Cross-correlated TIRF/AFM shows Self-assembled Synthetic Myosin Filaments are Asymmetric - Implications for Motile Filaments

Myosin-II’s rod-like tail drives filament assembly with a head arrangement that should generate equal and opposite contractile forces on actin–if one assumes that the filament is a symmetric bipole. Self-assembled myosin filaments are shown here to be asymmetric in physiological buffer based on cross-correlated images from both atomic force microscopy (AFM) and total internal reflection fluorescence (TIRF). Quantitative cross-correlation of these orthogonal methods produces structural information unavailable to either method alone in showing that fluorescence intensity along the filament length is proportional to height. This implies that myosin heads form a shell around the filament axis, consistent with F-actin binding. A motor density of ~50 - 100 heads/micron is further estimated but with an average of 32% more motors on one half of any given filament compared to the other, regardless of length. A purely entropic pyramidal lattice model is developed that qualitatively captures this lack of length dependence and the distribution of filament asymmetries. Such strongly asymmetric bipoles are likely to produce an imbalanced contractile force in cells and in actin-myosin gels, and thereby contribute to motility as well as cytoskeletal tension.

💡 Research Summary

Myosin‑II is a fundamental motor protein that generates contractile forces in both muscle and non‑muscle cells. Classical structural models depict myosin filaments as symmetric bipolar assemblies, with heads arranged at opposite ends of a rod‑like tail, thereby producing equal and opposite forces on actin filaments. However, the actual architecture of self‑assembled synthetic myosin filaments under physiological conditions had not been directly examined with sufficient spatial resolution. In this study, the authors combined atomic force microscopy (AFM) and total internal reflection fluorescence (TIRF) microscopy to obtain orthogonal, high‑resolution measurements of the same filaments. AFM provided nanometer‑scale topography (height) while TIRF delivered quantitative maps of fluorescently labeled myosin heads. By precisely aligning the two data sets and performing pixel‑by‑pixel cross‑correlation, they discovered a robust linear relationship between fluorescence intensity and filament height along the entire length of each filament. This proportionality indicates that myosin heads are not distributed uniformly throughout the filament cross‑section but rather form a thin shell surrounding the central axis, consistent with the known actin‑binding geometry of the heads.

Quantitative analysis of the fluorescence signal, calibrated against labeling efficiency, yielded an estimated motor density of roughly 50–100 heads per micron of filament length. More strikingly, when each filament was bisected longitudinally, the half containing more heads possessed on average 32 % more motors than the opposite half. This asymmetry was observed across a broad range of filament lengths (approximately 0.5–2 µm) and showed no statistically significant dependence on length, suggesting that the asymmetry is an intrinsic property of the assembly process rather than a size‑related artifact.

To rationalize these findings, the authors introduced a purely entropic “pyramidal lattice” model. In this conceptual framework, myosin tail domains occupy sites on a three‑dimensional lattice and, driven by entropy maximization, tend to form a pyramidal arrangement that naturally yields a non‑uniform distribution of heads along the filament axis. Monte‑Carlo simulations of the model reproduced the experimentally observed distribution of asymmetry values and the lack of length dependence, supporting the notion that stochastic, entropy‑driven packing can generate the observed structural bias.

The biological implications of asymmetric myosin filaments are profound. A filament that exerts unequal forces at its two ends will generate a net directional contractile bias when incorporated into an actin‑myosin network. In cellular contexts, such biased filaments could contribute to localized tension gradients, facilitating processes such as cell migration, shape changes, and tissue morphogenesis. In reconstituted actin‑myosin gels, the presence of strongly asymmetric bipoles may explain previously reported spontaneous flows and directional movements that cannot be accounted for by symmetric filament models. Moreover, the shell‑like arrangement of heads inferred from the AFM‑TIRF correlation aligns with the idea that myosin filaments interact with actin primarily through peripheral binding, potentially influencing filament‑filament cross‑linking and network elasticity.

In conclusion, by integrating AFM and TIRF imaging through cross‑correlation analysis, the authors provide the first quantitative evidence that self‑assembled synthetic myosin‑II filaments are structurally asymmetric, with a consistent ~32 % head density bias irrespective of filament length. The entropic pyramidal lattice model offers a plausible mechanistic explanation for this phenomenon. These insights challenge the prevailing symmetric bipolar paradigm and suggest that filament asymmetry is a key factor shaping contractile force distribution, cellular tension, and motile behavior in both living cells and biomimetic actomyosin systems. Future work should aim to visualize such asymmetry in native cellular filaments, explore its regulation by accessory proteins, and incorporate asymmetric filament mechanics into computational models of cytoskeletal dynamics.