Investigating ATRT cell of origin and tumor heterogeneity via hiPSC-derived models

Atypical teratoid rhabdoid tumors (ATRTs) are rare, highly malignant pediatric brain cancers that almost always result from biallelic inactivation of SMARCB1 , a core subunit of the SWI/SNF chromatin remodeling complex. Despite presenting a remarkably simple genome defined by SMARCB1 loss, ATRTs are molecularly diverse, consisting of three subgroups with distinct DNA methylation profiles, transcriptomes, and clinical outcomes, suggesting differences in cells of origin and unique mechanisms of oncogenesis. Neural progenitor cells (NPCs) and neural crest cells (NCCs) have been proposed as potential cells of origin, with NPCs aligning more with the ATRT-SHH subgroup. On the other hand, NCCs may account for the molecularly identical extracranial malignant rhabdoid tumors as well as the remaining intracranial subgroups, ATRT-TYR and ATRT-MYC, which have suspected extra-CNS origins. To study SMARCB1 loss in a genetically defined neural progenitor cellular context, our lab previously engineered human induced pluripotent stem cells (hiPSCs) with doxycycline (DOX)-inducible SMARCB1 knockdown. NPCs derived from these hiPSCs, that were differentiated without SMARCB1 expression, exhibited an ATRT-SHH subgroup transcriptome and formed orthotopic tumors. Building upon these findings, these engineered hiPSCs were differentiated into neural crest cells. In comparison with isogenic controls, neural crest cells differentiated without SMARCB1 expression acquired a proliferative phenotype, enhanced clonogenic potential, and became arrested in a “progenitor” state by maintaining expression of neural crest differentiation pathway genes. Furthermore, SMARCB1 knockdown during neural crest differentiation upregulated target genes of REST, a transcription factor that is enriched in ATRT-MYC. In contrast, REST target genes were found to be downregulated in SMARCB1-depleted NPCs, highlighting the impact of cell identity upon phenotypes and the utility of hiPSC-derived models for investigating mutations within different clinically relevant cellular contexts. Future work aims to orthotopically engraft SMARCB1-depleted NCCs and characterize chromatin accessibility of SMARCB1 loss within NPCs versus NCCs via ATACseq. Furthermore, since recent findings suggest that the p53 pathway is often suppressed in rhabdoid tumors via MDM2 overexpression, MDM2 and a dominant-negative p53 mutant will be introduced individually into this model to determine whether p53 pathway inactivation cooperates with SMARCB1 loss within these cellular contexts.