Biomedical Graduate Studies

Description of Research Expertise

The De Raedt Lab focuses on elucidating the mechanisms by which pediatric High Grade Gliomas develop and progress. We aim to understand which pathways are crucial in these processes, how they interact with each other, and how we can exploit these insights to develop novel, paradigm-shifting therapies. To gain insight into what drives these pediatric High Grade Gliomas, we molecularly analyze human tumor samples, perform in depth cellular studies and develop accurate in vitro and in vivo models. Additionally, whenever possible we perform high quality pre-clinical studies in our animal models that, if successful, can be quickly translated to the clinic. Importantly, our work has inspired multiple clinical trials that are currently ongoing.

Pediatric High Grade Glioma (pHGG) is a devastating disease with a median survival of ~12 months. Human hemispheric pHGG are often driven by an aberrantly activated RAS-pathway, for example by loss of NF1. Therefore, a major interest of ours lies in the RAS pathway, one of the major oncogenic signaling pathways in cancer, plays a crucial role in the development of these gliomas. Intriguingly, in pediatric High Grade Glioma, mutations in the epigenetic machinery often co-occur with RAS pathway mutations. Our goal is not only to study and understand the RAS pathway itself, but also to use advanced in vivo models and tools to functionally identify and understand how epigenetic mutations cooperate with RAS pathway activation. Additionally, we have a keen interest in developing novel therapeutic approaches, including cellular and immunotherapies.

A. ACCURATE IN VIVO AND IN VITRO MODELING OF GLIOMA
(1) Validation of tumeroid (organoid) cultures of NF1 low grade glioma: We observed a metabolic shift form FA oxidation to glycolysis when we grow low grade glioma tumeroids in vitro. These tumeroid cultures loose their immune microenvironment by week 4. We are currently exploring other media compositions to avoid loss of immune cells and the metabolic switch.
(2) Validation of NF1 HGG allograft models: In vitro cultures inherently loose the complexity observed in in vivo tumors. We show that the full complexity of these tumors is restored when we re-inject mouse glioma stem cell lines orthotopically in the brains of wildtype and Nf1 heterozygous mice. Moreover, we identified the in vivo upregulation of an antigen presentation phenotype in glioma tumor cells upon scRNAseq analysis. It is currently unclear if/how this antigen presentation phenotype contributes to tumor formation, but we can speculate on disruption of antigen recognition by immune cells or aberrant activation of signaling pathways.
(3) Functional validation and identification of new (epigenetic) drivers: developing mouse models that are genetically faithful to human central nervous system tumors. In the current era of next generation sequencing, identifying potential new drivers for cancer is no longer a bottleneck. A major challenge, however, remains the rapid functional validation of the vast number of genes that were found mutated or lost. With regards to functional validation, in vivo modeling remains a state of the art discovery tool. However, the development of classical genetically engineered mouse models is cumbersome and time consuming, especially when several genetic drivers need to be combined. We stereotacticly inject in vitro modified neuronal stem cells and glioma stem cells to screen cooperating candidate tumor suppressor and oncogenes that drive central nervous system tumors in vivo. Hereby we are generating more genetically accurate mouse models for central nervous system tumors; moreover the well-controlled nature and in vivo setting of these experiments enables me to investigate and understand the interaction between different mutational events in human central nervous system tumors. Currently we are focusing on the epigenetic regulator ATRX and how they cooperate with NF1 loss (RAS pathway activation).

B. Development of therapeutic strategies for glioblastoma
(1) Explore new therapies for central nervous system tumors:
We performed in vitro drug screens on spheroid HGG cell lines and identified potential combination therapies to be evaluated in our mouse models.

(2) Development of interneurons as a cellular delivery vehicle of therapeutic proteins to glioblastoma: Cellular therapies, like CAR-T therapy, have revolutionized the treatment of some cancers. Unfortunately, CAR-T has shown limited efficacy for brain tumors, in part due to the immune suppressive tumor microenvironment (TIME). To circumvent these brain tumor specific obstacles, we developed a cellular immunotherapy delivery system, derived from post-mitotic Migratory Inhibitory Interneuron Precursors (MIPs), that induces a cytotoxic tumor response in glioblastoma independent of unique tumor antigens.
In all mammals studied, the inhibitory interneurons of the cerebral cortex originate entirely, or at least predominantly, in the ventral/subcortical portion of the telencephalic neural tube. The embryonic and early postnatal migration into and through the cerebral cortex is guided by a variety of chemorepulsant and chemoattractant factors. Remarkably, MIPs maintain their migratory capacity when transplanted into the postnatal cerebral cortex, striatum, cerebellum, and spinal cord. This capacity has led to many preclinical and a first human trial of MCIP transplantation as cell based therapy for intractable epilepsy.
Intriguingly, glioblastoma frequently secrete the same chemoattractant factors, such as CXCL12, used by MIPs to guide their migration during development. Indeed, our in vitro and in vivo data show that MIPs robustly migrate to the majority of glioblastoma evaluated. This inspired us to modify MCIPs to serve as a delivery vector for cytotoxic agents that can eliminate glioblastoma. For example, EGFR-BiTE (Bispecific T-cell Engager) secreting MCIPs eliminate EGFR expressing glioblastoma both in vitro and in vivo. Bispecific Engagers are in essence two antibodies linked together, where one antibody binds to a tumor antigen and the other binds to an antigen on, for example, a T-cell or macrophage. Engagement with a bispecific antibody activates the immune cell.

(3) PC-CAR therapy for NF1 associated tumors: Cellular therapies have been promising for other cancers but have so far not provided relief for NF1 associated tumors. The paucity of good targets on the cell surface of HGG and MPNST remains a major bottleneck for developing an effective CAR-T therapy. To address these challenges, we are developing Peptide-Centric Chimeric Antigen Receptors (PC-CARs) to NF1 associated tumors. Unlike traditional CARs that target surface proteins, PC-CARs focus on intracellular tumor antigens/peptides presented on the cell surface by Human Leucocyte Antigen receptors, offering a unique and potentially transformative approach to glioma immunotherapy.
In an immunopeptidomics campaign, we identified over 108k unique peptides presented by HLA-I receptors of NF1 associated tumors. Because we identified a large number of potential peptides to target, we have the luxury to be highly selective and only retain those peptides derived from essential or lineage-defining genes, reducing the likelihood of antigen escape. Prioritization of candidate genes/peptides, which focuses on the peptide binding affinity, parent gene properties (oncofetal, lineage defining and/or essential gene), and uniform expression within and across HGG and MPNST, is ongoing. So far 2 peptides passed our selection criteria and we are currently performing scFv panning experiments to design binders. Subsequently, we will develop and de-risk PC-CARs through rigorous in vitro and in vivo testing to eliminate cross-reactivity with normal cells. Extensive in vitro and in vivo experiments will evaluate tumor clearance, persistence of CAR-T cells, and safety.