Neuroblastoma (NB) arises from oncogenic disruption of neural crest (NC) differentiation. Treatment with retinoic acid (RA) to induce differentiation has improved survival in some NB patients, but not all patients respond, and most NBs eventually develop resistance to RA. Loss of the chromatin modifier chromatin assembly factor 1 subunit p150 (CHAF1A) promotes NB cell differentiation; however, the mechanism by which CHAF1A drives NB oncogenesis has remained unexplored. This study shows that CHAF1A gain-of-function supports cell malignancy, blocks neuronal differentiation in three models (zebrafish NC, human NC, and human NB), and promotes NB oncogenesis. Mechanistically, CHAF1A upregulates polyamine metabolism, which blocks neuronal differentiation and promotes cell cycle progression. Targeting polyamine synthesis promotes NB differentiation and enhances the anti-tumor activity of RA. The authors' results provide insight into the mechanisms that drive NB oncogenesis and suggest a rapidly translatable therapeutic approach (DFMO plus RA) to enhance the clinical efficacy of differentiation therapy in NB patients.

During embryonic development, cells gradually restrict their developmental potential as they exit pluripotency and differentiate into various cell types. The POU transcription factor Oct4 (encoded by Pou5f1) lies at the center of the pluripotency machinery that regulates stemness and differentiation in stem cells, and is required for reprogramming of somatic cells into induced pluripotent stem cells (iPSCs). Several studies have revealed that Oct4 and other stemness genes are also expressed in multipotent cell populations such as neural crest cells (NCCs), and are required to expand the NCC developmental potential. Transcriptional regulation of Oct4 has been studied extensively in stem cells during early embryonic development and reprogramming, but not in NCCs. Here, we review how Oct4 is regulated in pluripotent stem cells, and address some of the gaps in knowledge about regulation of the pluripotency network in NCCs.

Developmental potential is progressively restricted after germ layer specification during gastrulation. However, cranial neural crest cells challenge this paradigm, as they develop from anterior ectoderm, yet give rise to both ectodermal derivatives of the peripheral nervous system and ectomesenchymal bone and cartilage. How cranial neural crest cells differentiate into multiple lineages is poorly understood. Here, we demonstrate that cranial neural crest cells possess a transient state of increased chromatin accessibility. We profile the spatiotemporal emergence of premigratory neural crest and find evidence of lineage bias toward either a neuronal or ectomesenchymal fate, with each expressing distinct factors from earlier stages of development. We identify the miR-302 miRNA family to be highly expressed in cranial neural crest cells and genetic deletion leads to precocious specification of the ectomesenchymal lineage. Loss of mir-302 results in reduced chromatin accessibility in the neuronal progenitor lineage of neural crest and a reduction in peripheral neuron differentiation. Mechanistically, we find that mir-302 directly targets Sox9 to slow the timing of ectomesenchymal neural crest specification and represses multiple genes involved in chromatin condensation to promote accessibility required for neuronal differentiation. Our findings reveal a posttranscriptional mechanism governed by miRNAs to expand developmental potential of cranial neural crest.