Session R45: Focus Session: Physics of the Cytoskeleton II

Biological Physics Dissertation Award Talk - Self-organization in cytoskeletal mixtures: from synthetic cilia to flowing networks

Inspired by biological functions such as ciliary beating and cytoplasmic streaming, we have developed a highly tunable and robust model system from biological components that self-organizes to produce a broad range of far-from-equilibrium materials with remarkable emergent properties. Using only simple components - microtubules, kinesin motor clusters, and a depletion agent that bundles MTs -- we reproduced several essential biological functions, including cilia-like beating, the emergence of metachronal waves in bundle arrays, and internally generated flows in active cytoskeletal gels. The occurrence of these biomimetic functions as self-organized processes provides unique insight into the mechanisms that drive these processes in biology. Beyond these biomimetic behaviors, we have also used the same components to engineer novel active materials which have no biological analogues: active streaming 2D nematics, and finally self-propelled emulsion droplets. These observations exemplify how assemblages of animate microscopic objects exhibit highly sought-after collective and biomimetic properties, challenging us to develop a theoretical framework that would allow for a systematic engineering of their far-from-equilibrium material properties.   

Nonlinear force propagation, anisotropic stiffening and non-affine relaxation in a model cytoskeleton

Forces are generated heterogeneously in living cells and transmitted through cytoskeletal networks that respond highly non-linearly. Here, we carry out high-bandwidth passive microrheology on vimentin networks reconstituted in vitro, and observe the nonlinear mechanical response due to forces propagating from a local source applied by an optical tweezer. Since the applied force is constant, the gel becomes equilibrated and the fluctuation-dissipation theorem can be employed to deduce the viscoelasticity of the local environment from the thermal fluctuations of colloidal probes. Our experiments unequivocally demonstrate the anisotropic stiffening of the cytoskeletal network behind the applied force, with greater stiffening in the parallel direction. Quantitative agreement with an affine continuum model is obtained, but only for the response at certain frequency $\sim$ 10-1000 Hz which separates the high-frequency power law and low-frequency elastic behavior of the network. We argue that the failure of the model at lower frequencies is due to the presence of non-affinity, and observe that zero-frequency changes in particle separation can be fitted when an independently-measured, empirical nonaffinity factor is applied.   

Conformational phases of membrane bound cytoskeletal filaments

Membrane bound cytoskeletal filaments found in living cells are employed to carry out many types of activities including cellular division, rigidity and transport. When these biopolymers are bound to a membrane surface they may take on highly non-trivial conformations as compared to when they are not bound. This leads to the natural question; What are the important interactions which drive these polymers to particular conformations when they are bound to a surface? Assuming that there are binding domains along the polymer which follow a periodic helical structure set by the natural monomeric handedness, these bound conformations must arise from the interplay of the intrinsic monomeric helicity and membrane binding. To probe this question, we study a continuous model of an elastic filament with intrinsic helicity and map out the conformational phases of this filament for various mechanical and structural parameters in our model, such as elastic stiffness and intrinsic twist of the filament. Our model allows us to gain insight into the possible mechanisms which drive real biopolymers such as actin and tubulin in eukaryotes and their prokaryotic cousins MreB and FtsZ to take on their functional conformations within living cells.   

Athermal Fluctuations of Probe Particles in Active Cytoskeletal Networks

A reconstituted active cytoskeletal networks consisting of an actin filament network coupled to myosins (motor proteins) have been shown to display rich in dynamical and mechanical behaviors that is often in contrast to passive, equilibrium system. The motor proteins, which spontaneously generate forces, kept the active cytoskeletal network out of equilibrium. The athermal fluctuations observed in the network are linked to the active force generation process by motor proteins which give more relevant information including the interaction with the surrounding materials. In prior studies, only the second moment of the athermal fluctuations has been investigated while the full displacement distribution of the athermal fluctuations in active cytoskeleton recently is found to be far from Gauss when observed with video microrheology. Here, we investigated the nonequilibrium statistics and dynamics of the active network by analyzing the athermal fluctuations of different probe sizes embedded in the same active system. The model developed here is based on truncated L\'{e}vy statistics which is generally observed for the force generators whose impact decays as 1/$r^{\mathrm{2}}$.   

Spontaneous Motion in Hierarchically Assembled Active Cellular Materials

With exquisite precision and reproducibility, cells orchestrate the cooperative action of thousands of nanometer-sized molecular motors to carry out mechanical tasks at much larger length scales, such as cell motility, division and replication. Besides their biological importance, such inherently far-from-equilibrium processes are an inspiration for the development of soft materials with highly sought after biomimetic properties such as autonomous motility and self-healing. I will describe our exploration of such a class of biologically inspired soft active materials. Starting from extensile bundles comprised of microtubules and kinesin, we hierarchically assemble active analogs of polymeric gels, liquid crystals and emulsions. At high enough concentration, microtubule bundles form an active gel network capable of generating internally driven chaotic flows that enhance transport and fluid mixing. When confined to emulsion droplets, these 3D networks buckle onto the water-oil interface forming a dense thin film of bundles exhibiting cascades of collective buckling, fracture, and self-healing driven by internally generated stresses from the kinesin clusters. When compressed against surfaces, this active nematic cortex exerts traction stresses that propel the locomotion of the droplet. Taken together, these observations exemplify how assemblies of animate microscopic objects exhibit collective biomimetic properties that are fundamentally distinct from those found in materials assembled from inanimate building blocks. These assemblies, in turn, enable the generation of a new class of materials that exhibit macroscale flow phenomena emerging from nanoscale components. ~~   

Mechanical Models of Microtubule Bundle Collapse in Alzheimer's Disease

Amyloid-beta aggregates initiate Alzheimer's disease, and downstream trigger degradation of tau proteins that act as microtubule bundle stabilizers and mechanical spacers. Currently it is unclear which of tau cutting by proteases, tau phosphorylation, or tau aggregation are responsible for cytoskeleton degradation., We construct a percolation simulation of the microtubule bundle using~a molecular spring model for the taus and including depletion force attraction between microtubules and membrane/actin cytoskeletal surface tension. The simulation uses a fictive molecular dynamics~to model the motion of the individual microtubules within the bundle as a result of~random tau removal, and calculates the elastic modulus of the bundle as the tau concentration falls. We link the tau removal steps to kinetic tau steps in various models of tau degradation.   

Properties of intracellular transport: the role of cytoskeleton topology

The eukaryotic cytoskeleton is composed of polarized filaments forming a complex, intertwined network. Various motor proteins such as kinesins or myosins convert ATP into mechanical work and are able to walk processively or even diffuse along the cytoskeleton. Large organelles such as vesicles or mitochondria can randomly bind and unbind to one or several motors and their transport in the cell can be described as alternating phases of diffusion in the cytoplasm and phases of directed or diffusive transport along the cytoskeletal network. Intracellular transport has been the focus of extensive studies both experimentally and theoretically. However, the impact of the cytoskeleton network structure on transport properties, which is expected to be significant, is not fully understood. We develop a computational model of intracellular transport, and explore the impact of the cytoskeletal structure on transport properties. We show that transport can be enhanced even by diffusional motion along the cytoskeleton after memory effects due to cytoskeletal structure are taken into account. We also explore the influence of the network structure on the first passage time distributions for a cargo to reach the cell membrane after being exported from the nucleus and for transport from the membrane to the nucleus.   

Active Stresses Drive Random Fluctuations in the Cytoplasm of Cells

The cytoplasm of living cells is a highly dynamic environment with continuous intracellular motion that is essential for life. Some intracellular movements appear directional, and are clearly actively transported. However, most intracellular movement appears random in nature. These random movements have often been interpreted as Brownian motion, and have been used to infer cellular mechanics. Here we describe direct quantifications of the random intracellular motion by using sub-micron beads, and independent micromechanical measurements of the local cellular environment using optical tweezers. We demonstrate that the random intracellular motion is driven by active stress fluctuations in a nearly elastic cytoskeletal matrix. The combination of our two measurements allows us to quantify the frequency spectrum of the intracellular forces, and directly shows that non-thermal active stresses dominate thermal forces in the cellular interior at long time scales (t\textgreater 0.1s), which results in the random intracellular motion. By using the photoconvertible fluorescent protein Dendra2, we also show that the movement of very small particles ($\sim$ nm) are also accelerated by active fluctuations. These active-stress driven movements may be an essential part of rapid transport in life.   

Microrheology of highly crosslinked microtubule networks is dominated by force-induced crosslinker unbinding

We determine the viscoelastic responses of reconstituted networks of microtubules that have been strongly bonded by labile crosslinkers using a magnetic tweezers device to apply localized forces. At short time scales, the networks respond nonlinearly to applied force, with stiffening at small forces, followed by a reduction in the stiffening response at high forces, which we attribute to the force-induced unbinding of crosslinks. At long time scales, force-induced bond unbinding leads to local network rearrangement and significant bead creep. Interestingly, for rigidly crosslinked networks, the material retains its elastic modulus even under conditions of significant plastic flow, suggesting that crosslinker breakage is balanced by the formation of new bonds. To better understand this effect, we developed a finite element model of such a stiff filament network with crosslinkers obeying force-dependent Bell model unbinding dynamics. The coexistence of dissipation, due to bond breakage, and the elastic recovery of the network is possible because each filament has many crosslinkers. Recovery can occur as long as a sufficient number of the original crosslinkers are preserved under the loading period. When these remaining original crosslinkers are broken, plastic flow results.   

Microtubules contribute to maintain nucleus shape in epithelial cell monolayer

INTRODUCTION: Tissue strains can result in significant nuclear deformations and may regulate gene expression. However, the precise role of the cytoskeleton in regulating nuclear mechanics remains poorly understood. Here, we investigate the nuclear deformability of Madin-Darky canine kidney cells (MDCK) under various stretching conditions to clarify the role of the microtubules and actin network on the mechanical behavior of the nucleus. METHODS: A custom-built cell-stretching device allowing for real time imaging of MDCK nuclei was used. Cells were seeded on a silicone membrane coated with rat-tail collagen I. A nuclear stain, Hoechst-33342, was used to image nuclei during stretching. We exposed cells to a compressive and non-compressive stretching strain field of 25{\%}. Nocodazole and cytochalasin-D were used to depolymerize the microtubules and actin network. RESULTS: Nuclei in control cells stretched more along their minor axis than major axis with a deformation of 5{\%} and 2{\%} respectively. This anisotropy vanished completely in microtubule-deprived cells and these cells showed a very high nuclear deformability along the minor axis when exposed to a compressive stretching strain field. CONCLUSIONS: The microtubules drive the anisotropic deformability of MDCK nuclei in a monolayer and maintain nuclear shape when exposed to compressive strain. Such intrinsic mechanical behavior indicates that microtubules are essential to maintain nuclear shape and may prevent down regulation of gene expression.