Nonequilibrium Microtubule Fluctuations in a Model Cytoskeleton

Biological activity gives rise to nonequilibrium fluctuations in the cytoplasm of cells; however, there are few methods to directly measure these fluctuations. Using a reconstituted actin cytoskeleton, we show that the bending dynamics of embedded microtubules can be used to probe local stress fluctuations. We add myosin motors that drive the network out of equilibrium, resulting in an increased amplitude and modified time dependence of microtubule bending fluctuations. We show that this behavior results from steplike forces on the order of 10 pN driven by collective motor dynamics.

Figure 1

(a) Fluorescently labeled intracellular MTs showing highly bent shapes. The inset is a schematic showing a MT bending under the action of myosin motors pulling on the actin network. (b) Localized MT bending fluctuations in the motor-driven in vitro network, time between frames is 78 msec.

Figure 2

(a) For thermal fluctuations, $⟨\Delta a_q^2{⟩}_{\max }\sim 1/q^2$, with $l_p\sim 1\mathrm{mm}$ (solid line). (b) $\alpha$, for thermal fluctuations, as a function of $\Delta t$ and $n$. Upper inset shows a higher magnification cut along $\Delta t=0.4\sec$. (c) Motor proteins ($1:100$) lead to enhanced fluctuations on short length scales. (d) $\alpha$, for motor-driven fluctuations. Example filaments in (b) and (d) have $L\sim 30\mathrm{mm}$. Data in (a) and (c) are averages over 10–20 filaments.

Figure 3

Subdiffusive dynamics, $\sim \Delta t^{0.6}$, for thermally fluctuating MTs, solid symbols. Squares, circles, triangles, inverted triangles, diamonds, pentagons, correspond to $q=0.097$, 0.194, 0.291, 0.388, 0.484, $0.58\mu {\mathrm{m}}^{-1}$, respectively. At high myosin concentration, the fluctuations evolve diffusively, $\sim \Delta t$, open symbols. Squares, circles, triangles, inverted triangles, diamonds, pentagons, correspond to $q=0.18498$, 0.369 96, 0.554 94, 0.739 92, 0.924 91, $1.10989\mu {\mathrm{m}}^{-1}$, respectively. Lower inset shows the transition to diffusive time dependence with increasing myosin concentration; $q\sim 0.29\mu {\mathrm{m}}^{-1}$, squares, circles, triangles, pentagons correspond to no myosin, $1:200$, $1:100$, and $1:50$ (molar ratio, myosin:actin).

Figure 4

(a) Shape of a section of the MT shown in Fig. 1b, along with fits to the predicted shape. (b) Distribution of $f^{\max }$ obtained for various myosin concentrations (squares: $1:50$; circles: $1:100$; triangles: $1:200$). (c) Examples of $y_0(t)$.