Congratulations to @stefanosala89.bsky.social , with help from @shreya-c.bsky.social al‬ on the optogenetic experiments, for the awesome story! Read the full preprint here: www.biorxiv.org/content/10.1... and thanks for listening!

Tension on vinculin does change in response to Y27 treatment, which alters both network tension AND actin architecture. This highlights how molecular tension doesn't always directly correlate with network tension and has big implications on how to interpret data from other molecular biosensors.

Our data also highlights the difference between molecular tension (as on vinculin) and network tension (i.e. traction stresses) in adhesions. Zyxin is sensitive to changes in tension on the adhesion, while vinculin isn't.

We think this same process is happening with other tension sensitive LIM domain proteins in adhesions too, like paxillin, LIMD1, and Trip6. They're all recognizing to different degrees the strained actin in adhesions, and contributing to signaling in these structures.

Putting this all together, our data suggests that zyxin is performing the same roles in strain sites as in FAs - it's helping to build a stress fiber while it's under tension by recognizing strained actin filaments

To test this we looked at how VASP responds to our laser ablation in WT and zyxin KO MEFs. We only saw a change in VASP intensity at the FA when zyxin was present! We also measured retrograde flow of actin at FAs and found that it was slower in the zyxin KO cells!

Now zyxin is known to play a key role in stress fiber repair (PMID: 20833360). It recognizes strained actin in stress fibers, recruits VASP to help polymerize new actin filaments, and alpha-actinin to crosslink them in place. Could zyxin be doing the same things in FAs?

If we looked at cell-cell adhesions we saw the exact same thing! Both Trip6 and LIMD1, tension sensitive LIM domain proteins, showed sudden large drops in intensity in response to changes in tension on the adhesion.

So we next looked at other FA proteins, including paxillin, kindlin2, and PINCH. Only paxillin, another tension sensitive LIM domain protein showed changes in response to drops in tension!

Vinculin clearly stretches in response to forces though (PMID: 20613844), so we looked at molecular tension using the VinTS tension sensor. Shockingly it also was unperturbed by the change in tension on the adhesion, but it did change in response to Y-27 (ROCK inh.)

We took the same optogenetic approach and saw the same results. Vinculin intensity in FAs was completely unresponsive to changes in traction stresses!

We next tried the same experiments with vinculin, a known mechanosensitive FA protein. What we found was a total surprise - while zyxin was responsive to changes in tension, vinculin accumulation was completely unchanged!

We then used an optogenetic RhoA probe to modulate tension in the cell. We identified FAs where forces increased, remained stable, or decreased and looked at what happened to zyxin intensity. The responses were perfectly coordinated!

We used a laser to damage a stress fiber (SF) upstream of an adhesion and see what happened to zyxin at the adhesion. We saw both zyxin intensity and traction stresses drop immediately in response to the laser ablation.

This machine learning model also directly predicted a relationship between zyxin intensity and traction stress on the focal adhesion (FA). Stefano set out to explicitly test this prediction.

The literature on the relationship between traction stress and adhesion composition/morphology has been mixed. Recently, we built a neural network to predict traction stresses from images of adhesions and the best predictions came from the FA protein zyxin www.cell.com/cell/fulltex...