Non-equilibrium microtubule fluctuations in a model cytoskeleton

Biological activity gives rise to non-equilibrium 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 step-like forces on the order of 10 pN driven by collective motor dynamics.

💡 Research Summary

This paper introduces a novel, direct method for quantifying non‑equilibrium fluctuations inside a cell‑like environment by exploiting the bending dynamics of embedded microtubules (MTs). The authors reconstitute a three‑dimensional actin network using purified actin filaments cross‑linked with scruin, then embed fluorescently labeled MTs at low density. To drive the network out of equilibrium, they add myosin II motor proteins and supply ATP, which generates continuous contractile stresses. High‑speed spinning‑disk confocal microscopy records MT shapes at sub‑10 ms intervals, allowing precise extraction of the filament centerline, curvature, and transverse displacement as a function of time and spatial wavelength.

In the passive (no‑motor) condition, the MT fluctuations obey the classic thermal equipartition result: the mean‑square amplitude ⟨|u(q)|²⟩ scales as kBT/(κq⁴) with κ≈2 × 10⁻²³ N·m², and the temporal autocorrelation decays exponentially with a characteristic relaxation time of ~0.2 s, reflecting the balance of filament bending rigidity and the viscous drag of the surrounding actin gel.

When myosin motors are active, the fluctuation spectrum changes dramatically. Low‑q (long‑wavelength) modes exhibit amplitudes that are five‑ to ten‑fold larger than thermal predictions, while high‑q modes also show excess power. The temporal autocorrelation no longer follows a simple exponential; instead it displays an initial rapid drop followed by a long power‑law tail, indicating that the MTs are being intermittently kicked by external forces rather than merely thermally agitated.

To interpret these observations, the authors construct a stochastic “step‑force” model. In this framework, discrete force events of magnitude F occur with a Poisson rate λ, each event applying a sudden transverse load to the MT. Fitting the model to the experimental spectra yields F≈10 pN and λ≈0.3 s⁻¹. The 10 pN magnitude exceeds the stall force of a single myosin II dimer (~2 pN), implying that multiple motors cooperate to generate localized stress spikes. The rate λ corresponds to roughly one kick every three seconds per filament segment, consistent with the observed temporal correlations.

The mechanical properties of the actin gel further shape the response. Rheological measurements reveal a storage modulus G′≈50 Pa and a loss modulus G″≈20 Pa, indicating a viscoelastic medium that stiffens under strain (strain‑hardening). This non‑linear background modifies the effective drag on the MT and contributes to the non‑exponential decay of the autocorrelation function. The combination of a viscoelastic matrix and intermittent motor‑driven forces produces the observed anomalous scaling of both spatial and temporal fluctuation spectra.

Overall, the study demonstrates that MT bending fluctuations serve as a highly sensitive, nanoscale probe of local stress fluctuations in a non‑equilibrium cytoskeletal network. By quantifying the amplitude and timing of motor‑generated force steps, the authors provide direct evidence that collective myosin activity creates ~10 pN, step‑like forces that continuously perturb the network. This insight bridges the gap between molecular motor energetics and macroscopic cellular mechanics, suggesting that cells constantly experience a background of active “noise” that can influence processes such as intracellular transport, mechanosensing, and shape remodeling.

Finally, the authors discuss the broader implications and future directions. The methodology can be extended to living cells, where MTs naturally polymerize and depolymerize, and where actin networks are more heterogeneous and dynamically regulated. With advances in live‑cell imaging and automated filament tracking, it should become possible to map spatiotemporal stress landscapes in real time, offering a powerful tool to study disease‑related mechanical dysregulation (e.g., cancer metastasis, neurodegeneration) and to test theoretical models of active matter in biologically relevant settings.