Human cerebellar organoids as an in vitro 3D model of Group 3 Medulloblastoma

Medulloblastoma (MB) is a heterogeneous tumor that represents the most common malignant brain tumor of childhood. It stands as a cause for a high percentage of morbidity and mortality among cancer patients. Thanks to genome-wide analyses, MB can be divided into four significant subgroups, different from each other for diagnosis, prognosis, and metastatic recurrence. WNT subtype has the best prognosis; SHH subtype has an intermediate prognosis; Group 3 subtype is characterized by a high percentage of metastases and worst prognosis; Group 4 MB is the most common subtype, but the less understood. Willing to increase the knowledge about the aggressiveness of the Group3 subtype, this work will focus on developing a reliable Human Group 3 MB model based on cerebellar organoids derived from human induced pluripotent stem cells (iPSC). Three-dimensional (3D) cell culture systems have gained increasing interest in modeling, drug discovery, and tissue engineering due to their evident advantages in providing more reliable information and more predictive data before in vivo tests. The field of cell development, differentiation, and cell organization was the first to make use of cerebellar organoids, but these 3D structures are starting to be a novelty in the cancer field. One of the innovative points of this work is the setup of a new way to modify wild type human cerebellar organoids, electroporating them with strong Group 3 MB inducers, derived from in vivo patient-specific NGS data screen. We validate that Gfi1/c-Myc and Otx2/c-Myc oncogenes give rise to MB-like organoids, which (in nude mice) can develop tumors harboring a DNA methylation signature that clusters specifically with human patient Group 3 tumors. Moreover, we identify Smarca4 as an oncosuppressor gene and discover that treatment with an EZH2 specific inhibitor, called Tazemetostat, reduces Otx2/c-Myc tumorigenicity in human organoids. We speculate that our Medulloblastoma 3D culture system holds great promises for applications in infant tumor research, cancer cell biology, and drug discovery, being a novel human 3D reliable tool for developing personalized therapies.