Multi-scale Modeling to Predict Cancer Cell Mechanics and Migration based on Transcriptional State and Microenvironment

Introduction: Cancer is a disease characterized by increasing heterogeneity as it progresses. Single-cell RNA sequencing has shed light on the variety of states that cancer cells can adopt, but connecting transcriptional signatures to functional outcomes remains a major challenge. Here, we have taken a multi-scale modeling approach to connect heterogeneous cytoskeletal gene expression programs exhibited by breast cancer cells to biochemical signaling networks and extracellular matrix conditions that regulate the cellular mechanical state. This in turn is used to predict cell migration behavior.

Materials and

Methods: Our model, xgene-BoHyMin3D, comprises three key components: transcriptomic data, Boolean-Hybrid-Modular Model (BoHyM), and 3D stochastic cell simulations. We identified 119 signaling nodes that connect various extracellular and gene expression signals to key cytoskeletal proteins. The signaling network is solved using a Boolean Hybrid Modular (BoHyM) approach specifically developed to deal with the distinct timescales and complexities of biochemical signaling processes. In the initial step, we employ transcriptomic data to determine the initial states of key cytoskeletal signaling proteins within the cell. Next, the states of various cytoskeletal elements determined using BoHyM are mapped to specific input parameters of mechanistic physical models of cell mechanics. Based on this mapping, we predict the migration dynamics of cells in heterogeneous 3D matrix-based environments

Results, Conclusions, and Discussions: Results & Discussion
This integrated framework accurately predicted differences in cell migration behavior of MDA-MB-231 breast cancer cells based on single-cell transcriptional differences. Further development will offer a versatile and user-modifiable tool for investigating how both extracellular and intracellular signaling mechanisms regulate cellular cytoskeleton components which in turn influence cell-substrate interactions, force generation, invasion, and migration.
Conclusion
By integrating RNA sequence data with signaling and mechanical models, we aspire to identify target genes and proteins responsible for specific cancer cell migration modes. This approach promises to provide novel insight into intricate molecular signaling systems underlying cell phenotypic diversity associated with cell mechanical properties. Eventually, such a modeling framework may facilitate the prediction of metastatic potential and the identification of targeted treatments.

Acknowledgements (Optional): The research reported was supported by the National Cancer Institute of the National Institutes of Health under award numbers: U54CA285117 (SDSU) & U54CA285115 (UC San Diego).