Multiscale Sensing of the Physical Cellular Environment: Phase-field Modeling and Experiments

Abstract

Cells constantly interact with their physical environment by sensing and responding to mechanical and topographical cues. These cues span multiple scales, from subcellular interactions with the extracellular matrix to population-scale confinement in morphogenesis. Central to this process is the actin cytoskeleton, which serves as both a local force generator and a medium for signal integration and propagation. In this dissertation, I combine multiscale computational modeling with quantitative imaging to investigate how actin dynamics drive physical sensing from the scale of single protrusions to collective cellular behavior.

The actin cytoskeleton is the primary mechanism for generating forces that cause cell protrusions and guided migration. Using 12Z cells as a model of endometriosis, we examine how exposure to the biochemical signal estradiol alters actin organization and, consequently, cell morphology. High-resolution 3D imaging reveals that estradiol treatment increases protrusion size and disorder in actin dynamics, consistent with enhanced cellular invasiveness. These findings highlight how chemical signals modulate mechanical output through actin-based protrusions, reinforcing the role of actin as a key transducer of biochemical cues into physical motion.

At the subcellular scale, we use a 3D phase-field model to illustrate how cells exhibit unidirectional migration on asymmetric nanotopographies, with directionality controlled by actin polymerization rate and topographic scale. In this model, an asymmetric substrate alone can cause spontaneous polarization and reproduce the shape and guidance morphologies observed experimentally. These predictions align with a reanalysis of \textit{D. discoideum} experiments, revealing that guidance on subcellular sawteeth depends on both cell velocity and feature height. This agreement indicates that membrane deformation and local curvature sensing, driven by actin forces, are sufficient to bias migration in complex microenvironments.

To study how the actin cytoskeleton responds to chemical cues found in the extracellular matrix, we analyze epithelial cell migration on collagen-coated nanoridges. On nanoridges, collagen IV enhances actin alignment and cell elongation; however, actin guidance remains decoupled from the direction of migration. Therefore, additional mechanisms, such as focal adhesion dynamics, may contribute to directional sensing.

At larger scales, we develop a scalable 2D multicellular phase-field model incorporating excitable actin dynamics. This framework enables simulations of thousands of deformable cells on consumer hardware. The model spans a range of scales, allowing cell interaction, long-distance wave propagation, and information exchange. In this model, excitable intracellular mechanics, along with local physical interactions, can lead to emergent synchronization and local sensing of the shape of large-scale confinement.

Together, these findings suggest that actin serves as a mechanochemical interface for multiscale environmental sensing by driving local protrusions, integrating physical signals, and enabling collective coordination through excitable dynamics. Actin polymerization additionally acts as an upstream regulator of other cell functions, which opens avenues for future experimental studies on the effects of actin synchronization and pulsing on biophysical behavior. By linking cytoskeletal signaling to large-scale coordination, this work lays the foundation for identifying new physical mechanisms underlying biological processes that require coordination, such as metastasis and tissue morphogenesis.