Tension Anisotropy Induces Mechanical Memory and Long-term Activation of Fibroblasts

Introduction: : Mechanical factors such as stress in the extracellular matrix (ECM) affect the phenotypic commitment of cells. Stress fields experienced by cells in tissues are multiaxial, but how cells integrate such information is largely unknown. Here, we demonstrate that the directional imbalance of mechanical tension, referred to as tension anisotropy, is a critical factor triggering a phenotypic transition in fibroblast cells, outweighing the role of stress amplitude and ECM stiffness, factors previously described to modulate such a transition [1].

Combining experimental and computational approaches, we identified a self-reinforcing mechanism in which cellular protrusions interact with collagen fibers to establish tension anisotropy. This anisotropy, in turn, stabilizes the protrusions and enhances their contractile forces. Disruption of this self-reinforcing process, either by reducing tension anisotropy or by inhibiting contractile protrusions, prevents the phenotypic conversion of fibroblasts to contractile myofibroblasts.

Building on these findings, we further investigated whether fibroblast activation induced by tension anisotropy is transient or sustained. By systematically varying the magnitude and duration of anisotropic tension in 3D collagen matrices, we found that prolonged mechanical stimulation induces a long-term cellular response. Even after tension was removed, fibroblasts maintained high levels of contractility and activation markers, revealing the presence of mechanical memory [2]. These results suggest that not only does tension anisotropy serve as a potent activator of fibroblast phenotype, but it also leads to the induction of cellular memory and long-term cell activation.

Materials and

Methods: : Human dermal fibroblasts and NIH/3T3 mouse fibroblasts were used to investigate the effects of tension anisotropy on fibroblast activation.

To assess the effect of tension anisotropy at the cellular scale, fibroblasts were cultured on micropatterned substrates of tunable stiffness with either circular or rectangular geometries (Figure 1). This setup enabled the independent evaluation of tension magnitude (by tuning substrate stiffness) and tension anisotropy (by changing cell geometry) effects on fibroblast activation. F-actin and phosphorylated myosin light chain (pMLC) staining were used to quantify contractility.

To assess the role of tension anisotropy at the tissue level, cruciform-shaped fibroblast-collagen tissues were created (Figure 2). This geometry naturally produced spatial gradients of tension anisotropy: Cells near the center and arm ends experienced more isotropic loading, while those in the mid-arm regions experienced uniaxial, anisotropic tension. To investigate the cytoskeletal basis of anisotropy-induced activation, we repeated the experiments with fibroblasts pretreated with nocodazole, a microtubule destabilizer.

To examine whether anisotropy-induced activation is sustained, fibroblast-collagen tissues were subjected to anisotropic tension using a custom stretching device at two stretching levels and for either 1 or 24 hours (Figure 3). Tissues were then relaxed for one day and returned to reference configuration to assess long-term cell activation.

A coupled bio-chemo-mechanical model was developed and calibrated to simulate the feedback between ECM mechanics and cell-generated forces. The model incorporated the effect of tension magnitude and tension anisotropy on cell contractility to predict changes in actomyosin activity over time (Figure 1-3).

Results, Conclusions, and Discussions:: At the single-cell level, fibroblasts cultured on micropatterned substrates exhibited increased contractility with increasing substrate stiffness, indicated by elevated F-actin and pMLC levels (Figure 1). This confirmed that tension magnitude promotes cytoskeletal activation. However, at all stiffness levels, rectangular cells, which were subjected to anisotropic tension, were significantly more activated than circular cells, which experienced isotropic tension. More importantly, anisotropy even outweighed stiffness: rectangular cells on soft ECMs displayed higher activation level than circular cells on stiff ECMs. These observations were quantitatively consistent with our model predictions, confirming that tension anisotropy is a dominant mechanical regulator of fibroblast activation.

At the tissue level, fibroblasts in regions of anisotropic tension (mid-arm areas of cruciform constructs) exhibited markedly higher F-actin and pMLC intensity compared to cells in isotropic regions (near the center and arm ends) (Figure 2a,b). These results confirmed that tension anisotropy regulates fibroblast activation at the tissue level. This tension-driven pattern was disrupted when fibroblasts were pretreated with nocodazole to inhibit microtubule polymerization; in these tissues, activation was reversed, with cells in isotropic regions showing higher contractility. The same inversion was captured in our model by removing the effect of tension anisotropy, supporting the role of microtubule-guided protrusions in anisotropy sensing (Figure 2c,d).

To assess the persistence of activation, tissues were subjected to anisotropic tension at two stretching levels and two stretching durations (1 or 24 hours) (Figure 3). Fibroblasts in tissues stretched for 24 hours retained significantly higher actin intensity after one day of relaxation compared to those stretched for only 1 hour (Figure 3). This revealed a form of mechanical memory, whereby prolonged anisotropic loading led to sustained fibroblast activation even after tension was removed.

Together, these findings identify two key mechanobiological mechanisms: (1) anisotropic tension enhances fibroblast activation, (2) prolonged anisotropic loading induces a mechanical memory in fibroblasts, maintaining high contractility after unloading. Through integrated experimental and modeling approaches, we establish tension anisotropy as a dominant mechanical factor capable of encoding long-term cellular responses.

Acknowledgements and/or References (Optional):: Alisafaei, F., Shakiba, D., Hong, Y., Ramahdita, G., Huang, Y., Iannucci, L. E., Davidson, M. D., Jafari, M., Qian, J., Qu, C., Ju, D., Flory, D. R., Huang, Y.-Y., Gupta, P., Jiang, S., Mujahid, A., Singamaneni, S., Pryse, K. M., Chao, P. G., Burdick, J. A., Lake, S. P., Elson, E. L., Huebsch, N., Shenoy, V. B. & Genin, G. M. Tension anisotropy drives fibroblast phenotypic transition by self-reinforcing cell–extracellular matrix mechanical feedback. Nat. Mater. 1–11 (2025) doi:10.1038/s41563-025-02162-5.

2. Hong, Y., Peng, X., Yu, H., Jafari, M., Shakiba, D., Huang, Y., Qu, C., Melika, E. E., Tawadros, A. K., Mujahid, A., Huang, Y.-Y., Sandler, J. A., Pryse, K. M., Sacks, J. M., Elson, E. L., Genin, G. M. & Alisafaei, F. Cell–matrix feedback controls stretch-induced cellular memory and fibroblast activation. Proceedings of the National Academy of Sciences 122, e2322762122 (2025).