Balance between cell-substrate adhesion and myosin contraction determines the frequency of motility initiation in fish keratocytes

Abstract

Cells are dynamic systems capable of spontaneously switching among stable states. One striking example of this is spontaneous symmetry breaking and motility initiation in fish epithelial keratocytes. Although the biochemical and mechanical mechanisms that control steady-state migration in these cells have been well characterized, the mechanisms underlying symmetry breaking are less well understood. In this work, we have combined experimental manipulations of cell-substrate adhesion strength and myosin activity, traction force measurements, and mathematical modeling to develop a comprehensive mechanical model for symmetry breaking and motility initiation in fish epithelial keratocytes. Our results suggest that stochastic fluctuations in adhesion strength and myosin localization drive actin network flow rates in the prospective cell rear above a critical threshold. Above this threshold, high actin flow rates induce a nonlinear switch in adhesion strength, locally switching adhesions from gripping to slipping and further accelerating actin flow in the prospective cell rear, resulting in rear retraction and motility initiation. We further show, both experimentally and with model simulations, that the global levels of adhesion strength and myosin activity control the stability of the stationary state: The frequency of symmetry breaking decreases with increasing adhesion strength and increases with increasing myosin contraction. Thus, the relative strengths of two opposing mechanical forces--contractility and cell-substrate adhesion--determine the likelihood of spontaneous symmetry breaking and motility initiation.

Keywords: adhesion; cell migration; myosin; symmetry breaking.

Fig. 1.

Traction force, actin flow, adhesion, and myosin distributions in motile and stationary cells. (A and B) Traction stress maps (A) and actin flow maps (B) for motile (Top) and stationary (Bottom) keratocytes. Arrows indicate the direction and magnitude of the traction forces (A) and actin network movement with respect the underlying substrate (B); hot colors indicate faster flow. (C−E) Images of motile (Top) and stationary (Bottom) keratocytes immunolabeled for the adhesion protein vinculin (C, total internal reflection fluorescence images) or myosin (D, epifluorescence images) or labeled for actin with fluorescent phalloidin (E, epifluorescence images). (F−I) Average traction force (F), actin flow (G), vinculin (H), and myosin (I) distributions in motile (black line) and stationary (red line) cells are plotted versus cell boundary position. Fluorescent intensities were normalized for each cell by subtracting the mean intensity and dividing by the SD. Error bars are SEM. The images of motile cells in A−E are oriented with the leading edge pointed toward the top of the page.

Fig. 2.

Simulated traction force, actin flow, adhesion, and myosin distributions in response to transient asymmetries in adhesion strength or myosin density. (A) Model simulations incorporate two feedback loops: positive feedback between actin network flow and myosin density and negative feedback between actin flow and adhesion strength. (B) Negative feedback between actin flow and adhesion strength is modeled as an actin flow-dependent, nonlinear switch in adhesion strength. Adhesion strength decreases from ζ₀ to ζ₁ when actin flow rates exceed a critical threshold, u*. (C and D) Simulated adhesion, myosin, actin flow, and traction force distributions after transient perturbation of adhesion (C) or myosin (D). The details of the model simulations are described in SI Text. In brief, the model simulations were initiated with a circular cell with a fixed boundary and uniform adhesion (ζ = ζ₀) and myosin densities. The cell boundary was fixed for the first 2 min of the simulation to allow myosin densities and actin flow patterns to equilibrate. Then, the strength of adhesion was reduced (C) or the myosin density was increased (D) on one side of the cell (t = 0 s). Adhesion, myosin, and actin flow patterns were then allowed to evolve according the rules governing feedback among actin flow, adhesion strength, and myosin density for 1 min. The external adhesion and myosin asymmetries were removed, and the cell boundary was released and allowed to move according to the simulated actin flow patterns (t = 60 s). Transient perturbation of either adhesion strength or myosin density resulted in persistent, rapid movement of fan-shaped cells (t = 350 s).

Fig. 3.

Traction forces decrease in the prospective rear before motility initiation. (A and C) Simulated (A) and experimental (C) traction force maps; arrows indicate direction and magnitude of traction forces. (B and D) Traction force measurements from six simulations (B) and four real cells (D). Traction forces in the prospective cell rear and sides are plotted for the start of the simulation or the first imaging frame (S) and 5 min and 1 min before the onset of stable motility (−5 and −1, respectively). For the simulations, stochastic fluctuations were added to actin flow in the dynamic equations for myosin localization (blue lines), adhesion strength (orange lines), or both (black lines).

Fig. 4.

Local inhibition of cell−substrate adhesion can drive motility initiation. (A and B) Soluble RGD peptides were applied near stationary cells using a micropipette. The cell in B was pretreated with 25 μM blebbistatin for 20 min before RGD application. The cells broke symmetry after RGD application (at t = 0 s) and migrated away from the microneedle; both cells maintained polarization and continued to migrate after the RGD peptides were removed (at 500 s in A and 275 s in B). Red pseudocolor indicates the flow of the RGD peptides away from the pipette. (C) The fraction of stationary cells that remained stationary (red), broke symmetry and then lost polarity (yellow), or initiated stable motility (green) after application of Texas Red (TR) dextran alone or TR dextran plus RGD peptides.

Fig. 5.

The balance between myosin contraction and adhesion strength determines the frequency of motility initiation. (A) The stability of the symmetric, stationary state was assessed experimentally by measuring the fraction of stationary cells that initiated motility following a temperature shift. (B) Numerical simulations of cell boundary dynamics were carried out at the indicated values for ζ₀ (adhesion friction coefficient at low actin flow) and M (myosin concentration). At low adhesion or high myosin strength, the symmetric state is unstable (green circles—cells initiate motility instantly in simulations due to very small fluctuations), whereas at high adhesion or low myosin strength, the symmetric state is stable (red crosses). Open green circles indicate the adhesion and myosin strengths at which symmetry breaking requires finite-amplitude fluctuations and takes tens of minutes.