Our work confirms that slow tissue movements can generate forces that are significant enough to deform an organ, as the timescale of tissue relaxation is large.

This suggests dynamical forces may be playing a role in many other developmental processes, too. We should look!
13/n

In addition to altering lumen shape changes, are dynamical forces sufficient to change individual cell shapes to drive KV remodeling involved in LR patterning?

Yes, notochord ablation reduces the AP distribution as compared to controls. The 3D vertex model predicts this.
12/n

These shape changes are generic – they can also be seen in a simpler hydrodynamic model of a membrane surrounded by a highly viscous medium, also with anterior pushing forces and posterior pulling forces.

11/n

The experiments match the simulation predictions! In notochord ablation experiments, the lumen elongates along the anterior-posterior axis (RgAP/RgLRincreases), while posterior cells ablations extend the lumen along the left-right axis (RgAP/RgLR decreases).

10/n

We quantify KV shape by the ratio of the radius of gyration of the lumen along the AP and LR directions (RgAP/RgLR).
In simulations, we mimic notochord (posterior cell) ablations by removing the pushing (pulling) forces, and find that the organ changes it shape significantly.
9/n

In simulations, we identify a set of model parameters (star in phase diagram) that generate the lumen shape seen in control experiments, where KV is pushed from the anterior by the notochord (orange cells) and pulled by posterior cells (purple cells).

Laser ablation of the notochord reduces the speed of KV in comparison to control experiments. That suggests that we were able to reduce the forces on the KV at the anterior part of KV.
We next ask how these reduced forces affect the shape of the KV lumen and KV cells
7/n

To test our hypothesis, we developed 3D models and laser ablation experiments to:
1) quantify how perturbing structures around KV impact its motion
2) measure cell and organ shape in these cases
3) show that observed in vivo shape changes match those predicted from the 3D model
6/n

Our hypothesis: The high-viscosity tailbud tissue exerts drag forces on KV as the organ is pushed through the tissue by convergent extension in the notochord, and pulled through the tissue by posteriorly migrating cells. These forces are sufficient to change KV shape.
5/n

Recent work has demonstrated that many tissues, including zebrafish tailbud, operate close to a fluid-solid transition. Near such transitions, the effective viscosity diverges. In these cases, very slow motions can give rise to large forces.
nature.com/articles/s41...
4/n

To answer this question, we study Kupffer’s Vesicle (KV). KV is composed of a single layer of epithelial cells surrounding a fluid-filled lumen, located in tailbud of zebrafish, and it moves through surrounding tailbud tissue slowly, at about a micron per minute.
3/n

A key question in developmental biology is how organisms robustly control the morphology of tissues and organs. Mechanical forces can help provide such a control mechanism, but because tissue motion is so slow, most studies have assumed local force balance. Is that true?
2/n

This was a fun collaboration spearheaded by @rajphys.bsky.social and Emma Retzlaff, in collaboration with the fabulous Amack Lab and @lovelessradio.bsky.social, in the great environment of @syracuseu.bsky.social

Sure, thanks! www.bigmarker.com/cellpress-so...

What the paper shows is that -- for some key observables -- it doesn't seem to matter which model you choose. The cell shape always predicts the rigidity transition, and the viscoelastic response is the same on either side of the transition.

But the key point is that we're not sure which of these models is "right" -- some might be right for one cell but not another. So what I mean is that different versions of vertex models are likely "wrong" for certain cell types.

One way of rationalizing the P^2 term is to say some cell types have a contractile actin ring; another way is to say that there's a limited pool in the cell of, e.g. cadherins or myosin.

The linear term is a line tension (2D version of a surface tension) but the quadratic term is more mysterious, and some other versions of vertex models remove it. Other versions of vertex models have a spring-like restoring force on each edge, instead of a restoring force on the perimeter.

We know that cell cortex mechanics is really complex (viscoelastic, with feedback loops, etc.) A standard vertex model makes a specific assumption about that mechanics -- with a term proportional to the perimeter P and a second term proportional to P^2,