Programmable morphogenesis
Research
We aim to decode and program morphogenesis by uncovering and rewriting the rules that turn signal inputs into multicellular structure.
Our question
How does multicellular structure emerge?
Developmental biology analyzes how nature builds tissues. Synthetic biology engineers biological systems. We work at the intersection of the two. Our fundamental question is: what rules do cells follow to build tissues? We answer it not only by observing natural development but by reconstructing and modifying it, because changing a system is a direct test of whether we understand it. We then use those rules to rebuild natural development, and to model how it goes wrong in disease.
The morphogen code
We call the mapping between signaling-field dynamics and cell-state transitions the morphogen code. Our central hypothesis is that tissue architecture is encoded by a quantitative morphogen code, defined by the combination of signal types, signal geometry, temporal dynamics, and feedback between progenitor state and signaling sources. If this code can be decoded, it can be used to program tissue organization, and to understand how its failure drives disease.
Three directions
01
Decode: How natural systems read signals
We break the "signal → fate" mapping into separable layers and measure each one in stem-cell-derived embryo models.
- Input: ligand identity, inhibitor type, and the spatial geometry of the source
- Field: gradient range, steepness, and temporal dynamics
- Interpretation: cell state (naive vs. primed) and feedback from responding cells
02
Reprogram: Rewriting how tissues are patterned
We use synthetic organizer cells and engineered niches to change the signals cells receive (their identity, timing, and spatial arrangement) and to rewrite how cells interpret them.
- Perturb the inputs: swap or combine signals (Wnt, Nodal, mechanical cues, synthetic proteins), change their timing, and shift boundaries or domain number
- Rewrite interpretation: threshold tuning, signal memory, synthetic cross-talk, feedback rewiring
- Build synthetic niches with ECM, stromal context, and state-responsive organizers
03
Explore: Beyond natural development
Using the rules we decode, we explore forms that nature has not made, and design the signals to build them.
- Evolutionary rewiring: the same input, interpreted differently, gives a different structure
- New morphologies beyond natural development
- Inverse design: from a target structure to the signals that produce it
How we work: Analyze & Synthesize
Breaking morphogenesis down into genes, proteins, and pathways explains the parts, but not why cell populations spontaneously generate spatial pattern. Building organoids and watching what happens finds conditions that are sufficient, but not why they work. Neither approach is enough on its own.
We deliberately alternate between the two. We analyze by perturbing single circuit elements, then synthesize by putting the perturbation back into a self-organizing tissue to test whether our pattern-level prediction holds. Computational models (reaction–diffusion, agent-based, Bayesian inference, deep learning) connect the two halves. They let us predict the whole pattern without losing the mechanism.
Where we are heading
- 1Measure
- 2Decode
- 3Control
- 4Rewire
- 5Expand
We start by building a system in which morphogenesis can be measured as a quantitative input–output system. This means multi-reporter stem cell lines, controlled delivery of defined signals from synthetic signaling centers, and time-resolved imaging with lineage tracking. On that foundation we work out the minimal rules that map signal history to fate. Then we use those rules to control pattern formation, and finally to explore which multicellular forms are possible beyond natural development.