One of the most fundamental mysteries in biology is how a single fertilized egg gives rise to hundreds of different cell types — brain cells, muscle cells, liver cells — each with its own identity and job. We tackled that question using zebrafish embryos and cutting-edge genomic tools, making four major contributions:
1. Mapping the developing brain: We created detailed "census" data for the zebrafish brain at different points in development, cataloguing the many cell types present and identifying the molecular signatures that define each one. We also developed better tools to track which cells are descended from which, essentially reconstructing family trees of brain cells as they diversify.
2. Understanding how cells become specialized: As cells specialize, thousands of genes switch on and off in coordinated waves. We built a computational tool (MIMIR) to identify groups of genes that work together during this process. Applying it to cells that become glands versus cells that produce structural scaffolding revealed both familiar and new biological mechanisms — including how cells manage to secrete very different products while using shared and distinct regulatory mechanisms.
3. Connecting gene regulation to cell specialization: Beyond which genes are active, we examined how genes get turned on — specifically, which regions of DNA are physically "open" and accessible in different cell types. We built an AI model to predict this from DNA sequence alone, and found that surprisingly few master regulator proteins (transcription factors) are responsible for giving each cell type its unique identity. Remarkably, many of these same regulators are borrowed and repurposed from their original roles in organ formation.
4. Seeing gene activity in space and time: We developed a way to measure the activity of hundreds of genes simultaneously across an entire embryo while preserving its physical structure. This allowed us to track not just what genes are active in a cell, but where that cell sits in the embryo and where it came from, capturing how gene activity changes as cells physically move and reorganize during development. The resulting dataset covers over 25,000 genes and is publicly available for other scientists to explore.
Taken together, these tools and discoveries move us from snapshots of individual genes toward a comprehensive "biography" of how every cell in an embryo acquires its identity. Such global views of development will help understand birth defects, regeneration, and ultimately what it means to build a body.
