Les lymphomes T périphériques (ou PTCL) sont des lymphomes malins non Hodgkiniens ayant pour cellules d’origine des lymphocytes T (LT) ou Natural Killer matures. Ces lymphomes sont rares, hétérogènes et méconnus. Des arguments issus de la littérature suggérant l’implication de la stimulation chronique du récepteur T à l’antigène (TCR) dans la transformation des LT, nous ont conduits à développer un modèle murin basé sur la stimulation chronique du TCR pour adresser spécifiquement cette question. Dans ce modèle, le transfert de LT p53-/- dans des souris CD3e-/- entraine l’apparition de lymphomes T périphériques (PTCL) clonaux dans 60% des cas avec une médiane de survenue de 230 jours alors que les souris transférées avec des LT wt ne développent pas de lymphomes. Ces PTCL présentent un phénotype T effecteur-mémoire CD62LLo-CD44hi-CD122lo-CD25lo ainsi qu’une profonde downrégulation de l’expression des gènes impliqués dans la voie du TCR illustrant l’impact de la stimulation chronique dans la lymphomagénèse. L’étude de ces lymphomes a révélé qu’ils ne dépendent plus, pour la plupart, de l’engagement du TCR pour leur survie et qu’ils acquièrent des caractéristiques « innate-like » avec notamment l’expression de récepteurs NK inhibiteurs (NKiR) et de récepteurs NK activateurs (NKaR) ainsi que des protéines adaptatrices DAP12 et FceRIg. Cette expression est associée à celle de Syk et PLC?2, impliquées dans la signalisation des NKaR. Nous montrons que les NKaR et leurs voies de signalisation associées sont fonctionnelles et participent à la survie des cellules lymphomateuses, le blocage de certains NKaR retardant notamment le développement lymphomateux in vivo. Nous avons par la suite exploré l’expression des NKR, de Syk et de PLCg2 au sein des PTCL humains et nous montrons ou confirmons que certaines entités expriment des panels variés de NKR ainsi que les effecteurs Syk et PLCg2 suggérant l’existence de mécanismes de lymphomagénèse similaires à ceux identifiés dans notre modèle au sein d’un certain nombres de PTCL humains / Peripheral T-cell lymphomas (PTCL) are rare non Hodgkin malignant lymphomas emerging from mature T or NK cells. PTCL are highly heterogeneous and mainly misunderstood. As several evidences pointed the potential role of TCR chronic stimulation in human T-cell lymphomagenesis, we developed a murine model based on chronic TCR stimulation to address this question. In this model, transfer of p53-/- T-cells into T-cell deficient mice (CD3e-/-) triggered PTCL development in 60% of cases with a median survival of 230 days while transfer of wt T-cells in CD3e-/- mice did not lead to PTCL development. These PTCL exhibited an effector-memory phenotype CD62LLo-CD44hi-CD122lo-CD25lo associated with a dramatic downregulation of TCR pathway genes expression consistent with a chronic TCR stimulation highlighting it’s implication in lymphomagenesis. The analysis of these PTCL revealed that a large majority of cases (80%) do not depend anymore on TCR stimulation for their growth and survival and that they acquire innate-like features with expression of inhibitory NKR (NKiR) and activating NK receptors (NKaR) as well as the adaptor proteins DAP12 or FceRIg. Expression of these receptors is associated with the expression of SYK and PLC?2, which are classical key effectors downstream of NKaR. We show that these NKaR are functional and can mediate TCR-independent activation in mPTCL and that this signaling is involved in cell survival/proliferation as in vivo blockade of NKG2D and NKp46 delays PTCL development in PTCL transplantation experiments. In parallel, we studied NKR, Syk and PLCg2 expression in human PTCL and found that some entities express a large range of these receptors as well as Syk and PLCg2, suggesting similar lymphomagenesis mechanisms in some human PTCL

Current usage of human embryonic stem cells (hES) and induced pluripotent stem cells (iPS) in clinical therapies and personalized medicine are limited as a result of ethical, technical and medical problems that arise from isolation and generation of these cells. Isolation of hES cells faces ethical problems associated with their derivation from human pre-implantation embryos. The most controversial aspect of hES cell isolation targets the generation of autologous hES cell lines which requires the transfer of a somatic-cell nucleus from the patient to an enucleated oocyte. While already established embryonic stem cell lines from IVF embryos can be used in a similar manner, lack of genetic identity can cause therapy rejection from the host, and prevent their use in personalized medicine. Induced pluripotent stem cells on the other hand, are generated from somatic cells that have been reprogrammed in vitro to behave like stem cells. While these cells can potentially be used for personalized medicine without the risk of rejection by the host system, derivation methods prevent their therapeutic use. The most efficient method used to generate iPS cells involves usage of viral particles which can result in viral DNA being integrated in the host cell’s genome and render these cells non-compliant for clinical therapies. Other methods not involving viral particles exist as well, but the reprogramming efficiency is too low and technical problems with generating large enough numbers of cells prevent these methods from being feasible approaches for clinical therapies. Direct reprogramming of a differentiated cell into a developmentally more plastic cell would offer alternatives to applications in regenerative medicine that currently depend on either embryonic stem cells (ES), adult stem cells or iPS cells. We hypothesize that Xenopus laevis egg cytoplasmic extract contains critical factors needed for reprogramming that may allow for non-viral, chemically defined derivation of human induced pluripotent/multipotent cells which can be maintained by addition of exogenous FGF2. In this thesis we investigated a new method for generation of multipotent cells through determining the ability of select fractions of Xenopus laevis egg extract to induce multipotency in already differentiated cells. We were able to identify select fractions from the extract that in combination with exogenously added FGF2 can reprogram and maintain the reprogrammed cells in an undifferentiated state. The findings of this work also determined that Xenopus laevis egg extract mRNA is required for achieving full reprogramming. The body of work presented in this thesis showed the ability of FGF2 isoforms to bind and activate select FGF receptor tyrosine kinases, act as extracellular mitogenic factors to support growth of hES cells in an undifferentiated state as well bind to nuclear DNA and affect expression of endogenous genes. Moreover, we showed that all FGF2 isoforms can induce expression of stem cell specific proteins in human dermal fibroblasts as well as extend lifespan of human dermal fibroblasts in vitro. In this work we identified HECW1, the gene coding for E3 ubiquitin ligase NEDL1, as a novel nuclear target for all FGF2 isoforms and showed that overexpression of recombinant FGF2 isoforms in human dermal fibroblasts can down regulate expression of HECW1 gene.

The complexity of life exhibited in humans and other living creatures has drawn many to investigate the principles associated with organismal growth and development. A few broad questions: How do tissues develop into specified organs? How are these tissues maintained? Do they become different tissues? Scientific research has incessantly been seeking answers to these as well as a plethora of other questions. While on a quest to better understand developmental biology, investigators discovered unique populations of stem cells within a variety of tissues, which retain both varying degrees of developmental plasticity and their potential for self-regeneration. This thesis provides a brief review discussing the development and history of stem cells in medicine and associated research on these cells and their potential clinical applications. Substantial attention has been paid to pluripotent embryonic stem cells (ESC) and induced pluripotent stem cells (iPSC) which are able to be recapitulate ESC properties through the in vitro reprogramming of somatic cells. While, the ethical and legal issues have greatly hindered the use of ESCs this has made the benefit of iPSCs so great. An interconnected network of pluripotency-associated genes, integrates external signals and exerts control to maintain the state of pluripotency. Recent research has proven the pluripotency regulatory network to be flexible such that the underlying principles promise unprecedented opportunities for scientific study and regenerative medicine. Additional topics reviewed here include vast clinical applications of stem cells as well as their notable limitations.

Somatic cell nuclear reprogramming (SCNR) by eggs is a way to forcibly transform the nuclei of terminally differentiated somatic cells to an embryonic state and gain totipotency (Gurdon et al., 1958). Additionally, induced pluripotency is applied to transform identities of somatic cells to induced pluripotent stem cells by overexpression of combinatorial Yamanaka factors (iPS, Takahashi et al., 2006). Although both approaches aim to derive cells with highest plasticity, the mechanisms and differences between these procedures are not yet clear. In my thesis, I used quantitative polymerase chain reaction (QPCR) and RNAseq plus 5-bromouridine 5'-triphosphate (BrUTP) pulldown to evaluate the transcriptional reprogramming by maternal factors and overexpressed transcription factors during SCNR by Xenopus oocytes, which are inactive in DNA replication and cell division. QPCR measures changes in the steady-state levels of transcripts within 2 days of nuclear transfer to Xenopus oocytes (Oocyte-NT). Three pairs of Yamanaka factor homologs were tested by QPCR and Yamanaka factor homologues regulated similar sets of pluripotency genes in mouse embryonic fibroblasts (MEFs). Pioneer factor mFoxA1 could not up-regulate most pluripotency genes and their binding targets of neurogenic genes in MEFs while pioneer factors are proposed to bind to their targets even if they may reside in inaccessible chromatin. This shows that the existence of other factors is needed at specified developmental stages. Hence, gene activation by transcription factors in the Oocyte-NT system requires not only corresponding binding on regulatory elements of linked genes but transcription cooperators to exert effective gene activation. Additionally, RNA-seq plus BrUTP pulldown measures the extent to which oocytes change the transcriptional activity of nuclei transplanted to oocytes. Through RNA-seq plus BrUTP pulldown, I compared the reprogrammed transcriptomes of embryonic and somatic cells, including mouse embryonic stem cells, mouse embryonic fibroblasts and mouse myoblasts, to demonstrate the effects of maternal factors and overexpression of transcription factors on gene activities during SCNR by oocytes. Importantly, I find that maternal factors of oocytes and the overexpression of transcription factors exert different strategies to reprogram somatic cells. Oocyte factors reprogram the donor cell nuclei to an oocyte-steady state except for the SCNR resistance genes and xklf2-HA overexpression enhances expression of reprogrammable genes and activates SCNR resistance genes.

Estudos recentes relataram oncogenes induzindo a reprogramação metabólica no câncer. Essa reprogramação é fundamental para que as células cancerosas tenham nutrientes e biomoléculas suficiente para manter sua alta taxa proliferativa. A mitocôndria tem um papel central no metabolismo energético da célula e alterações no seu genoma, tanto em relação a mutações como em número de cópias, já foram bastante observados em vários tipos tumorais. Além disso, deficiência no fator de transcrição mitocondrial A (TFAM), fundamental para a transcrição e estabilidade do mtDNA, já foi associada com o crescimento tumoral. Diante disso, nosso estudo teve como objetivo avaliar o papel da instabilidade do genoma mitocondrial no metabolismo energético e crescimento do melanoma. Para isso, nós medimos a instabilidade do mtDNA utilizando como parâmetros: o acúmulo de mutações no mtDNA, alterações no mtDNAcn e a expressão do TFAM. O impacto da instabilidade do mtDNA foi avaliado em três modelos diferentes de melanoma: um modelo in vitro de linhagens celulares, dados de expressão gênica de tumores de melanoma metastático proveniente do TCGA e um modelo murino induzível de melanoma (BrafV600E/Ptennull), adicionado a um background alternativo de deficiência para o TFAM/mtDNAcn. Esse modelo murino também nos permitiu avaliar a deficiência do TFAM limitada a células tumorais (Tfamflox) e tanto em células tumorais, como no seu microambiente (Tfam+/-). Nas análises in vitro, nós observamos correlações positivas entre o mtDNAcn e a expressão do TFAM com a taxa de consumo de glicose e produção de ATP, indicando um impacto desses parâmetros na bioenergética celular. Análises de expressão gênica, utilizando tanto as linhagens de melanoma como tumores de melanoma metastático, nos sugeriram que o TFAM regula genes indutores de angiogênese, a resposta imunológica humoral e vias metabólicas de aminoácidos. Nas análises in vivo, nós observamos um aumento dos tumores em camundongos Tfam+/-, indicando que a deficiência de TFAM/mtDNAcn em células tumorais e no seu microambiente induz a tumorigênese, o que confirma os dados de expressão gênica encontrados com linhagens e tecido de melanoma. Além disso, análises de metabolômica e transcriptômica combinadas nos sugeriram que as células de melanoma com deficiência no TFAM/mtDNAcn são mais dependentes do metabolismo de glutamina. Diante disso, nós concluímos que a deficiência do TFAM/mtDNAcn tem um papel importante no crescimento do melanoma, induzindo a expressão de genes pro-tumorigênicos e aumentando o consumo da glutamina para suprir as necessidades proliferativas das células cancerosas. Esses dados são relevantes e podem nos ajudar a entender melhor o papel da mitocondrial na progressão do melanoma. / Recent studies have shown many oncogenes triggering metabolic reprogramming in cancer. The metabolic switch in cancer cells is necessary to supply the high demand for nutrients and biomolecules for proliferative cells. In this context, mitochondria play a central role in the energetic metabolism of the cell and changes in its genome, such as an increased load of mutations and alterations in mtDNA content, have been reported in several cancers. In addition, deficiency in the Mitochondrial Transcription Factor A (TFAM), responsible for transcription and maintenance of mtDNA stability, was previously associated with tumor growth. Based on that, our goal was to evaluate the impact of the mitochondrial genome instability in the energetic metabolism and melanoma growth. mtDNA instability was inferred measuring mtDNA mutations load and content, as well as TFAM expression. Its impact was evaluated in three different melanoma models: an in vitro model using melanoma cell lines, gene expression data from metastatic melanoma tumors, publicly available at TCGA, and an inducible murine model of melanoma (BRAFV600E/PTENnull), crossed onto different TFAMdeficient backgrounds. The murine model also provides us a tractable model to examine the consequences of mtDNA instability limited to cancer cells (Tfamflox) and in both cancer cells and tumor microenvironment (Tfam+/-). In vitro analysis showed us a positive correlation between mtDNA copy number (mtDNAcn) and TFAM expression with glucose consumption and ATP production, pointing an impact of these parameters in cellular bioenergetics. Further gene expression analysis, using both cell lines and metastatic melanoma data, suggested that TFAM could regulate the expression of angiogenesis genes, humoral immunity and amino acid metabolism. In vivo analysis confirmed the gene expression data, and revealed a higher melanoma growth in Tfam+/-. Also, combined metabolomics and transcriptomics data suggested that TFAM/mtDNAcn deficient melanoma cells rely mostly on glutamine metabolism to supply their energetic requirements. In conclusion, these data indicate that TFAM/mtDNAcn influences melanoma growth by triggering pro-tumorigenic signals and inducing metabolic reprogramming towards glutamine metabolism. These results are relevant and might help us understand how mitochondria affect melanoma progression.

L'hétérochromatine péricentromérique semble jouer un rôle dans la régulation de l'expression génique et par conséquent dans le potentiel de développement des embryons. Nous avons fait l'hypothèse qu'une marque épigénétique, H3S10P, pourrait être un nouveau marqueur permettant le suivi des régions péricentromériques dans l'embryon préimplantatoire de souris. Par des techniques d'immunofluorescence et d'immuno-FISH couplées à de la microscopie en haute résolution, nous avons montré que la distribution de H3S10P dans les embryons de souris est différente de celle observée dans les cellules somatiques. Durant les stades 1 à 4-cellules, H3S10P est détectée en interphase autour des précurseurs des nucléoles (NPB), où elle colocalise avec les sondes ADN reconnaissant l'hétérochromatine péricentromérique, puis marque les bras chromosomiques sur toute la durée des phases de mitose. Après le stade 4-cellules, la distribution de H3S10P redevient similaire à ce qui est connu dans les cellules somatiques, avec un marquage au niveau des chromocentres seulement en fin d'interphase et sur les chromosomes mitotiques seulement jusqu'à la télophase. Cette cinétique particulière observée semble liée à l'absence de la kinase Aurora B aux stades les plus précoces. Nous avons également comparé la localisation de H3S10P avec celle d'autres marqueurs associés à l'hétérochromatine péricentromérique comme H3K9me3, HP1 β et la double modification H3K9me3S10P et en avons conclu que H3S10P est un meilleur marqueur pour l'hétérochromatine péricentromérique des deux génomes parentaux. Enfin, comme les embryons clonés obtenus par transfert nucléaire à partir de cellules somatiques (SCNT) montrent une redistribution anormale de l'hétérochromatine péricentromérique ainsi qu'un développement altéré, nous avons utilisé H3S10P pour détecter les remaniements de l'hétérochromatine après SCNT. Nos résultats montrent que, contrairement aux autres marqueurs, H3S10P n'est présente que sur la portion de l'hétérochromatine qui est correctement remaniée, tandis que l'hétérochromatine incorrectement reprogrammée conserve la signature épigénétique de la cellule donneuse. / Pericentromeric heterochromatin appears to be involved with gene regulation and therefore with the developmental potential of embryos. We hypothesized that an epigenetic modification, H3S10P, could be a new marker to follow pericentromeric heterochromatin in preimplantation mouse embryos. Using immunofluorescence, immunoFISH and high resolution microscopy, we observed that H3S10P shows a different distribution pattern in mouse embryos than in somatic cells. It is detected early in interphase around the Nucleolar-Precursor Bodies from 1- to 4-cell, in co-localization with the DNA probes for pericentromeric heterochromatin, and is seen in the chromosome arms throughout mitosis. In fact, H3S10P shows a similar kinetic as seen in somatic cells only after the 4-cell stage: being solely observed in the chromocenters during late interphase and on the mitotic chromosomes until telophase. This distribution seems related to the absence of Aurora B kinase in the earlier stages. We have also compared H3S10P to other related pericentromeric heterochromatin markers such as H3K9me3, HP1β and the double modification, H3K9me3S10P, and concluded that H3S10P is a better marker for pericentromeric heterochromatin of both parental origins. Finally, as cloned embryos often show abnormal pericentromeric heterochromatin remodelling and impaired development after Somatic Cell Nuclear Transfer (SCNT), H3S10P was used to track down heterochromatin reprogramming after SCNT. Our results show that H3S10P underlines only the portion of heterochromatin which is remodelled when compared with the other related markers and that the unremodelled portion maintains the epigenetic signature of the donor cell.

Glioblastoma multiforme (GBM) is one of the most common and aggressive primary brain tumours, with a median patient survival time of 6-12 months in adults. It has been recently suggested that a typically small sub-population of brain tumour cells, in possession of certain defining properties of stem cells, is responsible for initiating and maintaining the tumour. More recent experiments have studied the interactions between this subpopulation of brain cancer cells and tumour microenvironmental factors such as hypoxia and high acidity. In this thesis a computational approach (based on Gillespie’s algorithm and cellular automata) is proposed to investigate the tumour heterogeneities that develop when exposed to various microenvironmental conditions of the cancerous tissue. The results suggest that microenvironmental conditions highly affect the characterization of cancer cells, including the self-renewal, differentiation and dedifferentiation properties of cancer cells.

Breast cancer is the most common cancer in Taiwanese women and the invasive ductal carcinoma (IDC) is the most common type. Increasing evidence shows that cancer stem cells (CSCs) have been implicated in tumorigenesis, tumor progression, and drug-resistance. In addition, four reprogramming factors (Octamer-binding Protein 4 (Oct4), Sex-determining Region Y (SRY)-related Box 2 (Sox2), Nanog and Lin28) employed to induce induced pluripotent stem (iPS) cells are associated with CSCs formation. The purpose of this study was to investigate the relationship of the protein expression levels of the reprogramming factors (Oct4, Sox2, Nanog and Lin28) with the tumorigenesis, clinicopathologic outcomes and prognosis of breast IDC patients. Immunohistochemistry (IHC) assay of tissue microarrays, made by 309 IDC and 20 breast fibrosis paraffin embedded samples, were used to examine the protein expression levels of Oct4, Sox2, Nanog and Lin28 in normal mammary ductal tissues, tumor adjacent normal mammary ductal tissues, ductal carcinoma in situ (DCIS), IDC and recurrence tissues. Our IHC results showed that Sox2 and Lin28 were expressed in half of breast IDC patients¡¦ tumor tissue (49.6% and 49.7%, respectively), but Oct4 and Nanog are less expressed (13.5% and 24.7%, respectively). The protein expression levels of the four proteins were positively correlated with each other. In addition, the expression levels of the four proteins were upregulated in tumor adjacent normal tissue as compared to breast fibrosis pateints¡¦ normal mammary ductal tissue. To compare the expression levels of the four proteins in different tissues; such as tumor adjacent normal, DCIS, IDC and recurrence tissues, the expression levels of the four protiens gradually decreased when tumor developed and progressed. However, their expression levels were comparable between IDC and recurrence tissues. Additionally, the high expression levels of four proteins were high in two good clinicopathological characteristics and a biomarker of breast cancer; such as nuclear Sox2 and Lin28 in those with pathology stage I; nucleus expression of the four proteins in those with well and moderate cell differentiation; and Sox2 in those with positive estrogen receptor. However, the four proteins¡¦ expression levels were not correlated with IDC patients¡¦ survival. In conclusion, the reprogramming factors: Oct4, Sox2, Nanog and Lin28 may play an important role in tumorigenesis of breast IDC, but their impacts on tumor progression were quite small.

Cellular fate, once established, is usually stable for the lifetime of the cell. However, the mechanisms that restrict the developmental potential of differentiated cells are in principle reversible, as demonstrated by the success of animal cloning from a somatic genome through somatic cell nuclear transfer (SCNT). An increased understanding of the molecular determinants of cell fate has also enabled the reprogramming of cell fate using defined transcription factors; recently, these efforts have culminated in the discovery of four genes that convert somatic cells into induced pluripotent stem cells (iPSCs), which resemble embryonic stem cells (ESCs) and can give rise to all the cell types in the body. As a first step toward generating clinically useful iPSCs, we identified a small molecule, RepSox, that potently and simultaneously replaces two of the four exogenous reprogramming factors, Sox2 and cMyc. This activity was mediated by the inhibition of the Transforming Growth Factor-\(\beta\) \((Tgf-\beta)\) signaling pathway in incompletely reprogrammed intermediate cells. By isolating these stable intermediates, we showed that RepSox acts on them to rapidly upregulate the endogenous pluripotency factor, Nanog, allowing full reprogramming to pluripotency in the absence of Sox2. We also explored lineage conversion as an alternative approach for producing a target cell type in a patient-specific manner, without first generating iPSCs. A combination of pro-neural as well as motor neuron-selective factors could convert fibroblasts directly into spinal motor neurons, the cells that control all voluntary movement. The induced motor neurons (iMNs) displayed molecular and functional characteristics of bona fide motor neurons, actuating muscle contraction in vitro and even engrafting in the developing chick spinal cord when transplanted. Importantly, functional iMNs could be produced from fibroblasts of adult patients with the fatal motor neuron disease, amyotrophic lateral sclerosis (ALS). Given the therapeutic value of generating patient-specific cell types on demand, defined-factor reprogramming is likely to serve as an important tool in regenerative medicine. It is hoped that the different approaches presented here can complement existing technologies to facilitate the study and treatment of intractable human disorders.

Somatic cells can be reprogrammed into induced pluripotent stem (iPS) cells by over expression of the transcription factors OCT4, SOX2, KLF4 and c-MYC. Using serial live cell immunofluorescence imaging of human fibroblasts undergoing reprogramming, we traced the emergence of nascent iPS cell colonies among heterogeneous cell populations and defined the kinetics of marker expression. We identified distinct colony types that morphologically resemble embryonic stem (ES) cells yet differ in molecular phenotype and differentiation potential. By analyzing expression of pluripotency markers, methylation at the OCT4 and NANOG promoters, and differentiation into teratomas, we determined that only one colony type represented bona fide iPS cells, whereas the others represented reprogramming intermediates. Proviral silencing and expression of TRA-1-60, DNMT3B, and REX1 distinguished the fully reprogrammed state, whereas Alkaline Phosphatase, SSEA-4, GDF3, hTERT and NANOG proved insufficient as markers. Reprogramming in chemically defined medium favored formation of bona fide iPS cell colonies relative to partially reprogrammed colonies. These data highlight the need for rigorous characterization and standardization of putative iPS cells.