ID #153 MYC-driven pediatric brain tumors share a common photoreceptor identity

Abstract

Medulloblastoma (MB), pineoblastoma (PB), and retinoblastoma (RB) are malignant pediatric neural tumors arising in distinct anatomical regions of the central nervous system, yet they share elevated expression of a photoreceptor-associated transcriptional program. MYC-amplified MBs and PBs are often histologically indistinguishable, grow aggressively, frequently relapse after standard therapy, and have poor prognosis. Trilateral RB is rare but typically presents with aggressive retinal tumors followed by PB. Understanding the developmental origins of these photoreceptor-positive tumors is critical to determine whether they share common biology and therapeutic vulnerabilities.

CRX, a transcription factor essential for retinal and pineal gland development, is also a master regulator of MYC-driven Group 3 MB maintenance. To investigate whether MYC-driven MB, PB, and RB arise from common photoreceptor-positive progenitors, we performed Crx-lineage tracing using a tamoxifen-inducible Cre system during postnatal development. As expected, Crx-positive cells were detected in the retina and pineal gland progenitors. Unexpectedly, Crx-lineage tracing also labeled granule neurons in the flocculonodular lobe of the cerebellum, a region recently proposed as a site of origin for Group 3 MB derived from the developing rhombic lip.

To determine whether MYC activation in Crx-positive cells can drive tumor formation, we generated XMYC and XMYCT58A mouse models by crossing Crx-CreERT2 mice with Myc or stabilized MycT58A strains, respectively, and induced oncogene expression postnatally with tamoxifen. Myc-overexpressing mice developed tumors within 2–3 months, significantly earlier than MycT58A mice, which developed tumors at 4–6 months. Tumors arose from the pineal gland and cerebellum in both models. Histologically, tumors were non-glial (GFAP−, OLIG2−), exhibited neuronal features (NeuroD1+), and expressed photoreceptor markers (CRX+, OTX2+), closely resembling human MYC-driven Group 3 MB and PB.

These findings suggest that MYC-driven MB, PB, and RB may originate from photoreceptor-positive progenitor populations, highlighting shared developmental origins and potential therapeutic targets.