La leucémie Myéloïde Chronique est un néoplasme myéloprolifératif causé par l'expression de la kinase oncogène BCR-ABL. Les Inhibiteurs de Tyrosine Kinase (ITK) spécifiques de BCR-ABL ont révolutionné la prise en charge de la maladie. Les ITK ne sont cependant pas curatifs ; en effet, certaines cellules souches leucémiques (CSL) sont résistantes aux ITK, et persistent dans la moelle osseuse des patients même en rémission prolongée. Ces CSL sont probablement responsables de la rechute chez 60% de ces patients après arrêt des ITK. 30% des patients développent une résistance aux ITK via des mécanismes inconnus. Dans un contexte sain, les Bone Morphogenetic Proteins (BMP) régulent différentes propriétés des cellules souches hématopoïétiques. Nous avons mis en évidence que les patients atteints de LMC présentent une altération de la voie BMP avant leurs mises sous traitement, avec une hausse de l'expression du récepteur dans les cellules leucémiques immatures, amplifiée par de forts taux de BMP2/4 produits par le microenvironnement des CSL, la niche. Ici, nous démontrons que ces altérations sont maintenues chez les patients sous traitement, et sont activement impliquées dans la résistance aux ITK. Les patients résistants présentent une surexpression de BMPR1b dans les CSL et un maintien de forts taux de BMP produits à la fois par les cellules leucémiques mais aussi par les cellules stromales. Les BMP permettent la survie des CSL via l'expression du récepteur BMPR1b et induisent l'expression de TWIST-1, un facteur de transcription précédemment identifié par l'équipe comme induisant la résistance / Chronic Myeloid Leukemia (CML) is a myeloproliferative neoplasm caused by the expression of the oncogenic protein kinase, BCR-ABL. The Tyrosine Kinase Inhibitors (TKI) specifics of BCR-ABL kinase dramatically changed the outcome of CML, turning a life-threatening disease into a chronic illness. However, TKI are not yet curative since most CML patients still retain progenitors and leukemic stem cells (LSC) in bone marrow permanently. Thus, approximately 60% of patients that achieve Complete Molecular Remission =2 years relapse following TKI withdraw. Moreover, some patients develop true resistance to TKI, with ~30% due to unknown mechanisms. In chronic phase CML (CP-CML), LSC survive, sustain interactions with their niche where resistance mechanisms can occur, responsible for disease persistence and relapse following treatment cessation. In normal bone marrow, Bone Morphogenetic Proteins (BMP) pathway regulate the fate and proliferation of normal hematopoietic stem cells, as well as interactions with their niche. The deregulations of this pathway drive early steps of CML development. In newly diagnosed CP-CML patients, high concentration of BMP2/4 in the leukemic niche allows LSC maintenance and sustains a permanent pool of leukemic progenitors expressing elevated levels of BMPR1b receptor. Here, we report that alterations of the BMP pathway persist in TKI-CML resistant patients. As compared to patients in Complete Cytogenetic Remission (CCyR), cells isolated from TKI-resistant patients display a high level of BMPR1b expression in immature cells and high levels of BMP2/4 in bone marrow, provided by the niche and by the leukemic immature cells themselves. BMP allow leukemic stem cells resistance to treatments through binding to BMPR1b. Interestingly, BMP2/4-treated cells overexpressed TWIST-1, a transcription factor that we previously identified as a predictive factor of CML resistance

Les tumeurs mammaires sont connues pour présenter une grande hétérogénéité intratumorale qui contribue à l’échec thérapeutique et à la progression de la maladie. L’origine de dette hétérogénéité s’explique principalement par l’organisation hiérarchique des tissus tumoraux où plusieurs sous-populations de cellules souches de cancer du sein (bCSC) sont capables de s’auto-renouveler et de maintenir l’architecture oligoclonale de la tumeur. Dans la mesure où les bCSC stimulent la croissance tumorale, résistent aux thérapies conventionnelles et initient le développement des métastases, il est indispensable de développer des thérapies spécifiques ciblant ces cellules. L’élaboration d’une telle stratégie nécessite la compréhension des propriétés moléculaires intrinsèques des bCSC. Pour mieux comprendre leur biologie, nous avons isolé les bCSC de différentes xénogreffes dérivées de tumeurs de patientes et établit leurs profil d’expression génique. Nous avons identifié un programme transcriptionnel pouvant être impliqué dans la réduction du stress réplicatif (SR) des bCSC . Nous avons montré que comparé aux non-bCSC, les bCSC présentent une sur-activation de la recombinaison homologue qui leur permet de réduire leur niveau de SR. Nous avons ensuite montré en réalisant un essai clinique que l’inhibition de cette voie permet de les sensibiliser à des agents génotoxique. Ces travaux identifient le SR comme le talon d’Achille des bCSC et mettent en évidence la recombinaison homologue comme cible potentielle pour sensibiliser les BCSC aux thérapies conventionnelles. / Breast tumors are known to present a major intratumoral heterogeneity that contributes to therapy failure and disease progression. The origin of this cellular heterogeneity is mainly explained by a hierarchical organization of tumor tissues where several subpopulations of self-renewing breast cancer stem cells (bCSCs) sustain the long-term oligoclonal maintenance of the neoplasm. bCSCs drive tumor growth, resist to conventional therapies and initiate metastasis development. Thus, developing bCSC-targeting therapies is becoming a major challenge requiring the understanding of the unique molecular circuitry of bCSC as compared to non-bCSC. To better understand the biology of these cells, we isolated bCSCs from different patient–derived xenografts (PDXs), derived fom breast tumors, and established their gene expression profiles. We identified a bCSC core transcriptional program that may be implicated in the reduction of the replicative stress in CSC: overexpression of genes implicated in dNTP metabolism and homologous recombination (HR). Our results show that HR plays a major role in SR regulation of bCSC and that bCSC are more resistant to RS than non-bCSC, We realized a preclinical assay in PDX and showed that HR inhibition prevent bCSC expansion Cisplatin-induced, suggesting a sensitization of the bCSC to the chemotherapy. Our results identify replication stress as the Achilles’ heel of bCSC and highlights HR as potential targets for anti-bCSC therapy.

Chronic Myelogenous Leukemia (CML) is a blood cancer that arises when hematopoietic cells acquire an abnormal protein known as BCR-ABL. Current therapies for CML include drugs that inhibit BCR-ABL. However, these drugs only suppress the disease and do not cure it. One reason is that BCR-ABL drugs fail to kill the primitive population of CML cells, referred to as leukemia stem cells (LSCs), which are responsible for initiating and propagating CML. Since LSCs are not killed, the cancer is not cured and many affected patients eventually relapse. Recent studies suggest that LSCs are protected from current therapies by the bone marrow micro-environment where they reside. There, cytokine signaling molecules are present, which mediate processes that protect LSCs from BCR-ABL drugs. The stem cell factor (SCF) is one of these signaling molecules. It activates the receptor c-Kit located on the surface of LSCs, and this activation in turn allows proliferating LSCs to resist BCR-ABL drugs, even without prior exposure to these drugs, i.e., innate resistance is observed. In this thesis, the mechanism of this innate resistance is investigated, so that a suitable treatment strategy can be developed. To this end, a co-agent approach based on synthetic antibodies (sABs) is proposed to inhibit the receptor c-Kit, with the goal of disrupting its activation by the ligand SCF. This disruption should in turn block the SCF-mediated innate resistance, thus potentially restoring BCR-ABL drug apoptotic activity. The method for this disruption involves targeting the c-Kit structural susceptibility. Specifically, the sABs are designed via antibody phage display technology to target the D1–D2–D3 domains representing the SCF binding sites, hence preventing downstream pathway activation. The hypothesis is that, by blocking the SCF-mediated innate resistance, a suitable combination of such an sAB co-agent and a BCR-ABL drug should be conducive to suppressing LSCs, thereby providing a potential means to improve CML treatment. In addition, to assess the performance of the proposed treatment strategy, a set of in vitro tests is conducted, focusing on performance behaviors such as cell binding, cell death, and the progenitor inhibition. The experimental results support the hypothesis that the proposed combinatorial strategy is indeed a promising approach to mitigate the innate resistance, thus restoring BCR-ABL drug apoptotic activity.

Une nouvelle théorie du cancer s’est récemment imposée dans la communauté scientifique. Selon cette dernière, les cancers se développeraient à partir d’une sous-population bien particulière de cellules cancéreuses, appelées « cellules souches cancéreuses » (CSC). Les partisans de la théorie des CSC soutiennent que les rechutes seraient causées par ces cellules, plus aptes à échapper aux thérapies classiques. En conséquence, ils soutiennent que l’élimination de toutes les CSC, dans un cancer donné, est nécessaire et suffisante pour guérir le patient. Dans cette thèse, je propose d’examiner cette stratégie thérapeutique de ciblage des CSC et je montre que sa capacité à guérir les cancers dépend de la façon dont on envisage la nature de la propriété souche. En effet, les cellules souches cancéreuses sont définies par la possession de la propriété souche, c’est-à-dire par leur capacité à s’auto-renouveler et à se différencier. Cependant, cette propriété elle-même reste obscure quant à sa nature. S’agit-il d’une propriété catégorique ou d’une disposition ? Une cellule non-souche (cancéreuse ou non) peut-elle acquérir la propriété souche et sous quelle condition ? En me basant sur une analyse de la littérature scientifique, je montre que quatre conceptions distinctes de la nature de la propriété souche sont aujourd’hui possibles et que, si la théorie des CSC est vraie, déterminer la nature exacte de la propriété souche est capital pour le traitement des cancers. / A new theory of cancer has recently gained importance in the scientific community. According to this theory, cancers develop from a particular sub-population of cancer cells, named “cancer stem cells” (CSCs). The proponents of the CSC theory argue that relapses are caused by CSCs because they escape classical therapies. Consequently, they claim that eliminating all the CSCs of a given cancer is a necessary and sufficient condition to cure the patient. In this dissertation, I scrutinize this therapeutic strategy and I argue that its ability to cure cancers will depend on our understanding of the nature of stemness. Indeed, cancer stem cells are characterized by this property, that is, the capacity to self-renew and to differentiate. However, the nature of stemness is rather obscure. Is it a categorical property or a disposition? Can a non-stem cell (cancerous or not) acquire stemness, and under which conditions? On the basis of analysis survey of the scientific literature, I distinguish four possible concepts of the nature of stemness. I contend that if the CSC theory is true, determining the exact nature of stemness is essential for cancers treatments.

Submitted by Fabrizio Grandi (fgrandivet@gmail.com) on 2016-03-10T14:56:58Z No. of bitstreams: 1 Doutorado.pdf: 5138229 bytes, checksum: cf3859209f1985d48832ef7e89af5b94 (MD5) / Approved for entry into archive by Sandra Manzano de Almeida (smanzano@marilia.unesp.br) on 2016-03-10T17:08:07Z (GMT) No. of bitstreams: 1 grandi_f_dr_bot.pdf: 5138229 bytes, checksum: cf3859209f1985d48832ef7e89af5b94 (MD5) / Made available in DSpace on 2016-03-10T17:08:07Z (GMT). No. of bitstreams: 1 grandi_f_dr_bot.pdf: 5138229 bytes, checksum: cf3859209f1985d48832ef7e89af5b94 (MD5) Previous issue date: 2016-02-29 / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) / O tumor venéreo transmissível (TVT) canino é uma neoplasia transplantável, considerada um alo-enxerto. Entretanto, pouco se sabe a respeito da origem e processo de cacinogênese. Atualmente, postula-se que alguns tumores originem-se de células iniciadoras tumorais, classicamente descritas nas leucemias mielóide humanas. As características intrínsecas do TVT fornecem indícios de uma possível participação de células iniciadores tumorais no processo de carcinogênese nesse tumor. Foi realizado estudo de fenotipagem do TVT canino para avaliar a marcação das proteínas CD44, CD133, CD90 e CD34, comumente associadas ao potencial iniciador tumoral. Para tanto utilizou-se a citometria de fluxo, imuno-histoquímica e RTq-PCR. Foram analisados também as frações de crescimento pelo Ki-67) e o número de células em apoptose pela caspase-3 clivada. Trinta e oito amostras de TVT foram obtidas de pacientes sem tratamento quimioterápico prévio. As amostras foram classificadas em plasmocitóide ou mistas, de acordo com o subtipo citológico; as células positivas na citometria de fluxo foram representadas em termos percentuais para os marcadores CD44, CD34, CD90 e CD133; a fração de crescimento foi representada pela técnica do H-Score; a quantidade de células apoptóticas foi representada pelo somatório de células positivas para a caspase-3 clivada; as imuno-marcações das proteínas CD44 e CD34 foram representadas por escores semi-quantitativos baseados na intensidade e percentual de células positivas; as expressões de RNAm foram calculadas em termos relativos; ainda, os pacientes foram divididos em grupos resistente e não resistente e comparados quanto a expressão dos marcadores de células iniciadoras supracitados. Não foram observadas diferenças significativas entre os marcadores e os grupos citológicos plasmocitóide e misto; o percentual de células CD44+ comumente foi superior a 90%, enquanto que o percentual de células CD34+, menor que 0,5%; o percentual de células CD90+ e CD133+ variou amplamente; houve uma tendência em termos de diferença estatística entre os grupos quimiorresistente e não resistente; houve correlação forte entre o percentual de células CD133+ e CD90+ na citometria de fluxo. O estudo permitiu verificar diferentes níveis de expressão protéica e gênica nas amostras de TVT canino; ainda, os grupos citológicos aparentam não possuir diferenças com relação a expressão dos marcadores CD44, CD90, CD133 e CD34; os grupos quimiorresistentes e não-resistentes parecem diferir com relação a expressão dos marcadores de células iniciadoras tumorais. / The canine transmissible venereal tumor (CTVT) is a transplantable neoplasia considered an allograft. Information about the origin and carcinogenesis process is scarcely known. Currently, some neoplasms are believed to arise from a tumor-initiating cell (TIC´s) classically described in human myeloid leukemia. TVT intrinsic characteristics provide evidence of a possible TIC´s participation in carcinogenesis process of this malignancy. Thus, a phenotyping study of CTVT was conducted to assess the immunophenotyping properties of the proteins CD44, CD133, CD90 and CD34, already known to be associated to tumor initiator potential. The use of flow cytometry and immunohistochemistry contributed to this purpose. In addition, growth fractions and cells undergoing apoptosis were examined by Ki-67 and caspase-3 cleaved, respectively. Thirty-eight samples were chosen from patients having no previous chemotherapy and cytological diagnosis of CTVT. Samples were classified into plasmacytoid or mixed according to cytological subtype. Positive cells in the flow cytometry were expressed in percentage for the markers CD44, CD34, CD90 and CD133. H-score technique helped to represent growth fractions. Apoptotic cell quantity was calculated by summing positive cells. Immunohistochemical marking of CD44 and CD34 proteins were determined by semiquantitative scores based on the intensity and percentage of positive cells. Moreover, specimens were divided into resistant and non-resistant groups and compared according cell marker expressions cited before. No significant differences appeared between the markers, and cytological plasmacytoid and mixed groups. The CD44 + cells and CD34 + percentages showed up high and low values, respectively. CD90 + and CD133 + cell percentages presented variable values. Amplitudes of the gene expression values among markers were similar to those observed in flow cytometry with a low expression of CD34, and a high expression of CD44. There was a positive statistical tendence between chemo-resistant and non-resistant groups, as well as a strong correlation between the percentage of CD133 + and CD90 + in flow cytometry. Besides, cytological groups apparently have no differences with the marker expression of CD44, CD90, CD133 and CD34. Resistant and non-resistant groups to chemotherapy seems to differ with respect to the marker expression of TIC´s.

INTRODUÇÃO: Os múltiplos passos da carcinogênese do câncer de cólon envolvem a existência de subpopulações de células tronco tumorais (CSC), responsáveis pela transformação, crescimento e proliferação das células tumorais. As proteínas CD44 e CD166 são marcadores de CSC associados a sinalização celular, adesão, migração, metástase e resposta linfocitária. Alguns fatores podem modular a expressão CSC como a mutação KRAS. OBJETIVO: correlacionar a expressão dos marcadores CD44 e CD166 em carcinoma de cólon metastático e status do oncogene KRAS (selvagem/mutado) com as características clínico-patológicas e desfecho do paciente ao final do seguimento. MATERIAL E MÉTODOS: Foram coletadas 58 amostras de tecido tumoral de pacientes com neoplasia de cólon metastático, tratados com CapeOx no Serviço de Oncologia Clínica do HCFMRPUSP de 2003 a 2012. Foram coletadas informações do prontuário sobre status do gene KRAS, características clínico-patológicas e desfecho clínico, sendo também realizada imunohistoquímica para marcação CD44 e CD166 através da técnica de TMA. Utilizado software SPSS 17 para análise estatística e considerado valor de p<0,050 para significância dos dados. RESULTADOS: A expressão de CD44 e CD166 foi positiva em 41,4% e 43%, respectivamente, e o status KRAS mutado em 48,3%. No subgrupo kAs selvagem e nos idosos (>65 anos), houve associação entre CD44 e CD166, p=0,042 e p=0,001, respectivamente. Pacientes CD166 negativo tiveram 3 vezes mais chances de progressão de doença (p=0,02) do que CD166 positivo. Pacientes Kras mutado e CD166 negativo tiveram 8 vezes risco de progressão (p<0,01). Pacientes CD44 positivo tiveram 4 e 5 vezes mais chances de evoluir com metástases hepática e pulmonar (p<0,01) em relação aos CD44 negativo. Pacientes com a combinação KRAS mutado e CD44 positivo tiveram 7 vezes mais chance de evoluir com metástase pulmonar (p=0,02) em relação a pacientes KRAS selvagem e CD44 negativo. DISCUSSÃO: Na amostra estudada observamos a influência das expressões dos marcadores de CSC e suas combinações com o status de mutação do gene KRAS, de modo que pacientes com CD166 negativo no tumor primário apresentam um desfecho de maior recorrência e o CD44 positivo favorece a evolução para metástases pulmonar e hepática. A mutação do gene KRAS atua modulando a via do EGF influenciando o comportamento biologico do tumor e os desfechos (recidiva e metastases) diretamente relacionados com a expressão dos marcadores de CSC no cancer de colon metastatico. CONCLUSÃO: Este estudo demonstrou interação entre a expressão imuno-histoquímica dos marcadores CSC de cólon (CD166 e CD44) e o status KRAS, podendo carcterizar subgrupos de pacientes com maiores chances de evolução desfavorável e assim propor um modelo de tratamento e seguimento mais individualizado. / BACKGROUND: Colon cancer carcinogenesis has been recently correlated with specific cancer stem cell (CSC) subpopulations which are associated with transformation, growth and spread process of tumor cells. CD44 and CD166 are CSC markers correlated with cell signalization, adhesion, migration, metastasis, and lymphocyte response. Some factors as KRAS mutation could modulate CSC. OBJECTIVE: Analyze CD44 and CD166 expressions in metastatic colon carcinoma and its correlation with KRAS status, clinicopathological features, disease recurrence and patient survival. MATERIAL AND METHODS: Tissues were obtained from 58 patients with confirmed metastatic colon cancer, treated with CapeOx at FMRP-USP from 2003 to 2012. Clinical and outcomes informations and KRAS gene status were obtained from medical records. KRAS status was analyzed with RT-PCR. CD44 and CD166 were analyzed with TMA immunohistochemistry. Statistical analyses were performed using SPSS 17.0. A p-value <0,050 was considered to be statistically significant. RESULTS: CD44 and CD166 expressions were positive in 41,4% and 43%, respectively, and KRAS status was mutate in 48,3%. Wild-type KRAS in elderly patients had statistical association between CD44 and CD166, p=0,042 and p=0,001, respectively. Patients with CD166 negative had 3 fold increase in progression disease (p<0,01). Patients with CD44 positive had 4 and 5 fold increase in liver and lung metastasis (p<0,01), respectively. Patients with combined mutated KRAS and CD44 positive had 7 fold increase in lung metastasis (p=0,02) compared with wildtype KRAS and CD44 negative. DISCUSSION: In this study, the influence of markers expression of colon CSC (CD44 and CD166) and its combinations with status KRAS were proven. Patients with CD166 negative in primary colon tumor are more likely to present higher recurrence and, CD44 positive have a higher chance to develop lung and liver metastasis. KRAS mutation contributed, associated with studied CSC expressions, to cancer biological behavior and agressivness. CONCLUSIONS: This study demonstrated interaction between imunohistochemical expression of colonic CSC markers (CD166 and CD44) and KRAS gene status. Subgroups of patients with worse outcomes could be identified and this biological information contributed to personalized treatments and follow ups that should be proposed for these patients.

Ce projet explore les mécanismes de la morphogenèse des glandes mammaires, comme modèle de progression du carcinome du sein. La morphogenèse de la glande mammaire résulte de la coordination de réponses cellulaires distinctes (prolifération, différenciation, motilité, invasivité, apoptose) régulées par de nombreuses voies, y compris Wnt, EGF, FGF, Notch, SHH, Myc et l'activation hormonale. Nous estimons qu'il est essentiel d'analyser individuellement l'impact de ces voies dans la modulation de la prolifération, de la différenciation, de la motilité, de l'invasivité, de l'apoptose, de la cohésion intercellulaire et de la polarité dans les cellules impliquées dans une migration morphogénétique cohérente.Nous avons développé des modèles en 3D améliorés pour analyser l'impact d'EMT-TF dans un environnement 3D. Notre système permet de surveiller simultanément les voies mentionnées au niveau cellulaire pendant trois semaines, une période ajustée pour tester les médicaments de chimiothérapie. Notre premier modèle décrit l'émergence primaire des cellules envahissantes de carcinome mammaire de l'épithélium mammaire. Les cellules sont traitées avec des médicaments définis ou seront transfectées avec diverses constructions (en cours de validation) améliorant ou réprimant des voies spécifiques telles que Slug, en plus des constructions permettant de suivre l'évolution des structures cellulaires par marquage GFP en vidéomicroscopie. / This project explores the mechanisms of mammary gland morphogenesis, as a model for breast carcinoma progression. Mammary gland morphogenesis results from the coordination of distinct cell responses (proliferation, differentiation, motility, invasiveness, apoptosis) integrated by numerous pathways, including Wnt, EGF, FGF, Notch, SHH, Myc and hormonal activation. For the purpose of this study, we feel it is critical to analyze individually the impact of theses pathways in modulating proliferation, differentiation, motility, invasiveness, apoptosis, intercellular cohesion, and polarity in cells involved in a coherent morphogenetic migration.We have designed improved 3D models to analyze the impact of EMT-TF in a 3D environment. Our system allows monitoring simultaneously the mentioned pathways at a cellular level for three weeks, a period adjusted to test chemotherapy drugs.Our first model describes the primary emergence of invading breast carcinoma cells from mammary epithelium. Cells are treated with defined drugs or will be transfected with various constructs (under validation) enhancing or repressing specific pathways such as Slug, in addition to constructs allowing the monitoring of cell structures by GFP labeling for video microscopy..

Le carcinome hépatocellulaire est un grand problème de santé publique et la troisième de mortalité lié au cancer dans le monde. Il a été démontré qu'au sein des tumeurs se trouve un petite population des cellules cancéreuses avec des propriétés de cellules souches cancéreuses. Elles sont responsables de l'initiation des tumeurs ainsi que de la récidive post-traitement et résistance aux thérapies. Peu de choses sont connues par rapport à la biologie de ces cellules mais l'identification des facteurs favorisant leur existence pourrait conduire vers des nouvelles pistes thérapeutiques.Nous avons trouvé que le facteurs de transcription p73 est surexprimé chez les patients dans les tumeurs du carcinome hépatocellulaire sous forme de deux types d'isoformes, les isoformes complets et les isoformes tronqués. Les isoformes complets sont des suppresseurs de tumeurs et corrèlent avec un meilleur taux de survie alors que les isoformes tronqués agissent comme dominants négatifs de ces premiers et favorisent la récidive post-chirurgie.Les résultats in vitro ont montré que les isoformes tronqués de p73 sont surexprimés dans les cellules souches cancéreuses du carcinome hépatocellulaire et favorisent leur emergence / Hepatocellular carcinoma (HCC) is a major public health problem, being the second most lethal cancer with an increasing incidence around the world. The only approved systemic drug is the multikinase inhibitor Sorafenib, which prolongs patients’ survival by only three months. HCC is refractory to known chemotherapeutic drugs and more than 50% of patients relapse after surgical tumor removal. These phenomena are thought to be due to the existence of a population of poorly differentiated cancer cells, largely known as liver cancer stem cells (CSCs). Recent studies revealed that CSCs activate similar pathways as normal stem cells. They are therefore highly resistant to therapies and are thought to be capable of self-renewal and generation of tumor’s heterogeneous cell mass. The understanding of mechanisms proper to liver CSCs should allow the development of innovative drugs with original mechanism of action against liver CSC, likely to improve patients’ outcome. However, the development of new therapies against HCC is penalized by the limited number of experimental models.According to these current challenges in the field of HCC research, my PhD thesis project covers three main axes: Development of novel models of disease (IMODI consortium)The Innovative Models of Disease (IMODI) consortium is mainly dedicated to the development of innovative experimental models for 7 different types of cancer. Our participation to the project concerned 3 main objectives i) development of HCC patient-derived xenografts ii) development of new HCC cell lines and iii) set up a cryoconservation method of primary human hepatocytes (PHHs) in the aim to employ them in humanizing murine livers. 30 patients planned for HCC tumor resection were recruited and their clinicopathological data were collected. Fresh tumor specimens were subcutaneously xenografted in immune-deficient mice and dissociated for in-vitro tumor cell culture. One tumor led to the development of a moderately differentiated HCC PDX model, as confirmed by histological characterization. Several studies showed the importance of PDX models in drug discovery as they recapitulate the drug-sensitivity patterns seen in patients from which they derive but very few models have been described in the literature for HCC. In vitro, primary HCC cells could be maintained in culture for a limited period of time, in average 30 days. No HCC cell lines developed due to cells entering replicative senescence, as previously described

Les glioblastomes (GBM) sont des tumeurs astrocytaires au pronostic défavorable. L’échec des thérapies actuelles (chimio et radiothérapies) est principalement lié à la résistance des cellules souches cancéreuses (CSCs). Ces cellules ont besoin de communiquer en permanence avec leur microenvironnement pour leur survie et pour maintenir une niche favorable à leur développement. Le transfert de matériel entre les CSC, les cellules tumorales et le microenvironnement contribue à l’échappement thérapeutique. Des travaux récents révèlent l’importance des récepteurs aux neurotrophines TrkB et TrkC dans la survie et la croissance des CSC de GBM. Nos travaux préliminaires dans le cancer bronchique démontrent que les récepteurs aux neurotrophines sont transférés aux cellules du microenvironnement via les exosomes afin de les contrôler. Cependant, le mécanisme de diffusion de récepteurs oncogéniques à partir de CSC n’a jamais été étudié. Notre objectif principal était donc de déterminer l’implication des récepteurs des neurotrophines dans le transfert du phénotype agressif des CSC vers les cellules du microenvironnement afin de favoriser la résistance thérapeutique du glioblastome. Nos résultats ont permis d’établir un lien entre le stade de différenciation des cellules tumorales, l’expression des neurotrophines et leur interaction avec le microenvironnement tumoral via les exosomes. Le transfert de TrkB au sein des exosomes joue un rôle clé dans la progression tumorale du GBM et dans l’agressivité cellulaire. Néanmoins, le transfert des récepteurs aux neurotrophines via les exosomes pourrait également être impliqué dans les mécanismes de radiorésistance. Des études menées sur des cellules de GBM humain irradiées et traitées par des exosomes démontrent l’implication de ces derniers dans l’échappement thérapeutique. Parmi les cellules du microenvironnement ciblées par les exosomes, les CSM sont celles qui ont été les moins étudiées bien qu’elles possèdent un tropisme spécifique pour le GBM. Nos travaux démontrent que les exosomes de GBM modifient le phénotype des CSM et augmentent leurs capacités prolifératives et migratoires. La fonction exacte du transfert des récepteurs des neurotrophines devra être analysée dans ces différents modèles afin de préciser son importance dans la physiopathologie du glioblastome et sa progression. L’expression des récepteurs aux neurotrophines dans ces exosomes permet d’envisager leur utilisation en tant que biomarqueurs diagnostiques et/ou pronostiques dans le GBM. Mots clés : Glioblastomes, cellules souches cancéreuses, neurotrophines, TrkB, radiothérapie, cellules souches mésenchymateuses, exosomes. / Glioblastoma are tumors derived from astrocytes with a dark prognosis. Current therapies fail to inhibit relapses due to radioresistant properties of cancer stem cells (CSC). Communication between CSC and their microenvironment is required for maintain “stem cells niche” and cell survival . The transfer of materials between CSC, tumor cells and microenvironment contributes to therapeutic resistance. In glioma, recent studies reveal the major role of TrkB and TrkC in survival of CSC. Our previous work, in lung cancer, have shown that neurotrophin receptors exhibits a control on microenvironment cells and angiogenesis through exosome transfer. However, similar mechanism of oncogenic receptor transfer from CSC has never been studied. Our main goal was to determine the involvement of neurotrophin receptors in the transfer of biological aggressiveness to microenvironment cells in order to promote therapeutic resistance in glioblastoma. Our findings suggest a relationship between cell differentiation status, expression of neurotrophin receptors and their interaction with the microenvironment through exosomes. TrkB-containing exosomes play a key role in the control of glioblastoma progression and cell aggressiveness. Mechanisms of radioresistance might also be dependent of the transfer of neurotrophin receptors through exosomes. Indeed, our results on irradiated human GBM cells and treated by exosomes demonstrate the involvement of exosome in radioresistance mechanisms. Although mesenchymal stem cells (MSCs) are considered as stromal components of glioblastoma, their communication with CSC, particularly through exosomes, remain largely undefined. Our results show that GBM-derived exosomes modify the phenotype of MSCs and increase their proliferative and migratory abilities. The putative function of neurotrophin receptors transfer should be analyzed in these models to determine their prime role in glioblastoma pathogenesis and progression. This finding suggest that the neurotrophin receptor expression in exosomes could be used as diagnostis and prognosis biomarkers of GBM.

(EN) Tumours are heterogeneous and consist of multiple populations of cells. The population of cells with tumour-initiating capability is known as cancer stem cells (CSC). Cells with increased stemness properties and elevated resistance to anti-cancer treatment have been shown to be highly affected upon decline of mitochondrial respiration, linking the concept of CSCs to deregulated bioenergetics. Consistently, functional electron transport chain (ETC) is crucial in tumorigenesis. Expression of HER2 oncogene, associated with resistance to treatment in breast cancer, has been connected with regulation of mitochondrial function. We therefore investigated the possibility that manipulation of mitochondrial bioenergetics via disruption of ETC eliminates the conventional therapy-resistant populations of tumour, such as CSCs and HER2high cells. We demonstrate that HER2high cells and tumours have increased complex I-driven respiration and increased assembly of respiratory supercomplexes (SC). These cells are highly sensitive to MitoTam, a novel mitochondria-targeted derivative of tamoxifen, acting as a CI inhibitor and SC disruptor. MitoTam was able to overcome resistance to tamoxifen, and to reduce the metastatic potential of HER2high cells. Higher sensitivity of HER2high cells to MitoTam is dependent on...