parathyroid tumours. Paper I demonstrated frequent expression of LRP5Δ in breast tumours and substantiated that breast tumour cell growth was dependent on continuous activation of the Wnt/β-catenin pathway by LRP5Δ. A LRP5 antibody reduced the levels of active β-catenin, inhibited tumour cell growth and caused apoptosis in breast cancer cells. Antibody therapy may have a significant role in the treatment of breast cancer. Paper II revealed lost expression of the tumour suppressor gene APC in parathyroid carcinomas, likely due to CpG methylation. Also accumulation of nonphosporylated active β-catenin was observed, indicating activation of Wnt/β-catenin signaling. Treatment of primary parathyroid carcinoma cells with the demethylating agent 5-aza-2’-deoxycytidine reduced the levels of active β-catenin, inhibited cell growth and caused apoptosis, suggesting that adjuvant epigenetic therapy could be considered in patients with metastatic or recurrent parathyroid carcinoma. In paper III we showed that the expression of the tumour suppressor gene HIC1 was generally reduced in parathyroid tumours of primary and secondary origin, and parathyroid carcinomas. Overexpressing HIC1 reduced cell viability and suppressed colony formation, supporting a tumour suppressor role in the parathyroid gland. Results suggested that the observed underexpression of HIC1 could be explained by epigenetic deregulation involving histone methylation rather than CpG methylation. Paper IV demonstrated increased expression of the histone methyltransferase EZH2 in parathyroid tumours of primary and secondary origin, and most apparent in parathyroid carcinomas. Decreasing EZH2 resulted in reduced cell viability and colony formation capacity suggesting that EZH2 may function as an oncogene in parathyroid tumours. Furthermore, depletion of EZH2 also reduced the amount of active β-catenin. EZH2 may represent a novel therapeutic target in parathyroid tumours. The fact that HIC1 was underexpressed and EZH2 overexpressed in parathyroid tumours regardless of the hyperparathyroid disease state may represent a possibility for a common pathway in parathyroid tumour development.

Characterized by a wide histological spectrum, myxofibrosarcoma ranges from deceptively bland-appearing lesions to frankly pleomorphic sarcomas, representing a suitable model to elucidate the molecular aberrations in multistep disease progression. Using genome-wide array comparative genomic hybridization, we have profiled DNA copy number alterations in myxofibrosarcoma samples and cell lines in coupled with expression profiling data and identified prominent SKP2 amplification on 5p, frequent homozygous deletion of MTAP on 9p, as well as loss expression of ASS1, a candidate tumor suppressor. As a predominant driving mechanism, SKP2 gene amplification was detected in one-third of cases in independent cohort validation and associated with SKP2 immunohistochemical expression, adverse prognosticators, and worse patient survival. Besides the classical attribute in promoting cell proliferation and tumor growth, we have confirmed the pro-metastatic oncogenic function of SKP2 and identified differentially expressed motility-promoting genes as its potential mediators, including ITGB2, ACTN1, IGF1, and ENAH. Since MTAP and ASS1 are key enzymes in either salvage pathway of nucleologenesis and amino acid biosynthesis, their deficiency suggests alterations involving metabolic homeostasis is important in sarcomagenesis. We in our studies not only validated their deletion and/or epigenetic silencing in myxofibrosaroma, evaluating the therapeutic responses of L-alanosine and ADI-PEG20 in MTAP or ASS1 deficient cells in vitro and in vivo, but also confirmed their tumor suppressor functions with special focus on tumor angiogenesis. Our study provides further insight into the molecular pathogenesis in tumor progression and highlights the prognostic, biological, and potential therapeutic relevance in myxofibrosarcoma.

Epigenetic regulation of gene expression by DNA methylation plays a central role in the maintenance of cellular homeostasis. Here we present evidence implicating the DNA methylation program in the regulation of hypoxia-inducible factor (HIF) oxygen-sensing machinery. We show that DNA methyltransferase 3a (DNMT3a) methylates and silences the HIF-2alpha gene (EPAS1) in normal cells. Epigenetic silencing of EPAS1 prevents activation of the HIF-2alpha gene program associated with hypoxic cell growth, thereby limiting the proliferative capacity of cells under low oxygen tension. Naturally occurring defects in DNMT3a, observed in primary tumours and malignant cells, cause the unscheduled activation of EPAS1 in early dysplastic foci. This enables incipient cancer cells to exploit the HIF-2alpha pathway in the hypoxic tumour microenvironment, which is necessary for the formation of cellular masses larger than the oxygen diffusion limit. Reintroduction of DNMT3a in DNMT3a-defective cells restores EPAS1 epigenetic silencing, prevents hypoxic cell growth, and suppresses tumour growth in vivo. In addition, restoring HIF-2alpha expression in DNMT3a-reintroduced cancer cells restores full tumorigenic potential, including the capacity to traverse the hypoxic barrier. These data support a tumour-suppressive role for DNMT3a as an epigenetic regulator of the HIF-2alpha oxygen-sensing pathway and the cellular response to hypoxia.

Muscle tissue is grown and maintained by muscle stem cells termed satellite cells. Activated satellite cells become myoblasts, which must proliferate then differentiate into functional muscle. This process, known as myogenesis, is controlled by a cascade of epigenetic regulatory events. One facet of this regulation is histone acetylation, which can be influenced by the availability of metabolites within a cell. In this study, the ability of glucose, pyruvate, or glutamine to change histone acetylation levels in cultured myoblasts was investigated. Changing concentrations of glucose or pyruvate had no effect but decreasing the availability of glutamine in cell culture from 2mM to 0.2mM resulted in proliferating myoblasts accruing a hyperacetylated histone phenotype. However, when the same concentration of glutamine was used on differentiating myoblasts the hyperacetylated phenotype was lost and no change to differentiation was observed. This study demonstrates the potentials and limitations of altering epigenetic acetylation with metabolic circumstance. -- Le développement du tissu musculaire est soutenu par les cellules souches musculaires, communément appelées cellules satellites. Les cellules satellites activées se transforment en myoblastes qui doivent ensuite proliférer et se différencier en muscle fonctionnel. Ce processus, connu comme myogenèse, est contrôlé par une cascade de régulation épigénétique. Un aspect de ce processus est l’acétylation d’histones, qui peut être influencée par la disponibilité de métabolites dans la cellule. Dans cette étude de cas, la capacité du glucose, pyruvate, ou glutamine à changer les niveaux d’acétylation d’histones a été examinée. Le changement des concentrations de glucose ou de pyruvate n’a généré aucun effet, mais la diminution de la disponibilité de la glutamine dans la culture cellulaire de 2mM à 0.2mM a eu pour résultat une prolifération de myoblastes présentant un phénotype d’histones hyper-acétylées. Pourtant, quand la même concentration de glutamine a été utilisée pour différencier les myoblastes, le phénotype hyper-acétylé n’a pas été observé et aucun changement de différenciation n’a pu être détecté. Cette étude démontre le potentiel et les limites des modifications de l’acétylation épigénétique selon les circonstances métaboliques.

Yes / During development, multipotent progenitor cells establish lineage-specific programmers of gene activation and silencing underlying their differentiation into specialized cell types. We show that the Polycomb component Cbx4 serves as a critical determinant that maintains the epithelial identity in the developing epidermis by repressing nonepidermal gene expression programs. Cbx4 ablation in mice results in a marked decrease of the epidermal thickness and keratinocyte (KC) proliferation associated with activation of numerous neuronal genes and genes encoding cyclin-dependent kinase inhibitors (p16/p19 and p57). Furthermore, the chromodomain- and SUMO E3 ligase-dependent Cbx4 activities differentially regulate proliferation, differentiation, and expression of nonepidermal genes in KCs. Finally, Cbx4 expression in KCs is directly regulated by p63 transcription factor, whereas Cbx4 overexpression is capable of partially rescuing the effects of p63 ablation on epidermal development. These data demonstrate that Cbx4 plays a crucial role in the p63-regulated program of epidermal differentiation, maintaining the epithelial identity and proliferative activity in KCs via repression of the selected nonepidermal lineage and cell cycle inhibitor genes.

Alzheimer’s disease is a chronic, neurodegenerative disease characterised by amyloid plaque accumulation, neurofibrillary tangles and eventual neuronal cell loss. The complex aetiology exhibited in late-onset Alzheimer’s disease presents a considerable challenge in the field of genetics, with identified variants from genome-wide association studies collectively only explaining about a third of disease incidence. As such, new avenues are being explored to elucidate underlying mechanisms associated with disease onset and progression. In 2014, the first epigenome-wide association studies in Alzheimer’s disease were published, identifying several, novel differentially methylated loci in the nuclear genome in cortical brain samples, highlighting that epigenetic mechanisms may play a role in disease aetiology. Further, a growing body of evidence has implicated mitochondrial dysfunction as an early feature of disease pathogenesis. Despite this, few studies have investigated the role of mitochondrial DNA epigenetics in Alzheimer’s disease. Indeed, the relatively nascent field of mitochondrial epigenetics has largely been restricted to candidate-based gene approaches to identify differential methylation associated with disease. The main aim of this thesis was therefore to design an experimental and bioinformatic pipeline for the analysis of mitochondrial DNA methylation in post- mortem human brain tissue; first in healthy non-demented control donors, and subsequently in individuals with Alzheimer’s disease. Our work therefore represents the first epigenome wide studies of mitochondrial DNA methylation at single nucleotide resolution, providing a framework not only for mitochondrial DNA methylation in Alzheimer’s disease, but also in a number of complex diseases characterised by mitochondrial dysfunction.

No description available.

It is well established that epigenetics and chromatin modifications are important factors that can govern gene activity and nuclear architecture. They are also proven to be essential for normal embryonic development and cell differentiation. One important event during mouse development is the establishment of epigenetic reprogramming which is believed to be essential for normal growth and development, however; the mechanism is still poorly understood. The general objective of this PhD study was to investigate the profiles and mechanisms of epigenetic and chromatin modifications during normal mouse development and in embryonic stem cells. Mouse pre- and postimplantation embryos and ES cells were used in experiments employing a range of different methodologies. The dynamics of epigenetic DNA and histone methylation were captured using laser confocal immunofluorescent microscopy and western blotting. The activity of epigenetic modifiers was monitored by real-time PCR and candidate genes were validated using siRNA technology. The present studies demonstrate that heterochromatin markers H3K9me3, H3K9me2, H4K20me2, H4K20me3, HP1α and HP1β are reprogrammed during early development. Demethylation of H3K9me2, H3K9me3 and H4K20me3 took place at two-cell stage and remethylation occurred at four-cell stage except for H4K20me3. The reestablishment of H4K20me3 was initially observed in early postimplantation embryos in extraembryonic tissue, specifically in the mural trophectoderm. In embryonic tissue, H4K20me3 was not clearly detected until in mid to late postimplantation development. The mechanism of H3K9me2 and H3K9me3 demethylation might be due to either an imbalance of epigenetic modifiers or the presence of Jmjd2a and Jmjd1a histone demethylase postfertilisation. We have also report evidence that HP1α and Suv4-20h are required in heterochromatin before the recruitment of H4K20me3 during mouse development and in ES cells. Therefore H4K20me3 removal was believed to involve the lack of prerequisite heterochromatin complexes such as HP1α and Suv4-20h enzymes. Furthermore, the presence and levels of H4K20me3 and HP1α might be strongly associated with cell differentiation and tissue maturation in mouse in vivo development but not in vitro early differentiated ES cells. Surprisingly, the results showed that chromatin modifications and their modifiers in ES cells are different from ICM and epiblast. Chromatin modifications H4K20me3 and HP1α were absent from ICM and epiblast, but were detected in ES cells. Notably, H4K20me3 and HP1α were established after early incubation of ICM into ES cell medium, but this change was not dependent on the presence of serum and leukaemia inhibiting factor. Epigenetic modifier Jmjd2a but not Jmjd1a was found in ICM. Conversely, Jmjd1a is highly expressed in ES cells while Jmjd2a was inactivated. In addition, the present studies revealed the substantial role of histone demethylases in development, as it may be important for epigenetic reprogramming. The results demonstrated that inhibition of demethylase Jmjd2a and Jmjd1a caused preimplantation embryos to arrest at the twocell stage while Jmjd2c deficient embryos failed to reach blastocyst. Thus it is possible that Jmjd2a and Jmjd1a were essential for epigenetic reprogramming while Jmjd2c is critical for cell fate establishment during blastocyst formation. In conclusion, the global chromatin signature in ES cells differs from ICM and epiblast; heterochromatin reprogramming occurs at two-cell stage; maturation of heterochromatin occurs at postimplantation; and histone demethylases Jmjd1a, Jmjd2a and Jmjd2c are important in preimplantation development. Results from the present studies could provide crucial information for developmental biology and stem cell research, and provide as a model for improvement of reproductive biotechnologies such as somatic cell reprogramming, and diagnosis of epigenetic abnormalities in early development.