Watching a cell sheet transform from soft to stiff

The physical properties of the interface between a cell monolayer and its substrate have been hard to measure directly and are risky to infer from mechanical measurements done on the top of the cell. A new method using local contraction of a hydrogel substrate allows measurements of basal surface stiffness and shows how an epithelial cell monolayer becomes stiffer as it initiates the epithelial-mesenchymal transition.

When cells contact an underlying substrate, such as the basement membrane in an epithelial tissue, the contacts made between transmembrane protein adhesion receptors and their ligands in the extracellular matrix initiate signals that lead to cell contractility, cytoskeletal remodeling, and changes in cell morphology and mechanical properties. These effects have been well documented in single-cell studies, often using indentation of the apical surface of the cell to calculate cell stiffness. What is happening on the basal surface, where cells apply traction stresses to their substrate and link their cytoskeleton to the viscoelastic support below, has been much harder to measure. The report by Sutton et al. (1) describes a remarkably simple, in principle, method, called cell monolayer deformation microscopy (CMDM), to apply well-controlled deformations to the substrate below an adherent cell monolayer and compute the planar mechanical resistance of the monolayer. Direct measurement of viscoelasticity in the plane of the monolayer is hard for conventional techniques, because this material is so thin, and the adhesive monolayer-substrate interface further complicates attempts to measure stiffness. Furthermore, cells are heterogeneous and have nonlinear elastic responses, so it is not obvious that measurements of cell stiffness made by indentation or twisting of the apical surface can be easily related to the mechanical resistance, or apparent stiffness, of the basal plane of the cell or the monolayer.

The new method of Sutton et al. exploits the well-known capacity of hydrogels formed by N-isopropylacrylamide (NIPAM) to contract when heated above a critical temperature near 32°C. A sudden change in local temperature can produce local contractile strains in a hydrogel substrate with a well-characterized elastic modulus, allowing inference of the stiffness of the cell monolayer from the effect it has on the deformation of the hydrogel beneath it.

The method uses principles developed for cell traction force microscopy in which gels with a known elastic modulus are embedded with fluorescent beads or other fiducial markers (2, 3, 4). As cells pull on the gel beneath them, the strain they exert on the gel causes movements of the beads, and the displacement of these beads from their resting position enables quantification of a strain field from which the corresponding stresses can be computed, knowing the elastic modulus, which relates stress to strain. In this study, however, the system is used not to measure the active forces that the cells exert on the substrate, but rather the passive resistance that the attached cell monolayer places on the substrate when it is deformed by some other force. Here the substrate is subjected to a strain generated by a localized increase in temperature that drives the phase transition in the hydrogel, causing local contraction and consequent movement of beads from outside the contraction zone into it. The strain field resulting from such a temperature pulse is measured for the hydrogel alone and for the hydrogel with the cell monolayer above it (Fig. 1). The extra stiffness provided by the attached cell monolayer results in smaller strains within the hydrogel compared with the strains without the cells, and from this difference the elastic properties of the cell monolayer can be computed.

Figure 1.

Schematic of CMDM methodology. Experimental settings: (A) An N-isopropylacrylamide (NIPAM) substrate with an attached cell monolayer. (B) When the sample surface is illuminated with near infrared (NIR) laser light from the bottom, the NIPAM substrate contracts due to temperature changes, and the resulting stress and strain are transmitted to the cell monolayer. The displacements of the microbeads at the sample surface are tracked to measure the laser-induced strain. (D) The cell monolayer is then removed, and (E) the measurements are repeated without the cell monolayer in place to reinforce the substrate and resist the substrate deformation. Working principles: (C) The experimental setting in (B) may be represented by a pair of springs in parallel, where spring 1 is the NIPAM substrate with a known spring constant $k_1$ measured from standard rheology tests, and spring 2 is the cell monolayer with an unknown spring constant $k_2$ to be determined. The system is subjected to a temperature-induced force $F$, and the resulting displacement, $d$, of the system is measured. (F) The temperature-induced force $F$ can be deduced from experimental setting (E) by measuring the displacement $d_0$ of the NIPAM substrate. In particular, $F=k_1{d}_0$. Given that the temperature-induced forces are the same in (C) and (F), one can estimate the spring constant of the cell monolayer using $k_2$ = $k_1(d_0-d)/d$. To see this figure in color, go online.

The local temperature change is made by illuminating a 5-μm-diameter spot of a hydrogel containing nanoparticles that efficiently absorb photons and convert this energy to heat. By optimizing the hydrogel stiffness to approximately 10 kPa, the strain field generated for the contraction within the 5-μm illuminated spot extends to approximately 25 μm, enabling strain measurements over a length scale larger than a single cell. This method is used to study the difference in monolayer stiffness as an epithelial cell sheet undergoes epithelial-mesenchymal transition (EMT), in which cells typically reorganize their cytoskeletons and their cell-cell contacts, with the eventual movement of some cells out of the originally confluent sheet. To study this change, mouse mammary epithelial cells were cultured to confluence and then stimulated to undergo EMT by addition of TGF-β. Using this method, Sutton et al. conclude that the mammary epithelial sheet becomes nine times stiffer after initiation of EMT. This finding is remarkable because it is in contrast to earlier studies of single cells, which suggest that well-spread stationary epithelial cells are likely to become softer as they become mesenchymal and begin to move. The absolute values of the computed Young’s moduli at the cell basal surface are also informative. The epithelial monolayer has a Young’s modulus of ∼10 kPa, which increases to ∼100 kPa after EMT. These values are significantly higher than the apparent Young’s moduli computed from indentation or twisting measurements done on the apical surface of most cell types, including mammary epithelial cells (5). The difference between apical and basal stiffness reinforces the importance of stiffness heterogeneities as cells are studied in more physically relevant settings than single cells on a surface.

Because the temperature change causing substrate strain is modest, and the phase transition of NIPAM is reversible, the local stress in the substrate can be held constant or rapidly applied and removed, allowing different possible types of mechanical characterizations. For instance, Sutton et al. measure the creep response of the cell monolayer, a behavior that has been hard to assess previously, by keeping the stress in the substrate constant. They show that after EMT the cell monolayers are initially much stiffer, but gradually soften over time, exhibiting significant creep behavior. In addition, by tracking local bead displacements, the new method can measure cell monolayer properties at a microscale spatial resolution. In particular, Sutton et al. study the local mechanical properties of the cell monolayers before and after EMT, concluding that the transformed monolayers exhibit more heterogeneous viscoelastic properties.

As with all new methods, there are interesting issues to consider. For example, more remains to be learned about how large and stable the temperature-induced strains are in the substrate and how best to calculate monolayer material constants like the cell monolayer Young’s modulus from the raw data, which are the measured displacement fields and the elastic modulus of the substrate. Similar to calculations to derive stresses for traction force microscopy, the stress values calculated by CMDM also necessitate some choices of models and assumptions about the monolayer-substrate system. Due to the highly nonlinear and viscous behavior of cell monolayers, the simple linear elasticity model is unlikely to work, and more sophisticated models are needed. To circumvent this complication, Sutton et al. use instead pseudo-stresses and strains, which are essentially the raw displacement data, to analyze the mechanical imprint of cell monolayers. Although this method allows the investigation of more complex viscoelastic phenomena that go beyond linear elasticity, it cannot provide the absolute values of the true physical quantities, like stress supported by the monolayers and creep compliance of the monolayers. These open issues provide excellent topics for future research.

Overall, CMDM developed by Sutton et al. offers a novel platform to study the viscoelastic properties of cell monolayers at high spatial-temporal resolution in a physiologically relevant setting. It opens new avenues to the biophysical characterization of cell clusters and has potential applications to monitor multicellular mechanical changes in disease progression.

Editor: Timo Betz.