Base Teórica: O câncer de mama é provavelmente o mais temido pelas mulheres devido a sua alta freqüência e, sobretudo, pelos seus efeitos psicológicos que afetam a percepção da sexualidade e a própria imagem pessoal. Ele é relativamente raro antes dos 35 anos de idade, mas acima desta faixa etária sua incidência cresce rápida e progressivamente. Estudos indicam que fatores genéticos e ambientais são responsáveis pela incidência do câncer de mama, sendo que a hereditariedade provavelmente tenha participação restrita no desenvolvimento deste tipo de tumor. Os principais genes associados ao desenvolvimento do câncer de mama, BRCA1 e BRCA2, são responsáveis por cerca de 80% desses casos, conferindo um risco de 71 a 85% de chance de desenvolver a neoplasia em alguma fase da vida. Mutações nesses genes, classificados como supressores tumorais, demonstram que a perda de suas funções não pára o ciclo celular, não permite a ação do sistema de reparo, e não estimula a apoptose (morte celular programada), culminando em replicação anormal e câncer. A observação epidemiológica de que mulheres judias de origem Ashkenazi parecem ser mais vulneráveis ao câncer de mama está sendo explicada através de estudos moleculares dos genes BRCA1 e BRCA2, onde encontramos a prevalência de três mutações específicas: 185delAG e 5382insC, no gene BRCA1 e 6174delT, no gene BRCA2. Métodos: Utilizou-se um banco de DNA pré-existente, extraído de 209 mulheres da comunidade judaica Ashkenazi da cidade de Porto Alegre. A amplificação do DNA foi realizada por PCR, através da técnica PSM (PCR Mediated site-direct) seguida de digestão dos produtos de PCR com enzimas de restrição. Os objetivos foram verificar se as freqüências das mutações 185delAG e 5382insC, no gene BRCA1 são significativas nesta população e compará-las com demais freqüências encontradas. Resultados: Foram encontradas três pacientes com a mutação 185delAG e duas pacientes com a mutação 5382insC, com as freqüências de 1,435% (95% IC: 0,366; 3,856) e 0,957% (95% IC: 0,161; 3,125), respectivamente. / Introduction: Breast cancer is probably the worst diagnosed cancer for women due to its high frequency and furthermore by its psychological problems that affect the perception of sexuality and the self image. It is relatively rare before 35 years of age, but beyond this age its incidence increases rapidly and progressively. Studies show that genetic and environmental factors are responsible for breast cancer incidence, but heredity may play a restrict role in the development of this kind of tumor. The main genes associated to the development of breast cancer, BRCA1 and BRCA2, are responsible for almost 80% of these cases, reaching a chance between 71 and 85% of developing the disease at any life stage. Mutations in these genes, classified as tumor suppressors, do not allow the repair mechanisms of DNA to perform its action and do not stimulate apoptosis, culminating in abnormal replication and cancer. The epidemiological observation in which Ashkenazi Jewish women seems to be more vulnerable to breast cancer is explained through molecular studies of BRCA1 and BRCA2 genes, where three specific mutations have been found (185delAG and 5382insC, in the BRCA1 gene and 6174delT, in the BRCA2 gene). Methods: A pre-existent bank of DNA extracted from 209 women of the Ashkenazi Jewish community of Porto Alegre city has been used. The DNA amplification was performed through PCR, using the PSM (PCR Mediated Site-Direct) technique followed by the digestion of PCR products with restriction enzymes. The objectives of this study was to identify the frequencies of mutations 185delAG and 5382insC at the BRCA1 gene and verify if they are significantly different in this population when compared to frequencies found in other studies. Results: We found three patients with 185delAG mutation and two patients with 5382insC mutation, with frequencies of 1.435% (95% CI: 0,366; 3,856) and 0,957% (95% IC: 0,161; 3,125), respectively.

Base Teórica: O câncer de mama é provavelmente o mais temido pelas mulheres devido a sua alta freqüência e, sobretudo, pelos seus efeitos psicológicos que afetam a percepção da sexualidade e a própria imagem pessoal. Ele é relativamente raro antes dos 35 anos de idade, mas acima desta faixa etária sua incidência cresce rápida e progressivamente. Estudos indicam que fatores genéticos e ambientais são responsáveis pela incidência do câncer de mama, sendo que a hereditariedade provavelmente tenha participação restrita no desenvolvimento deste tipo de tumor. Os principais genes associados ao desenvolvimento do câncer de mama, BRCA1 e BRCA2, são responsáveis por cerca de 80% desses casos, conferindo um risco de 71 a 85% de chance de desenvolver a neoplasia em alguma fase da vida. Mutações nesses genes, classificados como supressores tumorais, demonstram que a perda de suas funções não pára o ciclo celular, não permite a ação do sistema de reparo, e não estimula a apoptose (morte celular programada), culminando em replicação anormal e câncer. A observação epidemiológica de que mulheres judias de origem Ashkenazi parecem ser mais vulneráveis ao câncer de mama está sendo explicada através de estudos moleculares dos genes BRCA1 e BRCA2, onde encontramos a prevalência de três mutações específicas: 185delAG e 5382insC, no gene BRCA1 e 6174delT, no gene BRCA2. Métodos: Utilizou-se um banco de DNA pré-existente, extraído de 209 mulheres da comunidade judaica Ashkenazi da cidade de Porto Alegre. A amplificação do DNA foi realizada por PCR, através da técnica PSM (PCR Mediated site-direct) seguida de digestão dos produtos de PCR com enzimas de restrição. Os objetivos foram verificar se as freqüências das mutações 185delAG e 5382insC, no gene BRCA1 são significativas nesta população e compará-las com demais freqüências encontradas. Resultados: Foram encontradas três pacientes com a mutação 185delAG e duas pacientes com a mutação 5382insC, com as freqüências de 1,435% (95% IC: 0,366; 3,856) e 0,957% (95% IC: 0,161; 3,125), respectivamente. / Introduction: Breast cancer is probably the worst diagnosed cancer for women due to its high frequency and furthermore by its psychological problems that affect the perception of sexuality and the self image. It is relatively rare before 35 years of age, but beyond this age its incidence increases rapidly and progressively. Studies show that genetic and environmental factors are responsible for breast cancer incidence, but heredity may play a restrict role in the development of this kind of tumor. The main genes associated to the development of breast cancer, BRCA1 and BRCA2, are responsible for almost 80% of these cases, reaching a chance between 71 and 85% of developing the disease at any life stage. Mutations in these genes, classified as tumor suppressors, do not allow the repair mechanisms of DNA to perform its action and do not stimulate apoptosis, culminating in abnormal replication and cancer. The epidemiological observation in which Ashkenazi Jewish women seems to be more vulnerable to breast cancer is explained through molecular studies of BRCA1 and BRCA2 genes, where three specific mutations have been found (185delAG and 5382insC, in the BRCA1 gene and 6174delT, in the BRCA2 gene). Methods: A pre-existent bank of DNA extracted from 209 women of the Ashkenazi Jewish community of Porto Alegre city has been used. The DNA amplification was performed through PCR, using the PSM (PCR Mediated Site-Direct) technique followed by the digestion of PCR products with restriction enzymes. The objectives of this study was to identify the frequencies of mutations 185delAG and 5382insC at the BRCA1 gene and verify if they are significantly different in this population when compared to frequencies found in other studies. Results: We found three patients with 185delAG mutation and two patients with 5382insC mutation, with frequencies of 1.435% (95% CI: 0,366; 3,856) and 0,957% (95% IC: 0,161; 3,125), respectively.

Base Teórica: O câncer de mama é provavelmente o mais temido pelas mulheres devido a sua alta freqüência e, sobretudo, pelos seus efeitos psicológicos que afetam a percepção da sexualidade e a própria imagem pessoal. Ele é relativamente raro antes dos 35 anos de idade, mas acima desta faixa etária sua incidência cresce rápida e progressivamente. Estudos indicam que fatores genéticos e ambientais são responsáveis pela incidência do câncer de mama, sendo que a hereditariedade provavelmente tenha participação restrita no desenvolvimento deste tipo de tumor. Os principais genes associados ao desenvolvimento do câncer de mama, BRCA1 e BRCA2, são responsáveis por cerca de 80% desses casos, conferindo um risco de 71 a 85% de chance de desenvolver a neoplasia em alguma fase da vida. Mutações nesses genes, classificados como supressores tumorais, demonstram que a perda de suas funções não pára o ciclo celular, não permite a ação do sistema de reparo, e não estimula a apoptose (morte celular programada), culminando em replicação anormal e câncer. A observação epidemiológica de que mulheres judias de origem Ashkenazi parecem ser mais vulneráveis ao câncer de mama está sendo explicada através de estudos moleculares dos genes BRCA1 e BRCA2, onde encontramos a prevalência de três mutações específicas: 185delAG e 5382insC, no gene BRCA1 e 6174delT, no gene BRCA2. Métodos: Utilizou-se um banco de DNA pré-existente, extraído de 209 mulheres da comunidade judaica Ashkenazi da cidade de Porto Alegre. A amplificação do DNA foi realizada por PCR, através da técnica PSM (PCR Mediated site-direct) seguida de digestão dos produtos de PCR com enzimas de restrição. Os objetivos foram verificar se as freqüências das mutações 185delAG e 5382insC, no gene BRCA1 são significativas nesta população e compará-las com demais freqüências encontradas. Resultados: Foram encontradas três pacientes com a mutação 185delAG e duas pacientes com a mutação 5382insC, com as freqüências de 1,435% (95% IC: 0,366; 3,856) e 0,957% (95% IC: 0,161; 3,125), respectivamente. / Introduction: Breast cancer is probably the worst diagnosed cancer for women due to its high frequency and furthermore by its psychological problems that affect the perception of sexuality and the self image. It is relatively rare before 35 years of age, but beyond this age its incidence increases rapidly and progressively. Studies show that genetic and environmental factors are responsible for breast cancer incidence, but heredity may play a restrict role in the development of this kind of tumor. The main genes associated to the development of breast cancer, BRCA1 and BRCA2, are responsible for almost 80% of these cases, reaching a chance between 71 and 85% of developing the disease at any life stage. Mutations in these genes, classified as tumor suppressors, do not allow the repair mechanisms of DNA to perform its action and do not stimulate apoptosis, culminating in abnormal replication and cancer. The epidemiological observation in which Ashkenazi Jewish women seems to be more vulnerable to breast cancer is explained through molecular studies of BRCA1 and BRCA2 genes, where three specific mutations have been found (185delAG and 5382insC, in the BRCA1 gene and 6174delT, in the BRCA2 gene). Methods: A pre-existent bank of DNA extracted from 209 women of the Ashkenazi Jewish community of Porto Alegre city has been used. The DNA amplification was performed through PCR, using the PSM (PCR Mediated Site-Direct) technique followed by the digestion of PCR products with restriction enzymes. The objectives of this study was to identify the frequencies of mutations 185delAG and 5382insC at the BRCA1 gene and verify if they are significantly different in this population when compared to frequencies found in other studies. Results: We found three patients with 185delAG mutation and two patients with 5382insC mutation, with frequencies of 1.435% (95% CI: 0,366; 3,856) and 0,957% (95% IC: 0,161; 3,125), respectively.

Neste trabalho, medimos o impacto da avaliação do risco de mutações dos genes BRCA1/2 na qualidade de vida de pacientes com câncer de mama, avaliada pelos questionários EORTC QLQ-C30 e QLQ-BR23. Convidamos 282 pacientes a participar, respondendo aos questionários antes e depois da avaliação do risco pelos métodos de Frank, Evans e BRCAPRO. Consideramos risco elevado pelo menos 10%. 272 foram incluídas e 198 completaram o estudo. Nas 180 avaliáveis, a idade mediana das pacientes foi de 53 anos com desvio padrão de 11,5 anos e, em 89, o tempo desde o diagnóstico de menos de 36 meses. 40 pacientes estavam em seguimento e 137 em hormonioterapia. Não detectamos alterações significativas da qualidade de vida com a determinação do risco para mutações. Houve diferença significativa entre imagem corporal negativa e cirurgia conservadora da mama (p<0,001). Classificamos 45 como risco elevado pelo método de Frank, 35 pelo BRCAPRO e 21 por Evans, sendo que em 12 dessas pacientes houve concordância dos três métodos juntos. Concluímos que, apesar do grande interesse demonstrado pelas pacientes em participar no estudo, a determinação do risco não interferiu na qualidade de vida dessas pacientes. Aparentemente, as informações sobre hereditariedade são desejadas, mas não acarretam estresse adicional e deveriam ser prestadas, pois o número de pacientes com risco elevado é semelhante ao indicado em outras populações. Além disso, a baixa concordância entre os métodos utilizados indica a necessidade de definir parâmetros para determinação de risco em nosso meio. / Here we evaluated the “impact of breast cancer hereditary cancer risk evaluation’ on the quality of life in a population of breast cancer patients, as measured by the EORTC questionnaires QLQ-C30 and QLQ-BR23. Of the 282 invited patients, 272 agree to participate and answered QLQ before and after the risk determination by Frank, Evans and BRCAPRO methods. High risk was defined as at least 10%. Overall 198 pts completed the study. In the 180 evaluable patients, median age was 53 (+11,5) years old and time since diagnosis was less than 36 months in 89. We did not detected significant differences in quality of life parameters after risk determination, except for negative body image and mastectomy/conservative surgery (p<0.001) There were 45 patients classified as high risk by Frank, 35 by the BRCAPRO and 21 by Evans, agreement being reached in 12. We conclude that pts wish to know about their hereditary breast cancer risk, and this do not cause necessarily more stress. Apart from that, there is a need for local methods of risk calculation.

La poly(ADP-ribosyl)ation est une modification post-traductionnelle des protéines par les poly(ADP-ribose) polymérases (PARPs). PARP3 a été identifiée comme un nouvel acteur de la réparation des cassures double-brin (DSBs). Nous avons évalué la contribution de PARP3 dans les différentes voies de réparation (HR, C-NHEJ ou A-EJ). Les résultats obtenus définissent PARP3 comme un modulateur de l’étape de résection d’ADN simple-brin permettant d’engager le choix de la voie de réparation. Nous avons montré que PARP3 favorise le recrutement du complexe Ku70/Ku80 aux sites de cassures et module la balance BRCA1/53BP1. Ces deux événements limitent l’étape de réparation de la voie HR et A-EJ et oriente la réparation vers la voie du C-NHEJ. Par immunoprécipitation de la chromatine, nous avons étudié les conséquences de l’absence de PARP3 sur les modifications d’histones, connues pour moduler la décision entre les différentes voies de réparation. Nos résultats actuels ne nous ont pas permis d’établir de lien entre PARP3 et les modifications d’histones en réponse aux DSBs. Nous avons toutefois observé qu’en absence de dommages, l’absence de PARP3 induit un enrichissement de H3K36me2 une marque d’histone connue pour réguler les gènes transcriptionnellement actifs. Dans un second projet, nous avons étudié l’impact de l’absence de PARP3 sur la viabilité cellulaire et la progression tumorale de cellules cancéreuses mutées en BRCA1. Nous avons montré par des approches in vitro et in vivo que l’absence de PARP3 induit une diminution de la survie et de la prolifération cellulaire plus marquée, une amplification exacerbée des centrosomes, ainsi qu’un ralentissement plus important de la progression tumorale, faisant de PARP3 une cible prometteuse en thérapie du cancer. / Poly(ADP-ribosyl)ation is a post-translational modification of proteins catalyzed by poly(ADPribose) polymerases (PARPs). PARP3 was identified as a novel actor of the double-strand break (DSBs) repair pathway. We evaluated the contribution of PARP3 in these repair pathways(HR, C-NHEJ ou A-EJ). Our results defined PARP3 as a modulator of the single strand DNA resection process which plays a role in driving the repair pathway choice. We showed that PARP3 enhances the recruitement of the Ku70/Ku80 complexe to damaged sites and modulates the BRCA1/53BP1 balance. These two events prevent the DNA end resection step initiating HR and A-EJ and drives the repair towards the C-NHEJ. By chromatin immunoprecipitation, we studied the consequences of the absence of PARP3 on histone modifications, known to modulate the decision of the DSBs repair pathways. Our current results didn’t allow us to establish a link between PARP3 and histone modifications in response to DSBs. However, in absence of DNA damage and PARP3, we observed an accumulation of H3K36me2, a histone mark known to regulate transcriptionally active genes. In a second project, we studied the impact of the absence of PARP3 on cell viability and tumor progression in breast cancer cell lines mutated in BRCA1. By in vitro and in vivo approaches, we showed that the absence of PARP3 induces an important decrease in cell survival and proliferation, an increase in centrosomal amplification and a strong delay in tumor progression. The roles of PARP3 in both cellular response to DNA damage and mitotic progression introduce PARP3 as a possible promising therapeutic target in cancer therapy.

Environ 6000 à 8000 maladies rares différentes existent aujourd’hui, affectant environ 6 à 8% de la population mondiale. La grande majorité d’entre elles correspond à des maladies génétiques, pour lesquelles il n'existe pas de traitement curatif. La révolution génomique a augmenté l’espoir d’obtenir des traitements spécifiques du défaut génétique pour de nombreuses maladies. Dans ces conditions, le traitement et l’analyse des données ne sont pas triviaux et s’éloignent de la simple routine.Cette thèse rapporte la création de Saut d'exon, capables d’aider les chercheurs et les cliniciens à identifier les mutations responsables de certaines maladies et développer de nouvelles stratégies thérapeutiques. Ainsi, les systèmes Human Splicing Finder et UMD-Predictor permettent respectivement de prédire l’effet d’une mutation sur l’épissage et la fonction de la protéine. Ils ont été validés grâce à des jeux de données de référence et/ou issus de la littérature, et peuvent aider les cliniciens à annoter correctement les variations de signification inconnue. De plus, cette thèse propose deux outils à visées thérapeutiques : Skip-E, un outil d’identification des AON candidats à la thérapie par saut d’exon, et NR-Analyser, un système de prédiction des codons de terminaison prématurés candidats à la thérapie par translecture des codons stop.Ces différents systèmes s'intègrent dans un projet plus global dédié à la recherche translationnelle. Par ces deux volets, prédictif et thérapeutique, cette thèse s’inscrit dans une stratégie de recherche en adéquation avec les objectifs du Consortium International pour la Recherche contre les Maladies Rares, l’IRDiRC. / About 6,000 to 8,000 distinct rare diseases exist today and are estimated to affect 6-8% of the world population. The vast majority of them are genetic and for most of them there is no cure. The genomic revolution has increased the hope of specific treatments based on the gene for many diseases. New technologies have emerged, changing drastically data scale produced in biomedical research. In these conditions, treatment and analysis of data are far from trivial and mere routine, despite spectacular advances in computer technology.This thesis reports the creation of bioinformatics systems, capable of helping researchers and clinicians to identify mutations responsible for certain diseases and to develop new therapies. Thus, the Human Splicing Finder and UMD-Predictor systems predict the effect of a mutation on splicing and protein, respectively. Both bioinformatics systems have been validated through high quality reference datasets, and may help clinicians to properly annotate variations of unknown significance. In addition, this thesis offers two new systems for therapeutic purposes: the Skip-E system identifies optimal candidates AONs for exon skipping therapies, and NR-Analyser, a system that predicts premature termination codons potentially candidates to nonsense readthrough therapies.These different systems are part of a larger project dedicated to translational research. With its predictive and therapeutic aspects, this thesis is part of a research strategy matching with the objectives of the IRDiRC (International Rare Diseases Research Consortium).

Dans les cancers du sein triple négatif sporadiques, BRCA1 est fréquemment inactivé au niveau transcriptionnel, et il a été rapporté que cette inactivation peut être réalisée par une méthylation du promoteur. Plus récemment, il a été constaté que BRCA1 peut également être régulée au niveau post-transcriptionnel par les microARNs. L'accumulation de preuves indique que les miARNs ont un rôle causal dans la tumorigenèse. Nos travaux se sont axés sur l'étude de l’expression et des fonctions des microARN in vitro, in silico et ex vivo.Basé sur nos résultats de profilage de l'expression, quatre miARN candidats (miR-10b, miR-26a, miR-146a et miR-153) ont été choisis comme étant potentiellement impliqués dans le développement du cancer du sein triple négatif. Des essais d'expression exogènes ont révélé que miR-10b et miR-26a, mais pas miR-146a, peuvent réguler négativement l'expression du gène BRCA1 dans les cellules cancéreuses triple négatif MDA-MB-231 et luminales MCF7, alors que miR-153 pourrait réguler négativement l'expression du gène BRCA1 uniquement dans les cellules MCF7. L'analyse in silico des données de Cancer Genome Atlas (TCGA) a confirmé que miR-146a est significativement plus exprimé dans les tumeurs du sein triple négatif par rapport à d'autres tumeurs (non triple négatif) mammaires. L’étude ex vivo a montré que le niveau élévé d’expression de miR-146a et de miR-26 est associé à l’absence des métastases ganglionnaires dans le cancer du sein triple négatif. Aussi une corrélation entre l’expression de 4 miARNs est révélée permettant l’identification de différentes voies de signalisations impliquées dans le cancer du sein triple negatif.Nos travaux fournissent des preuves de l'implication des miARNs spécifiques comme des biomarqueurs potentiels dans le développement du cancer de sein triple négatif. / In sporadic triple-negative breast cancers BRCA1 is frequently inactivated at the transcriptional level, and it has been reported that this inactivation may be brought about by promoter methylation. More recently, it was found that BRCA1 may also be regulated at the post-transcriptional level by miRNAs. Accumulating evidence indicates that miRNAs have a causal role in tumorigenesis. Our work focused on the study of microRNAs expression and functions in vitro, in silico and ex vivo.Based on our expression profiling results, four candidate miRNAs (miR-10b, miR-26a, miR-146a and miR-153) were selected as being potentially involved in triple-negative breast cancer development. Exogenous expression assays revealed that miR-10b and miR-26a, but not miR-146a, can down-regulate the expression of BRCA1 in both triple-negative MDA-MB-231 and luminal epithelial MCF7 breast cancer-derived cells, whereas miR-153 could down-regulate BRCA1 expression only in MCF7 cells. In silico analysis of The Cancer Genome Atlas (TCGA) data confirmed that miR-146a is significantly higher expressed in triple-negative breast tumors compared to other (non triple-negative) breast tumors. The ex vivo study showed that the high level expression of miR-146a and miR-26 is associated with the absence of lymph node metastasis in triple negative breast cancer. Also a correlation between the expression of the 4 miRNAs was revealed, allowing the identification of different signaling pathways involved in the triple negative breast cancer.Our work provides evidence of the involvement of specific miRNAs as potential biomarkers in breast cancer triple negative development.

Des mutations germinales dans les gènes majeurs du cancer du sein BRCA1 et BRCA2 sont responsables de la maladie chez les patientes cumulant histoire familiale et apparition du cancer à un jeune âge. Environ 15% des femmes testées pour les mutations de BRCA1 et BRCA2 sont porteuses d’une mutation clairement pathogénique dans un des deux gènes. Cependant, des variants de signification clinique incertaine (VUS pour "variants of uncertain clinical significance") sont détectés dans 5% à 15% des cas testés. Pour évaluer la signification clinique des VUS, le Breast cancer Infomation Core (BIC) a développé un modèle Bayésien intégré, basé sur des données d'observations. Align-GVGD, un algorithme d'évaluation des substitutions faux-sens basé sur l'histoire évolutionnaire de la protéine fournit la probabilité a priori du modèle. Cependant, lorsqu'une substitution silencieuse est détectée, elle sera jugée comme neutre par l'évaluation in silico. Pourtant, une mutation au niveau de l'ARNm peut perturber la mécanique de l'épissage, par deux moyens principaux: endommagement des sites sauvages d'épissage, ou la création de sites exoniques d'épissage de novo. Notre premier objectif est de rassembler les variants déjà publiés, de les re-analyser avec le modèle d'évaluation intégrée. Nous voulons extraire le plus de variants publiés premièrement sous le statut de VUS vers un statut plus informatif, avec des recommandations cliniques associées. Par la suite, nous voulons étendre le modèle pour évaluer plus de variants, plus précisément, en intégrant l'évaluation des perturbations de l'épissage. Finalement, nous serons capable de présenter et de fournir ces informations librement sur Internet, via une interface web populaire, une Leiden Open Variation Database (LOVD) / Germline mutations in major breast cancer susceptibility genes BRCA1 and BRCA2 are responsible for the disease for high-risk patients (patients with early onset and familial history of breast cancer). Around 15% of screened women for BRCA1 and BRCA2 mutations carry one clearly pathogenic mutation in one of those two genes. However, variants of uncertain clinical significance (VUS) are detected in 5% to 15% of tested patients. To assess clinical significance of VUS, the Breast cancer Information Core (BIC) has developed a Bayesian integrated model, based on observational data. Align-GVGD, an algorithm evaluating damage of missense substitutions based on the evolutionary history of the protein, is providing the prior probability of the model. However, whenever a silent substitution arise, it is firstly treated as neutral by the in silico assessment. Indeed, a mutation at the mRNA level can disrupt the splicing machinery by two main means: damaging wild-type splice sites, or creating exonic de novo splice sites. Our first goal is to be a central repository of already published variants, to re-analyze them using the unified integrated evaluation model. We would like to extract the most variants from the original published status of VUS to a more informative status, with associated clinical recommendations. Then we would like to extend the model to be able to evaluate more variants more precisely by adding the splicing damages assessment in the integrated evaluation. In the end, we will be able to provide these informations freely on Internet, via a widely use web interface, a Leiden Open Variation Database (LOVD)

Environ 5 à 10 % des cancers du sein et de l’ovaire sont liés à des prédispositions génétiques héréditaires. Les principaux gènes responsables de telles prédispositions sont BRCA1 et BRCA2. Depuis plusieurs années, l’analyse de ces gènes est proposée dans un cadre clinique. Aux Hôpitaux Universitaires de Genève (HUG) en Suisse, une consultation d’oncogénétique a été mise sur pied dès 1994 pour les personnes concernées par leurs antécédents personnels et familiaux de cancer. Jusqu’en 2009, le seul laboratoire suisse assurant l’analyse des gènes BRCA1/BRCA2 était établi aux HUG. Ce travail de thèse intègre, d’une part, des études en lien avec la démarche clinique de conseil génétique pour les formes familiales et héréditaires de cancer du sein et de l’ovaire et, d’autre part, une évaluation détaillée des données moléculaires résultant des analyses (n= 1'163) des gènes BRCA1/BRCA2 réalisées aux HUG entre 1996 et 2009. Des perspectives quant au développement de l’oncologie prédictive aux HUG et en Suisse, et à l’activité de conseillère en génétique particulièrement dans ce domaine, sont finalement présentées. / Genetic predispositions are responsible for 5 to 10 % of all breast and ovarian cancers. The main breast/ovarian cancer predisposing genes are BRCA1 and BRCA2. For some years, the screening of pathogenic mutations in BRCA1/BRCA2 genes is provided in a clinical setting. At the Hôpitaux Universitaires de Genève (HUG, Geneva, Switzerland), a consultation in predictive oncology has been set up since 1994 for individuals concerned by the evaluation of their familial cancer risk and the probability to carry a genetic predisposition to cancer. Until 2009, the single national laboratory for BRCA1/BRCA2 testing was established in the HUG. The objectives of this work were to evaluate different aspects of the consultation process for breast/ovarian cancer predisposition syndromes provided in our Unit and to review all BRCA1/BRCA2 complete screenings (n=1’163) performed between 1996 and 2009. Results of the present study will certainly influence future activity in predictive oncology, particularly regarding the role of the genetic counselor.

L’objectif de ce travail a consisté à évaluer le rôle d’un microenvironnement avec une haplo-insuffisance hétérozygote du gène BRCA1 dans les événements précoces de la transformation tumorale du cancer du sein. Dans ce but, nous avons modélisé un stroma BRCA1-muté en utilisant des cellules souches / stromales mésenchymateuses (MSCs) obtenues par différenciation de cellules souches pluripotentes induites (iPSCs) issues d’une patiente porteuse de la mutation (MSCs BRCA1+/-). Ces cellules mutées pour BRCA1 ont été comparées à des MSCs sans la mutation (MSCs BRCA1+/+) générées à partir d’iPSCs BRCA1+/+. Ce travail de thèse a porté sur l’influence du stroma BRCA1-muté à travers deux axes : le caractère pro-angiogénique des MSCs BRCA1+/- et l’induction d’une transition épithélio-mésenchymateuse (TEM) sur des cellules mammaires normales (HME1).Nous montrons que les MSCs BRCA1-muté présentent des propriétés pro-angiogéniques significativement augmentées en surexprimant le facteur hypoxique HIF-1α et des facteurs de la famille du VEGF, PDGF et Angpt se traduisant par des capacités augmentées à former des structures vasculaires in vitro et in vivo. Les MSCs BRCA1-muté présentent également des capacités migratoires supérieures en produisant et sécrétant la périostine (POSTN), une protéine de la matrice extracellulaire impliquée dans l’adhésion, la motilité et la migration cellulaires. Ces capacités ont été validées par une approche de siRNA spécifique pour la POSTN. In vivo, nous montrons que la co-injection de MSCs BRCA1-muté et de cellules malignes mammaires murines (4T1-Luc-GFP) a permis d’augmenter significativement la croissance tumorale et la formation de métastases pulmonaires. Ces résultats sont corrélés avec la détection de la POSTN in situ et avec la formation d’un réseau vasculaire tumoral développé, quantifié par marquage du CD34. Par ailleurs nous avons démontré qu’un surnageant de MSCs BRCA1+/- peut induire une TEM des cellules HME1 en favorisant l’acquisition d’un phénotype souche cancéreux (CD24Low/CD44High) et en accélérant leur migration. Enfin nous avons initié la production in vitro d’organoïdes mammaires en utilisant des MSCs et des HME1 afin d’étudier plus précisément les mécanismes moléculaires de cette TEM après contact et des possibles événements précoces de la transformation maligne. Nos résultats indiquent que les MSCs peuvent participer à l’initiation tumorale et à la progression métastatique dans un contexte d’une mutation hétérozygote du gène BRCA1. La POSTN pourrait représenter à la fois un marqueur pronostique mais également une cible thérapeutique pour ces cancers du sein héréditaires. / The aim of this study was to evaluate the role of a BRCA1 heterozygous haplo-deficient microenvironment in the early events of tumour transformation of breast cancer. For this purpose we modeled a BRCA1-mutated stroma using mesenchymal stem / stromal cells (MSCs) obtained by differentiation of induced pluripotent stem cells (iPSCs) from a patient carrying the mutation (MSCs BRCA1+/-). These BRCA1-mutated cells were compared to MSCs without the mutation (MSCs BRCA1+/+) generated from iPSCs BRCA1+/+. This study focuses on two aspects of BRCA1-mutated stroma, namely the pro-angiogenic properties of BRCA1+/- MSCs and the induction of an epithelial-mesenchymal transition (EMT) on normal breast cells (HME1).We have shown that BRCA1-mutated MSCs exhibit enhanced pro-angiogenic properties by overexpressing the hypoxic factor HIF-1α and factors from VEGF, PDGF and Angpt families resulting in increased capacities to form vascular structures in vitro and in vivo. BRCA1-mutated MSCs exhibit also higher migratory capabilities by production and secretion of periostin (POSTN), an extracellular matrix protein, which is involved in cell adhesion, motility and migration. These capacities have been validated by a specific siRNA approach for POSTN. In vivo, the coinjection of BRCA1-mutated MSCs with murine breast cancer cell line (4T1-Luc-GFP) promotes tumour growth and the formation of lung metastases. These results are correlated with in situ POSTN detection and with the formation of a developed tumour vascular network, quantified by CD34 staining. We also demonstrated that supernatant of BRCA1+/- MSCs can induce an EMT on HME1 cells by promoting the acquisition of stemness properties (CD24Low/CD44High) and accelerating their migration. Finally we initiated the in vitro production of mammary organoids using MSCs and HME1 in order to study more precisely the molecular mechanisms of this EMT after contact and possible early events of the malignant transformation. These results indicate that MSCs can participate to tumour initiation and metastatic progression in heterozygous BRCA1-mutated background. POSTN could represent a prognostic marker and a therapeutic target for these hereditary breast cancers.