3D Visualization of compressed actin filaments simulated in ReaDDy 1 and Cytosim 2. Filaments are aligned at the final time step of the simulation. ReaDDy filaments show notable directional biphasic out-of-plane behavior, indicative of filament twist only captured with monomer-scale resolution. Cytosim filaments showed very little out-of-plane behavior, and when they did, it was just as likely to be in either direction.
The dynamic bending and twisting of actin filaments mechanically drive many processes in cells.
Fundamental cellular processes such as endocytosis, cell motility, and cytokinesis are reliant on a cell's ability to produce force. The actin cytoskeleton plays a central role in force production in these processes.
Actin filaments bend and twist in response to mechanical forces, storing elastic energy. Bent actin filaments are found in diverse cellular structures, where this elastic energy can be used to do mechanical work.
The helical structure of actin filaments results in the coupling of filament twisting and bending, which impacts the geometry of the actin network in 3D space. This influences how actin networks respond to and participate in subcellular force production.
Diagram of a bending and twisting actin filament at a site of endocytosis.
Fluorescence micrograph of an actin filament (magenta) attached to a paramagnetic bead (cyan).
JP Bibeau, NG Pandit, S Gray, N Shatery Nejad, CV Sindelar, W Cao, EM De La Cruz. Twist response of actin filaments (Figure 2A). Proceedings of the National Academy of Sciences, 120(4):e2208536120, 2023. DOI: 10.1073/pnas.2208536120
Cryo-electron micrograph of purified actin filaments with curvature color coded in blue.
MJ Reynolds, C Hachicho, AG Carl, R Gong, GM Alushin. Bending forces and nucleotide state jointly regulate F-actin structure (Figure 3A). Nature, 611:380-386, 2022. DOI: 10.1038/s41586-022-05366-w
Segmented actin filaments from cryo-electron tomogram of Ptk2 cells infected with Listeria Monocytogenes.
M Jasnin, S Asano, E Gouin, R Hegerl, JM Plitzko, E Villa, P Cossart, W Baumeister. Three-dimensional architecture of actin filaments in Listeria monocytogenes comet tails (Figure S4G). Proceedings of the National Academy of Sciences, 110(51):20521-20526, 2013. DOI: 10.1073/pnas.1320155110
Segmented actin filaments from cryo-electron tomogram of a podosome in a human white blood cell.
M Jasnin, J Hervy, S Balor, A Bouissou, A Proag, R Voituriez, J Schneider, T Mangeat, I Maridonneau-Parini, W Baumeister, S Dmitrieff, R Poincloux. Elasticity of podosome actin networks produces nanonewton protrusive forces (Figure 1C). Nature Communications, 13:3842, 2022. DOI: 10.1038/s41467-022-30652-6
Segmented actin filaments and cell membrane from cryo-electron tomogram of an endocytic site in a human skin melanoma cell.
D Serwas, M Akamatsu, A Moayed, K Vegesna, R Vasan, JM Hill, J Schöneberg, KM Davies, P Rangamani, DG Drubin. Mechanistic insights into actin force generation during vesicle formation from cryo-electron tomography (Figure 2B). Developmental Cell, 57(9):1132-1145, 2022. DOI: 10.1016/j.devcel.2022.04.012
We developed two models of actin to compare twisting and bending at different spatiotemporal scales.
Many different simulation methods have been developed to model actin. Each simulation method has limitations depending on the spatial scale.
ReaDDy Actin filaments are composed of particles, one for each actin monomer, which are held together by potentials. (ReaDDy actin model code).
Cytosim actin filaments are represented by control points that define a mathematical line, which is acted upon by physical forces. (our Cytosim code version).
Our simulation, analysis, and visualization pipeline code is available here and simulation data is available here.
Diagram of actin filaments modeled at monomer-scale, in which the interactions of each monomer of the filament with its neighbors is explicitly simulated, and fiber-scale, in which the filament is modeled as a series of points connected by simple length and angle-limited springs that have no rotational constraints.
We simulated the compression of these filaments to 350 nm across a range of velocities.
We wanted to know how each simulator captured structural properties of actin in different situations. We simulated compression at five velocities, spanning the range of velocities that were computationally feasible and physiologically relevant.
Use the matrix to explore results from different simulators at different compression velocities.
Simulation replicates
| Compression velocity (µm/s) | |||||
|---|---|---|---|---|---|
| 0 | 4.7 | 15 | 47 | 150 | |
| ReaDDy | |||||
| Cytosim | |||||
Our quantitative analysis revealed a divergence particularly in terms of non-coplanarity.
ReaDDy actin filaments showed high non-coplanarity in one direction, especially in filaments simulated at higher compression velocities. Cytosim filaments showed very little out-of-plane behavior at any compression velocities, and when they did, it was just as likely to be in either direction.
Explore trajectories by metrics and in PC space using the interactive plots.
Trajectories of filaments in PC space for each simulator, velocity, and replicate. The start or end of each trajectory is marked by a diamond or circle, respectively.
Feature
Metrics for each trajectory over time for each simulator, velocity, and replicate.
Compression velocity
Compression metric
Monomer-scale simulations captured characteristics of helical structure more effectively than fiber-scale simulations.
Inverse PCA transforms of the filament shapes confirmed that PC1 captured filament compression and peak asymmetry, while PC2 scaled with filament non-coplanarity and twist. Inverse transforms also revealed that low values of PC2 correspond to twist in the opposite direction. Some Cytosim simulations show this behavior, likely because directional helicity is not enforced in Cytosim.
Use the slider to explore inverse transforms for values of each PC.
Latent walks for filament shape along PC 1 and PC 2.