Long terminal repeat (LTR) retrotransposons account for approximately 10% of mouse and 8% of human genomes and may play a role in modifying gene expression. Many species harbor retrotransposon families encompassing both autonomous and non-autonomous members. Specifically, the mouse Early Transposon (ETn) family members lack all retroviral genes but are transcriptionally and retrotranspositionally active, causing over 20 known insertional germline mutations. ETns owe their retrotransposition potential to proteins encoded by structurally intact MusD retrotransposons with whom they share LTRs. ETn elements are transcribed at a much higher level than MusD retrotransposons in embryos and undifferentiated cells, suggesting their evasion of host restriction mechanisms. However, mechanisms responsible for the replicative success of non-autonomous retrotransposon subfamilies over their coding-competent relatives are poorly understood. In the first stage of my research, I analyzed regulatory sequences in an ETn LTR responsible for its high promoter activity in the undifferentiated cell line P19. I found that three GC-boxes that may function as Sp1/Sp3 binding sites act synergistically and are indispensable for undifferentiated cell-specific promoter activity of the LTR. Sp1 binding partners may be responsible for the restricted ETn expression. Moreover, I have shown that unlike many retroviruses, ETn elements possess multiple transcription initiation sites and that they have amplified via intracellular retrotransposition in the P19 teratocarcinoma cell line. In the next step of my research, I performed analysis of epigenetic mechanisms as a means of ERV suppression. Specifically, I showed that in embryonic stem cells, autonomous MusD retrotransposons are epigenetically suppressed to a greater degree than non-autonomous ETn retrotransposons, illustrated by a higher level of DNA methylation and a lower level of active histone modifications. I hypothesize that MusD elements may be silenced by DNA methylation and repressive chromatin spreading into the LTR from the CpG-rich internal retroviral sequence absent in ETn elements. I propose that internal structure largely devoid of high CG content enables ETn elements to evade host-imposed transcriptional repression, contributing to their high mutagenic activity in the mouse germline. / Medicine, Faculty of / Medical Genetics, Department of / Graduate

No description available.

Background: Rheumatoid arthritis (RA) is a common complex autoimmune disorder which is influenced by both genetic and environmental factors. While multiple factors that influence susceptibility to and outcome of disease have been identified there is still a large proportion of missing heritability and limited understanding of disease pathogenesis. In recent years, biologic drug therapies have advanced treatment of RA; however good disease control is achieved in just 30% of patients, making identification of predictors of treatment response important. One area of research which is yet to be explored in relation to treatment response, and requires further evaluation in RA susceptibility, is epigenetics. Epigenetics is the study of modifications of the DNA which can influence gene expression but do not alter genetic sequence, and the most commonly studied epigenetic phenomenon, to date, is DNA methylation. Objectives: To identify DNA methylation signatures predictive of treatment response to anti-TNF biologics, to explore the role of DNA methylation in RA susceptibility using disease discordant monozygotic (MZ) twins, and to assess the effect of cryopreservation of cells on DNA methylation. Methods: Genome-wide DNA methylation levels were measured using the HumanMethylation450 BeadChip in pre-treatment whole blood DNA samples from individuals who had extremely good or extremely poor response to the anti-TNF therapies, etanercept and adalimumab, and in MZ twins discordant for RA (n=79 pairs). I also compared genome-wide methylation in cells which had been cryopreserved with fresh cells, to investigate if this technique is suitable for epigenetic investigations. Results: I identified four methylation sites which were significantly related to response to etanercept at a false discovery rate of 5%, the most significantly differentially methylated of which maps to the LRPAP1 gene (p=1.46E-8). Indeed, four other sites at the same locus also showed evidence for differential methylation indicating that this represents a differentially methylated region. No sites were significantly associated with response to adalimumab after correction for multiple testing. I identified subtle differences in DNA methylation between RA discordant twins. Although these were not statistically significant following adjustment for cell composition, one of the most differentially methylated positions mapped to the ZNF74 gene (p=4.97E-6), and replicated a methylation difference identified in the largest previous epigenome-wide association study of RA cases and unrelated healthy controls. I found that cryopreservation of cells does not significantly alter the methylome, an important observation that will impact upon design of future studies. Conclusions: In the largest studies of DNA methylation in RA treatment response and RA discordant MZ twins to date, I identified significant differential methylation in etanercept response, but not adalimumab response, and found small differences in methylation in RA discordant MZ twins. I also concluded that cryopreservation does not significantly alter the methylome.

Le mélanome cutané est le cancer de la peau le plus agressif. Il représente moins de 5% des cancers de la peau mais sa forme métastatique est responsable de 60 à 80% des décès. La médiane de survie des patients atteints d’un mélanome métastatique n’est que de 6 à 9 mois avec les chimiothérapies classiques ou ciblées. L’immunothérapie a été un grand progrès thérapeutique, néanmoins la prise en charge du mélanome métastatique souffre encore de trois points négatifs, tous les patients ne répondent pas aux différents traitements, l’efficacité des chimiothérapies ciblées est limitée par l’apparition rapide de résistances, les cliniciens manquent de marqueurs prédictifs de l’évolution dès les stades précoces pour la prise en charge. Dans ce contexte, notre groupe s’intéresse à la méthylation de l’ADN associées à l’agressivité du mélanome. La méthylation est catalysée par les méthyltransférases de l’ADN (DNMTs). Lorsque celle-ci est présente au niveau d’îlots CpGs localisés dans les promoteurs des gènes, elle empêche la machinerie transcriptionnelle de se mettre en place et inhibe l’expression du gène correspondant. Le profil de méthylation globale de l’ADN étant altéré dans le cancer, elle a donc un rôle fonctionnel dans le processus tumoral mais peut aussi être utilisée en tant que biomarqueur, chaque cancer ayant un profil de méthylation différent. De plus, au sein d’un même cancer, ce profil évolue avec la progression de la tumeur. Au cours de cette thèse, j’ai identifié des modifications du profil de méthylation de l’ADN associées à l’agressivité du mélanome et j'ai sélectionné 9 loci hyperméthylés candidats biomarqueurs. J’ai d’abord comparé les profils de méthylation au niveau du génome entier de lignées cellulaires de mélanome correspondant à des stades d’agressivité différents. À partir de ces informations, des loci candidats ont été choisis par une analyse de la répartition de ces loci hyperméthylés sur le génome couplée à une analyse bioinformatique des fonctions et interactions des gènes associés. Ensuite, leur statut de méthylation a été validé par une technique différente (collaboration avec le Dr. J. Tost, CNG, Evry). Une fois leur hyperméthylation confirmée, j’ai entrepris l’analyse de la méthylation de l’ADN au niveau de ces gènes dans des échantillons de tumeurs primaires issues de patients montrant des survies différentes (Collaboration : L. Lamant, N. Meyer, IUCT, Toulouse, France; L. Lanfrancone IFOM, Milan, Italie). Notre étude confirme le statut hyperméthylé des gènes retenus dans les échantillons métastatiques par rapport aux échantillons de tumeurs primaires. De plus il apparaît un lien entre le niveau de méthylation de la tumeur primaire et le délai d’apparition de métastases ainsi que la survie globale des patients. Parmi les loci sélectionnés, j’ai déterminé si le statut hyperméthylé de ces loci est corrélé à leur sous- expression dans le but d’étudier leur rôle dans l’agressivité du mélanome, et si ces loci peuvent être déméthylés et ré-exprimés par un inhibiteur de DNMTs. Cela a amené à l’identification d’un gène et d'un miR peu décrits dans la littérature, dont les expressions sont corrélées au statut de méthylation et modulées par un traitement déméthylant l'ADN. J’ai entrepris de comprendre leur rôle dans l’agressivité du mélanome et évaluer leur intérêt potentiel en tant que cibles anti-tumorales. Pour cela, j’ai utilisé des tests fonctionnels pour étudier les conséquences de la surexpression dans la lignée métastatique ou de l’inhibition dans la lignée primaire. Les résultats suggèrent une régulation épigénétique fine de ce miR et de ce gène pendant la progression du mélanome, retrouvée indépendamment par notre analyse des tumeurs de patients de la banque TCGA. / Cutaneous melanoma is the most aggressive skin cancer and represents less than 5% of skin cancers but its metastatic form is responsible for 60-80% of the deaths. The median survival for patients with metastatic melanoma is only 6 to 9 months with conventional chemotherapy or targeted therapy. Despite recent therapeutic advances with the immune-therapies, the treatment of metastatic melanoma still suffers from three negatives drawbacks: 1) All patients do not respond to the different treatments. 2) The effectiveness of the targeted chemotherapy is limited by the rapid emergence of resistance. 3) Predictive biomarkers of the evolution of the disease are lacking for the clinicians. In this context, we studied the epigenetic regulations associated with the aggressiveness of melanoma, particularly in DNA methylation. DNA methylation is catalyzed by DNA methyltransferases (DNMTs). When CpGs islands are methylated and located in the promoters of genes, expression of the corresponding gene is inhibited. Commonly, DNA methylation profile is altered in cancer and plays a role in tumorigenesis and tumor maintenance. In addition, the alterations of the DNA methylation profile can be used as biomarkers for prognostic and diagnostic. Here, I identified changes in the DNA methylation profile associated with the aggressiveness of melanoma and selected 9 candidates loci that are hypermethylated in the most aggressive forms of melanoma. First, I compared the methylation patterns genome-wide (450K BeadChip methylation) of melanoma cell lines bearing different aggressiveness. The loci biomarker candidates were selected by combining an analysis of the distribution of these hypermethylated loci on the genome and a bioinformatic analysis of the functions and interactions of the associated genes. Their methylation status was validated by a different technique in collaboration with Dr. J. Tost, CNG, Evry. Once confirmed their hypermethylation, I started to analyze the DNA methylation of the selected genes in the pairs of cell lines of different aggressiveness, such as the primary tumor compared to the metastasis from the same patient, and in patients samples of primary tumors (Collaboration L. Lamant, N. Meyer, iUCT, Toulouse, France; L. Lanfrancone IFOM, Milan, Italy). A total of twenty samples were analyzed. Our study confirms the status of selected hypermethylated genes in metastatic samples compared to primary tumors. Moreover there is a link between the level of methylation of the primary tumor and the overall survival. A patent is being filed in and new samples are collected in order to extend the patient cohort to validate the biomarkers in prognosis of the evolution of the disease. Among the selected loci, I determined whether their hypermethylated status is correlated with their under-expression. For this, I measured their level of expression in a couple of cell lines WM115 / WM266-4 by RT-PCR and RT-qPCR. Then, I chose the loci for which I observed a correlation between methylation and expression. In addition, I studied whether these loci can be demethylated and re-expressed by a DNMTs inhibitor. This led to the identification of a microRNA and gene not fully described in the literature, of which the expression is correlated to DNA methylation status and are reactivated upon treatment with demethylating agents. I explored the role of the microRNA and the gene in the aggressive features of the metastatic melanoma suggesting a tumor suppressive function. These data were further comforted by the data analysis of patient samples.

Lactate has been long thought of as a dead-end waste product of glycolysis. In the brain, recent evidence has revealed a key role of L-lactate creating a paradigm-shift in our understanding of the neuronal energy metabolism. The Astrocyte neuron lactate shuttle (ANLS) model, has shown L-Lactate as the main energy substrate delivered by astrocytes to neurons to sustain neuronal oxidative metabolism. This metabolic coupling is an essential mechanism for long-term memory formation. Experimental evidence indicates that the role of lactate in cognitive function is not limited to being a neuronal metabolic substrate, but rather it is also an important signaling molecule for synaptic plasticity. One of the new emerging roles of lactate is its effect on gene expression levels; however, our current understanding of the mechanism of lactate effect on gene expression is rudimentary. Here, I investigate the role of lactate as an epigenetic modulator in neuronal cultures. First, I explored the effect of lactate on the transcriptome and methylome of the neuronal cells using primary neuronal cell culture models. Our results reveal a significant role for lactate in inducing neuronal cell differentiation. Following, I characterized a neuroblastoma cell line as our neuronal differentiation cell model and assessed its metabolic features relative to other immortal cell lines. Further, using the cell line in vitro model, I looked into the metabolic reprograming that occurs in parallel with the first indications of differentiation, focusing on lactate production rates. Subsequently, I investigated the role of lactate in differentiation through transcriptomic analysis. We show that lactate induced histone acetylation and promoted expression of dopaminergic markers, with a stronger effect of D-lactate over L-lactate. Further studies to establish potential linkages between those two pathways will enhance our understanding of the effect of lactate.

Final publication is available at http://dx.doi.org/10.1016/j.nmd.2016.03.001 / 京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第19928号 / 医博第4148号 / 新制||医||1017(附属図書館) / 33014 / 京都大学大学院医学研究科医学専攻 / (主査)教授 萩原 正敏, 教授 羽賀 博典, 教授 松田 秀一 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 新制・課程博士 / 博士(医学) / 甲第23415号 / 医博第4760号 / 新制||医||1052(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 武藤 学, 教授 浅野 雅秀, 教授 川口 義弥 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Beyond the haploid genome, mammalian sperm carry a payload of epigenetic information with the potential to modulate offspring phenotype. Morphologically mature sperm exit the testes, but cannot swim or interact with the oocyte without extensive remodeling during epididymal transit; this includes modifications to the lipid composition of the sperm membrane, gain of necessary proteins, and a dramatic shift in sperm RNA content. Epididymal maturation has also been linked to changes in the sperm methylome suggesting that the epididymis might play a broader role in shaping the sperm epigenome. First, we characterized the genome-wide methylation landscape in seven germ cell populations from throughout the male reproductive tract. Our data emphasize the stability of cytosine methylation in mammalian sperm, and identify a surprising, albeit transient, period during which sperm are associated with extracellular DNA. Second, given our interest in the small RNA repertoire of sperm we set out to address known bias in sequencing protocols by comparing several small RNA cloning protocols. We found a protocol recently developed by Kathleen Collins’ lab (OTTR) to be superior to commercially available kits in providing an accurate representation of tRNA fragment levels as compared to Northern blotting. These results not only provide a more accurate representation of tRNA fragments, but also more complexity than previously seen allowing us to reassess the true sperm small RNA content. Taken together, these results provide significant insight into the mechanisms and factors modulating sperm epigenomics during post-testicular sperm maturation.

Hematopoietic stem cells (HSCs) give rise to all blood cells and are characterized by their ability for life-long self-renewal and multilineage differentiation. HSC function is regulated by complex cell-intrinsic and -extrinsic pathways, but these regulatory mechanisms are not completely understood. Recent work has demonstrated that the epitranscriptional modification N6-methyladenosine (m6A) has important roles in the regulation of many physiologic and pathologic processes in various cell types, but it was previously unknown if and how m6A may regulate adult HSC function. In this work, I uncover the role for m6A in HSC regulation, both cell-intrinsically in regulating HSC differentiation and cell–extrinsically by regulating the formation the HSC bone marrow niche.

Acute myeloid leukemia (AML) is a hematopoietic malignancy characterized by great heterogeneity and clonal nature. In recent years, rapidly evolving next-generation sequencing methods provided a deep insight into the mutational background of AML. It was shown that ~ 44 % of AML patients harbor mutations in genes that regulate DNA methylation. So far, many researchers have tried to evaluate the prognostic significance of DNA methylation changes in AML, however, due to a great inconsistency in these studies, none of the reported markers were implemented into clinical practice. The aim of this work was to further investigate the DNA methylation changes in AML patients with specific mutations and their prognostic effect. Next, we wanted to develop a new approach for a complex evaluation of prognostically significant DNA methylation aberrations. In our first project, we assessed the overall DNA methylation, hydroxymethylation, and gene expression in AML patients with mutations in either DNMT3A or IDH1/2 or their combinations. We discovered that each genetic aberration is connected with a distinct pattern of DNA hydroxy-/methylation changes that are not entirely reflected in altered gene expression. Patients with mutations in both genes exhibited a mixed DNA methylation profile most similar to healthy...