Finally, we disrupted apical patterned behaviors experimentally. The optic cup completely failed to start the invagination! (side note, the phenotype was so striking that I thought the microscope had crashed while imaging)

Our model became even cooler: we converted tissue segmentations into networks. Like this we could test the strain patterns in the REAL geometry of the tissue

But can the shape transitions at the apical surface be sufficient to drive basal surface invagination? Updating our model with basal surfaces: apical strain patterns alone are enough to initiate invagination—even if basal surfaces are passive

First, we needed to understand how are the apical surfaces curving. We went back and forth between experiments and modelling (inspired by shape programming). We found each apical surface displays different cell behaviors, generating its own in-plane strain patterns,

But the surprise came from the apical surfaces. They undergo shape transitions before invagination starts. Could they be driving the process?

Separating the basal surface of the two epithelia (NE and RPE), we saw that they change their shape at the same time!

We followed 3D shape changes of the forming optic cup, by tracking the basal surface

Many things happen at the same time: 2 epithelial layers differentiate, cells migrate, other cells flatten and others constrict their basal surface, and the lens grows in. We went through (a LOT of) negative results to try to understand which processes were triggering the onset of the invagination

We used the Optic Cup formation as a system. The shape transition from a flat vesicle to an hemispheric eye precursor is just mesmerizing