← Back to research

Epithelial cells on curved substrates

Epithelia line most organs, and many of them are curved. My PhD asked a deliberately narrow question: can the three-dimensional shape of an epithelial cell on a curved surface be explained by a handful of tensions, and if not, what is missing?

A minimal model

I grew epithelial monolayers on wavy substrates and measured cell thickness on the crests and in the valleys. On short wavelengths, cells are thicker in the valleys, as they are in the folding ventral furrow of the Drosophila embryo.

A foam-like model, in which each cell is described by the tensions of its faces, predicts the opposite. One extra ingredient corrects it: a line tension along the apical actin belt. Its fitted value barely changes when adhesion or myosin activity is perturbed.

Confocal image of an epithelium on a wavy substrate: diagonal stripes of cells, nuclei in cyan, junctions in green and white
MDCK cells on a wavy substrate, a single confocal plane cutting through the crests.
Cross-section of an epithelium on a wavy substrate showing thickness variations
Cross-section of the monolayer along the waves.

Where it breaks

The model cannot explain why the effect saturates at high curvature: a term that resists flattening is missing, and the nucleus is the most likely candidate. Describing the epithelium as a continuous film leads to the same place: a passive elastic film fails, and the data require an active stress in the apical layer.

Naming that boundary is the result as much as the fit: it says which mechanism a tension-based description of epithelia has to add.

Collaborations

Related publications