Introduction How tumors adapt to changing environmental conditions and treatment is a major unsolved problem frustrating effective therapy. Rhabdoid tumors are highly aggressive pediatric tumors with a low survival rate. They occur in multiple tissues, including brain and kidney, and likely originate as a consequence of aberrant differentiation during development. Understanding the origin of rhabdoid tumors and the relationship with normal development is crucial for investigating new treatment options. The almost certain appearance of therapy resistance, low mutation burden, and recent epigenetic profiling suggest the existence of epigenetic heterogeneity within rhabdoid tumors. We hypothesize that epigenetic heterogeneity within rhabdoid tumors underlies their aggressive behavior. Goal I previously exploited organoid models and CRISPR technology to study colorectal cancer progression. My lab now developed a protocol to, for the first time, efficiently grow rhabdoid tumor organoids from patient tissue. We are in the unique position to identify the cellular origin of rhabdoid tumors and the key molecular mechanisms driving disease progression and therapy resistance. Approach We will I) apply single-cell epigenomic and transcriptomic analyses on tumor tissue for in-depth characterization of the cellular identity and heterogeneity within rhabdoid tumors II) combine our unique rhabdoid tumor organoids with genetic lineage tracing technology to reveal clonal dynamics in rhabdoid tumor progression and therapy resistance III) perform retrospective lineage tracing using somatic mutations to track down the cell-of-origin of rhabdoid tumors. Innovation Our integrative use of state-of-the-art technologies on unique patient-derived tissue and tumor organoids will provide comprehensive insights into the origin, heterogeneity and progression of rhabdoid tumors. This will also establish novel approaches for other cancer research as well as new concepts for improving therapy.

Assessing mutagenicity, the capacity of a substance to cause permanent and potentially harmful changes to an organism's DNA, is a pivotal concern in drug development. This assessment is crucial because it helps to predict the carcinogenic potential of new pharmaceuticals, ensuring that they are safe for human use. The importance of evaluating mutagenicity in drug development cannot be overstated. Drugs are designed to interact with biological systems to treat or prevent diseases, but they must do so without causing harmful genetic alterations. The process of developing a new drug involves a complex, time-consuming, and costly sequence of research and testing to ensure efficacy and safety. If mutagenicity is not accurately assessed, the consequences can be severe. Addressing the complexities and limitations associated with assessing mutagenicity in drug development, an innovative approach utilizing artificial intelligence (AI) and whole genome sequencing (WGS) data presents a promising solution. MUTAPREDICT introduces an innovative AI-driven platform designed to revolutionize the field of drug development by accurately predicting the mutagenic potential of new chemical compounds. This platform is set to address the ethical, economic, and temporal challenges associated with conventional animal testing and in vitro analyses, promoting a more humane, cost-effective, and efficient approach to drug safety assessments. Leveraging advanced WGS data, our solution significantly accelerates the drug development process, ensuring both safety and efficacy. Our value proposition encompasses ethical and sustainable drug development by offering a viable alternative to animal testing, thus meeting the growing demands for ethical research practices.

How hematopoietic stem and progenitor cells (HSPCs) regenerate blood is a major unsolved question, frustrating their effective use for therapy. Every year, >40.000 patients receive an HSPC transplantation (HSCT), as a last-resort therapy for various diseases, including leukemia. However, ~40% of HSCT recipients die, due to poor outgrowth of the donor HSPCs, inflammatory complications or relapse. There is an unmet need for strategies to predict and prevent these adverse outcomes. In mice, single-cell methods have revolutionized our understanding of how hematopoiesis is organized, allowing us to control the outcome of murine HSCT in detail. In contrast, our understanding of human hematopoietic regeneration, and our ability to control this process, is lagging behind. As a clinician in pediatric HSCT and stem cell biologist, my mission is to change this. RESTART aims to comprehensively characterize the cellular and molecular mechanisms guiding hematopoietic regeneration in humans. I have pioneered single-cell methods to study human HSPC biology. Here, we will apply state-of-the-art multiomics to dissect the identities and functional states of thousands of HSPCs and their surrounding niche cells in human bone marrow (BM). Embedded in Europe’s largest pediatric cancer center, we will apply these methods to a unique, longitudinal collection of BM samples of pediatric HSCT recipients and their donors, collected before and up to a year after HSCT. Our objectives are: (1) Dissect the cellular and molecular composition of the HSPC population during successful hematopoietic regeneration in human HSCT recipients; (2) Determine how HSCT-induced alterations in BM niche cells affect HSPC fate; (3) Leverage this information to identify and validate single-cell states or trajectories predictive of adverse HSCT outcome (graft failure, relapse). This study will contribute to improved survival of human HSCT recipients and to increased fundamental knowledge on human tissue regeneration.

Neuroblastoma are pediatric tumors that respond poorly to chemotherapy and have a very poor prognosis. To improve treatment options, a global development towards precision medicine is ongoing. This strongly focusses on molecular genetic target identification in tumors and subsequent assigning patients to the best trials according to their molecular profile. Still it is difficult to predict which patients will benefit from these targeted compounds. In addition, if tumors do respond to single compound targeted therapy, they almost always relapse. These tumor evolution processes could be prevented by simultaneous intervention in different activated tumor pathways. We now want to study how we can select patients that will most likely respond to a targeted compound and what combinations of targeted compounds are most effective? This can’t be tested in a clinical setting since the number of neuroblastoma patients that can be included in Phase1/2 trials is very small. Recent research shows that tumor organoids mimic human tumors and can effectively be used as xenograft in nude mice as well. These in vitro and in vivo models could be used as an alternative selection system for optimal combination treatment in a personalized approach. The overall aim is now to test if combinations of targeted compounds can cause complete remission in neuroblastoma organoid model systems to select combination treatment options for personalized clinical trials For this purpose we will generate neuroblastoma organoids that properly represents the complexity and heterogeneity of individual tumors and build a repository that represents the various subtypes of neuroblastoma tumors. We will identify synergistic compound combinations that are effective in neuroblastoma tumors that are characterized by specific molecular genetic aberrations. Thereby we will build an efficient pipeline to generate personalized models that can be used in precision medicine programs to perform compound validation.

Extrachromosomal circular DNA (ecDNA) present in cancer cells stochastically amplifies oncogenes and drug resistance genes independent from the core genome and thereby contributes to tumor adaptability and heterogeneity. While ecDNA prevalence, diversity, and biological traits have been studied, we lack insights into their expression regulation. It is unclear how it differs from chromosomal genes and how it is linked to the non-random and dynamic nuclear localization patterns observed for many ecDNA. Moreover, we are unaware of the proteins that participate in interactions with nuclear compartments to control ecDNA transcription. Most studies using sequencing and FISH methods fall short in describing these dynamic processes. However, understanding their implications is vital for designing therapies that target ecDNA expression as a whole instead of targeting individual oncogenes. To advance our knowledge, we need tools to map ecDNA transcription at high temporal and spatial resolution and systematic methods to screen for associated proteins. The aim of this project is to identify crucial overarching mechanisms and players of ecDNA transcription through an interdisciplinary approach. The Medema group has devised protocols to induce extrachromosomal amplification of drug resistance genes in diverse cellular backgrounds and genetically tag them for imaging and perturbation. The present study will build on this work to for the first time reveal the spatiotemporal co-regulation of ecDNA transcription and localization using innovative imaging reporters for measurements in single live cells and use unbiased proteomic profiling to map factors that interact with ecDNA and regulate their expression. Besides addressing outstanding questions in the field, our study will create unique cell models and datasets as a basis for future studies. Our results will furthermore inform the design of more effective treatments for the many cancer patients affected in Europe and worldwide.