Elucidating Neurodevelopmental Trajectories in Cancer with Topic Modeling: Revealing Persistent External Granule...

Area of Science:

• Neuroscience
• Developmental Biology
• Oncology

Background:

• The cerebellar rhombic lip is crucial for generating neurons, forming over half of the adult brain's neuronal population.
• Standard clustering methods struggle to identify transient rhombic lip progenitor cells, hindering understanding of cerebellar development and medulloblastoma origins.
• Medulloblastoma, a common pediatric brain tumor, has varied survival rates, with Sonic hedgehog (SHH) subtype having particularly poor outcomes.

Purpose of the Study:

• To identify and characterize rhombic lip and external granule layer (EGL) progenitor states during cerebellar development.
• To investigate the persistence of these developmental signatures in SHH medulloblastoma.
• To develop a transferable framework for comparing developmental and disease states across different datasets.

Main Methods:

• Applied a novel topic modeling framework to analyze over one million fetal cerebellar nuclei.
• Identified distinct proliferative and differentiating cell states within the rhombic lip and EGL.
• Validated developmental signatures in SHH medulloblastoma samples.

Main Results:

• Identified proliferative rhombic lip and EGL states that bifurcate into glial and neuronal lineages.
• Captured the developmental spectrum from outer EGL (oEGL) proliferation to inner EGL (iEGL) differentiation.
• Demonstrated that developmental signatures persist in SHH medulloblastoma, confirming granule neuron precursor (GNP) origins and revealing age-specific molecular programs.

Conclusions:

• The developed topic modeling framework successfully identifies transient progenitor populations and developmental trajectories in the cerebellum.
• These developmental signatures are conserved in SHH medulloblastoma, providing insights into tumor origins and heterogeneity.
• The framework offers a transferable solution for comparing cross-technology genomic atlases of development and disease.