Modeling oncogenesis, phenotypic identity and fate in brain cancer avatars

Ziel

Despite decades of cancer genomics, recreating human cancers in the lab remains unattainable because we cannot control the normal-to-tumor trajectory.
I propose gaining control by functionally tracing developmental trajectories in human brain organoids. Using synthetic genetic tracers of brain tumors, engineered DNA reporters that switch on when a normal organoid cell runs a tumor-like program, we identify “best-in-class surrogate target” (BEST) cells that reproduce specific tumor phenotypes. Then, we genetically engineer BEST cells and lock in their phenotype with defined microenvironmental cues and targeted epigenetic editing to build clinically faithful tumor avatars, enabling predictive genetic and drug screens. This strategy uniquely suits rare pediatric brain cancers, which arise from immature cells and exhibit limited genetic complexity, features compatible with current organoid, synthetic genetics, and (epi)genome-editing technologies.
Building on our validated synthetic genetic tracers design platform, we will assemble a scalable atlas to functionally map BEST cells across region-specific brain organoids. As near-term proof-of-concept, we will deliver BEST avatars for embryonal tumors, leveraging tracers, target cells and organoids defined in our preliminary work. In parallel, we will extend the approach toward systematic modeling of other brain tumor entities. Genomic atlases, clinical DNA methylation classifiers and explainable AI will nominate precise genetic/epigenetic edits and supportive cell types to maximize fidelity, after which CRISPR-dependency and phenotypic drug screens will reveal actionable vulnerabilities and combinations.
By stepwise reconstructing tumorigenesis, MOIRA shifts the field from describing what cancer is to understanding how cancer arises, yielding predictive avatars for rare pediatric tumors that often lack robust models and accelerating translational discovery for underserved patients.