Le cancer ovarien est la 1ère cause de décès par cancer gynécologique. A ce jour, le traitement le plus efficace consiste en une exérèse la plus complète possible de la tumeur associée à une chimiothérapie à base de sel de platine et de taxanes. Cependant, la survie globale des patientes à 5 ans reste faible (environ 40%), due à un fort taux de récidive et au développement d'une résistance aux traitements. Les interactions entre les cellules tumorales ovariennes (CTO) et le microenvironnement semblent être impliquées dans la progression de la maladie et l'acquisition de cette chimiorésistance. Parmi les cellules du microenvironnement tumoral, nous avons pu isoler des cellules stromales mésenchymateuses (MSC) à partir de biopsies de carcinome ovarien. Ces MSC associées au cancer (CA-MSC) ont la capacité d'induire une résistance des CTO au traitement par carboplatine. Afin de mieux comprendre les mécanismes conduisant des MSC à sécréter des facteurs pro-tumoraux, dans un contexte de cancer ovarien, nous avons mis en place un modèle à partir de MSC issues de donneurs ne présentant pas de cancer. Nous avons observé que la présence d'un environnement tumoral ovarien modifie le phénotype de MSC de moelle osseuse (BM-MSC) physiologiques, en entrainant notamment la sécrétion par ces CA-MSC " induites " de facteurs chimioprotecteurs pour les CTO. Parmi les facteurs sécrétés, nous avons identifié des chimiokines telles que CXCL1, CXCL2 et IL-8 comme des cibles thérapeutiques intéressantes pour lutter contre la résistance à la chimiothérapie. En effet, les CA-MSC et les CA-MSC " induites " sécrètent plus de CXCL1, CXCL2 et d'IL-8 que les BM-MSC. En outre, l'inhibition des récepteurs de ces chimiokines (CXCR1 et CXCR2) sensibilise les CTO au carboplatine, même en présence des sécrétions des CA-MSC et des CA-MSC " induites ". Ces expériences in vitro ont été confirmées in vivo dans un modèle expérimental murin. En effet, la co-injection à des souris immunodéprimées de MSC avec des CTO humaines entraine une protection des CTO vis-à-vis du carboplatine comparée à l'injection de CTO seules. Le co-traitement avec un inhibiteur de CXCR1 et de CXCR2 a permis de sensibiliser les CTO au carboplatine et d'empêcher la chimiorésistance induite par les MSC. En effectuant une étude rétrospective évaluant la concentration de ces chimiokines au moment du diagnostic, nous avons confirmé que les patientes qui seront a posteriori " résistantes " au carboplatine ont une concentration sérique de ces chimiokines qui est plus élevées que les patientes du groupe " sensibles " au carboplatine. Outre leur rôle direct dans l'acquisition de la chimiorésistance, les chimiokines comme CXCL1, CXCL2 et IL-8 peuvent être impliquées dans la régulation du système immunitaire. Nous avons montré que les CA-MSC étaient capables de modifier le phénotype des macrophages vers un phénotype M2 décrit comme pro-tumoral. En effet, ces macrophages ainsi polarisés ont un pouvoir cytotoxique dirigé contre les CTO inférieur aux macrophages non stimulés. L'ajout d'un inhibiteur de CXCR1 et de CXCR2 restaure le pouvoir cytotoxique initial des macrophages, même en présence des sécrétions des CA-MSC. Ainsi nos travaux suggèrent que les CA-MSC pourraient provenir de MSC physiologiques qui, au contact d'un environnement tumoral ovarien, vont acquérir un phénotype capable d'induire la sécrétion de facteurs chimioprotecteurs pour les CTO et de polariser les macrophages vers un phénotype moins cytotoxique pour les CTO. Ces deux phénomènes pro-tumoraux peuvent être inhibés par l'utilisation d'un inhibiteur des récepteurs de CXCR1 et de CXCR2. Ainsi ces récepteurs des chimiokines, semblent être des cibles thérapeutiques intéressantes afin de sensibiliser les CTO au carboplatine et traiter plus efficacement la tumeur. Ceci pourrait permettre d'éviter les récidives des cancers ovariens qui sont, à l'heure actuelle, observées chez plus de 70% des patientes. / Ovarian cancer is the leading cause of gynecological cancer death. To date, the most effective treatment consists of the complete excision of the tumor associated with chemotherapy based on platinum salts and taxanes. However, the 5-year overall survival remains low (close to 40%) due to a high rate of recurrence and development of resistance to treatments. Disease progression and the acquisition of this chemoresistance seem to be due to interactions between ovarian tumor cells (OTC) and the microenvironment. Amidst the cells of the tumor microenvironment, we were able to isolate mesenchymal stromal cells (MSC) from tumor biopsies of patients with ovarian adenocarcinoma. These cancer-associated MSC (CA-MSC) have the ability to induce resistance to carboplatin in OTC. In order to understand the mechanisms leading to the secretion of pro-tumoral factors by the CA-MSC in the context of ovarian cancer, we have developed a model based on the in vitro MSC culture of from healthy donors in tumor conditioning media. We have observed that an ovarian tumor environment modifies the physiological phenotype of bone marrow MSC (BM-MSC), leading in particular to the secretion by these "induced" CA-MSC of chemoprotective factors for OTC. Among these secreted factors, we have identified chemokines such as CXCL1, CXCL2 and IL-8 as therapeutic targets in order to control drug resistance. In fact, CA-MSC and "induced" CA-MSC secrete more CXCL1, CXCL2 and IL-8 than BM-MSC and the use of an inhibitor of their receptors (CXCR1 and CXCR2) sensitized OTC to carboplatin even in the presence of CA-MSC and " induced " CA-MSC secretions. These in vitro experiments have been confirmed in an experimental mouse model in vivo. Indeed, the co-injection of MSC with OTC yielded a greater protection of OTC to carboplatin compared with the OTC injection alone. Co-treatment with a CXCR1 and CXCR2 inhibitor resulted in sensitization of OTC to carboplatin and prevention of MSC-induced chemoresistance. We conducted a retrospective study evaluating the concentration of these chemokines at the time of diagnosis. We thus showed that patients who are a posteriori "resistant" to carboplatin have a higher concentration of chemokines than patients belong to the "sensitive" group to carboplatin. In addition to their direct role concerning the acquisition of chemoresistance, chemokines such as CXCL1, CXCL2 and IL-8 may be involved in the immune system regulation. In this context, we showed that CA-MSC were able to modify the phenotype of macrophages into a M2 phenotype described in literature to have a pro-tumoral activity. Indeed, these polarized macrophages present a lower cytotoxic capacity against OTC than unstimulated macrophages. CXCR1 and CXCR2 inhibitor restores the initial cytotoxic activity of macrophages even in the presence of CA-MSC secretions. Thus, our work suggests that CA-MSC could originate from physiological MSC which, in contact with an ovarian tumor environment, acquire a phenotype capable of inducing the secretion of chemoprotective factors for CTO and of polarizing macrophages into a less cytotoxic phenotype for OTC. These two pro-tumoral mechanisms can be inhibited by the use of CXCR1 and CXCR2 receptor inhibitors emphasizing the role of these chemokines in the development of a chemoresistance and showing how important is to go further is this study. Finally, these chemokines receptors seem to be therapeutic targets in order to sensitize OTC to carboplatin and to potentialize actual treatments. This could prevent the recurrence of ovarian cancers that are presently observed in more than 70% of patients.

T-cells are present in the immune system to fight against invaders. Once their job is done, suppressing their activity is an important step in maintaining a proper immune response. Myeloid derived suppressor cells (MDSCs) are immune cells that suppress T-cell activity. Currently, MDSCs are defined as a heterogeneous population of immature cells that are derived in the bone marrow and travel to the site of inflammation or cancer. Two major subtypes of MDSCs have been identified in mice and humans, monocyte-like MDSCs (M-MDSC) and granulocyte MDSCs (G-MDSC). G-MDSCs typically make up the majority of the total population of MDSCs but are less T-cell suppressive than M-MDSCs. One of the major problems in the study of MDSCs is that the current marker system for subtypes does not differentiate between precursor MDSCs (lacking suppressive ability) and functional MDSCs (those with suppressive ability). Therefore, using cancer models in mice, we investigated the development and potential to classify precursor MDSCs from functional MDSCs. While MDSCs have been highlighted as a target cell to inhibit in cancer, in other conditions, such as pregnancy, MDSCs have been shown to be beneficial in maintaining a normal pregnancy. Therefore, targeting the increase of MDSCs in abnormal pregnancy conditions like pre-eclampsia may act as a prevention or therapeutic strategy. Finally, it is known that many dietary components can act as modulators of immune cells. Specifically, the polyphenol like phytochemical, curcumin has been shown to act as an anti-inflammatory agent with the potential to modulate multiple immune cells. Therefore, we propose two different studies to investigate the potential of curcumin as either an inhibitor and/or promotor of MDSCs in a disease-specific context. Together the role of phytochemicals as immunomodulators of MDSCs is still very young, in part due to the complexity of phytochemicals themselves, but the studies cited here provide evidence that the field is ripe for additional questions to be asked.

Tumor microenvironmental conditions play a vital role in promoting metastasis and tumor recurrence. Due to inefficient vasculature, cancer cells experience hypoxia, glucose deprivation and low pH even during the early stages of tumor growth. Tumor cells are proposed to adapt to these microenvironmental conditions by acquiring increased migratory and invasion potential and tumor initiating ability. Our research addresses the effect of these biochemical factors of the tumor microenvironment (TME) on motility, epithelial to mesenchymal transition (EMT) and stemness of non-small cell lung cancer (NSCLC). NCI-H292 and NCI-H1650 NSCLC cell lines were used to measure the effect of the above mentioned TME conditions. Apart from acidic pH, low glucose and hypoxia, the effect of high glucose conditions was also measured on H292 and H1650 cell lines. Acidic pH, high and low glucose conditions were observed to have no effect on the motility, EMT and stemness of H1650 cell line. Hence, use of this cell line was discontinued and no further treatment conditions were tested on this cell line. In H292 cell line, acidic pH, low glucose and tumor like conditions combined together (acidic pH + low glucose + hypoxia) [AP+LG+HYP] significantly decreased motility whereas hypoxia significantly increased the motility of H292 cells. High glucose did not affect the motility of H292 cells. Although N-cadherin, a mesenchymal marker, expression was significantly upregulated by acidic pH, high and low glucose conditions, no direct correlation was observed between N-cadherin expression and motility. E-cadherin expression was not affected by acidic pH, high and low glucose conditions. An increase in N-cadherin expression and no change in E-cadherin expression under these conditions might be an indication of partial EMT. Hypoxia and AP+LG+HYP did not alter the expression of E-cadherin and N-cadherin. Although expression of vimentin, another mesenchymal marker, and Sox2, a cancer stem cell marker (CSC), was observed at the mRNA level, no expression of vimentin and Sox2 proteins was observed in H292 cells under any of these treatment conditions. The expression of OCT4, another CSC marker, was also not observed at the protein level in H292 cells. HIF-1α expression was observed in H292 cells under normoxic conditions and was unaffected by hypoxia and AP+LG+HYP. Therefore our research indicates that the effect of these TME conditions might be different on different cancer cell lines or cancer types. Not all cancers may depend on EMT for metastasis. An increase in metastasis under hypoxia may be independent of HIF-1α.