(C-86) Role of actin-polymerization in the adherent cell’s stiffness sensing mechanism

Introduction:: Mechanical stiffness of the extracellular matrix (ECM) plays a critical role in proliferation, differentiation, and migration of adherent cells (1). The stiffness is sensed by integrin-based focal adhesions (FAs) (2), where molecules within FAs change shape and biochemical activities in response to the force, but its exact mechanism has been not clearly understood. For FA-triggered mechanotransduction to occur differently per each stiffness, different traction force levels should be transmitted through FAs which we termed as ‘stiffness-dependent differential traction’. Studies have shown that traction increases with an increase in ECM stiffness (3-5). This differential force-stiffness relationship has been previously confirmed experimentally and explained theoretically by a ‘molecular clutch’ model. In this model, F-actin’s retrograde flow dynamically ‘clutches’ with FAs resulting in the force transmission to the ECM (6,7). Conventionally, only Myosin-II contractility has been described as the sole source of retrograde flow where myosin’s muscle-like behavior of force-velocity could nearly explain the observed traction in response to stiffness (8). However, alternatively, actin polymerization via two actin nucleators, Arp2/3 and formin could also contribute to retrograde flow but have not been yet explored. To explore this hypothesis, we studied actin nucleators, independent of myosin activity to understand their role in stiffness-dependent differential force transmission. Our aim is to develop actin-polymerization powered molecular clutch which considers the role of actin nucleators in myosin-independent mechanosensing of ECM stiffness. Thus, studying the differential force transmission by the cell without myosin’s effect would elucidate the very initial stiffness sensing event by a cell.


Materials and Methods:: Traction force microscopy (TFM) and actin fluorescence-speckle experiments were performed on a high refractive index silicone gels (Q-gel, Quantum Silicones) substrate of different stiffnesses (0.6, 1.3, 2.7, 6 and 12.7 kPa) coated with 40nm diameter, far-red carboxylated beads as fiducial markers. The gels were fabricated by mixing different ratios of Q-gel 920-part A and B. The stiffness of gels was measured through shear storage moduli (G’) using DHR-2 hybrid rheometer. 3T3 fibroblasts stably expressing paxillin-mRuby and SNAP actin-expressing 3T3 fibroblasts were used for TFM and actin-speckle experiments respectively. For both TFM and actin-speckle imaging, cells were treated with treated with 1) blebbistatin (BBS), a myosin inhibitor, only, 2) BBS with smifh2, a formin inhibitor or 3) BBS with CK666, an arp2/3 inhibitor. In addition, cells without treatment were used as a control. For TFM experiments, 4 hours after seeding, cells and beads on the gels, were imaged under a total internal reflection fluorescence (TIRF) microscope. Another bead image, a reference image, was collected after lysing cells using 10% bleach. The images were analyzed using a MATLAB-based TFM software in which displacement was calculated by cross-correlation-based particle tracking velocimetry with retracking (cPTVR) method. Fast Boundary Element Method (FastBEM) with L2-norm-based regularization was used for traction reconstruction. The regularization parameter was chosen based on L-curve, L-corner method. Actin-speckle analysis performed using quantitative fluorescence speckle microscopy (qFSM) software in MATLAB in which flow analysis was performed using speckle tracking.


Results, Conclusions, and Discussions:: We observed that in control cells total force increased with the gel stiffness (Fig. 1a,b). Interestingly, in BBS-treated cells traction levels increases with increasing stiffness, suggesting differential force transmission in response to varying stiffness is myosin-independent (Fig. 1a,b). Using We also observed that myosin activity does not affect the actin flow as a function of the ECM stiffness implying that stiffness-dependent differential traction independent of myosin is flow-dependent (Fig. 1c,d). Inhibiting either of two actin nucleators, Arp2/3 and formin in addition to myosin i.e. CK666-BBS and smifh2-BBS treated cells still showed the increasing stiffness dependent traction trend (Fig. 1e,f). The actin flow in cells treated with CK666-BBS showed a decreasing trend from low to intermediate stiffness (0.6-2.6 kPa) with a constant trend in higher stiffness (6-12.7 kPa) (Fig. 1g,h). This trend was observed to be like BBS treated cells with only small flow speed difference. While SMIFH2-BBS treated cells showed diminished flow in most stiffness conditions except for a small increase at low stiffness (Fig. 1g,h). To understand myosin-independent, stiffness-dependent force trend, we modified molecular clutch model to accommodate actin polymerization driven retrograde flow. However, it could not alone explain the flow transmission and flow behavior differences in only BBS and CK666-BBS treated cells. Therefore, we further developed the model by considering polymerizing actin as viscoelastic material (Fig. 1i). In this actin-elasticity based clutch model, actin polymerizing end extends by adding discrete actin unit of length L (Fig. 1j). This addition leads to compression of actin network bound by individual clutches resulting in transmission of compressive elastic force which is then transmitted to the substrate as traction force (Fig. 1k). Depending on substrate stiffness and actin elasticity, force loading can either be fast or slow. As the spring compression is controlled by number of added actin units, upon clutch release, the relaxation of compressed actin by expansion creates retrograde flow. The simulations from this model recapitulate experimental observation (Fig. 1l,m) and explains the traction and flow behaviors in myosin-actin nucleator inhibitor treated cells. Thus, our work explains that stiffness-sensing mechanism in adherent cells depends on actin polymerization independent of myosin.


Acknowledgements (Optional): : Thanks to Dr. Martin A. Schwartz (Yale University) for SNAP-actin-expressing 3T3s, Dr. Mark H. Ginsberg and Dr. Ho-sup Lee (UCSD) for the 3T3 fibroblast cells stably expressing mRuby-Paxillin as a kind gift. Thanks to Dr. Tristan Driscoll (University of Florida) for advise on fluorescence speckle microscopy experiment. We also thank Dr. Pere Roca-Cusachs (Institute for Bioengineering of Catalonia and University of Barcelona) for their myosin-powered molecular clutch model.

This work was supported by a NIH R15GM135806 grant, Health Research Institute (HRI) graduate student fellowship at Michigan Technological University and Fleming & Skochelak Graduate Fellowship in Human Health.


References (Optional): : (1) Discher, D. E. et al., Science, 310:1139-1143, 2005.

(2) Hoffman, B. D. et al., Nature, 475:316-323, 2011.

(3) Han, S. J. et al., Biophys J, 103:640-648, 2012.

(4) Ghibaudo, M. et al., Soft Matter, 4:1836-1843, 2008.

(5) Oakes, P. W. et al., Biophys J, 107:825-833, 2014.

(6) Elosegui-Artola, A. et al., Nat Cell Biol, 18:540-548, 2016.

(7) Chan, C. E. et al., Science, 322:1687-1691, 2008.

(8) Mitrossilis, D. et al., Proc Natl Acad Sci U S A, 106:18243-18248, 2009.