Machine-Learning Provides Patient-Specific Prediction of Metastatic Risk Based on Innovative, Mechanobiology Assay

Mechanobiological cell adaptations to changing microenvironments determine cancer invasiveness: Experimentally validated finite element modeling

Metastases are the leading cause of cancer-associated deaths. A key process in metastasis is cell invasiveness, which is driven and controlled by cancer cell interactions with their microenvironment. We have previously shown that invasive cancer cells forcefully push into and indent physiological stiffness gels to cell-scale depths, where the percentage of indenting cells and their attained depths provide clinically relevant predictions of tumor invasiveness and the potential metastatic risk. The cell-attained, invasive indentation depths are directly affected by gel-microenvironment mechanics, which can concurrently modulate the cells' mechanics and force application capacity, in a complex, coordinated mechanobiological response. As it is impossible to experimentally isolate the different contributions of cell and gel mechanics to cancer cell invasiveness, we perform finite element modeling with literature-based parameters. Under average-scale, cell cytoplasm and nucleus mechanics and cell-applied force levels, increasing gel stiffness 1-50 kPa significantly reduced the attained indentation depth by >200%, while the gel's Poisson ratio reduced depths only by up to 20% and only when the ratio was >0.4; this reveals microenvironment mechanics that can promote invasiveness. Experiments with varying-invasiveness cancer cells exhibited qualitative variations in their responses to gel stiffness increase, for example large/small reduction in indentation depth or increase and then reduction. We quantitatively and qualitatively reproduced the different experimental responses via coordinated changes in cell mechanics and applied force levels. Thus, the different cancer cell capacities to adapt their mechanobiology in response to mechanically changing microenvironments likely determine the varying cancer invasiveness and metastatic risk levels in patients.

Modeling force application configurations and morphologies required for cancer cell invasion

We show that cell-applied, normal mechanical stresses are required for cells to penetrate into soft substrates, matching experimental observations in invasive cancer cells, while in-plane traction forces alone reproduce observations in non-cancer/noninvasive cells. Mechanobiological interactions of cells with their microenvironment drive migration and cancer invasion. We have previously shown that invasive cancer cells forcefully and rapidly push into impenetrable, physiological stiffness gels and indent them to cell-scale depths (up to 10 μm); normal, noninvasive cells indent at most to 0.7 μm. Significantly indenting cells signpost increased cancer invasiveness and higher metastatic risk in vitro and in vivo, as verified experimentally in different cancer types, yet the underlying cell-applied, force magnitudes and configurations required to produce the cell-scale gel indentations have yet to be evaluated. Hence, we have developed finite element models of forces applied onto soft, impenetrable gels using experimental cell/gel morphologies, gel mechanics, and force magnitudes. We show that in-plane traction forces can only induce small-scale indentations in soft gels (< 0.7 μm), matching experiments with various single, normal cells. Addition of a normal force (on the scale of experimental traction forces) produced cell-scale indentations that matched observations in invasive cancer cells. We note that normal stresses (force and area) determine the indentation depth, while contact area size and morphology have a minor effect, explaining the origin of experimentally observed cell morphologies. We have thus revealed controlling features facilitating invasive indentations by single cancer cells, which will allow application of our model to complex problems, such as multicellular systems.