UCSF

Nearly half of patients with neuroblastoma present with high-risk disease, which often has poor outcome. The proto-oncogene MYCN is amplified in half of all high-risk neuroblastoma. Patients with high-risk disease inevitably relapse, and show frequent mutational activation of MAPK signaling. High-risk disease remains a clinical enigma, and the lack of actionable targets has prevented sufficient advances in improving survival. To better understand the genetic requirements of high-risk neuroblastoma, we generated a human induced pluripotent stem cell (iPSC) model of the disease. Human iPSC models are superior to genetically engineered mouse models or cell line models at representing the chromosomal landscape and telomere biology of human neuroblastoma.

Human iPSCs were differentiated toward trunk neural crest, the cell of origin for neuroblastoma. We optimized protocols to generate this progenitor cell. We transduced candidate and established drivers of neuroblastoma including MYCN and hyperactive mutant ALK (F1174L) and orthotopically implanted the resulting trunk neural crest cells (NCCs) into the renal capsules of immunocompromised mice. The combination of ALK (F1174L) and MYCN significantly accelerated tumor formation. Tumors expressed markers typically found in neuroblastoma on histological and RT-PCR analyses, while lacking markers of other tumors. We found gene expression changes specifically enriched in MYCN/ALK vs MYCN alone-driven tumors, including increased integrin and PI3 kinase signaling. We also introduced CRISPRi/a gene expression modulation into the iPSC system to detangle TERT and MYCN biology that could regulate neuroblastoma tumorigenesis. We further found that MYCN P44L mutation significantly decreases tumor latency compared to WT MYCN misexpression, while CMYC misexpression results in tumors that progress with similar rapidity to MYCN. ChIP-seq and RNA-seq studies in MYCN/CMYC/P44L mutants clarify the differential transcriptional landscapes in various neuroblastoma subgroups.We also delved into alternative splicing of NF1 as potentially modulating a MYCN-driven axis in high-risk and recurrent neuroblastoma tumors. We have identified TIAL1 and TIA1 as alternative splicing proteins which are regulated by MYCN and which affect NF1 exon 23a inclusion to modulate overall NF1 activity and downstream RAS/MAPK signaling. Together, our models of high-risk neuroblastoma clarify the genetic requirements and identify multiple potential pathways which could be therapeutically targeted. This thesis work not only generated a workable model that will continue to be a platform for study in the years to come, but also opened the door to studying other neural tumors like medulloblastoma and low grade glioma.