Genome-wide CRISPR-Cas9 knockout screens decipher genetic dependencies and chemogenetic interactors in sonic hedgehog subgroup of medulloblastoma

Abstract

Medulloblastomas (MB) are the most common malignant embryonal tumors of the Central Nervous System (CNS), primarily diagnosed during childhood. This highly heterogeneous tumor entity encompasses four molecular consensus subgroups: wingless (WNT), sonic hedgehog (SHH), Group 3 and Group 4, which exhibit differential transcriptomic, epigenetic and proteomic signatures. SHH subgroup of MB (SHH-MB), which belongs to the scope of this thesis, accounts for approximately 30% of all MBs and displays a bimodal age distribution, as it more frequently affects infants and adults. As the name suggests, SHH-MB is characterized by constitutive activation of SHH signaling cascade, induced by genetic alterations in critical components of the SHH pathway. Inhibition of Smoothened (SMO), an upstream member of the SHH pathway, is a targeted therapy approach applied in several SHH-driven cancers, including MB, displaying promising anti-tumor results. However, due to primary or secondary resistance mechanisms, its clinical efficacy is limited. Hence, there is an urgent need for novel therapeutic interventions that will be able to inhibit SHH-MB growth regardless of the genetic mutations within the SHH pathway. In light of limited targeted therapies for SHH-MB, the primary aim of this thesis is to identify genetic vulnerabilities in SHH-MB potentially beyond the SHH pathway that could potentially serve as novel therapeutic targets. To this end, CRISPR-Cas9 genome-wide knockout screens were carried out in this study, in order to unravel genetic vulnerabilities in SHH-MB. Two distinct models murine SMB21 and human DAOY cells were screened and their suitability as SHH-representative model systems was addressed. Our functional genomics data provide evidence that SMB21 cells recapitulate SHH-associated hallmark features, while DAOY cells do not, emphasizing the potential of SMB21 cells to serve as a faithful in vitro model for SHH-MB. Among the top scored hits, members of the epigenetic machinery including DNA methyltransferase 1 Dnmt1 and chromatin remodeler Smarca5 showed robust depletion in SMB21-screened cells, indicating that they represent genetic dependencies for SHH-MB. Genetic and pharmacological inhibition experiments showed that DNMT1 inhibition suppresses SHH-MB proliferation in vitro, by blocking SHH signaling, as further supported by DNA methylation and RNA sequencing analyses. We also demonstrate that Dnmt1 plays a crucial role in proper formation of the murine cerebellum, while its genetic ablation prolongs survival outcome of tumor bearing SHH MB mice. Of note, by conducting an additional CRISPR-Cas9 knockout drug, we unraveled Smo knockout as a synergistic partner for DNMT1 inhibitor, 5-Azacytidine. We show that simultaneous combination treatment of 5-Azacytidine and Sonidegib effectively inhibited tumor growth in murine and human SHH-MB cell models. These findings were further evaluated in vivo using an established preclinical model of SHH-MB and we demonstrated that combination therapy of both inhibitors extends survival of SHH-MB mice by reducing tumor cell proliferative activity. Finally, we assessed the role of SNF2-family member of ATP-dependent chromatin remodeler Smarca5 in SHH-MB development. By genetically targeting Smarca5 in SMB cells, we could prove that its loss reduces SMB survival in vitro, by inhibiting SHH pathway activation. Furthermore, we present that loss of Smarca5 in granule cell neuron precursors (GCNPs) affects their viability and migratory activity, resulting in developmental abnormalities in regard to murine cerebellar growth. In parallel, our data indicate that Smarca5 is required for tumor progression in mouse models of SHH-MB. Taken together, we provide strong evidence that epigenetic regulators Dnmt1 and Smarca5 are essential for normal cerebellar, as well as GCNP-derived SHH-MB development. We unraveled DNMT1 as a novel therapeutic target for SHH-driven MB that acts downstream of the SHH signaling cascade, thus representing an efficacious treatment for this tumor entity irrespective of the genetic alterations within the SHH pathway. Finally, we propose that inhibiting DNMT1 alone or in combination with SMO inhibition could serve as a novel treatment modality for SHH-MB tumors.