Angiogênese é o processo de formação de novos vasos sanguíneos a partir dos vasos existentes. É um processo vital, mas muitas doenças também dependem deste mecanismo para obter nutrientes e progredir. Estas \"doenças dependentes de angiogênese\" incluem cânceres, retinopatias e degeneração macular. Alguns inibidores da angiogênese foram desenvolvidos na última década, com o objetivo de auxiliar no manejo dessas doenças e melhorar a qualidade de vida dos pacientes. A maioria destes compostos funciona inibindo a ligação de VEGFA/VEGFR2, que também é um elemento importante para a sobrevivência de células endoteliais quiescentes; e isso pode explicar parcialmente eventos adversos observados em alguns ensaios clínicos. Nossa hipótese é que a melhoria das terapias anti-angiogênicas depende de uma compreensão melhor e mais ampla desse processo, especialmente quando relacionada à progressão das doenças. Utilizando RNA-Seq e um modelo animal bem aceito de angiogênese, o modelo murino de Retinopatia Induzida por Oxigênio, exploramos o transcritoma e identificamos 153 genes diferencialmente expressos durante a angiogênese. Uma extensiva validação de vários genes realizada por qRT-PCR e hibridização in-situ confirmou a superexpressão de Esm1 em células endoteliais de tecidos com angiogênese ativa. A análise de enriquecimento desta lista de genes confirmou a ligação da angiogênese com genes frequentemente mutados em tumores, consistente com a conhecida ligação entre câncer e angiogênese, e forneceu sugestões de fármacos já aprovados que podem ser reutilizados para controlar a angiogênese em circunstâncias patológicas. Finalmente, com base neste panorama amplo da angiogênese, fomos capazes de criar um biomarcador molecular com poder prognóstico para a predição da recidiva de câncer de mama, com aplicações clínicas promissoras. Em resumo, este trabalho revelou com sucesso genes relacionados à angiogênese e forneceu novas alternativas terapêuticas, incluindo potenciais fármacos para reposicionamento. Esse conjunto de genes diferencialmente expressos é também um recurso valioso para investigações futuras. / Angiogenesis is the process of formation of new blood vessels based on existing vessels. It is a vital process but many diseases also rely on this mechanism to get nourishment and progress. These so called angiogenesis-dependent diseases include cancers, retinopathies and macular degeneration. Some angiogenesis inhibitors were developed in the past decade, aiming to help the management of such diseases and improve patients quality of life. Most of these compounds work by inhibiting VEGFA/VEGFR2 binding, which is also a key element to the survival of quiescent endothelial cells; this may partly explain unanticipated adverse events observed in some clinical trials. We hypothesize that the improvement of anti-angiogenesis therapies hinges on a better and broader understanding of the process, especially when related to diseases\' progression. Using RNA-seq and a well accepted animal model of angiogenesis, the murine model of Oxygen Induced Retinopathy, we have explored the transcriptome landscape and identified 153 genes differentially expressed in angiogenesis. An extensive validation of several genes carried out by qRT-PCR and in-situ hybridization confirmed Esm1 overexpression in endothelial cells of tissues with active angiogenesis, providing confidence on the results obtained. Enrichment analysis of this gene list endorsed a narrow link of angiogenesis and frequently mutated genes in tumours, consistent with the known connection between cancer and angiogenesis, and provided suggestions of already approved drugs that may be repurposed to control angiogenesis under pathological circumstances. Finally, based on this comprehensive landscape of angiogenesis, we were able to create a prognostic molecular biomarker for prediction of breast cancer relapse, with promising clinical applications. In summary, this work successfully unveiled angiogenesis-related genes, providing novel therapeutic alternatives, including potential drugs for repositioning. The set of differentially expressed genes is also a valuable resource for further investigations.

Quatorze femmes canadiennes meurent chaque jour du cancer du sein. Le cancer du sein triple-négatif (CSTN) détient un mauvais pronostic De nombreux efforts sont fournis afin d'offrir à ces patientes des traitements ciblés, comme les inhibiteurs de poly (adenosine diphosphate-ribose) polymerase inhibitors (PARPi) afin d’améliorer leur survie et de minimiser la toxicité liée à la chimiothérapie. Le sous-groupe de CSTN qui pourrait bénéficier des PARPi reste à être identifié. De plus, différentes stratégies d'administration des PARPi et de la chimiothérapie pourraient améliorer leur efficacité thérapeutique tout en diminuant la toxicité. Nous avons précédemment dérivé une signature génétique de 63 gènes prédisant la réponse aux PARPi avec une précision globale élevée. Nos objectifs sont 1) d'évaluer les implications cliniques de la signature génétique; et 2) de déterminer la séquence optimale d'administration du talazoparib et du carboplatin in vivo en cancer du sein triple-négatif BRCAWT. D'abord, nous avons évalué la fréquence mutationnelle des 63 gènes dans différents contextes cliniques. Deux bases de données publiques furent utilisées. Puis, nous avons comparé trois cohortes de xénogreffes orthotopiques: A) talazoparib en premier, combiné au carboplatin le jour 3; carboplatin en premier suivi du talazoparib B) un jour après; et C) sept jours après. La fréquence mutationnelle des 63 gènes était élevée chez les tumeurs luminales B et celles de mauvais pronostic. Les patientes luminales B mutées avaient une moindre survie que les patientes non mutées. Aussi, l'inhibition tumorale et métastatique était similaire pour les cohortes A et B, cependant la cohorte B avait moins de toxicité. Les PARPi pourraient avoir un rôle chez les tumeurs luminales B et celles de mauvais pronostic. Deuxièmement, le prétraitement avec le carboplatin semble améliorer la sensibilité au talazoparib et diminuer la toxicité. / Fourteen Canadian women die every day from breast cancer. Triple-negative breast cancer (TNBC) has a poor prognosis. Numerous efforts are made to offer these patients targeted therapies such as poly (adenosine diphosphate-ribose) polymerase inhibitors (PARPi) to improve survival and minimize chemotherapy-related toxicity. It is not well understood which subset of TNBC patients will benefit from PARPi; and if different sequencing strategies of PARPi and chemotherapy can improve therapeutic efficacy and decrease toxicity. We previously derived a 63-gene signature predicting response to PARPi with a high overall accuracy. Our objectives are 1) to evaluate the clinical implications of the 63-gene signature; and 2) to determine the optimal sequence of administration of talazoparib and carboplatin in vivo in BRCAWT TNBC. First, we evaluated the mutational frequency of the 63 genes in different clinical settings using two publically-available datatsets. Second, we compared three cohorts of orthotopic xenografts: A) talazoparib first, combined with carboplatin on day 3; carboplatin first, followed by talazoparib B) one day later; and C) seven days later. We found that the mutational frequency was high in breast cancer subtypes of poor prognosis. Mutated luminal B patients had a lower survival than non-mutated patients. We also found that tumoral and metastatic inhibition were similar between cohorts A and B, but cohort B had less toxicity. In conclusion, there is potential for PARPi efficacy in luminal B and poor prognosis tumors. Second, pretreatment with carboplatin may be an effective approach with less toxicity.