Adhesion, Actin and the production of Mechanical Forces

 
Cell shape is the outcome of a complex interaction between cell cytoskeleton and external conditions. When cells are plated on micropatterns, they tend to adopt the convex hull (or convex envelop) of the micropattern, ie they will be square on X, or K patterns and triangular on V or T patterns. This property is not artefactual and reflects an interesting physiological cell behaviour. Indeed, in vivo, most epthelial and mesenchymal cells, which constitute the majority of tissues, are convex.
We have studied the architecture of adhesions and actin cytoskeleton in response to different microenvironment geometries. We plated cells on concave micropatterns and found that cells first spread partially over and then fully cover non-adhesive region by contracting their free edges. The video below shows the spreading of a RPE1 cell on a fibronectin coated V shaped micropattern. It takes 30 minutes to the cell to spread on the V pattern and 1h30 to contract the free edge which stands above the non adherent area. The picture on the right shows cell adhesions (green) and actin cytoskeleton (red) when cell has reached the final steady-state. A large stress fiber stands between the extremities of the V bars. The contraction of this large actin bundle support the free edge. The width of this bundle and the size of the focal adhesion with which it is anchored to the fibronectin indicate a locally high level of tension.
 
RPE1 spreading
  
 
 
Since cells have this property to develop large and contractile stress fibers upon non-adhesive region, it is possible to control the number and location of stress fibers within cells by adapting the micropattern geometry. We found that if the length between the two anchoring areas is reduced the stress fibers becomes weaker. In addition, the group of images on the left shows that when stress fibers are more numerous (ie on the Y rather than on the T or the V) they are also weaker. Actin bundles are thinner and focal adhesions are smaller. So the strength of a fiber depends on both local and global parameters.  Interestingly, on adhesive regions, cells develop much smaller stress fibers, here actin is preferentially involved in branched network pushing on the membrane to form large protrusions. The group of images on the right show that cells form protusions (revealed by a cortactin staining) above these adherent regions and particularly at the extremities of the bar (ie at cell apices) where adhesion are larger. There seems to be a sort of mutual exclusion between stress fibers, the contractile form of actin network, and lamellipodia, the branched form of actin network, and that cell adhesions serve as an intermediate plateform to anchor the first and initiate the second.
 
 
                                                                       CONTRACTILE ACTIN NETWORK                                                                                    BRANCHED ACTIN NETWORK

 Bottom images represent average images of actin (on the left) or cortactin (on the right). Click here to get more details on the production of these averaged images.
 
 
 It is the amazing reproducibility of actin structures on micropatterns which prompted us to describe our observation as if they were some sort of biological laws based on define physical properties.
The first image sequence below shows actin stress fibers revealed by phaloidin staining in cells that have been first permeabilised and then fixed. Actin fibers are remarquably similar from one cell to the other. In particular, the prominence of the stress fibers along the hypothenuse of the L shape is strikingly conserved in all cells. The second image sequence shows actin polymerisation in membrane protrusions revealed by cortactin staining. Although cells still appear quite similar, with protrusions along the curved edges and at the extremity of the bar, the structures appear more variable. Click here to get more details on how to quantify the spatial distributions of staining variability.
 
Actin staining of RPE1 cells on L shaped micropatterns
 
 Cortactin staining of RPE1 cells on crossbow shaped micropatterns
 

The actin network can be decomposed into expansive and contractile regions whose respective distributions are coupled to cell adhesion pattern. Close to actin nucleation sites (plasma membrane and cell adhesion) the polymerization of actin filaments generates an expansive network which encounters cell adhesions and form bundles of filaments. Bundles contraction and actin flow move the network toward cell interior where it is disassembled. All fibers do not disassemble. Local adhesion and specific dynamics support large force production in non-disassembling bundles called stress fibers. The overall distribution of mechanical forces is quite complex. It results from  network expansion pushing on cell borders, and network contraction in moving or stationnary fibers. The balance between all these contribution ensures cell mechanical equilibrium. It is therefore very important to investigate force production in all these subcellular elements and the way they interact with each others. It is also necessary to measure the corresponding forces.

Force relaxation on deformable substrate







We used a new micropatterning method applicable to soft, deformable substrate (poly-acrylamide) (see Protocols) to measure the overall traction force field cells exert on their micro-environment. When cells are attached to the micropattern they exert traction forces on the substrate which get deformed. Disruption of these forces, by actin disassembly or myosinII inactivation, induces substrate relaxation. It recovers its original shape. See video and corresponding paper (Tseng et al., Lab On a Chip, 2011).











The spatial distribution of forces can be quite complex. It depends on the overall organization of the actin cytoskeleton. By normalizing actin architecture, the traction force pattern can also be normalized. See (Tseng et al., Lab On a Chip, 2011).





 
 







These pictures show actin network organisation on deformable micropatterns and the corresponding traction force field in one or several overlaid and averaged cells.
Forces are all localized at the extremity of the bar of the crossbow-shaped micropattern.




It is thus possible to measure global cell contraction level simply by looking at local micropattern deformation
as it is the case for an archer pulling on the arrow.










Comments