Cytosine methylation of CpG dinucleotides is an epigenetic modification that cells use to regulate gene expression, largely to promote transcriptional silencing. Focal hypermethylation of tumor suppressor genes (TSGs) accompanied by genomic hypomethylation are epigenetic hallmarks of malignancy. DNA methyltransferase 1 (DNMT1) is the principle vertebrate enzyme responsible for maintenance of DNA methylation and its dysregulation has been found to lead to aberrant methylation in cancer. In addition, recent findings demonstrated that the ten-eleven translocation 1 (TET1) protein functions as a 5-methylcytosine dioxygenase that converts 5-methylcytosine (5mC) bases to 5-hydroxymethylcytosine (5hmC) to mediate active DNA demethylation. Emerging evidence suggests that TET1 might function as a TSG. To understand the dynamic regulation of DNA methylation during cellular transformation, my work focused on intrinsic regulation of DNMT1 and how TET1 regulates DNA demethylation in generating a cancer methylome. The replication foci targeting sequence (RFTS) is an N-terminal domain of DNMT1 that inhibits DNA-binding and catalytic activity, suggesting that RFTS deletion would result in gain of DNMT1 function. However, other data suggested that RFTS may be a positively acting domain. To test biochemical and structural predictions that the RFTS domain of DNMT1 is inhibitory, we established cellular systems to evaluate the function of DNMT1 alleles. The data indicate that deletion of RFTS is necessary and sufficient to promote cellular transformation, focal hypermethylation of specific TSGs, and global hypomethylation. These data and human mutation data suggest that RFTS domain is a target of tumor-specific dysregulation. RAS mutations are frequently in multiple malignancies. Methylation-associated silencing of TSGs is a hallmark of RAS-driven-tumorigenesis. I discovered that suppression of TET1 by the ERK signaling cascade is responsible for promoter hypermethylation and the malignant phenotype in KRAS-transformed cells. Restoration of TET1 expression reactivates silenced TSGs and reduces colony formation. Moreover, TET1 knockdown in a cell depleted for KRAS is sufficient to rescue the inhibition of colony formation by KRAS knockdown. My findings suggest that dysregulated TET1-mediated DNA demethylation is a target responsible for epigenetic changes in cancers with KRAS activation.

Non-alcoholic fatty liver disease (NAFLD) now affects around one in four adults in the human population and parallels the global increase in obesity. Within the spectrum of NAFLD, simple steatosis is associated with insulin resistance and type 2 diabetes while progression to steatohepatitis (NASH) is associated with an increased risk of liver cirrhosis and all-cause mortality. The molecular pathology of NAFLD is incompletely understood, however observational studies in human cohorts suggest the regulation of DNA methylation may play a role. 5-hydroxymethylcytosine (5hmC) is a cytosine modification generated from 5- methylcytosine (5mC) by the Ten eleven translocase isoenzymes (Tets) as part of a demethylation process. The aim of this project was to examine the role of Tet enzyme activity on the pathogenesis and progression of NAFLD. Detailed characterisation of two established murine dietary interventions allowed the selection of a NAFLD mouse model which broadly recapitulated the metabolic, histological and transcriptional features of human disease. Using DNA immunoprecipitation coupled with whole genome next generation sequencing and RNA micro expression arrays I examined the effect of high fat diet feeding (HFD) on hepatic DNA 5hmC levels within annotated gene regions. Whilst the global 5hmC profile was not altered by HFD, there was profound genic enrichment of 5hmC in upregulated mediators of cholesterol synthesis and transport (Lss, Sc4mol, Fdps, Hsd17b7, Cyp17a1, Mvd, Cyp1a2, Dhcr7 and Apoa4) with no enrichment in genes with other pathological functions (drug detoxification, inflammation, cell cycle regulation). Induced peaks of 5hmC enrichment were subsequently abolished following rescue of the NAFLD phenotype by conversion to control diet. Cross species validation was performed in vitro utilising embryonic stem cell derived hepatocytes challenged with a cocktail of high energy substrates. My in vivo findings were broadly replicated with specific 5hmC enrichment in genes synthesising lipotoxic molecules (PLIN2, CIDEC, APOA4, ACADVL, HMGCS2, APOA5, CYP2J2, IGFBP1, PPAP2C, ACSL1, APOC3, ANGPTL4, NRG1) with no enrichment in upregulated genes of alternative function. To determine whether or not the 5hmC enrichment seen is of functional relevance, I studied Tet1-/- C57BL/6J mice. Tet1-/- mice are grossly normal in appearance, however loss of Tet1 conferred a striking resistance to diet induced obesity with reduced body fat mass, improved insulin-sensitivity and near complete absence of NAFLD compared to wild type littermates. Furthermore, the HFD fed Tet1-/- liver transcriptome showed a ‘protective’ profile, with suppression of genes for lipid synthesis, inflammation and fibrosis. Thus, in multiple cross-species models of NAFLD, over nutrition induces genic hydroxymethylation specifically within activated genes driving the synthesis and transport of lipid molecules. Such changes are reversible with resolution of the NAFLD phenotype strengthening functional association. Tet1 deficiency conveys an obesity and NAFLD resistant phenotype. I therefore introduce Tet1 mediated hydroxymethylation as a novel mechanism for NAFLD pathogenesis.

DNA hypermethylation is known to contribute to the formation of cancer and this process has been widely studied. However, DNA hypomethylation has received far less attention and the factors controlling the balance between hypo and hypermethylation and its impact on tumorigenesis remains unclear. TET1 is a DNA demethylase that regulates DNA methylation, hydroxymethylation and gene expression. Full length TET1 (TET1FL) has a CXXC domain that binds to un-methylated CG islands (CGIs). This CXXC domain allows TET1 to protect CGIs from aberrant methylation but it also limits its ability to regulate genes outside of CGIs. This dissertation reports a novel isoform of TET1 (TET1ALT) that has a unique transcription start site from an alternate promoter in intron 2, yielding a protein with a unique translation start site. Importantly, TET1ALT lacks the CXXC domain but retains the catalytic domain. TET1ALT is repressed in ESCs but becomes activated in embryonic and adult tissues while TET1FL is ex / Biomedical Sciences

Glioblastome machen etwa 47% aller intrinsischen Tumore des zentralen Nervensystems aus. Sie sind durch ein aggressives und invasives Wachstumsverhalten gekennzeichnet. Als erster wegweisender Schritt zur Therapie der Glioblastome gilt die frühe und möglichst vollständige Resektion, gefolgt von einer simultanen Radiochemotherapie. Dennoch sind Tumorrezidive binnen weniger Monate die Regel und es bestehen trotz intensiver Forschung bis heute kaum alternative Behandlungsoptionen. Das Verständnis für die Pathogenese der Glioblastome erfuhr in den letzten Jahren tiefgreifende Änderungen. Nach Berücksichtigung der molekularen Marker in der WHO- Klassifikation, wurden Glioblastome in zwei molekulare Gruppen unterteilt: die IDH-Wildtyp (95% der Fälle) und die IDH-mutierten Glioblastome. IDH-Wildtyp Glioblastome treten bei Patienten mit einem medianen Alter von 64 Jahren auf und gehen mit einer ungünstigen Prognose (medianes Überleben 14,2 Monate) einher. IDH-mutierte Glioblastome kommen vor allem bei jüngeren Patienten mit einem medianen Alter von 45 Jahren vor und weisen eine vergleichsweise bessere Prognose mit einem medianen Überleben von 4-5 Jahren auf. Bei IDH-Wildtyp Glioblastomen wurden am häufigsten TERTp-, und PTEN-Mutationen sowie EGFR-Amplifikationen beschrieben. Dabei stellt die TERTp-Mutation die häufigste somatische Alteration im Genom der IDH-Wildtyp Glioblastome dar. Die oben genannten molekularen Marker bieten eine solide Grundlage für die molekulare Diagnose der IDH- Wildtyp Glioblastome. Dennoch bleibt die Frage unbeantwortet, warum IDH-Wildtyp Glioblastome vor allem bei älteren Patienten auftreten und jüngere Patienten eine bessere Prognose besitzen. Unter der Annahme, dass IDH-Wildtyp Glioblastome ein altersspezifisches molekulargenetisches Profil aufweisen, welches wohlmöglich die Prognose beeinflusst, wurden daher die Tumor- und korrespondierenden Blutproben von 55 Patienten mittels Whole Exome Sequencing untersucht. Nach Filterung der Rohdaten wurden verschiedene Mutationen und Copy Number Variations identifiziert. Zur Validierung der Methode wurde zum einen ein Literaturabgleich der detektierten Alterationen durchgeführt und zum anderen einzelne Kandidatengene mittels Sanger Sequenzierung manuell bestätigt. Insgesamt wurden 1841 Mutationen auf 1544 verschiedenen Genen detektiert. Obwohl viele der 1544 Mutationen ohne Relevanz für die Pathogenese waren, konnte eine enorme Vielzahl an verschiedenen Treibermutationen nachgewiesen werden. Die manuell sequenzierte TERTp-Mutation war mit 76,4% am häufigsten aufgetreten. Weitere Treibermutationen, beispielsweise EGFR, TP53, PTEN, PI3K-Gruppe, NF1 und PDGFRA zeigten mit der Literatur vergleichbare Prävalenzen und betonten die Validität der Methode. Die aufgetretenen Copy Number Variations belegten sowohl auf chromosomaler (Chromosom 7-Amplifikation und Chromosom 10-Deletion), als auch auf genspezifischer Ebene (EGFR-, CDK6-, MET-, PDGFRA-Amplifikation und CDKN2A/B-, und PTEN-Deletion) die molekulargenetischen Charakteristika des IDH-Wildtyp Glioblastoms. Über eine Clusterung dieser Alterationen und Gegenüberstellung mit klinischen Eigenschaften konnten die typischen, publizierten Glioblastomsignaturen (proneural, klassisch, mesenchymal) beschrieben und mit dem Erkrankungsalter in Verbindung gebracht werden. Besonders eindrücklich zeigten unsere Daten, übereinstimmend mit der Literatur, dass eine proneurale Signatur mit einem jungen Erkrankungsalter und günstiger Prognose assoziiert war. Unerwartet zeigte ein Patient mit pädiatrischer Signatur und hohem Erkrankungsalter dennoch ein überdurchschnittlich vorteilhaftes Überleben, verglichen mit seiner Altersgruppe. Unabhängig von molekulargenetischen Alterationen, war ein junges Erkrankungsalter alleinstehend mit einer günstigeren Prognose verknüpft. Für einzelne molekulargenetische Alterationen konnte kein Zusammenhang mit dem Erkrankungsalter oder dem (Progressionsfreien-) Überleben hergestellt werden. Ein alterspezifisches, prognosebeeinflussendes Mutationsmuster konnte demnach nicht identifiziert werden. Limitierend muss dabei die geringe Kohortengröße (n=55) angemerkt werden. Eine Vergrößerung der Studienpopulation war aufgrund der geringen Inzidenz von jungen Patienten mit IDH-Wildtyp Glioblastomen in diesem Studiendesign nicht möglich und könnte perspektivisch in einem multizentrischen Ansatz oder einer langen Akquirierungsphase verwirklicht werden. Die Fülle an identifizierten Treibermutationen verdeutlichte nichtsdestotrotz die große intratumorale Heterogenität des Glioblastoms. Zusätzlich ermöglichte die enorme diagnostische Tiefe der verwendeten Methode die Identifikation der bisher nicht im Zusammenhang mit dem IDH-Wildtyp Glioblastom beschriebenen TET1-Deletion. Obwohl die detaillierte Rolle der TET1-Deletion für das Glioblastom nicht verstanden ist, liefern unsere Daten einen vielversprechenden Hinweis, dass ein Funktionsverlust des TET1-Enzyms, in Kombination mit EGFR-Amplifikation oder Deletion von PTEN oder CDKN2A/B, eine Auswirkung auf die Pathogenese des IDH-Wildtyp Glioblastoms besitzt und die Prognose negativ beeinflusst. Zukünftige molekulargenetische Sequenzierungen sind indiziert, um die Rolle der TET1- Deletion zu bestätigen und darüber hinaus die Pathogenese des Glioblastoms auf molekulargenetischer Ebene noch besser zu verstehen und weitere individualisierte Therapieansätze abzuleiten.:Abkürzungsverzeichnis 5 1 Einleitung 8 1.1 Klinische Grundlagen des Glioblastoms 8 1.1.1 Epidemiologie/Ätiologie 8 1.1.2 Symptomatik und Diagnostik 9 1.1.3 Therapie 10 1.1.4 Prognose 12 1.2 Pathologie und Molekulargenetische Veränderungen des Glioblastoms 12 1.2.1 Treibergene 13 1.2.2 Subgruppen 17 1.3 Fragestellung 18 2 Materialien 19 3 Methoden 23 3.1 Patientenrekrutierung 23 3.1.1 Erweiterte Einschlusskriterien 25 3.1.2 Ausschlusskriterien 25 3.2 Kohortendesign 25 3.3 Klinische Daten 26 3.4 Materialgewinnung 27 3.4.1 Proben-Lagerung 27 3.4.2 DNA-Extraktion 27 3.5 Sanger Sequenzierung Prä-WES 29 3.5.1 Primer-Design 29 3.5.2 Polymerase Kettenreaktion 29 3.5.3 Elektrophorese 31 3.5.4 Aufreinigung PCR-Produkt 33 3.5.5 Sequenzierreaktion 35 3.6 WES 38 3.6.1 Datensatz 41 3.6.2 CNV-Analyse 42 3.7 Sanger Sequenzierung Validierung 42 3.8 Statistische Auswertung 45 4 Ergebnisse 47 4.1 Deskriptive klinisch-pathologische Beschreibung der Kohorte 47 4.1.1 Klinisch-pathologische Zusammenhänge 54 4.2 TERTp-Sequenzierung 58 4.3 Datensatz WES 58 4.3.1 Somatische Mutationen 59 4.3.2 CNV-Analyse 67 4.4 Altersbezogene molekulargenetische Unterschiede 74 4.5 Zusammenhang zwischen Genotyp und OS sowie PFS 77 4.5.1 Somatische Mutationen 77 4.5.2 CNV-Analyse 78 5 Diskussion 82 5.1 Klinische Charakteristika von Patienten mit IDH-Wildtyp Glioblastom 82 5.2 Der prognostische Stellenwert von Tumoreigenschaften 84 5.3 Molekulargenetisches Profil des IDH-Wildtyp Glioblastoms 85 5.3.1 Somatische Mutationen und Copy Number Variations 85 5.3.2 Chromosomale Aberrationen 94 5.4 10q21.3-Deletion 95 6 Zusammenfassung 97 6.1 Deutsch 97 6.2 Englisch 99 Anlagen 101 Darstellung zur Geschlechtsneutralität im geschriebenen Wort 101 Erklärung zur Eröffnung des Promotionsverfahren 102 Erklärung zur Einhaltung der gesetzlichen Vorgaben 104 Anhang 105 Literaturverzeichnis 113 Abbildungsverzeichnis 125 Tabellenverzeichnis 127 Danksagung 128 / Glioblastomas account for approximately 47% of all intrinsic central nervous tumors. They are characterized by an aggressive and invasive growth pattern. Early and, if possible, complete resection followed by simultaneous radiochemotherapy is crucial for treatment success. However, tumor recurrence within a few months are very frequent. Despite intensive research, there are still hardly any alternative treatment options. The understanding of the pathogenesis of glioblastomas underwent profound changes in recent years. After considering molecular markers in the WHO classification, glioblastomas have been divided into two molecular groups: IDH-wild-type (95% of cases) and IDH-mutated glioblastomas. IDH-wild-type glioblastomas occur in patients with a median age of 64 years and are associated with an unfavorable prognosis (median survival 14.2 months). On the other hand, IDH-mutated glioblastomas are mainly associated with young patients with a median age of 45 years and have a rather good prognosis with a median survival of 4-5 years. In IDH- wild-type glioblastomas, TERTp-, and PTEN-mutations as well as EGFR-amplifications have been described most frequently. Among them, TERTp-mutation represents the most frequent somatic alteration in the genome of IDH-wild-type glioblastomas. The above molecular markers provide a solid basis for molecular diagnosis of IDH-wild-type glioblastomas. To date, however, the reason why IDH-wild-type glioblastomas appear primarily in older patients with younger patients having a better prognosis remains unclear. Assuming that IDH-wild-type glioblastomas have an age-specific molecular signature with possible impact on the outcome, we analyzed tumor and corresponding blood samples from 55 patients by whole exome sequencing. After filtering the raw data, various mutations and copy number variations were identified. To validate the method, a literature review as well as Sanger Sequencing of selected candidate genes was performed. A total of 1841 mutations on 1544 different genes were detected. Although many mutations appeared to be background mutations with no relevance to pathogenesis, an enormous number of different driver mutations remained. The manually sequenced TERTp-mutation was the most abundant at 76.4%. Other driver mutations, for example EGFR, TP53, PTEN, PI3K group, NF1 and PDGFRA showed prevalence comparable to published data and emphasized the validity of the method. The copy number variations that occurred concurred previously described molecular genetic characteristics of IDH wild-type glioblastoma at both chromosomal (chromosome 7 amplification and chromosome 10 deletion) and gene-specific levels (EGFR-, CDK6-, MET-, PDGFRA-amplification and CDKN2A/B-, and PTEN-deletion). Via clustering of these alterations and juxtaposition with clinical features, the typical glioblastoma signatures (proneural, classic, mesenchymal) could be described and related to age of diagnosis. In line with the literature, our data showed that a proneural signature was associated with younger patients and favorable prognosis. Unexpectedly, a patient with a pediatric signature and high age of diagnosis, showed a survival above average compared with his age group. Regardless of molecular alterations, young age was an independent characteristic of favorable prognosis. For individual genetic alterations, no association with age of diagnosis or (progression-free) survival could be established. Thus, an age-specific mutational pattern could not be identified. The relatively small cohort size (n=55) must be noted as a limiting factor. An increase of the study population was not possible in this study design due to the low incidence of young patients with IDH wild-type glioblastoma and could be realized in a multicenter approach or a long acquisition phase in the future. Nevertheless, the abundance of identified driver mutations highlighted the large intratumoral heterogeneity of glioblastoma. In addition, the tremendous diagnostic depth of the method used enabled the identification of the TET1-deletion not previously described in the context of IDH wild-type glioblastoma. Although the detailed role of TET1-deletion in glioblastoma is not yet understood, our data provides promising evidence that loss of function of the TET1-enzyme in combination with EGFR-amplification or deletion of PTEN or CDKN2A/B has an impact on the pathogenesis of IDH wild-type glioblastoma and negatively affects prognosis. Future genetic sequencing is indicated to confirm the role of TET1-deletion and moreover to get further understanding of the genomic pathogenesis of glioblastoma with the aim to derive individualized therapeutic approaches.:Abkürzungsverzeichnis 5 1 Einleitung 8 1.1 Klinische Grundlagen des Glioblastoms 8 1.1.1 Epidemiologie/Ätiologie 8 1.1.2 Symptomatik und Diagnostik 9 1.1.3 Therapie 10 1.1.4 Prognose 12 1.2 Pathologie und Molekulargenetische Veränderungen des Glioblastoms 12 1.2.1 Treibergene 13 1.2.2 Subgruppen 17 1.3 Fragestellung 18 2 Materialien 19 3 Methoden 23 3.1 Patientenrekrutierung 23 3.1.1 Erweiterte Einschlusskriterien 25 3.1.2 Ausschlusskriterien 25 3.2 Kohortendesign 25 3.3 Klinische Daten 26 3.4 Materialgewinnung 27 3.4.1 Proben-Lagerung 27 3.4.2 DNA-Extraktion 27 3.5 Sanger Sequenzierung Prä-WES 29 3.5.1 Primer-Design 29 3.5.2 Polymerase Kettenreaktion 29 3.5.3 Elektrophorese 31 3.5.4 Aufreinigung PCR-Produkt 33 3.5.5 Sequenzierreaktion 35 3.6 WES 38 3.6.1 Datensatz 41 3.6.2 CNV-Analyse 42 3.7 Sanger Sequenzierung Validierung 42 3.8 Statistische Auswertung 45 4 Ergebnisse 47 4.1 Deskriptive klinisch-pathologische Beschreibung der Kohorte 47 4.1.1 Klinisch-pathologische Zusammenhänge 54 4.2 TERTp-Sequenzierung 58 4.3 Datensatz WES 58 4.3.1 Somatische Mutationen 59 4.3.2 CNV-Analyse 67 4.4 Altersbezogene molekulargenetische Unterschiede 74 4.5 Zusammenhang zwischen Genotyp und OS sowie PFS 77 4.5.1 Somatische Mutationen 77 4.5.2 CNV-Analyse 78 5 Diskussion 82 5.1 Klinische Charakteristika von Patienten mit IDH-Wildtyp Glioblastom 82 5.2 Der prognostische Stellenwert von Tumoreigenschaften 84 5.3 Molekulargenetisches Profil des IDH-Wildtyp Glioblastoms 85 5.3.1 Somatische Mutationen und Copy Number Variations 85 5.3.2 Chromosomale Aberrationen 94 5.4 10q21.3-Deletion 95 6 Zusammenfassung 97 6.1 Deutsch 97 6.2 Englisch 99 Anlagen 101 Darstellung zur Geschlechtsneutralität im geschriebenen Wort 101 Erklärung zur Eröffnung des Promotionsverfahren 102 Erklärung zur Einhaltung der gesetzlichen Vorgaben 104 Anhang 105 Literaturverzeichnis 113 Abbildungsverzeichnis 125 Tabellenverzeichnis 127 Danksagung 128

DNA methylation plays a pivotal role in governing cellular processes including genomic imprinting, gene expression, and development. Recently, the Tet family of methylcytosine dioxygenases(Tet1, Tet2 and Tet3) was found to catalyze the oxidation of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC), an intermediate in the pathway of DNA demethylation. Tet enzymes catalyze this hydroxylation in a 2-oxoglutarate and Fe2+ dependent manner. We have recently reported significant levels of 5mC and 5hmC modification in immunoprecipitates of mammalian mitochondrial DNA(mtDNA). We provide the first evidence that a DNA Methyltransferase-1 isoform (mtDNMT1) translocates to the mitochondria using an N-terminal mitochondrial targeting sequence. mtDNMT1 expression is upregulated by NRF1 and PGC1α, master regulators of mitochondrial biogenesis and function, as well as by loss of p53. Altered mtDNMT1 expression asymmetrically affects mtDNA transcription. We are now pursuing the role of Tet proteins in generating 5hmC in mtDNA. Using an in vitro enzyme assay, we have successfully detected Tet activity in crude and percoll purified mitochondrial fractions of HCT116 cells. Mitoprot analysis on Tet family predicts that Tet1 may be translocated to the mitochondria. Immunoblot analysis indicates that a band of expected size(235kDa) is present on immunoblots of mitochondrial fraction from mouse embryonic stem cells with an antibody directed against Tet1. This band, however, is not protected from trypsin treatment of mitochondria indicating that Tet1 may not be transported to the mitochondrial matrix. The putative Tet1 mitochondrial targeting sequence (MTS) fails to carry heterologous protein to the mitochondria. Knock out of Tet1 in mouse ES cells also does not alter 5hmC signal in hydroxyMeDIP assay. We now seek to determine if Tet2/Tet3 may be involved in 5hmC generation. In the nucleus, 5hmC serves as an intermediate in the process of DNA demethylation through the combined action of cytidine deaminases and the base excision repair pathway. We plan to investigate if 5hmC holds the same functional significance in the mitochondria as it does in the nucleus. Our overall goal is to understand epigenetic regulation of normal mitochondrial function and changes that occur in diseases involving mitochondrial dysfunction such as ischemic heart disease, neurodegenerative diseases like Parkinsons disease, and cancer.

Ten-Eleven-Translocation (TET) proteins form a family of enzymes responsible for active DNA demethylation by oxidation of 5-methylcytosine. TET proteins play a key role in genomic reprogramming in vitro and in vivo. Although TET proteins are expressed in embryonic stem cells (ESCs), their role in regulating pluripotency remains unclear. In addition, the mechanisms by which TET proteins are recruited to chromatin are largely unknown. To visualise TET protein dynamics during pluripotency and differentiation, the endogenous Tet1/2/3 alleles were fused to epitope tags in ESCs using CRISPR/Cas9. Characterisation of these cell lines showed that TET1 is the highest expressed TET protein in both naïve and primed pluripotent cells. In contrast, TET2 is expressed heterogeneously in ESCs and marks cells with a high self-renewal capacity. To assess the function of Tet genes in pluripotent stem cells, the endogenous Tet1/2/3 ORFs were removed using CRISPR/Cas9. Comparative analysis of single and combined Tet gene knockout ESC lines indicated that Tet1 and Tet2, but not Tet3, play redundant roles to promote loss of pluripotency. Furthermore, Tet-deficient cells retained a naïve morphology in differentiating conditions, suggestive of a LIF-independent self-renewal phenotype. To characterise physiological TET1 protein-protein interactions, TET1 protein partners were identified in ESCs by mass spectrometry and co-immuno-precipitations. This revealed that TET1 interacts with multiple epigenetic and pluripotency-related factors in ESCs. Moreover, detailed characterisation of the interaction between TET1 and NANOG identified three regions of TET1 involved in protein-protein interactions that are conserved in evolution. To investigate TET1 chromatin binding in ESCs, both at the molecular and cellular levels, TET1 was characterised by ChIP-seq analysis and live imaging experiments. Interestingly, TET1 is targeted to chromatin by two different mechanisms, involving distinct protein regions. The interaction with multiple protein partners, including NANOG, might enable TET1 to be targeted to specific chromosomal locations. Additionally, TET1 has the unusual ability to bind mitotic chromatin through its N-terminus, independently of its interaction with NANOG. Together these analyses provide a new understanding of the role of TET proteins in pluripotent cells, as well as a detailed map of TET1 residues involved in protein-protein interactions and mitotic chromatin binding.

DNA methylation is an epigenetic mechanism critical for tissue development, cell specification and cellular function. Mammalian brains consist of millions to billions of neurons and glial cells that can be subdivided into many distinct types of cells. We hypothesize that brain methylomes are heterogeneously methylated across different types of cells and the transcription factors play key roles in brain methylome programming. To dissect brain methylome heterogeneity, in Chapter 2, we first focused on the identification of cell-subset specific methylated (CSM) loci which demonstrate bipolar DNA methylation pattern, i.e., hypermethylated in one cell subset but hypomethylated in others. With the genome-scale hairpin bisulfite sequencing approach, we demonstrated that the majority of CSM loci predicted likely resulted from the methylation differences among brain cells rather than from asymmetric DNA methylation between DNA double strands. Importantly, we found that putative CSM loci increased dramatically during early stages of brain development and were enriched for GWAS variants associated with neurological disorder-related diseases/traits. It suggests the important role of putative CSM loci during brain development, implying that dramatic changes in functions and complexities of the brain may be companied by a rapid change in epigenetic heterogeneity. To explore epigenetic regulatory mechanisms during brain development, as described in Chapter 3, we adopted unbiased data-driven approaches to re-analyze methylomes for human and mouse frontal cortices at different developmental stages. We predicted Egr1, a transcriptional factor with important roles in neuron maturation, synaptic plasticity, long-term memory formation and learning, plays an essential role in brain epigenetic programming. We performed EGR1 ChIP-seq and validated that thousands of EGR1 binding sites are with cell-type specific methylation patterns established during postnatal frontal cortex development. More specifically, the CpG dinucleotides within these EGR1 binding sites become hypomethylated in mature neurons but remain heavily methylated in glia. We further demonstrated that EGR1 recruits a DNA demethylase TET1 to remove the methylation marks at EGR1 binding sites and activate downstream genes. Also, we found that the frontal cortices from the knockout mice lacking Egr1 or Tet1 share strikingly similar profiles in both gene expression and DNA methylation. Collectively, the study in this dissertation reveals EGR1 programs the brain methylome together with TET1 during postnatal development. This study also provides new insights into how life experience and neuronal activity may shape the brain methylome. / Ph. D. / DNA methylation is a widespread epigenetic mark on DNA, serving as a “switch” to turn on or off gene expression. It plays essential roles in cellular functions, tissue development. Mammalian brains contain millions to billions of neurons and glial cells, which can be further divided into many different types of cells. We hypothesize that brain cells have different methylation profiles across the genome, and transcriptional factors play important roles in programming methylation in the mammalian brain genome. To study the diversity of methylation profiles across the genomes of different brain cells, in Chapter 2, we first focused on the identification of cell-subset specific methylated (CSM) genomic regions which show bipolar DNA methylation pattern, i.e., hypermethylated in one type of cell but hypomethylated in others. By applying a technique called the genome-scale hairpin bisulfite sequencing to mouse frontal cortices, we demonstrated that the majority of CSM genomic regions predicted likely resulted from the methylation differences among brain cells, rather than from methylation differences between DNA double strands. Surprisingly, we found that these predicted CSM genomic regions increased dramatically during early stages of brain development and were enriched for GWAS variants associated with neurological disorder-related diseases/traits. It suggests the importance of predicted CSM genomic regions, implying that dramatic changes in brain function and structure may be companied by a rapid change in DNA methylation diversity during brain development. To explore underlying epigenetic mechanisms during brain development, as described in Chapter 3, we re-analyzed methylomes for human and mouse frontal cortices at different developmental stages, and predicted Egr1, a transcriptional factor with important roles in neuron maturation, synaptic plasticity, long-term memory formation and learning, plays an essential role in brain methylome programming. We found thousands of EGR1 binding sites showed cell-type specific methylation patterns, and were established during postnatal frontal cortex development. More specifically, the methylation level of these EGR1 binding sites was low in mature neurons but pretty high in glial cells. We further demonstrated that EGR1 recruits a DNA demethylase TET1 to remove the methylation marks at EGR1 binding sites and activate downstream genes. Also, we found that the frontal cortices from the Egr1 knockout or Tet1 knockout mice show strikingly similar profiles in both gene expression and DNA methylation. Collectively, the study in this dissertation reveals EGR1 works together with TET1 to program the brain methylome during postnatal development. This study also provides new insights into how life experience and neuronal activity may shape the brain methylome.

Deregulation of the epigenome is associated with the initiation and progression of various types of human cancers. Here we investigated the level of 5-hydroxymethylcytosine (5hmC), expression and function of TET1 and TET2, and DNA methylation in parathyroid tumors and small intestinal neuroendocrine tumors (SI-NETs). In Paper I, an undetectable/very low level of 5hmC in parathyroid carcinomas (PCs) compared to parathyroid adenomas with positive staining, suggested that 5hmC may represent a novel biomarker for parathyroid malignancy. Immunohistochemistry revealed that increased tumor weight in adenomas was associated with a more aberrant staining pattern of 5hmC and TET1. A growth regulatory role of TET1 was demonstrated in parathyroid tumor cells. Paper II revealed that the expression of TET2 was also deregulated in PCs, and promoter hypermethylation was detected in PCs when compared to normal parathyroid tissues. 5-aza-2′-deoxycytidine treatment of a primary PC cell culture induced TET2 expression and further supported involvement of promoter hypermethylation in TET2 gene repression. TET2 knockout demonstrated a role for TET2 in cell growth and migration, and as a candidate tumor suppressor gene. In Paper III, variable levels of 5hmC, and aberrant expression of TET1 and TET2 were observed in SI-NETs. We demonstrated a growth regulatory role for TET1, and cytoplasmic expression with absent nuclear localization for TET2 in SI-NETs. In vitro experiments supported the involvement of exportin-1 in TET2 mislocalization, and suggested that KPT-330/selinexor, an orally bioavailable selective inhibitor of exportin-1 and nuclear export, with anti-cancer effects, could be further investigated as a therapeutic option in patients with SI-NETs. In Paper IV, DNA methylation was compared between SI-NET primary tumors and metastases by reduced representation bisulfite sequencing. Three differentially methylated regions (DMR) on chromosome 18 were detected and chosen for further analyses. The PTPRM gene, at 18p11, displayed low expression in SI-NETs with high levels of methylation in the presumed CpG island shores, and in the DMR rather than the promoter region or exon 1/intron 1 boundary. PTPRM overexpression resulted in inhibition of cell growth, proliferation, and induction of apoptosis in SI-NET cells, suggesting a role for PTPRM as an epigenetically deregulated candidate tumor suppressor gene in SI-NETs.

IKAROS est un facteur de transcription majeur dans l’hématopoïèse qui agit en recrutant à la chromatine de nombreux partenaires décisifs dans le renouvellement cellulaire et l’engagement vers des lignages spécifiques. Il est notamment requis dans les cellules lymphoïdes pour réprimer les gènes cibles de la voie de signalisation NOTCH. IKAROS est aussi important dans le développement des cellules érythroïdes dans lesquelles il facilite le passage d’une globine fœtale à adulte chez l’embryon grâce au recrutement des complexes remodeleurs de la chromatine NuRD et BAF. En condition normale, la voie de signalisation NOTCH réprime la différenciation en cellules érythroïdes. Il est donc important que les gènes cibles de la voie NOTCH soient finement régulés afin d’amener une cellule progénitrice à se différencier en érythrocyte énucléé. Dans les cellules hématopoïétiques, incluant les cellules érythroïdes, IKAROS est un régulateur important du gène Hes1, cible effectrice majeure de la voie NOTCH. En effet, IKAROS participe activement à la répression du gène Hes1, permettant le développement des cellules érythroïdes. Nous avons donc émis l’hypothèse que dans ces cellules, IKAROS pourrait avoir une action plus généralisée sur le contrôle des gènes ciblés par NOTCH, comme observé dans les cellules lymphoïdes. Il pourrait ainsi agir en recrutant les complexes enzymatiques nécessaires à la régulation de ces gènes comme NuRD et BAF afin d’assurer le développement des cellules érythroïdes. Étant donné que la régulation des gènes est aussi dépendante du motif de méthylation de l’ADN, nous avons étendu notre questionnement à cet autre aspect de la régulation qu’IKAROS pourrait utiliser pour contrôler les gènes de la voie NOTCH. Pour ce faire, nous avons procédé à l’analyse bio-informatique d’un séquençage d’ARN de cellules érythroïdes murines préalablement réalisé au laboratoire afin d’en extraire les gènes régulés par IKAROS, mais aussi par NOTCH. L’analyse nous a permis d’extraire deux motifs d’expression intéressants observés dans les cellules érythroïdes pour lesquels IKAROS réprime ou active des gènes qui sont normalement réceptifs à l’activation de la voie NOTCH. Parmi les gènes réprimés par IKAROS en sont ressortis les gènes cibles de NOTCH Cdkn1a (P21WAF1/CIP1) et Trp53 (TP53), dont l’expression augmente fortement quand IKAROS est muté et que NOTCH est actif. Parmi les gènes activés par IKAROS en sont ressortis les gènes cibles de NOTCH Prdm16 et Nrarp, dont l’expression diminue fortement quand IKAROS est muté et que NOTCH est actif. IKAROS est donc un régulateur pouvant être répresseur, mais aussi activateur d’une multitude de gènes ciblés par NOTCH dans les cellules érythroïdes. Par des expériences d’immunoprécipitation de la chromatine, nous avons pu observer qu’IKAROS semblait toujours agir en partenariat avec le complexe NuRD et que la présence du complexe BAF était plutôt dépendante de la voie NOTCH. L’association IKAROS-NuRD semble servir de plateforme pour imposer un état de chromatine bivalente (avec co-présence de H3K4me3 et de H3K27me3) associée à une pause transcriptionnelle. Dans ce contexte, les éléments nécessaires à l’initiation de la transcription (présence de la marque H3K4me3) des gènes ciblés par NOTCH sont recrutés mais, l’élongation transcriptionnelle est affectée. L’état de chromatine bivalente peut être associé à l’activité des déméthylases de l’ADN Ten-Eleven-Translocation (TET) qui empêchent alors l’hyperméthylation de ces régions. Nos résultats démontrent qu’IKAROS peut utiliser la protéine TET1 pour réguler des gènes cibles de la voie NOTCH, en y formant l’hydroxyméthylcytosine (5-hmC). Celle-ci peut aussi marquer les régions de régulation génique caractérisées par une chromatine bivalente et une pause transcriptionnelle. Ces travaux décrivent IKAROS comme un facteur agissant de façons multiples dans la régulation des gènes cibles de NOTCH dans les cellules érythroïdes. Nous proposons qu’IKAROS et son partenaire NuRD soient requis pour mettre en place un état de chromatine bivalente et de pause transcriptionnelle pour faciliter l’activation physiologique des gènes cibles de NOTCH lors de la signalisation. IKAROS peut ainsi prendre part à l’activation ou la répression de gènes cibles de NOTCH, tout en facilitant la déméthylation de l’ADN ainsi que le recrutement d’éléments transcriptionnels qui favorisent un état de pause transcriptionnelle. NOTCH ainsi que d’autres éléments de régulation sont alors requis pour induire l’activation ou la répression des gènes cibles. / IKAROS is a critical transcription factor in hematopoiesis. It facilitates the chromatin binding of many important co-factors required for chromatin organization during cell renewal and lineage commitment. IKAROS is particularly important in lymphoid cells whereby it is involved in the repression of target genes of the NOTCH signaling pathway. IKAROS is also important in the development of other hematopoietic lineages, including the erythroid cells, in which it facilitates the passage of a fetal to adult globin in the embryo through the recruitment of the chromatin remodeling complexes NuRD and BAF. Under normal conditions, the NOTCH signaling pathway represses development of erythroid cells. It is therefore important to precisely understand how the NOTCH target genes are regulated during passage from hematopoietic progenitor to the enucleated circulating erythrocyte. IKAROS has been demonstrated to be an important regulator of Hes1 gene expression in hematopoietic cells of different lineages. Hes1 is the major effector target of the NOTCH pathway and IKAROS actively participates in its repression. In erythroid cells, the regulation of Hes1 imposed by IKAROS is required for terminal differentiation. We therefore investigated the importance of IKAROS in the regulation of NOTCH-targeted genes in erythroid cells. The combined effect of the mutation of IKAROS with NOTCH signaling was particularly investigated in these cells. To define how IKAROS influences the regulation of NOTCH target genes, we performed the bioinformatics analysis of a RNA-sequencing performed in murine erythroid cells activated or not for NOTCH signaling and whereby IKAROS is absent. We identified genes influenced by IKAROS expression and by NOTCH, and defined the effect of the combination of the absence of IKAROS expression and NOTCH pathway activation. Two particular expression patterns were identified and characterized the combined effect of the absence of IKAROS and NOTCH pathway activation in erythroid cells. Indeed, the absence of IKAROS either favors the overexpression of NOTCH target genes or prevents their response to NOTCH pathway activation. To understand how IKAROS could have an opposite effect on different NOTCH target genes we analysed the effect of IKAROS on their regulation. Among the genes repressed by IKAROS are the target genes of NOTCH Cdkn1a (encoding the P21WAF1/CIP1 protein) and Trp53 (encoding the TP53 protein), whose expression increases strongly when IKAROS is mutated and the NOTCH pathway is activated. Prdm16 and Nrarp are, instead, requiring IKAROS expression for their activation by NOTCH. The characterization of these NOTCH target genes suggests that IKAROS can work in partnership with the NuRD complex to influence the expression. The chromatin characterization of these genes led us to posit that the IKAROS-NuRD could act as a ‘platform’ to impose a bivalent chromatin organization associated with poised transcription. Then, the regulation imposed by IKAROS-NuRD would be required for the physiological activation of NOTCH targets upon external signaling. Finally, since in embryonic stem cells the Ten-Eleven Translocation (TET) enzymes are reported to be frequently associated to bivalent chromatin in order to prevent DNA hypermethylation, we assessed whether IKAROS could interact and ‘use’ TET enzymes to regulate NOTCH target genes. We determined that IKAROS can co-immunoprecipitate with the TET1 enzymes. We show that IKAROS influences both recruitment and activity of TET1 to different NOTCH target genes and favors the accumulation of hydroxymethylcytosine (5-hmC) to these genes. 5-hmC can be considered as a mark of transcriptional pausing/bivalence. Thus, these studies bring new information on the mechanism used by IKAROS to influence gene regulation in hematopoietic cells. Our results suggest that IKAROS primary function is to organize a bivalent chromatin and to promote transcriptional pausing to multiple NOTCH target genes. IKAROS is required to set the epigenetic and promoter organization for rapid activation upon NOTCH signaling.

Endogenous retrovirus ERVWE1 is an integral part of the human genome. In the course of evolution, a protein encoded by the env gene of this retrovirus - Syncytin-1 - has gained unique function in human development. It mediates cell-to-cell fusion of placental cytotrophoblasts. Receptor that binds to Syncytin-1 is expressed in different cell types. Syncytin-1-mediated fusion is essential in placenta, but it can cause disruption of tissue integrity in other cell types. ERVWE1 expression is regulated by promoter DNA methylation, transcription factor GCM1 and efficient mRNA splicing. This thesis concerns the ERVWE1 expression and its regulation in non-placental tissues. It was found out that the moderate GCM1 overexpression was not sufficient to induce Syncytin-1 expression. Neither treatment with DNA demethylation agent 5-azacytidine nor with Syncytin-1 activator forskolin was able to manage Syncytin-1 expression. This thesis extends previous findings concerning high syncytin-1 expression in seminomas. In same tissues, there was found elevated TET1 expression on mRNA level in comparison with controls. The presence of the TET1 demethylation enzyme can influence ERVWE1 promoter DNA methylation. Previously unreported splicing variant of TET1 has been found during the construction of human TET1 expression...