Force production and regulation by the cytoskeleton

Get Complete Project Material File(s) Now! »

Stress fibers formation and dynamics

One of the major contractile elements in cells are the stress fibers (Kassianidou & Kumar, 2015; Pellegrin & Mellor, 2007; Vallenius, 2013). They were observed decades ago as bundles of microfilaments that crossed the cell body (Abercrombie, Heaysman, & Pegrum, 1971), terminating in dense plaques at the base of the cell and classified as contractile structures (Isenberg, Bielser, Meier-Ruge, & Remy, 1976; Kreis & Birchmeier, 1980). They were found to be actomyosin bundles crosslinked by α-actinin, with the actin in bi-polar arrangement (Lazarides & Burridge, 1975; Mitchison & Cramer, 1996). These contractile structures are often anchored to focal adhesions, connecting the cytoskeleton to the ECM (Mitchison & Cramer, 1996; Naumanen, Lappalainen, & Hotulainen, 2008; Pellegrin & Mellor, 2007).
The mechanical and biochemical interactions between the ECM and cells modulates stress fiber abundance, structure and organization. Hence, stress fibers are important features of cells that encounter mechanical resistance (S. Tojkander, Gateva, & Lappalainen, 2012). Fibroblasts develop prominent stress fibers during wound closure through generation of tension and ECM remodeling (Sandbo & Dulin, 2011). Epithelial cells involved in wound closure also display stress fibers and then differentiate into myoepithelial cells (Pellegrin & Mellor, 2007). During embryogenesis, stress fiber like cables that span multiple cells are present in epithelial cells during dorsal closure (Jacinto et al., 2002). The endothelial cells which experience mechanical strain due to the hydrostatic pressure caused by blood flow also display prominent stress fibers (Wong, Pollard, & Herman, 1983). The presence of these structures in such varied cell types that experience mechanical stimuli shows their importance in regulating the mechanical signals that cells receive. Understanding the factors involved in generation and maintenance of these stress fibers is hence important in building models that are physiologically relevant.
Myosins were found to be part of the stress fiber of non-muscle cells along with actin a few decades ago and it was also found that myosin II forms periodic striations along these stress fibers (Weber & Groeschel Stewart, 1974). Electron microscopy studies of platinum replicas showed without a doubt that myosin II filaments are present in these cells (Svitkina, Verkhovsky, McQuade, & Borisy, 1997; Verkhovsky, Svitkina, & Borisy, 1995). These early studies showed that, in fibroblasts, myosin II filaments have uniform length and form super-structures where multiple parallel filaments are organized in registry forming filament stacks. Such structures were also found later in the interphase state of other non-muscle cells and contractile rings of dividing cells (Fenix et al., 2016; Goeckeler et al., 2008). In more recent studies, utilizing techniques such as TIRF (Total Internal Reflection Microscopy) and SIM (Structured Illumination Microscopy) (Gustafsson, 2000; Kner, Chhun, Griffis, Winoto, & Gustafsson, 2009), the visualization of the individual bipolar filaments in non-muscle cells has improved significantly (Beach et al., 2014; Burnette et al., 2014; Shutova, Spessott, Giraudo, & Svitkina, 2014). The filaments have been found not only in the ventral stress fibers, but also in the transverse arcs (these structures to be discussed later) and it was observed that they are present at the cell periphery (Burnette et al., 2011, 2014), sometimes near focal adhesions (Pasapera et al., 2015). The stacks that Myosin II filaments form are perpendicular to the actin filament orientations and the myosin II stacks are spaced regularly, with the regions in between the stacks being taken by α-actinins (Gordon, 1978; Langanger et al., 1986; Lazarides & Burridge, 1975). The organization of NMM’s in cells has been shown to effect traction forces exerted by cells and the contractility of the actin network (Dasbiswas, Hu, Bershadsky, & Safran, 2019; Hu et al., 2017, 2019). Over time, these structures were classified into different types of stress fibers, each with specific roles that they undertake (Naumanen et al., 2008; S. Tojkander et al., 2012). The four main classes of stress fibers are called the dorsal stress fibers, the transverse arcs, the ventral stress fibers and the perinuclear actin cap (Figure 10).

Adapted from: (Kemp & Brieher, 2018)

Model depicting one way in which α-actinin-4 might suppress stress fiber formation. In the presence of α-actinin, due to restriction of filament sliding by crosslinking, a highly interconnected network of disordered actin filaments forms at the basal surface of cells that undergoes myosin mediated biaxial contraction through filament buckling. In the absence of α-actinin, network connectivity is lost, favoring stress fiber formation by filament condensation. Loss of α-actinin also allows tropomyosin to bind to F-actin, which stiffens the filaments to further inhibit buckling and favors stress fiber formation. Pink arrows show the directions of contractile forces.
These results suggests that high actinin crosslinking of stress fibers leads to a jammed system where forces are dissipated within the actin-myosin-actinin network, rather than the forces being to the substrate. While the relative levels of proteins such as tropomyosin and cofilin can be quantified, actinin and myosin interact with them and compete for actin binding. As such, quantifying force transmission to the substrate and comparing the values to the relative ratios of actin-myosin, myosin-actinin and actinin-actin should shed light on the role of these different interactions between proteins and traction force variations in cell populations.
Along with actin filament nucleation at the focal adhesions, there will be an equivalent disassembly of filaments in the existing fiber. In the absence of this disassembly, the fibers should keep getting thicker. Studies that analyzed the turnover rates along the length of the stress fibers show that filaments turnover with a characteristic lifetime of 1 minute compared to the lifetime of the stress fiber which is around 1 hour (Hu et al., 2017). The myosin and actinin turnover is faster near the center of the fiber compared to the ends (Peterson et al., 2004). These dynamics, which can lead to reorganization of the stress fibers, make it difficult to relate the forces exerted by them to their observed composition. The protein zyxin has been shown to be recruited to damaged regions of stress fibers and help in the recovery process by recruiting VSAP, α-actinin and inducing filament formation and bundling (Smith et al., 2010). This might be a method through which new actin filaments that are more aligned to the stress fiber are generated around it. Conclusively identifying the parameters regulating this dynamic exchange of filaments and the sharing of the available actin monomer pool between these contractile structures should help elucidate the link between force production at the cellular level and actin network remodeling.
Stress fibers are not isolated structures within cells (Kassianidou, Brand, Schwarz, & Kumar, 2017; Marek, Kelley, & Perdue, 1982; Vignaud et al., 2020). While the connection of dorsal to transverse stress fibers has already been discussed, ventral stress fibers are also not isolated structures. Laser ablation of single stress fibers have been shown to compromise the entire traction force field (S. Kumar et al., 2006; Tanner et al., 2010) and lead to changes in the tension at focal adhesions not connected to these fibers (C. W. Chang & Kumar, 2013). These studies show that contractile forces generated in one stress fiber can propagate to other stress fibers through the connections between them (Kassianidou et al., 2017). Kurzawa et al., showed that contractile bundles not only possess interconnections to other stress fibers by are also embedded in a continuous meshwork of cortical filaments along their entire length (Figure 14). The cortical actin filaments close to the stress fiber were more aligned to it compared to those further away from them, which suggests that randomly oriented filaments were being realigned and incorporated into the stress fiber. They showed that this cortical network has myosin motors attached to it and is also contractile. These results are in agreement with previous studies that showed cortical actin connections to stress fiber (A. Kumar et al., 2019; Marek et al., 1982; Svitkina, 2018). It is possible that the interconversion of filaments between the cortical structure and the stress fiber is essential for the modulation of the production of traction forces.

Actomyosin cortex

The animal cell shape is controlled primarily by the cell cortex (Salbreux et al., 2012). This is a thin network of actin filaments, myosin proteins and actin –binding proteins that are situated immediately below the plasma membrane of most eukaryotic cells that lack a cell wall. It is the main determinant of the stiffness of cellular surface and it resists external mechanical stress (D. Bray & White, 1988). It is thought to oppose intracellular osmotic pressure (Stewart et al., 2011). Local changes in the mechanical properties of the cortex drives many cellular deformations, like those occurring during mitotic cell rounding, cytokinesis, cell migration and morphogenesis (Bergert et al., 2015; D. Bray & White, 1988; Heisenberg & Bellaïche, 2013; Levayer & Lecuit, 2012; Maddox & Burridge, 2003; Sedzinski et al., 2011).
Quantifying the ultrastructure of the cortex has been difficult in part due to its location below the plasma membrane which makes it difficult to image using electron microscopy compared to flat actin structures such as lamellipodial and Filopodia (Svitkina et al., 2003). However, using fast scanning AFM and super resolution microscopy has provided important details of the apical and basal cortex architecture and composition in recent years (Xia et al., 2019; Yanshu Zhang et al., 2017). They have shown that the cortical mesh size and orientation of actin filaments diverge significantly depending on cell type and the various actin binding proteins involved in the maintenance suggesting the cells tune their cortex based on their function.
Cortical tension is regulated by myosin II motors as they create contractile forces by pulling actin filaments with respect to one another (Clark, Wartlick, Salbreux, & Paluch, 2014; Vicente-Manzanares et al., 2009). It has been shown that inhibition of myosin contraction reduces bleb growth while increase in myosin activity through RhoA transfection resulted in increased bleb size (Tinevez et al., 2009). In mitotic cells, increased recruitment of myosin II to the cortex has been shown to correlate with increasing intracellular pressure (Ramanathan et al., 2015). Tinevez et al., also showed that by impairing actin turnover is leads to changes in the cortical tension. Depletion of ADF/cofilin and the barbed end capping protein CAPZB leads to increases in the tension and cortical thickness increases(Chugh et al., 2017), whereas cytochalasinD treatment resulted in ten-fold decrease in the tension compared to control cells. Chemical inhibition of Arp2/3, formins and myosin also lead to changes in cortical filament length, architecture (Shirai et al., 2017; Xia et al., 2019) and tension (Chugh et al., 2017). These studies suggest that the type of actin network, the length of actin filaments, actin dynamics and myosin activity plays important roles in maintaining cortical architecture. Reduction in α-actinin has also been shown to reduce mitotic cell tension suggesting that crosslinks in the cortex helps maintain its strength. As it has been discussed earlier, α-actinin crosslinking has been demonstrated to coordinate long range contraction of actomyosin systems. The protein helps to maintain connections between all the filaments in a mesh, allowing the contractile stresses to propagate. However, a high concentration of actinin inhibits contraction, suggesting that the excessive crosslinks can jam the system, preventing forces from being transmitted across the system (Bendix et al., 2008; Chugh et al., 2017; Ennomani et al., 2016).

Traction force measurement methods:

The technique of traction force microscopy (TFM), provides a powerful tool that can be used to measure the forces exerted by cells on their substrate. In the last few decades, different methods have been created to measure the forces generated by cells. Each method comes with positives and negatives that are discussed below in brief.

Substrate deformation based:

One of the earliest methods used to measure forces generated by individual cells was using wrinkles generated on a silicone layer(Harris, Wild, & Stopak, 1980). However, getting quantitative information regarding cellular forces was difficult as the wrinkles were nonlinear and irregularly shaped. In later work, beads were introduced onto a non-wrinkling elastic film made of silicone and the deformation of the layer observed through the displacement of the beads (J. Lee et al., 1994; Tim Oliver et al., 1995). However, these substrates were nonporous, poorly adhesive and their mechanical properties could not be tuned easily to match the force generation capacity of many mammalian cells (Dembo & Wang, 1999). These limitations were overcome by using polyacrylamide (PA) hydrogels as the substrate to plate cells. These substrates are non-toxic, optically transparent, elastic, can be embedded with fluorescent beads, the protein density can be independently controlled and the stiffness can be tuned to the physiological range of many cell types (Wang & Pelham, 1998). They can also be micropatterned with proteins with micrometer precision allowing for control of cellular morphology and actomyosin architecture (Thery, 2010; Tseng et al., 2011; Vignaud, Ennomani, & Théry, 2014). The calculations to obtain the forces from the deformation of these substrates has been well characterized (Figure 15) and optimized (Dembo & Wang, 1999; Stricker, Sabass, Schwarz, & Gardel, 2010; Tambe et al., 2011; Tseng et al., 2011). However, careful application of the regularization procedure used to add additional constraints to the force estimation or filter the image data is required to avoid error in measurement (Kulkarni, Ghosh, Seetharaman, Kondaiah, & Gundiah, 2018; Martiel et al., 2015).

Cantilever based methods

Micropillar based traction force detection is another useful method in which, microfabricated arrays of polydimethylsiloxane (PDMS) (du Roure et al., 2005; Tan et al., 2003) or acrylamide (S. W. Moore, Biais, & Sheetz, 2009) serve as substrates and force sensors. As the micropillars are made out of elastic material attached to a stiff substrate, a force exerted on the free end causes the pillar to bend (Figure 17). If this force is sufficiently small compared to the pillar height, the force is proportional to the displacement allowing simple calculation of traction forces exerted by cells. The spring constant of these pillars can be quantified using elastic theory and the dimensions of the pillar or by quantifying thermal fluctuations of the pillar (Hutter & Bechhoefer, 1993). If the pillars are closely spaced and only their top surfaces are selectively coated with proteins, cells attach and exert forces only on the tips of the pillars. This method has a few advantages. Displacements can be calculated without a reference image. The displacement of a pillar depends only on the force applied on it and not on the other pillars. The stiffness of the pillar can be modulated by changing the geometry of the pillars allowing generation of steep and heterogeneous mechanical environments without altering material properties (Sujin Lee, Hong, & Lee, 2016; Saez, Ghibaudo, Buguin, Silberzan, & Ladoux, 2007). The disadvantages include having cells attach to discrete regions and topology, instead of a uniform adhesive surface which influences the morphology of cell-ECM adhesions. As the topography of the substrate has been shown to affect biological behavior of cells such as differentiation (Trappmann et al., 2012) and focal adhesion formation (Huang et al., 2009; Malmström et al., 2010), the availability of discrete regions of adhesion in micropillar assays is a concern. Focal adhesions attached to one pillar can extend beyond the adhesive area (Sarangi et al., 2017). Fabrication technology restricts the stiffness range to approximately one order of magnitude compared to continuous substrates where the stiffness range is around two orders of magnitude (Roca-Cusachs, Conte, & Trepat, 2017).

Table of contents :

1 General introduction
2 Force production and regulation by the cytoskeleton
2.1 The actin network
2.2 Role of molecular motors
2.3 α-actinin
2.4 Focal adhesions
2.5 Stress fibers formation and dynamics
2.6 Actomyosin cortex
3 Traction force measurement methods:
3.1 Substrate deformation based:
3.2 Cantilever based methods
3.3 Molecular sensors
4 Positioning of the study
5 Methods
5.1 Pattern design
5.2 Mask design
5.3 Silanized coverslip preparation
5.4 Passivation of fluorescent beads
5.5 Polyacrylamide micropatterning
5.6 Cell culture
5.7 Imaging
5.8 Protocol for fixed cell assay
5.9 Fluorescent labelling of cells
5.10 Image processing of fluorescent images
5.11 Translation and rotation correction functions
5.12 Traction force microscopy analysis
5.13 Automated acquisition
5.14 Transfection protocol
5.15 Photoconversion experiment protocol
5.16 Quantification of photoconverted signal in cells
5.17 Data plotting and statistical analysis
6 Results
6.1 Cell cycle analysis
6.2 Actin dynamics results
6.3 Biochemical characterization (Paper draft)
6.4 Extra data
7 General conclusions and perspectives about the work presented in this thesis manuscript.
7.1 The modified TFM assay:
7.2 Biochemical composition study
7.3 Actin dynamics results cannot explain difference in cell forces.
7.4 Perspectives
8 Abbreviations
9 Bibliography 

GET THE COMPLETE PROJECT