Wilms tumor (WT) is the most common kidney cancer in childhood. It is a genetically heterogeneous tumor and several genetic alterations have been identified in WT patients. Recurrent mutations were found in the homeo-domain of SIX1 and SIX2 in high proliferative tumors (18.1% of the blastemal-type tumors) as well as in the microprocessor genes DROSHA and DGCR8 (18.2% of the blastemal-type tumors), indicating a critical role of the SIX-SALL pathway and aberrant miRNA processing in WT formation. Underlined by the fact that a significant overlap between mutations in DROSHA and SIX1 was found, indicating a synergistic effect. To characterize the in vivo role of DROSHA and SIX mutations during kidney development and their oncogenic potential, I analyzed mouse lines with either a targeted deletion of Drosha or an inducible expression of human DROSHA or SIX1 carrying a tumor-specific E1147K or Q177R mutation, respectively. The DROSHA mutation E1147K was predicted to act in a dominant negative manner. Six2-cre mediated deletion of Drosha in nephron progenitors led to a lethal phenotype with apoptotic loss of progenitor cells and early termination of nephrogenesis. Mosaic deletions via Wt1-creERT2 resulted in a milder phenotype with viable offspring that developed proteinuria after 2-4 weeks, but no evidence of tumor formation. Activation of the DROSHA-E1147K transgene via Six2-cre, on the other hand, induced a more severe phenotype with apoptosis of progenitor cells, proteinuria and glomerular sclerosis. The severely growth-retarded mice died within the first two months. This strong phenotype was consistent with the predicted dominant-negative effect of DROSHA-E1147K. Analysis of the SIX1-Q177R mutation suggested that the mutation leads to a shift in DNA binding specificity instead of a complete loss of DNA binding. This may end up in subtle changes of the gene regulatory capacity of SIX1. Six2-cre mediated activation of SIX1-Q177R lead to a viable phenotype with no alterations or shortened life span. Yet a global activation of SIX1-Q177R mediated by Zp3-cre resulted in bilateral hydronephrosis and juvenile death of the mice. To mimic the synergistic effect of DROSHA and SIX1 mutations, I generated compound mutants in two combinations: A homozygous deletion of Drosha combined with an activation of SIX1-Q177R and a compound mutant with activation of DROSHA-E1147K and SIX1-Q177R. Each mouse model variant displayed new phenotypical alterations. Mice with Six2-cre mediated homozygous deletion of Drosha and activation of SIX1-Q177R were not viable, yet heterozygous deletion of Drosha and activation of SIX1-Q177R led to hydronephrosis, proteinuria and an early death around stage P28. Combined activation of DROSHA-E1147K and SIX1-Q177R under Six2-cre resulted in proteinuria, glomerulosclerosis and lesions inside the kidney. These mice also suffered from juvenile death. Both mouse models could confirm the predicted synergistic effect. While these results underscore the importance of a viable self-renewing progenitor pool for kidney development, there was no evidence of tumor formation. This suggests that either additional alterations in mitogenic or antiapoptotic pathways are needed for malignant transformation, or premature loss of a susceptible target cell population and early lethality prevent WT formation.

Identifying novel driver genes in cancer remains a crucial step towards development of new therapeutic approaches and the basic understanding of the disease. This work describes the impact of the AP1 transcription activator component FOSL1 on melanoma maintenance. FOSL1 is strongly upregulated during the progression of melanoma and the protein abundance is highest in metastases. I found that the regulation of FOSL1 is strongly dependent on ERK1/2- and PI3K- signaling, two pathways frequently activated in melanoma. Moreover, the involvement of p53 in FOSL1 regulation in melanoma was investigated. Elevated levels of the tumor suppressor led to decreased FOSL1 protein levels in a miR34a/miR34c- dependent manner. The benefit of elevated FOSL1 amounts in human melanoma cell lines was analyzed by overexpression of FOSL1 in cell lines with low endogenous FOSL1 levels. Enhanced levels of FOSL1 had several pro-tumorigenic effects in human melanoma cell lines. Besides increased proliferation and migration rates, FOSL1 overexpression induced the colony forming ability of the cells. Additionally, FOSL1 was necessary for anchorage independent growth in 3D cell cultures. Microarray analyses revealed novel downstream effectors of FOSL1. On the one hand, FOSL1 was able to induce the transcription of different neuron-related genes, such as NEFL, NRP1 and TUBB3. On the other hand, FOSL1 influenced the transcription of DCT, a melanocyte specific gene, in dependence of the differentiation of the melanoma cell line, indicating dedifferentiation. Furthermore, FOSL1 induced the transcription of HMGA1, a chromatin remodeling protein with reprogramming ability, which is characteristic for stem cells. Consequently, the influence of HMGA1 on melanoma maintenance was investigated. In addition to decreased proliferation and reduced anoikis resistance, HMGA1 knockdown reduced melanoma cell survival. Interestingly, the FOSL1 induced pro-tumorigenic effects were demonstrated to be dependent on the HMGA1 level. HMGA1 manipulation reversed FOSL1 induced proliferation and colony forming ability, as well as the anchorage independent growth effect. In conclusion, I could show that additional FOSL1 confers a clear growth benefit to melanoma cells. This benefit is attributed to the induction of stem cell determinants, but can be blocked by the inhibition of the ERK1/2 or PI3K signaling pathways.

The transcription factor NRF2 is considered as the master regulator of cytoprotective and ROS-detoxifying gene expression. Due to their vulnerability to accumulating reactive oxygen species, melanomas are dependent on an efficient oxidative stress response, but to what extent melanomas rely on NRF2 is only scarcely investigated so far. In tumor entities harboring activating mutations of NRF2, such as lung adenocarcinoma, NRF2 activation is closely connected to therapy resistance. In melanoma, activating mutations are rare and triggers and effectors of NRF2 are less well characterized. This work revealed that NRF2 is activated by oncogenic signaling, cytokines and pro-oxidant triggers, released cell-autonomously or by the tumor microenvironment. Moreover, silencing of NRF2 significantly reduced melanoma cell proliferation and repressed well-known NRF2 target genes, indicating basal transcriptional activity of NRF2 in melanoma. Transcriptomic analysis showed a large set of deregulated gene sets, besides the well-known antioxidant effectors. NRF2 suppressed the activity of MITF, a marker for the melanocyte lineage, and induced expression of epidermal growth factor receptor (EGFR), thereby stabilizing the dedifferentiated melanoma phenotype and limiting pigmentation markers and melanoma-associated antigens. In general, the dedifferentiated melanoma phenotype is associated with a reduced tumor immunogenicity. Furthermore, stress-inducible cyclooxygenase 2 (COX2) expression, a crucial immune-modulating gene, was regulated by NRF2 in an ATF4-dependent manner. Only in presence of both transcription factors was COX2 robustly induced by H2O2 or TNFα. COX2 catalyzes the first step of the prostaglandin E2 (PGE2) synthesis, which was described to be associated with tumor immune evasion and reduction of the innate immune response. In accordance with these potentially immune-suppressive features, immunocompetent mice injected with NRF2 knockout melanoma cells had a strikingly longer tumor-free survival compared to NRF2-proficient cells. In line with the in vitro data, NRF2-deficient tumors showed suppression of COX2 and induction of MITF. Furthermore, transcriptomic analyses of available tumors revealed a strong induction of genes belonging to the innate immune response, such as RSAD2 and IFIH1. The expression of these genes strongly correlated with immune evasion parameters in human melanoma datasets and NRF2 activation or PGE2 supplementation limited the innate immune response in vitro. In summary, the stress dependent NRF2 activation stabilizes the dedifferentiated melanoma phenotype and facilitates the synthesis of PGE2. As a result, NRF2 reduces gene expression of the innate immune response and promotes the generation of an immune-cold tumor microenvironment. Therefore, NRF2 not only elevated the ROS resilience, but also strongly contributed to tumor growth, maintenance, and immune control in cutaneous melanoma.

Bei Melanomen handelt es sich um die gefährlichste Form von Hautkrebs mit der höchsten Mortalitätsrate. Deshalb sind Untersuchungen dieser Hautkrebsart von immenser Bedeutung. Es ist bekannt, dass der AP-1-Transkriptionsfaktorkomplex eine große Rolle für Melanomentstehung und -progression spielt. In der vorliegenden Arbeit wurde die Funktion der AP-1 Komponente FOSL1 in Melanomen untersucht. Hierbei konnte zunächst ermittelt werden, dass die FOSL1 Expression im humanen Melanom durch den MAPK-Signalweg vermittelt wird und von den Onkogenen BRAF und NRAS abhängig ist. Dies wird auch durch die Tatsache unterstützt, dass die Stabilität von FOSL1 durch MAPK reguliert wird. Des Weiteren konnte gezeigt werden, dass FOSL1 in vielen Melanomzellen die Proliferation verstärkt und auch an Migration beteiligt ist. Da diese Prozesse zur Krebsprogression beitragen, deutet dies darauf hin, dass FOSL1 bei der Melanomentwicklung eine wichtige Funktion besitzt. Weiterhin konnten SLUG, SNAI3, IL6 und MMP14 als FOSL1-Zielgene identifiziert werden, deren Regulierbarkeit durch FOSL1, jedoch abhängig von der jeweiligen Zelllinie war. Somit konnte mit dieser Arbeit gezeigt werden, dass FOSL1 nicht nur, wie zuvor für Brustkrebszellen beschrieben, an Migration beteiligt ist, sondern auch zur Proliferation humaner Melanome beiträgt. Zukünftige Arbeiten werden zeigen, ob die identifizierten Gene für die FOSL1-vermittelte Migration und Proliferation verantwortlich sind.

Molecular genetic analyses, such as mutation analyses, are becoming increasingly important in the tumor field, especially in the context of therapy stratification. The identification of the underlying tumor entity is crucial, but can sometimes be difficult, for example in the case of metastases or the so-called Cancer of Unknown Primary (CUP) syndrome. In recent years, methylome and transcriptome utilizing machine learning (ML) approaches have been developed to enable fast and reliable tumor and tumor subtype identification. However, so far only methylome analysis have become widely used in routine diagnostics. The present work addresses the utility of publicly available RNA-sequencing data to determine the underlying tumor entity, possible subgroups, and potential therapy options. Identification of these by ML - in particular random forest (RF) models - was the first task. The results with test accuracies of up to 99% provided new, previously unknown insights into the trained models and the corresponding entity prediction. Reducing the input data to the top 100 mRNA transcripts resulted in a minimal loss of prediction quality and could potentially enable application in clinical or real-world settings. By introducing the ratios of these top 100 genes to each other as a new database for RF models, a novel method was developed enabling the use of trained RF models on data from other sources. Further analysis of the transcriptomic differences of metastatic samples by visual clustering showed that there were no differences specific for the site of metastasis. Similarly, no distinct clusters were detectable when investigating primary tumors and metastases of cutaneous skin melanoma (SKCM). Subsequently, more than half of the validation datasets had a prediction accuracy of at least 80%, with many datasets even achieving a prediction accuracy of – or close to – 100%. To investigate the applicability of the used methods for subgroup identification, the TCGA-KIPAN dataset, consisting of the three major kidney cancer subgroups, was used. The results revealed a new, previously unknown subgroup consisting of all histopathological groups with clinically relevant characteristics, such as significantly different survival. Based on significant differences in gene expression, potential therapeutic options of the identified subgroup could be proposed. Concludingly, in exploring the potential applicability of RNA-sequencing data as a basis for therapy prediction, it was shown that this type of data is suitable to predict entities as well as subgroups with high accuracy. Clinical relevance was also demonstrated for a novel subgroup in renal cell carcinoma. The reduction of the number of genes required for entity prediction to 100 genes, enables panel sequencing and thus demonstrates potential applicability in a real-life setting.