Session P48: Motion and Jamming of Cells

The influence of cellular rosettes on epithelial tissue mechanics and the jamming transition.

Models for densely packed and confluent biological tissues often describe the network formed cells as a triple-junction network, similar to foams. However, higher order vertices or cellular Rosettes are prevalent in biology and is crucial in many important aspects of development, disease and physiology. In this work, we study the influence of cellular rosettes on the mechanics of a confluent tissue. We find that the existence of rosettes in a tissue can act to rigidify it. We characterize a new fluid-to-solid transition that is purely driven by the density of rosettes in a tissue. We propose a universal framework for the rigidity of a confluent tissue.   

Hydrodynamics of shape-driven rigidity transitions in motile confluent tissues

In biological tissues, it is now well-understood that mechanical cues are a powerful mechanism for pattern regulation. While much work has focused on interactions between cells and external substrates, recent experiments suggest that cell polarization and motility might be governed by the internal shear stiffness of nearby tissue, deemed ``plithotaxis''. Meanwhile, other work has demonstrated that there is a direct relationship between cell shapes and tissue shear modulus in confluent tissues. Joining these two ideas, we develop a hydrodynamic model that couples cell shape, and therefore tissue stiffness, to collective cell motility. Linear stability analysis indicates that the formation of aster-like and banding patterns in these tissues is controlled by a composite ``morphotaxis'' parameter which encapsulates the influence of inhomogeneities in cell shape on collective cell migration and vice versa.   

The Energy Landscape of 2D Cellular Matter: A Statistical Description

Cellular matter, consisting of space-filling domains, comprises foams, emulsions, and many granular and confluent tissue systems. The ground state, and in general the energy landscape, of cellular matter is difficult to determine, consisting of many metastable states separated by energy barriers. Using foam as a prototype, we show that the energy landscape of a large class of 2D cellular systems can be mapped using measures of geometric and topological statistics only. In the limit of monodisperse foam, the system energy is linearly correlated with the defect density. In polydisperse foams, the defect density and the cross-correlation between size and topology are necessary to reliably predict the energy. Surprisingly, metastable states of a larger class of systems with energy functionals more complex than that of foam are found to lie on a quantitatively equivalent energy surface. Thus, statistical information, obtained from visual information, is sufficient to evaluate the relative proximity of the cellular structure to its ground state. This work suggests that information on the mechanical state of cellular matter can be obtained by its morphology only, which makes statistics an important diagnostic tool in the study of engineering materials as well as of biological tissues.   

Role of morphological differences in cell sorting

Cell sorting (the spatial segregation of different cell types in a tissue or co-culture) is thought to play a vital role in development and disease. Although sorting in particle-model simulations is well studied, recent discoveries suggest that in confluent tissues (with no gaps or overlaps between cells), both fluid-solid transitions and surface tension are strongly influenced by cell shapes and topologies. Hence, we seek to analyze the behavior of cell sorting in a confluent model of tissues across the fluid-solid transition, and compare to experimental data for a 2D co-culture of breast carcinoma and non-malignant cell lines, which display robust sorting. Using a 2D Self-Propelled Voronoi (SPV) model, we investigate the role of disparity in cell shape and size in cell sorting for bidisperse mixtures. Quite surprisingly, we find very minimal segregation when cell shapes differ, but robust sorting when cell areas differ. This is consistent with experimental results, where carcinoma cells that are larger, more elongated and more motile robustly sort from non-cancerous cells in co-cultures. We will conclude by discussing possible mechanisms for size-based segregation.   

Relaxation of self-induced jamming in confined budding yeast population

In nature, fast growing cells like microbes or cancer cells inevitably experience spatial confinement either due to living in a limited in space and/or because their population reaches high densities.
When grown in deformable micro-chambers that mimic these conditions, budding yeast cells displays features of jammed systems, as they build up large mechanical pressures resulting of heterogeneous forces inside the cell population. Such high pressure are sufficient to impact the population’s growth rate and eventually stall the cells, as individuals do not commit to division anymore (blocked in G1). We will present evidence that this phenotype is reversible if confinement is suddenly released and look at the relaxation process in terms of physical and physiological observables.   

Dense packing of cell monolayers: Jamming of deformable polygons

Collective motion in dense packings of cells occurs in wound healing,
embryonic development, and cancerous tumor growth. Most current
computational models of dense cell packings either treat the system as
collections of spherical particles or assume that the system is
confluent, with no extracellular space. We have developed a new model
for dense cell packings in two spatial dimensions, where the cells are
modeled as deformable particles that have a preferred area and
perimeter. We measure the packing fraction, φ_j at jamming onset
as a function of the asphericity, α, which is the ratio of the
perimeter square to the area of the particle. We find that the jammed
packing fraction increases monotonically with α and that the system
becomes confluent with φ_j = 1 for α > 1.16. Using surface
Voronoi analysis, we show that this value for α corresponds to the
case when the cells completely fill their Voronoi-tesselated
regions. We also demonstrate that the free area per cell obeys a
k-gamma distribution, which has been found for jammed packings of
non-deformable particles. Finally, we will describe results
from our model concerning the mobility of deformable particles
subjected to applied forces, as well as diffusion of deformable
particles subjected to active forces to discuss the effect of geometry in active jamming of deformable particles.   

Physical modeling of cancer and immune cells in melanoma tumors

The emergence of two-photon imaging in live tissue allows us to study cell dynamics in three spatial dimensions over time. I will present quantitative image analysis of tumor and immune cells in live murine melanoma tumors. I will discuss how the structural and mechanical properties of cells impact the development and spatiotemporal evolution of melanoma tumors. The in vivo measurements will be compared to computational models of static packings of deformable polyhedra. In the simulations, we will study the cell shape, packing fraction, and mobility in response to applied forces as a function of the cell deformability. My talk will describe the advantages of physical modeling in understanding tumor formation and how the physical models can advance cancer diagnosis and treatment.
  

Anomalous glassy dynamics in simple models of biological tissue

The analogy between colloidal glasses and epithelial tissues provides a powerful framework for
understanding the interplay between rigidity and motility in biological tissues. Here we present a detailed study of the role of fluctuations on the disordered rigidity transition of a Voronoi model that describes a confluent tissue monolayer as a network of polygons determined by a Voronoi tessellation of the cell positions. We show that just as in particulate glasses, the dynamics slow as the system approaches the glass transition. However, in contrast to the super-Arrhenius behavior associated with growing energy barriers seen in particulate glasses, the Voronoi model displays sub-Arrhenius behavior. Moreover, the low-frequency spectrum indicates that the potential energy landscape becomes flatter at lower temperatures, with little evidence for quasi-localized excitations, which
is also in marked contrast to structural glasses. This suggests that glassy behavior in Voronoi models, and perhaps biological tissues, is distinct from that in many glasses, and may provide unique insight into the glass transition itself.   

Bacterial Colony Spreading via Surface Tension Driven Flows

Dispersal and range expansion is critical to the survival of bacteria. Bacterial colonies grown on surfaces have developed various strategies to spread, such as swarming, gliding or twitching. Here we report a novel spreading behavior in Pseudomonas aeruginosa that does not rely on single-cell motility but requires surface-tension driven flows maintained by cooperative synthesis of bio-surfactants. We combined fluorescence microscopy imaging, genetic engineering, and computational modeling to reveal the biophysical mechanism underlying this behavior. Our findings shed new light on the cooperative behavior and long-range transport of bacteria.   

A Neuronal Jam in a Primitive Brain

The neuronal packing in the brain does not conform to the norm of jamming, which typically includes a single type of particle. This is because the brain comprises a mix of different types of neurons, and same-type neurons avoid contacts between one another. This unique packing and its property remain to be explored. We simulate neuronal packing with particles that are colored differently but otherwise identical, with concolor particles repelling each other. We show that two levels of jamming can be reached simultaneously: one for all particles (“global jamming”) and another for each set of concolor particles (“sub-jamming”). Reaching this double-jamming requires a minimum number of colors, and that number increases with the range of repulsion. Experimentally, we characterize the packing of peptidergic neurons in the planarian, one of the most basal animals known to have a brain. We found that homotypic neurons pack independently of other neuronal types around the jamming point. Consistent with the simulation, this packing is driven by short-range, homotypic repulsion, mediated by conserved cell adhesion molecules. This study opens a new line of inquiry: how the two levels of jamming are related, and how the configurational space of global jamming projects onto sub-jammed states.