Cancers often resurrect embryonic molecular programs to promote disease progression. In melanomas, which are tumors of the neural crest (NC) lineage, a molecular signature of the embryonic NC is often reactivated. These NC factors have been implicated in promoting pro-tumorigenic features like proliferation, migration and therapy resistance. However, the molecular mechanisms that establish and maintain NC identities in melanomas are largely unknown. Additionally, whether the presence of a NC identity has any clinical relevance for patient melanomas is also unclear. Here, using comparative genomic approaches, I have a) identified a novel role for GDF6-activated BMP signaling in reawakening a NC identity in melanomas, and b) identified a NC signature as a clinical predictor of melanoma progression. Like the genomes of many solid cancers, melanoma genomes have widespread copy number variations (CNV) harboring thousands of genes. To identify disease-promoting drivers amongst such huge numbers of genes, I used a comparative oncogenomics approach with zebrafish and human melanomas. This approach led to the identification of a recurrently amplified oncogene, GDF6, that acts via BMP signaling to invoke NC identities in melanomas. In maintaining this identity, GDF6 represses the melanocyte differentiation gene MITF and the proapoptotic factor SOX9, allowing melanoma cells to remain undifferentiated and survive. Functional analysis in zebrafish embryos indicated a role of GDF6 in blocking melanocyte differentiation, suggesting that the developmental function of GDF6 is reiterated in melanomas. In clinical assessments, a major fraction of patient melanomas expressed high GDF6, and its expression correlated with poor patient survival. These studies provide novel insights into regulation of NC identities in melanomas and offer GDF6 and components of BMP pathway as targets for therapeutic intervention. In additional studies, I wanted to test whether a broader NC identity in melanomas had any clinical relevance. In these studies, I performed transcriptome analysis of zebrafish melanomas and derived a 15-gene NC signature. This NC gene signature positively correlated with the expression of SOX10, a known NC marker in human melanomas. Patients whose melanomas expressed this signature showed poor overall survival. These findings identify an important predictive signature in human melanomas and also illuminate the clinical importance of NC identity in this disease.

Melanoma is one of the most aggressive and deadly forms of skin cancer. Arising from melanocytes, a pigment cell population derived from the neural crest, melanomas often adopt characteristics associated with the neural crest – the ability to rapidly proliferate, migrate and invade throughout the body. Historically, these characteristics along with a baseline resistance to chemotherapy have made melanoma extremely difficult to treat. Improvements in targeted and immunotherapeutic options have improved patient outcomes, but many patients still experience limited durable responses to therapy. In order to improve patient outcomes, new potential avenues of therapy must be identified based on the underlying pathogenesis of the disease. We previously identified and characterized the function of a novel melanoma oncogene, GDF6, uncovering a role in promoting melanoma cell survival and dedifferentiation by activating a neural crest identity. Here, we have a) identified a role for GDF6-activated BMP signaling during melanocyte development that forms a basis for its oncogenic role in melanoma, b) determined BMP signaling may play a role in promoting a neural crest-like state during melanoma initiation, and c) assayed novel monoclonal antibodies targeting GDF6 for use as blocking antibodies to treat advanced melanoma. Previous work identified GDF6 as a melanoma oncogene that promotes melanoma progression through suppression of apoptosis and differentiation in melanoma cells, by regulating neural crest factor expression and neural crest identity, suggesting a potential role for GDF6 in the embryonic neural crest. Additional studies had previously identified roles for GDF6 and its orthologous genes in specific biological contexts, including embryonic neuronal cell survival, bone and cartilage development, embryonic eye development, and bone and ligament repair in adult tissue. Furthermore, a study had indicated a role for a GDF6 ortholog, gdf6a, during zebrafish neural crest induction, but had not uncovered any specific role for gdf6a in further development of the neural crest or in any neural crest derivatives. We determined blocking gdf6a-activated BMP signaling acts to increase melanocyte development during embryogenesis by increasing the proportion of neural crest cells activating the pigment cell marker, mitfa. Furthermore, we showed the increase in melanocytes is at the expense of the iridophore population. These results indicate GDF6 function in melanoma is a reiteration of the normal physiological function of GDF6 during embryonic melanocyte development from the neural crest. Given these results and our previous findings of the role of GDF6-activated BMP signaling established melanomas, we hypothesized a potential role for GDF6-activated BMP signaling during melanoma initiation. Previous studies have determined neural crest identity and neural crest-like characteristics to be crucial during multiple phases of melanoma, including initiation, progression, and metastasis. We evaluated melanoma initiating lesions to determine the potential impact of BMP signaling on development and progression of these lesions. We found early lesions in our model to have active BMP signaling and that modulation of BMP signaling could alter the rate of development of these lesions in our animals. Furthermore, BMP modulation ultimately impacted the development of these lesions into melanomas. Together, these results indicate BMP signaling is a potential driving pathway during melanoma initiation and progression. Finally, we wanted to determine the therapeutic potential of targeting GDF6 in order to treat patients with advanced melanoma. Given our previous findings and mechanism of ligand-activated BMP signaling, we hypothesized a monoclonal antibody targeting GDF6 could block GDF6 activity at its receptor on melanoma cells, thus inhibiting GDF6-activated BMP signaling. Monoclonal antibodies have been widely used as therapy in cancer as well as many other rheumatologic and immunologic conditions. We established a panel of GDF6-targeting antibodies via a hybridoma approach. We then assessed the antibodies ability to identify mammalian GDF6 in vitro and performed functional assays to determine if anti-GDF6 antibody treatment yielded the expected results of inhibiting GDF6-activated BMP signaling. We observed decreased pathway activity, decreased cell viability, and increased cell death in melanoma cells treated with anti-GDF6 antibodies in vitro. We further investigated whether these antibodies could exert anti-melanoma effects in vivo. Together, these results indicate potential therapeutic value for our antibodies in treating GDF6-positive melanomas.

This diploma thesis dealt with the comparison of different methods of amination of polycaprolcatone in terms of their effectiveness for tissue engineering. A polycaprolactone membrane was prepared by an electrospinning method, which was subsequently modified by three different amination methods. Selected types of amination were plasma polymerization with cyclopropylamine monomer, hybrid modification using plasma and N-allylmethylamine monomer, and chemical amination using aminolysis with diaminohexane. Surface amines were subsequently characterized by electron scanning microscopy (SEM), X-ray photoelectron spectroscopy (XPS), attenuated total reflection infrared spectroscopy (ATR-FTIR) and contact angle measurement. A cell culture designated A375 (Human malignant melanoma cell lines A375® CRL-1619®) was cultured on the thus modified membranes, which was analyzed by optical microscopy, and a proliferation assay was performed by determining the relative amount of ATP. Based on the experimental results, we can confirm the success for all types of amination. In terms of efficiency for tissue engineering, the amination method by plasma polymerization with the monomer cyclopropylamine has the most satisfactory results.

Dans le glioblastome (GBM), nous avons recherché et comparé les facteurs de radiorésistance parmi 8 lignées cellulaires et 11 xénogreffes (5 issues de patients, et 6 issues de lignées cellulaires traitées par radiothérapie hypofractionnée 6x5Gy) en utilisant la technologie RPPA (Reverse Phase Protein Analysis). Nous avons exploré 89 marqueurs protéiques appartenant à 10 voies différents: réparation de l’ADN, cycle cellulaire, apoptose, adhésion/cytosquelette, stress et voies PI3K, tyrosine kinase, MAPK/ERK, SAPK/JNK et NFκB. Aucun marqueur de radiorésistance n’était commun entre les expériences in vitro (FAK, HSP90, HSF1, HSPA2, vimentine et integrin β4) et in vivo (EGFR, CHK1 et VCP). Nous avons ensuite, dans ces mêmes modèles de GBM étudié la potentielle radiosensibilisation par une classe innovante d’inhibiteurs de la réparation de l’ADN, Dbait. Les molécules Dbait sont de courts fragments d'ADN double brin mimant une cassure de l'ADN. Ces molécules agissent comme des leurres vis-à-vis des enzymes de signalisation des dommages à l'ADN notamment PARP et DNA-PK qui sont hyperactivés, ce qui empêche la détection des dommages réels induits par les traitements et inhibe la réparation. Les molécules Dbait se sont montrées capables de radiosensibilier 6/11 modèles de xénogreffes de GBM. Les marqueurs prédictifs de la résistance à Dbait étaient Phospho-H2AX/H2AX, Phospho-NBS1/NBS1 et cleaved-PARP/PARP. Nous avons également mené une étude préclinique étudiant le potentiel radiosensibilisant de Dbait/DT01 (DT01 = forme clinique de Dbait) dans un modèle de mélanome. Les souris xénogreffées sur flanc étaient traitées par un protocole de RT « palliatif » (10x3Gy) ou « curatif » (20x3Gy) en combinaison à des injections de Dbait/DT01. Les souris traitées par la combinaison Dbait/DT01 et RT avaient une inhibition significative de la croissance tumorale et une survie prolongée. Du fait de l’ensemble des résultats précliniques positifs sans toxicité ajoutée, les molécules Dbait/DT01 ont été testées lors d’un essai clinique de phase I en association à la RT dans le cadre des métastases cutanées de mélanome avec des résultats très encourageants. / We aimed to identify predictive biomarkers of GBM radioresistance in 8 GBM cell lines, 6 corresponding cell lines derived xenografts (CDX) and 5 patient derived xenografts (PDX) treated with hypofractionated radiotherapy (6x5Gy), using an RPPA (Reverse Phase Protein Array) approach. We explored 89 potential protein markersinvolved in DNA repair, PI3K pathway, apoptosis, tyrosine kinase signaling, stress signaling, cell cycle, MAPK/ERK signaling, SAPK/JNK signaling, NFκB signaling and adhesion/cytoskeleton. In vitro identified biomarkers (FAK, HSP90, HSF1, HSPA2, vimentin and integrin β4) were not found in vivo. The markers specific of in vivo resistance were Phospho-EGFR/EGFR, Phospho-Chk1/Chk1 and VCP. Then, in these GBM models, we assessed the potential radiosensitization of an innovative DNA repair inhibitor, Dbait. Dbait consists of 32 bp deoxyribonucleotides that mimics DNA lesions. They act as a bait for DNA damage signaling enzymes, the polyadenyl-ribose polymerase (PARP), and the DNA-dependent kinase (DNA-PK), inducing a “false” DNA damage signal and ultimately inhibiting DNA repair. 6/11 GBM models had been radiosensitized by Dbait.Phospho-H2AX/H2AX, Phospho-NBS1/NBS1 and cleaved-PARP/PARP were predictive markers of Dbait resistance. We assessed the efficacy and safety of combining radiotherapy with Dbait/DT01 (DT01 = clinical form of Dbait) in a preclinical model of human melanoma. Nude mice subcutaneously engrafted with human melanomas (SK28) were or were not treated with Dbait/DT01, “palliative” (10x3Gy) or “radical” (20x3Gy) RT, or a combination of Dbait/DT01 and RT. Mice treated with Dbait/DT01 and RT combination had significantly better tumor growth control and longer survival compared to RT alone with the “palliative” protocol or the “radical” protocol. The results of this preclinical study led to the conduction of a phase 1 study in the palliative management of melanoma in-transit metastases (DRIIM trial) with very encouraging results.

PKD1 est une sérine-thréonine kinase codée par le gène PRKD1. Elle appartient à la famille des protéines kinase D (PKD). La famille des PKD comprend également deux autres membres, PKD2 et PKD3 qui partagent une grande similarité structurale avec PKD1. Certaines études suggèrent que la PKD1 est impliquée dans de nombreuses voies oncogéniques telles que MAPK, NF-κB, HDAC ... Les données préliminaires de notre équipe avait montré que PKD1 est fortement exprimée dans les lignées cellulaires de mélanome ayant un phénotype mésenchymateux. Nous avons également montré que PKD1 participe à l’indépendance vis-à-vis des oestrogènes dans le cancer du sein et que la forte expression des transcrits PRKD1 était associée à une moins bonne survie dans les tumeurs traitées au Tamoxifène. L'objectif de ma thèse est d'analyser le rôle de la PKD1 dans le maintien du phénotype mésenchymateux dans le mélanome et de déterminer si PKD1 pourrait être un facteur pronostique et une cible thérapeutique pertinente pour certaines sous-populations de cancer du sein. Nous avons analysé les marqueurs moléculaires impliqués dans le phénotype mésenchymateux et épithélial (cadhérine, β-caténine ...) ainsi que les caractéristiques fonctionnelles des cellules de mélanome ayant un phénotype mésenchymateux après inhibition pharmacologique de PKD1. Les résultats ont montré que l'inhibition de PKD1 induit une réversion du phénotype mésenchymateux vers un phénotype pseudo-épithélial moins agressif. L'expression de PRKD1, PRKD2 et PRKD3 a été analysée par RT-qPCR dans 527 tumeurs de sein. Nous avons montré que les niveaux plus élevés d'ARNm de PRKD1 étaient associés à un mauvais pronostic dans la population globale de cancer du sein et plus particulièrement dans les tumeurs ERα- et les tumeurs triples négatives. Nous avons criblé des inhibiteurs pharmacologiques de PKD1 en collaboration avec la société pharmaceutique AB Science, et nous avons évalué l'activité anti-tumorale de plusieurs d’entre eux in vitro dans la lignée cellulaire de cancer du sein triple négative MDA-MB-436 et in vivo dans un modèle de PDX (xénogreffe dérivée du patient) de cancer du sein triple négatif. L’inhibition pharmacologique de PKD1 ou à l’aide de siRNA a réduit la capacité des cellules MDA-MB-436 à former des colonies. L'inhibition pharmacologique de PKD1 a également réduit la croissance tumorale in vivo dans le modèle de PDX. Finalement, nous avons analysé l'expression de PRKD1 dans différentes cohortes multi-cancers. Nos résultats ont montré qu'une forte expression de PRKD1 est retrouvée dans le mélanome et le glioblastome. En conclusion, les résultats ont montré que PKD1 pourrait être une cible thérapeutique pertinente pour le mélanome, le cancer du sein et le glioblastome. / PKD1 is a serine-threonine kinase encoded by the PRKD1 gene. It belongs to the protein kinase D (PKD) family. The PKD family also includes PKD2 and PKD3 which share a high structural similarity with PKD1. Some studies suggest that PKD1 is involved in many oncogenic pathways such as MAPK, NF-κB, HDAC... Recent data from our group showed that PKD1 is more expressed in melanoma cell lines with mesenchymal phenotype compared to melanoma cell lines with epithelial phenotype. We also showed that PKD1 participates in the development and/or maintenance of the estrogen-independent phenotype in breast cancer, and that high PKD1 mRNA levels were associated with a worse outcome in tamoxifen-treated tumors. The objective of my thesis is to analyze the role of PKD1 in the maintenance of mesenchymal phenotype in melanoma and to determine whether PKD1 could be a prognostic factor and/or a therapeutical target for the treatment of breast cancer subpopulations. We analyzed molecular markers involved in mesenchymal and epithelial phenotype (cadherin, β-catenin …) and the functional characteristics of mesenchymal melanoma cells after PKD1 pharmacological inhibition. Results suggest that PKD1 inhibition induced a mesenchymal to epithelial like transition in these melanoma mesenchymal cell lines. In addition, the expression of PRKD1, PRKD2 and PRKD3 was analyzed by RT-qPCR in 527 breast cancers. We showed that higher PRKD1 mRNA levels were associated with a lower metastasis-free survival in the overall breast cancer population. PRKD1 prognostic value was even stronger in ERα- and in triple negative tumors. Then, we tested PKD1 inhibitors in collaboration with the AB Science pharmaceutical company. We assessed the anti-tumoral activity of PKD1 inhibitors in vitro in TNBC (triple negative breast cancer) cell lines and in vivo in a TNBC PDX (patient derivated xenograft) model. PKD1 pharmacological inhibition or depletion by siRNA reduced colony forming units in MDA-MB-436 TNBC cells. PKD1 pharmacological inhibition also reduced tumor growth in vivo in TNBC PDX model. Following those results, we analyzed PRKD1 expression in different multi-cancer cohorts. Our results showed that higher PRKD1 expression is found in melanoma and glioblastoma. In conclusion, results showed that PKD1 could participate in the maintenance of mesenchymal phenotype in melanoma. And that PKD1 could be an interesting therapeutic target in melanoma, breast cancer and glioblastoma.

Up to date, only four cases of primary intradural extramedullary spinal cord melanoma (PIEM) have been reported. No previous reports have described a case of PIEM located in the lower thoracic spine with long-termfollow-up. Purpose. Demonstrating an unusual, extremely rare case of melanoma manifestation. Study Design. Case report. Methods. We report a case of a 57-year-old female suffering from increasing lower extremity pain, left-sided paresis, and paraesthesia due to spinal cord compression caused by PIEM in the lower thoracic spine. Results. Extensive investigation excluded other possible primary melanoma sites and metastases. For spinal cord decompression, the tumor at level T12 was resected, yet incompletely. Adjuvant radiotherapy was administered two weeks after surgery. The patient was recurrence-free at 104 weeks after radiotherapy but presents with unchanged neurological symptoms. Conclusion. Primary intradural extramedullary melanoma (PIEM) is extremely rare and its clinical course is unpredictable.

Im Rahmen der Dissertation wurde ein Verfahren zur Quantifizierung monosomer Zellpopulationen innerhalb eines disomen Normalgewebes auf Basis der Pyrosequenzierung von Einzelbasenmutationen etabliert und hinsichtlich seiner Genauigkeit untersucht. Dabei liegt ein besonderer Schwerpunkt auf der Entwicklung eines Verfahrens zur Festlegung von Grenzwerten für die Detektion monosomer Population sowie für genetisch heterogene Subpopulationen. Zur Bestimmung der Genauigkeit wurden Mischreihen von DNA zweier Genotypen angefertigt und das Allelverhältnis durch Pyrosequenzierung gemessen. Diese Ergebnisse wurden genutzt, um Grenzwerte für die Detektion von LOH3-positiven Zellen im UMM estzulegen. In diesen Vorversuchen konnte die Anwendbarkeit der Analysemethode für Proben aus UMM sowohl aus Enukleations wie auch aus Feinnadelaspirationspräparaten demonstriert werden. Es wurde dann in einem weiteren Schritt analysiert, wie viele differente Loci für eine korrekte Diskriminierung zweier Genotypen analysiert werden müssen. Hier wurde gezeigt, dass zum einen die Anzahl der untersuchten SNP aber auch das gemessene Allelverhältnis maßgeblichen Einfluss auf die Genauigkeit der Analyse haben. Basierend auf diesen Daten wurde ein Verfahren entwickelt, das aus der gewünschten Genauigkeit eine Berechnung des Umfangs eines zu etablierenden SNP-Panels ermöglichte.

Hyaluronan (HA) and its principal receptor CD44 are known to be involved in regulating tumor cell dissemination and metastasis. It is hypothesized that the CD44-HA interaction regulates proliferation and invasion of tumor cells in dependence on the molecular weight and the presentation form of HA. To address this hypothesis, we reconstituted 3D collagen (Coll I) matrices and functionalized them with HA of molecular weight of 30-50 kDa (low molecular weight; LMW-HA) and 500-750 kDa (high molecular weight; HMW-HA). A post-modification strategy was applied to covalently immobilize HA to reconstituted fibrillar Coll I matrices, resulting in a non-altered Coll I network microstructure and stable immobilization over days. Functionalized Coll I matrices were characterized regarding topological and mechanical characteristics as well as HA amount using confocal laser scanning microscopy, colloidal probe force spectroscopy and quantitative Alcian blue assay, respectively. To elucidate tumor cell behavior, BRO melanoma cell lines with and without CD44 receptor expression were used for in vitro cell experiments. We demonstrated that only soluble LMW-HA promoted cell proliferation in a CD44 dependent manner, while HMW-HA and immobilized LMW-HA did not. Furthermore, an enhanced cell invasion was found only for immobilized LMW-HA. Both findings correlated with a very strong and specific adhesive interaction of LMW-HA and CD44+ cells quantified in single cell adhesion measurements using soft colloidal force spectroscopy. Overall, our results emphasize the importance of presentation mode and molecular weight specificity in biomaterial studies on the impact of HA on cell behavior.

Les microARN (miARN) sont de petits ARN non-codants dont la principale fonction est de réprimer l’expression génique en s’hybridant par complémentarité de séquence à leurs cibles ARN. L’activité des miARN est également régulée par leurs cibles qui entrent en compétition pour leur liaison. Certains de ces ARN compétiteurs endogènes (ARNce) résistent à la répression induite par le miARN et vont alors les séquestrer. Ils sont appelés éponges à miARN. La dérégulation des réseaux d’ARNce et des éponges à miARN est impliquée dans des processus pathologiques tels que le cancer. Au cours de ma thèse, nous nous sommes intéressés à la séquestration des miARN dans le mélanome cutané. Le mélanome provient de la transformation maligne du mélanocyte, une cellule spécialisée dans la production de pigment. S’il n’est pas pris en charge à temps, des métastases apparaissent et se disséminent rapidement dans l’organisme (ganglions, foie, poumons, cerveau, etc.). Des solutions thérapeutiques existent mais une faible proportion de patients y répondent de manière efficace nécessitant de nouvelles stratégies de traitement. Nous avons mis en évidence que l’ARN messager (ARNm) de TYRP1, gène spécifiquement exprimé dans le mélanocyte et donc le mélanome, porte le rôle d’éponge à miARN dans le mélanome métastatique. Ce rôle est indépendant de la fonction protéique de TYRP1. Nous avons déterminé que l’ARNm de TYRP1 séquestre le suppresseur de tumeurs miR-16 via des sites de liaison (MRE-16) non-canoniques. Les MRE-16 non-canoniques permettent à l’ARNm de TYRP1 de ne pas être dégradé par le miR-16 et le rendent donc plus stable dans la cellule de mélanome. La majorité du pool de miR-16 est ainsi séquestrée et ne peut donc plus réprimer ses cibles intervenant dans la prolifération cellulaire et la croissance tumorale in vivo. Afin de remettre en activité le miR-16 au sein de la cellule de mélanome, nous avons utilisé la technologie du « target site blocker » (TSB), un oligonucléotide antisens modifié ayant une forte stabilité et affinité pour sa cible. Le TSB, spécifique du MRE-16 de l’ARNm de TYRP1, entre en compétition pour la liaison à l’ARNm de TYRP1 avec le miR-16 pour permettre sa libération et son action sur ses cibles effectrices. Nous avons montré in vitro et in vivo via un modèle murin de xénogreffe de tumeur dérivée de patient que la stratégie du TSB est efficace contre le mélanome métastatique. Ces travaux ont permis l’identification d’un nouveau mécanisme oncogénique basé sur la séquestration de miARN et proposent une nouvelle stratégie de thérapie ciblée contre le mélanome métastatique. / MicroRNAs (miRNAs) are small non-coding RNAs. They fine tune gene-expression through specific complementary interaction with their RNA targets. The miRNA repressive function towards a given RNA is highly regulated and in part dependent on the abundance of its other targets competing for miRNA’s binding. Some of these competing endogenous RNAs (ceRNAs) can resist to miRNA-mediated RNA decay thereby sequestering miRNAs. They are named miRNA sponges. Deregulation of ceRNAs and miRNA sponges networks are implicated in many pathologic processes including cancer. My PhD work focused on miRNA sequestration in cutaneous melanoma. Melanomas arise from the malignant transformation of melanocytes; the skin-cell specialized in pigment production. Most melanoma undergoes metastatic evolution, with metastatic cells spreading rapidly in the entire organism (lymph node, liver, lungs, brain, etc.). Early and complete resection of primary in situ melanoma is thus determinant for patient outcome. Since 2010, potent therapeutic options have been developed. Unfortunately, patients ultimately develop resistance while some are non-responders. There is thus an urgent need to develop new therapeutic strategies to treat metastatic melanoma. We have identified that the Tyrosinase Related Protein 1 (TYRP1) mRNA function as a miRNA sponge. TYRP1 is specifically expressed in the melanocytic lineage. TYRP1 mRNA governs melanoma growth endorsing thereby a non-coding function. We demonstrated that TYRP1 mRNA sequesters the tumor suppressor miR-16 via non-canonical miRNA binding sites (MREs-16). Non-canonical miR-16 binding lacks mRNA decay function favoring TYRP1 mRNA stability and miRNA sequestration. Sequestered miR-16 can no more repress its canonical targets involved in cell proliferation and tumor growth. To reset miR-16’s activity and block melanoma growth, we used “Target Site Blocker” (TSB). TSBs are modified antisense oligonucleotides with enhanced stability and affinity to its target. We designed a TSB, named TSB-T3, overlapping specially TYRP1 non-canonical MRE-16. We first showed that TSB-T3 binds to TYRP1 mRNA and competes with miR-16. Freed miR-16 binds to its canonical targets inducing their decay. TSB-T3 blocks melanoma cell growth in vitro and in vivo, using patient-derived tumor xenograft. We thus showed for the first time that TSB’s strategy redirecting a tumor suppressor miRNA is a potent tool to monitor metastatic melanoma growth. Together my PhD work brings out a new oncogenic mechanism based on miRNA sequestration and proposes an original strategy of targeted therapy against metastatic melanoma.

Cutaneous melanoma is a very aggressive cancer with increasing incidence. It originates from transformed pigmented skin cells (melanocytes). The main risk factor for melanoma development is exposure to UV light and repeated sunburns. In approximately 10 % of cases, melanoma occurs on hereditary basis. Patients with cutaneous melanoma diagnosed in early stages have very good prognosis, with surgical resection of the primary tumour being mostly sufficient for treatment. In contrast, the advanced melanoma stages with metastases are often progressive and refractory to conventional therapies. Cutaneous melanoma is referred to as an immunogenic tumour that is frequently infiltrated by cells of the immune system. Tumours with immune cell infiltration show better prognosis. Spontaneous regression may occur. Over the last few years, progress has been made in the treatment of melanoma using checkpoints molecules (anti-CTLA-4 and anti-PD-1) to activate patients own immune system to recognize tumour lesions. In the tumour microenvironment, cytokines play an important role, enabling communication between cells and regulation of cell proliferation and migration and thus the tumour development. Cytokines (IL-2, IFNα) can be used in adjuvant therapy of melanoma. This work analysed levels of expressed cytokines in...