Imagine trying to understand how a person reacts when they are squeezed into a tight space—but you cannot see them move, cannot feel how strongly they push back, and cannot tell whether they are actively resisting or simply being deformed.
That is surprisingly close to the problem we faced when studying cells.
Inside the body, cells rarely live in the wide-open world of a laboratory dish. They squeeze through narrow spaces, push against neighbouring cells, deform their surroundings and experience forces from many directions. These mechanical constraints can influence how cells move, develop, divide and survive.
A cell is not a passive object. When a cell has less and less room, the obvious expectation is that it should simply become compressed. But living cells are much more complicated than that.
A cell can reorganize its cytoskeleton, change its shape, generate forces and redistribute internal stresses. So it is not enough to know how much we squeeze a cell. We also need to know how strongly the cell pushes back.
This question led us to develop Confinement Force Microscopy (CFM), a method that combines controlled confinement, high-resolution live imaging and three-dimensional force measurement.
Building a "mechanical sandwich"
The basic idea behind CFM is surprisingly simple.
We place living cells between two thin polyacrylamide gels. Fluorescent beads embedded in the gels allow us to see how the gels deform. The upper gel is mounted on a precisely controlled piezo-driven stage, so we can actively change the amount of confinement while the experiment is running.
We can squeeze cells step by step, release the confinement, compress them repeatedly, or hold them at a particular height for hours.
By tracking the displacement of the fluorescent beads, we can determine the stresses that cells exert on their surroundings—not only laterally, but also vertically.
In other words, CFM lets us do three things together:
squeeze the cell → watch the cell → measure how it pushes back.
That combination is the key idea behind the method.
Could we really keep a cell confined for hours?
One of our first challenges was not squeezing cells, but keeping the distance between the two gels stable enough for long experiments.
For a cell only a few micrometres high, even a small change in confinement can matter. We therefore tested the stability of the system over many hours. The confinement distance remained remarkably stable, changing by only about 700 nm after 10 hours and approximately 1 µm after 20 hours.
This stability allowed us to move beyond observing an immediate response. We could begin asking how cells adapt to mechanical stress over time.
What does a cell actually do when you squeeze it?
We first tested CFM with HeLa cells.
By gradually reducing the distance between the two gels, we could follow the same cell as its available space decreased. We exposed cells to different levels of vertical confinement, from no confinement to 60% nominal strain.
The result was not simply a smaller cell.
The forces generated by the cell changed as confinement changed, and living and fixed cells behaved differently. A fixed cell responded largely as a passive material, whereas a living cell could actively reorganize itself in response to the mechanical challenge.
This distinction is important. It gives us a way to separate what happens to a cell because it is physically compressed from what the cell actively does in response.
Once we could control confinement and measure forces, we could begin asking broader biological questions.
We used CFM to apply repeated mechanical stimulation, to follow force generation during dynamic cellular processes, and to investigate the mechanical properties of cells and multicellular structures. We could even compress cancer spheroids and follow how their mechanical response relaxed over time, revealing their viscoelastic behaviour.
Different experiments asked different biological questions, but they shared one principle:
control the mechanical environment and measure the mechanical response.
And then the cells surprised us
Perhaps the most interesting part of the experiments was discovering how cells actively respond to extreme mechanical confinement. One striking example is blebbing—the formation of temporary membrane protrusions.
A living cell is an active mechanical system. When we suddenly compress a cell, there is an immediate mechanical response as the cell and its internal structures are deformed. But that is only the beginning. The cell then actively responds through its cytoskeleton. Actomyosin contractility builds up, internal cytoplasmic pressure increases over time, and the cytoplasm is redistributed toward the cell membrane, forming blebs. This process helps the cell release the excess pressure and mechanical stress, allowing it to adapt to the confined environment and survive.
Our force measurements showed this active response directly: the increase in cellular stress was followed by a reduction as blebbing developed. Thus, blebbing is not merely a visual consequence of squeezing—it is part of the cell's active mechanical response and a mechanism that cells actively use to cope with confinement and survive mechanical stress.
This leads to a much more interesting question than simply asking, "How much force does a cell generate?"
How does a living cell sense mechanical stress, generate its own forces, and change its shape to cope with that stress?
CFM gives us a way to watch this process unfold.
For us, that is what makes CFM exciting. Because confinement and force measurements are coupled in real time, we can capture the immediate mechanical response of a living cell and then follow how that response evolves over time.
The people behind the instrument
CFM was conceived with Timo Betz, in whose lab I built it during my PhD. Matthias adapted his traction force software for CFM. Katharina measured neutrophil traction stresses across confinement levels. Maja provided the Drosophila embryos, and some of the revision work was done in Eva Kiermaier's lab, where I am now a postdoc.
Build one yourself
We designed CFM to be reproducible in other labs. The LabVIEW program that controls the instrument and the MATLAB pipeline for traction reconstruction are both openly available on Zenodo, and the full component dimensions and material specifications are given in Supplementary Note 1 of the paper. If you build one, we would like to hear how it goes.