Jamming and arrest of cell motion in biological tissues
JJamming and arrest of cell motion in biological tissues
Elizabeth Lawson-Keister and M. Lisa Manning ∗ Department of Physics and BioInspired Institute,Syracuse University, Syracuse, NY 13244, USA
Collective cell motility is crucial to many biological processes including morphogenesis, woundhealing, and cancer invasion. Recently, the biology and biophysics communities have begun to usethe term “cell jamming” to describe the collective arrest of cell motion in tissues. Although this termis widely used, the underlying mechanisms are varied. In this review, we highlight three independentmechanisms that can potentially drive arrest of cell motion – crowding, tension-driven rigidity, andreduction of fluctuations – and propose a speculative phase diagram that includes all three. Sincemultiple mechanisms may be operating simultaneously, this emphasizes that experiments shouldstrive to identify which mechanism dominates in a given situation. We also discuss how specificcell-scale and molecular-scale biological processes, such as cell-cell and cell-substrate interactions,control aspects of these underlying physical mechanisms.
I. INTRODUCTION
Cell motility drives many biological processes, including morphogenesis and wound healing, andits disregulation is implicated in diseases like cancer. In the past, cell motility research has oftenfocused on the behavior of single cells, such as fibroblasts, in different environments [48, 62].Recently, there has been a growing understanding that in dense tissues cell motion can be drivenby collective effects, i.e. by cell-cell interactions instead of cell-autonomous properties. One par-ticular area of focus has been "cellular jamming", a term that researchers in biology and relatedfields have adopted to describe the collective arrest of cell motion in dense tissues [2, 24, 30, 46, 52].One reason the concept is useful is because it suggests new, non-cell autonomous mechanisms canalter cell motion, potentially identifying new targets for therapies for disease. For example, recentwork emphasizes that non-cell autonomous processes, such as cell-cell adhesion [30, 42, 47] andstress fluctuations driven by nearby cell division [13, 45, 51], impact cell motility and structuralrearrangements in dense tissues.While it is exciting that a widening group of researchers are studying the collective arrest of cellmotion, a challenge is that the term "cell jamming" is being used as a broad umbrella descriptionof such processes. Since there are multiple distinct mechanisms that can drive collective cell arrest,and the term "cell jamming" has been used to describe all of them, it often remains unclear whichmechanisms are actually operating in a given process. Therefore, the focus of this review is todescribe several distinct mechanisms for collective cell arrest, and highlight ideas for how one mightconfirm a given mechanism is operating in a given situation.To build intuition about mechanisms for collective arrest in cells, we turn to the physical sciences,where the collective arrest of particle or molecular motion is termed "solidification" or "rigidifica-tion". In introductory physical science classes, one learns that a material can be solidified by cooling– i.e. reducing fluctuations induced by temperature – or by increasing pressure – i.e. packing the ∗ Corresponding author: Manning, M. Lisa ([email protected]) a r X i v : . [ phy s i c s . b i o - ph ] F e b particles, atoms, or molecules closer together. When a material is cooled into a solid while re-maining disordered, it undergoes a glass transition. And in the physical sciences and engineering,jamming is a technical term reserved to describe the onset of solidification at zero temperature,driven specifically by changes to pressure or density. Researchers have developed a "jamming phasediagram" to unite various mechanisms that are responsible for solidification [34, 60].In this review, we take our cue from the physical sciences and adapt recent results from theliterature to construct a jamming phase diagram for cell collectives, focused on three mechanisms:crowding, active fluctuations, and tension-driven rigidity. We are not the first to conjecture such adiagram; several other phase diagrams have been proposed previously [8, 46]. Nevertheless, recentwork over the past two years has generated new explicit predictions for the onset of cell arrest andprovided experimental evidence for its validity. Perhaps more importantly, it has become clear thatin real tissues, multiple mechanisms that could drive cell arrest are often operating at the sametime in subtle ways. Therefore, it is not sufficient for scientists to measure a single quantity, suchas cell number density, and claim that changes in that quantity are driving cell arrest. Instead, ashighlighted in the work below, it is important to quantify multiple observables quantitatively inspace and time to confirm the dominant mechanisms driving cell arrest. II. PHYSICAL MECHANISMS FOR CELL ARREST
In this brief review, we will follow the existing literature and use the term "cell jamming" to referto a collective arrest of cell motion. In jammed or solid-like tissues, cells do not change neighbors,the cell-scale structure does not remodel, and the tissue resists changes to its shape. In unjammedor fluid-like tissues, cells do change neighbors, and the tissue flows and remodels in response tofluctuations or internally or externally applied forces.We next discuss in some detail three mechanisms that can drive jamming or unjamming: crowd-ing, tension-driven rigidity, and fluctuations.
A. Crowding
Crowding is a mechanism for cell jamming that is directly related to particle jamming in thephysical sciences. Crowding occurs when the fraction of available space taken up by particles,molecules, or cells becomes sufficiently high that the entire system becomes rigid. It is easiestto understand the math behind this mechanism in the simplest case of squishy spheres at zerotemperature. Each sphere can move left or right, forward or backward in two dimensions (and alsoup and down in three dimensions). Therefore, the number of degrees of freedom in the system isthe number of particles N p times the number of dimensions d : N DOF = N p d .When things get crowded, particles must start to contact other particles, and each contact addsanother constraint to the system, shared between the two particles. So if each particle has onaverage z contacts, the number of constraints in the system is N c = N p z/ . In line with intuition,the system rigidifies precisely when the number of constraints equals the number of degrees offreedom: N c = N DOF , which can also be written in terms of the average number of contacts as z critical = 2 d . Therefore, when the system is less crowded and the number of contacts is less thanthis critical value, the system is floppy and particles can change neighbors. When the system issufficiently crowded so that the number of contacts is above that critical value, the system is rigid.Of course, real materials deviate from this idealized case. Making the spheres slightly adhesivealters the nature of the constraints and changes the rigidity transition [60]. In the presence of finiteFigure 1: A speculative cell jamming phase diagram extrapolated from recent results in literature.The green line represents the tension-driven rigidity transition seen in confluent tissues due to acompetition between fluctuations and cell shape induced geometric frustration [8]. The blue linerepresents the glass transition seen in harmonic spheres [6], which models the behavior seen bynonconfluent rounded cells at low adhesions. The orange line represents the shear instability seenin partially confluent tissues at finite temperature as density and adhesion are tuned. At lowadhesion, the tissue becomes solid-like as density increases, reminiscent of crowding. At highadhesion, there is a density-independent transition similar to what is seen in completely confluenttissues [30].fluctuations, such as a non-zero temperature, the particles can still be "caged" by their neighbors,and that state is called a glass. In the presence of applied forces, a system rigidified by crowdingcan also begin to yield and flow. Understanding the precise nature of these glasses and yieldingtransitions is still a highly active field [5, 33].Given that many cell types round up in cell culture medium and respond much like sticky "activebubbles" [16], it is obvious how a crowding mechanism could generate rigidity in a tissue. Roughlyspherical cells could adhere, divide, or be compressed until the density of cells increases past thecritical threshold and they can no longer move past each other. Modeling cells as sticky spheresgenerate a jamming phase diagram that depends on density, adhesion, and applied stress [60] whichwas the inspiration for a conjectured phase diagram for cell arrest [52].A very clear example of this "sticky-sphere" transition was recently discovered in the zebrafishblastoderm [47], a tissue where the viscosity drops by more than an order of magnitude in a fewminutes, as the tissue transitions from a solid-like to a fluid-like state. During this transition, it 𝒔 𝒊 = 𝒔 𝒔 𝒇 ≠ 𝒔 𝒔 𝒊 𝒔 𝒇 Geometric IncompatibilityIncreasing FluctuationsSmall Fluctuations Large Fluctuations
Increasing
Density(a)(b)(c) Low Density High Density
Figure 2: General mechanisms for cell arrest. (a) Crowding: Compression of cells inside a boxincreases the number of contacts, and therefore constraints, on each cell which causes the cells tojam. (b) Tension-driven rigidity: Cells in a monolayer are fluid-like when their current cell shape, s i , is the same as their preferred shape, s . Then by altering the preferred cell shape, the cellsbecome geometrically frustrated and the tissue becomes rigid. (c) Fluctuations: Trajectories ofcells in a monolayer illustrate the caging effect of cells at low temperature. As the temperature isincreased, the cells have the energy to escape their cage and rearrange.is observed that the cell packing fraction decreases slightly, while the viscosity drops precipitously.This is precisely what is seen in particulate jamming, where a small change in packing fractiondrives a significant change in the contact network or connectivity. In the new work, the authorsmeticulously reconstruct cell connectivity networks to demonstrate the same effect as in jammedparticles, and then perturb E-cad expression and demonstrate that the resulting change in networkconnectivity completely explains the resulting change in tissue viscosity.The picture becomes more complicated in heterogeneous systems. For example, recent workdemonstrates that a fluid-to-solid transition occurs as a function of position along the body axis inzebrafish embryo [42]. Fluid-like cells in the mesodermal progenitor zone (MPZ) differentiate andare incorporated into the presomitic mesoderm (PSM). The MPZ has larger extracellular spacesand active fluctuations than the PSM leading to the PSM having a higher cell density. The MPZtissue is fluid-like with a lower tissue viscosity and lower yield stress than the PSM. Togetherthis is suggestive that the embryo experiences a jamming transition due to crowding along theanterior-posterior axis. B. Tension-driven rigidity
While the standard explanation for rigidity in materials is crowding, a new type of rigidity tran-sition has recently been discovered in biological systems, such as confluent tissues and biopolymernetworks, and also in so-called "mechanical metamaterials" like origami [12]. It is deeply connectedto the older idea of tensegrity structures as models for cells and tissues [28].In all of these systems, the contact network (sometimes called the network topology) remainsconstant, in direct contrast to jamming scenarios. Instead, a continuous parameter can be tuned sothat the system crosses a rigidity transition. The underlying mechanism has been termed geometricincompatibility [39, 43] and it is similar to how a guitar string becomes rigid once it is stretchedbeyond the initial length of the string. We will refer to this type of transition as "tension-drivenrigidity".In biopolymer networks, the tuning parameter is the applied strain (amount of external deforma-tion). For small strains, the network is floppy, and at a critical strain the stiffness of the networkchanges by several orders of magnitude. This behavior is predicted by simple models and observedin experiments [31, 50, 56].Similarly, this rigidity transition can also be seen in models for confluent tissues. In vertex orVoronoi models for such systems, each cell has a characteristic volume (or area in 2D cross-sectionsof monolayers). In addition, each cell has a preferred surface area (or perimeter in 2D), whichis generated by cell-cell adhesion due to cadherins and other adhesive molecules, a surface-areaminimizing cortical shell of actin, myosin and other cytoskeletal components, and non-linear effectssuch as contractile rings or saturation of molecules at adhesive contacts. In these models, thedimensionless preferred cell shape, which in 2D is just the ratio of the preferred perimeter to thesquare root of the preferred area, continuously tunes the model across the transition [7, 17], inagreement with what is seen in experiments [46]. Recent work has demonstrated that these modelsquantitatively predict, with no fit parameters, cell rearrangement rates in body axis elongationin the fruit fly [64, 67], if one also takes into account cell alignment and disorder in the packing.Similar effects are predicted in 3D [40].
C. Interpolating between crowding and tension-driven rigidity:
While crowding is often studied in particle-based models where the interaction depends on howmuch the particles overlap, and tension-driven rigidity has been studied on vertex networks that fillall of space, in real biological systems we are often interested in tissues that are nearly confluent,with small gaps between cells. Do such tissues rigidify due to crowding, tension-driven rigidity,both, or something else?One recent manuscript [26] showed that in nearly confluent mesenchymal tissues in Xenopusdevelopment, cells exhibited features of both types of systems – they could actively tug past othercells (more like particle-based crowding models) and they could also shorten and extend edges tochange neighbors more like vertex models.To explain such observations, a new set of partially confluent models have recently been developed.One version begins with a particle-based model so that cells can fully break apart, but whereoverlap between cells can create an interfacial edge [59]. Due to a competition between two-cellinteractions, three-cell interactions, and geometric constraints, this model exhibits several differenttissue phases, including a gas phase where cells behave as repulsive spheres and confluent phasesthat shares features with those found in vertex models. Another set of simulations investigates cellsmodeled as deformable rings that can change shape up to the confluent limit [9]. In this model, thetissue always behaves like an elastic solid with invaginations occurring after confluence. As of yet,neither model has been directly compared with experiments.Most recently, an exciting new study developed an active foam model, which specifies a foam-like interfacial tension on each edge of a cell, and allows gaps to open up spontaneously if theyare energetically favored.This model replicates both the crowding transition seen in passive foamswith increasing packing fraction and some aspects of the tension-driven rigidity transitions seenin vertex models [30]. However, the model includes only linear interfacial tensions and does notinclude nonlinear effects that stabilize the fluid phase of vertex models, so it will be interesting tosee how nonlinear effects might alter predictions.
D. Role of fluctuations:
Finally, we return to the familiar idea of a material becoming solid or glassy as the temperatureis decreased. Temperature fluctuations provide an energy source that allow a particle to explore itslocal environment, and a particle inside a dense system requires energy to escape the constraintsimposed by its neighbors – i.e. break out of its cage. As the temperature is decreased, it becomesless likely that a particle will have the energy required to do so.Similarly, in self-propelled particle models for tissues, where a particle generates its own forcesand momentum, the system can also undergo a glass transition [25] controlled by the packingfraction (as in crowding) as well as the persistence time (how long a particle moves in a straightline before changing direction) and the magnitude of the self-propulsion forces. Detailed work hasshown that the glass transition is different in thermal systems compared to self-propelled ones [5].Fluctuations also play a similar role in confluent models. In self-propelled confluent models, asthe magnitude of the propulsion force or the persistence time decreases, the system becomes moresolid-like [8]. Glassy behavior is observed in the low-temperature fluid phase of the vertex model,although a detailed analysis reveals that some features are interestingly different from what isobserved in particulate systems [57]. Another possible source of fluctuations is fluctuating tensionsalong cell-cell interfaces, which have been shown to be coupled with the cycle of cell division [29].Recent simulations [32, 66] and experiments [14] suggest that a confluent tissue can solidify asthe magnitude of tension fluctuations in the tissue decrease, although there is evidence that thepersistence of tension affects tissue fluidity in a complicated way [66].Finally, recent experiments on pharmacologically perturbed epithelial cells highlight the possibil-ity that cell arrest is not associated with an underlying rigidity transition at all. In the perturbed celltypes, it appears that the relevant fluctuations become so small that cells do not change neighborsand remodel even though the cellular structure is floppy [14]. In other words, the cells could moveeasily, but they do not because there are no fluctuations to drive them, similar to an unjammedparticulate system at absolute zero temperature. This highlights that even in dense tissues, theabsence of cell motion is not automatically a proxy for solidity or rigidity.
III. BIOLOGICAL DRIVERS OF JAMMING AND CELL ARREST
In the previous section we focused on the physical mechanisms that govern cell arrest, but clearlythose physical mechanisms are driven by specific cell-scale and molecular-scale features. In addition,individual cells can sense and respond to their environments in mechanosensitive feedback loops.In this section, we highlight how specific cell-scale biological features such as expression levels andsignaling pathways can be used to tune cell arrest in the scenarios discussed above.
A. Cell-cell interactions
Multicellular organisms have evolved a large number of redundant cell-cell interactions that actto preserve tissue cohesion and allow the robust formation of tissue-scale structures. Many of thesesignaling cascades begin with cell-cell adhesion molecules, such as cadherins. We first focus onhomotypic interactions between two cells of the same type. In non-confluent tissues, evidence isemerging that cell-cell adhesion performs a role very similar to that expected for adhesion in stickyspheres [60]; decreasing adhesion generates larger intracellular spaces and fewer cell-cell contacts,which increases the fluidity of the tissue via a decrease in crowding [30, 42, 47].In confluent tissues, the role of cell-cell adhesion is much more subtle and likely cell-type specific.In such tissues, the tuning parameter for tension-driven rigidity is cell shape, and evidence isemerging that cadherin expression levels can change the cell shape in context-dependent or non-monotonic ways. For example, cell doublet experiments suggest that, consistent with intuition,increasing cadherin expression increases the surface area of cell-cell contacts [65]. However, newexperiments in confluent monolayers show that knockdown of E-cadherins in keratinocytes resultsin an increase in cell shape index compared to wildtype [54], which indicates that cells with lowerE-cadherin expression prefer more surface area of cell-cell contact. In some ways, these subtlebehaviors are not surprising, as it is well-known that cadherin signaling significantly changes themechanics of the cortical cytoskeleton [1, 36] so that increases in adhesion are often balanced bychanges to cortical tension that have an opposite effect.A related phenomenon is heterotypic cell-cell interactions between two different cell types. Mix-tures of two cell types are often observed to sort, and the mechanisms driving such sorting shouldbe deeply related to crowding-based vs. tension-driven rigidity transitions discussed in the previoussection. For example, the differential adhesion hypothesis (DAH) is based on the assumption thatcells act like sticky spheres, with an interfacial tension directly proportional to differences in cell-celladhesion, so sorting occurs when cells rearrange to minimize that interfacial tension [35].In confluent models, on the other hand, adhesion-based changes in cell shape are not sufficientto drive macroscopic sorting [54]. Instead, confluent models require a specific heterotypic response– an explicit change to interfacial tension along heterotypic cell-cell contacts – in order to generatemacroscopic sorting [10, 58]. Heterotypic interfacial tension causes complete and rapid demixingbetween cell types [54] and there is a discontinuous restoring force for perturbations of the bound-ary [58]. For tissues near the fluid-solid transition, the final cell and interface shapes may be set bya competition between the interfacial forces and the shape-based forces governing tension-drivenrigidity [53].Another class of cell-cell interactions – particularly important in morphogenesis – are pathwayssuch as planar cell polarity that localize adhesion molecules and motor proteins along interfaceswith a specific orientation, generating large-scale anisotropic forces [4, 69]. It has recently beenshown that any anisotropic forces, including those generated by external stretching or pulling fromnearby tissues [15], alter tension-driven rigidity. Specifically, while previous work focused largely onisotropic systems where cells are not aligned, anisotropic forces generically lead to cell alignment,where the long axis of cells point in the same direction. By carefully measuring cell shape, cell-cellalignment, and the disorder [67], one can extend theories of tension-driven rigidity to predict ratesof cell rearrangement with no fit parameters. Remarkably, these predictions with no fit parametersquantitatively match experimental data for body axis elongation in Drosophila [64].In addition to changing the overall magnitude of tension on a cell-cell interface, cell-cell inter-actions can also induce fluctuations in intercellular tension, as well as feedback loops to regularizesuch fluctuations. For example, pulses of non-muscle myosin II have been shown to cause perma-nent junctional remodeling that can drive shape changes and increased cell rearrangements duringconvergent extension. A recent model captures this behavior by allowing for junctions to undergopermanent tension remodeling after surpassing the critical strain threshold. However, to avoidpermanent junctional shortening, there is continuous strain relaxation which allows the system toslowly lose deformation memory. Together this allows for large-scale irreversible deformations dur-ing convergent extension [55]. A different model allows for cytoskeletal remodeling through activerecruitment of myosin depending on the internal strain rate of its associated actin filament. In thiscase, myosin pulsation causes deformations in cell shape, which in turn stimulate myosin recruit-ment, which then stabilizes the deformation [44]. Additionally, the adhesion molecule Sidekick (sdk)is shown to localize at tricellular adherens junctions (tAJs) and disruptions to sdk cause abnormalcell shape changes and a decrease in rearrangements contributing to convergent extension. Onehypothesis for this behavior is that the sdk adhesion molecule is involved with the transition fromshortening to elongation which occurs at tricellular vertices during intercalation. To capture this,the authors developed a vertex model where higher-fold rosettes structures were stabilized and long-lived, and the simulations generated shape changes and intercalation rates that were quite similarto the sdk mutants. This suggests that sdk may exert feedback control of tension fluctuations atjunctions, and disruptions in sdk may significantly delay or halt cell rearrangements [18].A related observation is that many cell types exhibit active forces, such as cell motile forces ortension along stress fibers, that are polarized along a specific direction. Cell shape alignment andother types of cell-cell signaling can drive these polarizations to align in the same direction. Suchalignment interactions can lead to large-scale collective behavior [63] and can alter cell jamming indense tissues. For example, in crowding models, cells become more aligned as the packing fractionincreases preceding the onset of rigidity [25]. In confluent models, increasing alignment in cellpolarity drives the tissue towards a "solid-flocking" state [20], very similar to the behavior seenin epithelial monolayers with upregulated RAB5A [38]. This observation highlights that not allsolid-like states have arrested motion: the "solid-flocking" state corresponds to a group of cells thatis internally rigid so that the cells do not change neighbors, and yet the collective is still movingtogether in the same direction as a unit. Therefore, one needs to examine the relative displacementsbetween cells rather than the absolute displacement of cells to determine tissue fluidity.
B. Cell-substrate interactions
Cell-substrate interactions play a key role in cell jamming, as many cell types exert forces onthe substrate in order to locomote, via integrin-based traction forces [27]. Especially in epithelialmonolayers, these active propulsion forces generate fluctuations that can drive fluidization of thetissue. In less confluent systems, cell-cell contacts may decrease this propensity through contactinhibition of locomotion [70].The mechanical and biochemical properties of the substrate also provide cues that can alterjamming and cell arrest. For example, cells become more stiff on stiff substrates [22, 23, 49], whichwould be expected to lead to higher energy barriers in particle-based crowding models and generatechanges to preferred cell shape in confluent models. Simultaneously, cells tend to spread moreon stiffer substrates [21], which would decrease cell number density in particle-based models anddecrease cell shape index in confluent models, in the absence of other feedbacks. Adhesion matterstoo – increasing cell-substrate adhesion enhances cell spreading and drives high locomotion [11].A very valuable direction for future research would be to characterize how cell-cell adhesion andcell-substrate adhesion co-regulate one another and thereby alter cell arrest.
C. Cell Division
Historically, cell division has been highlighted as a mechanism for increasing cell densities intissues. Specifically, in particle-based models with fixed boundaries where cells grow before divid-ing, cell divisions increase crowding which in turn leads to cell arrest. However, there are otherbiophysical mechanisms triggered during cell division that can affect tissue rigidity.First, the act of cell division necessarily creates active stress fluctuations that can fluidize thetissue [13, 45, 51]. Second, tension fluctuations are often introduced, for example, by asymmetricdivision where daughter cells have different mechanics, as seen in mouse blastocysts [37], on insymmetric divisions which result in lower tension between mitotic cells and their neighbors comparedto other interfaces [41]. Lastly, there is emerging evidence that stereotyped changes to global corticaltension occur as a function of cell cycle [29]. Recent work has demonstrated that this last mechanismis likely the dominant source of fluctuations and tissue fluidization in MDCK monolayers [14].
D. Cell differentiation
Cell differentiation can also drive changes to tissue rigidity. As higlighted previous previously, thedifferentiation of cells from the MPZ to PSM in zebrafish embryos results in smaller gaps betweencells and jamming due to crowing [42]. This fluid-solid transition along the body axis guides themorphogenetic flows that shape the embryo [3].Similarly, in cancers, the epithelial to mesenchymal transition (EMT) causes epithelial cells tobecome less confluent and more migratory [61, 68]. In mixtures of these two cell types, whichoccur when only a fraction of the cells have transitioned to a mesenchymal type, it is observed thatincreasing the fraction of mesenchymal cells results in an increase in motility and cell shape of theepithelial cells. This frustrates jamming in the tissue [19].
IV. CONCLUDING REMARKS
The goal of this review is to reduce some of the ambiguity around the increasingly commonterm "cell jamming". We highlight that jamming and the collective arrest of cell motion canbe driven by multiple physical mechanisms, and emphasize the role of three specific mechanisms:crowding, tension-driven rigidity, and a reduction in fluctuations. We develop a "cellular jamming"phase diagram along these three axes extrapolating from recent results in the literature. Carefulquantitative measurements are required to distinguish which mechanism is dominant, and recentwork has begun to carefully test where in this phase space different tissues operate.In addition, we attempt to connect specific cell-scale features, such as the expression levels ofcadherins or frequency of cell divisions, to the physical mechanisms that appear in the cell jamming0phase diagram. However, given the scope, complexity, and cross-talk between these cell biologyprocesses, much work remains to be done to understand how specific molecular mechanisms areconnected to the physics of tissue rigidity. Moreover, cells can sense the rigidity of the surroundingtissue and alter molecular-scale properties in response. Using cell jamming as a lens to understandhow such feedback loops guide morphogenetic processes and how the disregulation of such feedbackloops drive disease will be an exciting direction for future research.
V. CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
VI. FUNDING SOURCES
This work was supported by the Simons Foundation grants and and theNational Science Foundation NSF-DMR-1951921.
VII. ANNOTATED REFERENCES
Papers of particular interest, published within the period of review, have been highlighted as ∗ of special interest ∗∗ of outstanding interest• ∗∗ Devany 2019 [14]: Cell division Rate Controls Cell Shape Remodeling in EpitheliaThis paper observes the collective motion and rheology of MDCK monolayers after reachingconfluence. On different ECM stiffnesses, the number density of the tissues where cell motionreaches its minimum varies significantly. This highlights that crowding alone isn’t always enoughto predict jamming behavior. Also, researchers observed that disruptions to the cell cycle causecollective cell arrest. Together with their measurements of cell-cell contact angles, this suggeststhat cell-cycle dependent fluctuations in interfacial tension are responsible for tissue remodeling.• ∗∗ Kim 2020 [30]: Embryonic Tissues as Active FoamsThis paper describes an active foam model for near-confluent tissues where gaps are allowed toopen spontaneously. The model replicates both the crowding transition seen in passive foams withincreasing packing fraction and some aspects of the tension-driven rigidity transitions seen in vertexmodels. Their results suggest that tension fluctuations rather than adhesion or cortical tension maycontrol tissue rheology in some embryonic tissues.• ∗ Mongera 2018 [42]: A fluid-to-solid jamming transition underlies vertebrate body axis elon-gation.This paper highlights the fluid-solid transition along the body axis in the zebrafish embryo. Thefluid-like cells in the mesodermal progenitor zone have larger extracellular spaces and fluctuationsthan the solid-like presomitic mesoderm. This is suggestive that the embryo undergoes the jammingtransition due to crowding along this anterior-posterior axis. Together with [3], the results suggestthat the fluid-solid transition may be responsible for guiding morphogenetic flows necessary forembryonic morphogenesis.1• ∗∗ Petridou 2021 [47]: Rigidity percolation uncovers the structural basis of embryonic tissuephase transitionsThis paper investigates the rigidity transition seen in the zebrafish blastoderm. The large dropin the tissue’s viscosity is explained by changes to the cell contact network, which are strikinglysimilar to those predicted by theories of rigidity percolation that explain jamming in soft spheres.The results indicated that when E-cadherin expression is reduced, the resulting decrease in cell-celladhesion reduced connectivity in the contact network and increases tissue fluidity. Together thisindicates that in this system cellular jamming is caused by increased crowding driven by increasesin cell adhesion. Interestingly, it’s also observed that synchronous cell divisions are required forhomogeneously distributed contact loss throughout the tissue, which is necessary to produce uniformchanges in viscosity.• ∗ Staddon 2019 [55]: Mechanosensitive Junction Remodeling Promotes Robust EpithelialMorphogenesisThis paper demonstrates that tissue-scale shape changes during convergent extension can bedriven by frequency-dependent actomyosin pulsation. Experimental observations indicate that ep-ithelial junctions behave elastically under low contractile stress but undergo reversible deformationsunder high contractile stress. The authors propose an extension to the vertex model in which epithe-lial junctions remodel after a critical strain is surpassed and impose a continuous strain relaxation.Together these additions capture the tissue remodeling seen in experiments.• ∗ Wang 2020 [64]: Anisotropy links cell shapes to tissue flow during convergent extensionThis paper examines cell anisotropy induced during convergent extension of the Drosophilagermband epithelium. It’s observed that tension-driven rigidity predictions from cell shape alonedo not predict the rearrangement rates inside the tissue. A combination of cell shape and cellanisotropy is shown to robustly predict cell rearrangement rates and tissue fluidity with no fitparameters body axis elongation in Drosophila.• ∗ Yan 2019 [67]: Multicellular Rosettes Drive Fluid-solid Transition in Epithelial TissuesThis work finds that the rigidity of epithelial tissue increases as the number of multicellularrosettes increases. It predicts a scaling relation describing how global tissue tension grows ashigher-order vertices are created and the tension of cell edges adjacent to rosettes increases. Theirpredictions of cell topology and edge tensions closely match with experimental observations inDrosophila embryos. They also develop a generalized Maxwell constraint counting approach toexplain the rigidity transition seen in confluent tissue. [1] Jeffrey D Amack and M Lisa Manning. Knowing the Boundaries : Extending.
Science , 212(2012):212–216, 2013.[2] Thomas E. Angelini, Edouard Hannezo, Xavier Trepatc, Manuel Marquez, Jeffrey J. Fredberg, andDavid A. Weitz. Glass-like dynamics of collective cell migration.
Proceedings of the National Academyof Sciences of the United States of America , 108(12):4714–4719, 2011.[3] Samhita Banavar, Emmet Carn, Payam Rowghanian, Georgina Stooke-Vaughan, Sangwoo Kim, andOtger Campàs. Mechanical control of tissue shape and morphogenetic flows during vertebrate bodyaxis elongation. bioRxiv , pages 5–7, 2020. [4] Claire Bertet, Lawrence Sulak, and Thomas Lecuit. Myosin-dependent junction remodelling controlsplanar cell intercalation and axis elongation. Nature , 429(6992):667–671, 2004.[5] Ludovic Berthier and Giulio Biroli. Theoretical perspective on the glass transition and amorphousmaterials.
Reviews of Modern Physics , 83(2):587–645, 2011.[6] Ludovic Berthier, Hugo Jacquin, and Francesco Zamponi. Microscopic theory of the jamming transitionof harmonic spheres.
Physical Review E - Statistical, Nonlinear, and Soft Matter Physics , 84(5):1–27,2011.[7] Dapeng Bi, J. H. Lopez, J. M. Schwarz, and M. Lisa Manning. A density-independent rigidity transitionin biological tissues.
Nature Physics , 11(12):1074–1079, 2015.[8] Dapeng Bi, Xingbo Yang, M. Cristina Marchetti, and M. Lisa Manning. Motility-driven glass andjamming transitions in biological tissues.
Physical Review X , 6(2):1–13, 2016.[9] Arman Boromand, Alexandra Signoriello, Fangfu Ye, and Mark D Shattuck. Jamming of DeformablePolygons.
Physical Review Letters , pages 1–5, 2018.[10] Laura Canty, Eleyine Zarour, Leily Kashkooli, Paul François, and François Fagotto. Sorting at embry-onic boundaries requires high heterotypic interfacial tension.
Nature Communications , 8(1), 2017.[11] Yuansheng Cao, Richa Karmakar, Elisabeth Ghabache, Edgar Gutierrez, Yanxiang Zhao, Alex Grois-man, Herbert Levine, Brian A. Camley, and Wouter Jan Rappel. Cell motility dependence on adhesivewetting.
Soft Matter , 15(9):2043–2050, 2019.[12] Bryan Gin Ge Chen and Christian D. Santangelo. Branches of Triangulated Origami Near the UnfoldedState.
Physical Review X , 8(1):11034, 2018.[13] Michael Czajkowski, Daniel M. Sussman, M. Cristina Marchetti, and M. Lisa Manning. Glassy dy-namics in models of confluent tissue with mitosis and apoptosis.
Soft Matter , 15(44):9133–9149, 2019.[14] John Devany, Daniel M Sussman, M Lisa Manning, and Margaret L Gardel. Cell division Rate ControlsCell Shape Remodeling in Epithelia Department of Physics , James Franck Institute and Institute forBiophysical Dynamics , The University of Chicago , Chicago , IL 60637 , USA , 2 Department ofPhysics and Soft Matter. bioRxiv , 2019.[15] Maria Duda, Natalie J. Kirkland, Nargess Khalilgharibi, Melda Tozluoglu, Alice C. Yuen, NicolasCarpi, Anna Bove, Matthieu Piel, Guillaume Charras, Buzz Baum, and Yanlan Mao. Polarizationof Myosin II Refines Tissue Material Properties to Buffer Mechanical Stress.
Developmental Cell ,48(2):245–260.e7, 2019.[16] E. Evans and A. Yeung. Apparent viscosity and cortical tension of blood granulocytes determined bymicropipet aspiration.
Biophysical Journal , 56(1):151–160, 1989.[17] Reza Farhadifar. Dynamics of Cell Packing and Polar Order in Developing Epithelia.
PhD Thesis,Dresden University of Technology , 2009.[18] Tara M. Finegan, Nathan Hervieux, Alexander Nestor-Bergmann, Alexander G. Fletcher, Guy B.Blanchard, and Bénédicte Sanson.
The tricellular vertex-specific adhesion molecule Sidekick facilitatespolarised cell intercalation during Drosophila axis extension , volume 17. PLoS Biology, 2019.[19] Marielena Gamboa Castro, Susan E. Leggett, and Ian Y. Wong. Clustering and jamming in epithelial-mesenchymal co-cultures.
Soft Matter , 12(40):8327–8337, 2016.[20] Fabio Giavazzi, Matteo Paoluzzi, Marta Macchi, Dapeng Bi, Giorgio Scita, M. Lisa Manning, RobertoCerbino, and M. Cristina Marchetti. Flocking transitions in confluent tissues.
Soft Matter , 14(18):3471–3477, 2018.[21] Ming Guo, Adrian F. Pegoraro, Angelo Mao, Enhua H. Zhou, Praveen R. Arany, Yulong Han, Dylan T.Burnette, Mikkel H. Jensen, Karen E. Kasza, Jeffrey R. Moore, Frederick C. Mackintosh, Jeffrey J.Fredberg, David J. Mooney, Jennifer Lippincott-Schwartz, and David A. Weitz. Cell volume changethrough water efflux impacts cell stiffness and stem cell fate.
Proceedings of the National Academy ofSciences of the United States of America , 114(41):E8618–E8627, 2017.[22] Mukund Gupta, Bryant L. Doss, Leyla Kocgozlu, Meng Pan, René Marc Mège, Andrew Callan-Jones,Raphaël Voituriez, and Benoît Ladoux. Cell shape and substrate stiffness drive actin-based cell polarity.
Physical Review E , 99(1):1–16, 2019.[23] Mukund Gupta, Bibhu Ranjan Sarangi, Joran Deschamps, Yasaman Nematbakhsh, Andrew Callan-Jones, Felix Margadant, René Marc Mège, Chwee Teck Lim, Raphaël Voituriez, and Benoît Ladoux.Adaptive rheology and ordering of cell cytoskeleton govern matrix rigidity sensing.
Nature Communi- cations , 6, 2015.[24] Anna Haeger, Marina Krause, Katarina Wolf, and Peter Friedl. Cell jamming: Collective invasionof mesenchymal tumor cells imposed by tissue confinement. Biochimica et Biophysica Acta - GeneralSubjects , 1840(8):2386–2395, 2014.[25] Silke Henkes, Yaouen Fily, and M. Cristina Marchetti. Active jamming: Self-propelled soft particles athigh density.
Physical Review E - Statistical, Nonlinear, and Soft Matter Physics , 84(4):1–4, 2011.[26] Robert Huebner, Abdul Naseer Malmi-Kakkada, Sena Sarikaya, Shinuo Weng, and John Wallingford.Mechanical heterogeneity along single cell-cell junctions is driven by lateral clustering of cadherinsduring vertebrate axis elongation. bioRxiv , 2020.[27] Anna Huttenlocher and Alan Rick Horwitz. Integrins in cell migration.
Cold Spring Harbor Perspectivesin Biology , 3(9):1–16, 2011.[28] Donald E. Ingber. Tensegrity I. Cell structure and hierarchical systems biology.
Journal of Cell Science ,116(7):1157–1173, 2003.[29] Matthew C. Jones, Janet A. Askari, Jonathan D. Humphries, and Martin J. Humphries. Cell adhesionis regulated by CDK1 during the cell cycle.
Journal of Cell Biology , 217(9):3203–3218, 2018.[30] Sangwoo Kim, Marie Pochitaloff, Georgina-Stooke-Vaughan, and Otger Campàs. Embryonic Tissuesas Active Foams. bioRxiv , page 2020.06.17.157909, 2020.[31] Gijsje H. Koenderink, Zvonimir Dogic, Fumihiko Nakamura, Poul M. Bendix, Frederick C. MacKintosh,John H. Hartwig, Thomas P. Stossel, and David A. Weitz. An active biopolymer network controlled bymolecular motors.
Proceedings of the National Academy of Sciences of the United States of America ,106(36):15192–15197, 2009.[32] Matej Krajnc. Solid-fluid transition and cell sorting in epithelia with junctional tension fluctuations.
Soft Matter , 16(13):3209–3215, 2020.[33] Andrea J. Liu and Sidney R. Nagel. The jamming transition and the marginally jammed solid.
AnnualReview of Condensed Matter Physics , 1:347–369, 2010.[34] Andrea J. Liu and R N Sindey. Jamming is not just cool any more.
Nature , 81(2):226–231, 1998.[35] M. S. Steinberg. Reconstruction of tissues by dissociated cells. Some morphogenetic tissue movementsand the sorting out of embryonic cells may have a common explanation.
Science , 141(3579):401–408,1963.[36] Jean Léon Maître and Carl Philipp Heisenberg. Three functions of cadherins in cell adhesion.
CurrentBiology , 23(14):626–633, 2013.[37] Jean León Maître, Hervé Turlier, Rukshala Illukkumbura, Björn Eismann, Ritsuya Niwayama, FrançoisNédélec, and Takashi Hiiragi. Asymmetric division of contractile domains couples cell positioning andfate specification.
Nature , 536(7616):344–348, 2016.[38] Chiara Malinverno, Salvatore Corallino, Fabio Giavazzi, Martin Bergert, Qingsen Li, Marco Leoni,Andrea Disanza, Emanuela Frittoli, Amanda Oldani, Emanuele Martini, Tobias Lendenmann, Gi-anluca Deflorian, Galina V Beznoussenko, Dimos Poulikakos, Ong Kok Haur, Marina Uroz, XavierTrepat, Dario Parazzoli, Paolo Maiuri, Weimiao Yu, Aldo Ferrari, Roberto Cerbino, and Giorgio Scita.Endocytic reawakening of motility in jammed epithelia Europe PMC Funders Group.
Nat Mater ,16(5):587–596, 2017.[39] Matthias Merkel, Karsten Baumgarten, Brian P. Tighe, and M. Lisa Manning. A minimal-length ap-proach unifies rigidity in underconstrained materials.
Proceedings of the National Academy of Sciencesof the United States of America , 116(14):6560–6568, 2019.[40] Matthias Merkel and M. Lisa Manning. A geometrically controlled rigidity transition in a model forconfluent 3D tissues.
New Journal of Physics , 20(2), 2018.[41] Yekaterina A. Miroshnikova, Huy Q. Le, David Schneider, Torsten Thalheim, Matthias Rübsam, Na-dine Bremicker, Julien Polleux, Nadine Kamprad, Marco Tarantola, Irène Wang, Martial Balland,Carien M. Niessen, Joerg Galle, and Sara A. Wickström. Adhesion forces and cortical tension couplecell proliferation and differentiation to drive epidermal stratification.
Nature Cell Biology , 20(1):69–80,2018.[42] Alessandro Mongera, Payam Rowghanian, Hannah J. Gustafson, Elijah Shelton, David A. Kealhofer,Emmet K. Carn, Friedhelm Serwane, Adam A. Lucio, James Giammona, and Otger Campàs. A fluid-to-solid jamming transition underlies vertebrate body axis elongation.
Nature , 561(7723):401–405, Physical Review Letters , 120(26):268105, 2018.[44] Nicholas Noll, Madhav Mani, Idse Heemskerk, Sebastian J. Streichan, and Boris I. Shraiman. Activetension network model suggests an exotic mechanical state realized in epithelial tissues.
Nature Physics ,13(12):1221–1226, 2017.[45] Linda Oswald, Steffen Grosser, David M. Smith, and Josef A. Käs. Jamming transitions in cancer.
Journal of Physics D: Applied Physics , 50(48), 2017.[46] Jin Ah Park, Jae Hun Kim, Dapeng Bi, Jennifer A. Mitchel, Nader Taheri Qazvini, Kelan Tantisira,Chan Young Park, Maureen McGill, Sae Hoon Kim, Bomi Gweon, Jacob Notbohm, Robert Steward,Stephanie Burger, Scott H. Randell, Alvin T. Kho, Dhananjay T. Tambe, Corey Hardin, Stephanie A.Shore, Elliot Israel, David A. Weitz, Daniel J. Tschumperlin, Elizabeth P. Henske, Scott T. Weiss,M. Lisa Manning, James P. Butler, Jeffrey M. Drazen, and Jeffrey J. Fredberg. Unjamming and cellshape in the asthmatic airway epithelium.
Nature Materials , 14(10):1040–1048, 2015.[47] Nicoletta I. Petridou, Bernat Corminas-Murtra, Carl Philipp Heisenberg, and Edouard Hannezo. Rigid-ity percolation uncovers the structural basis of embryonic tissue phase transitions.
Accepted to Cell ,2021.[48] Ryan J. Petrie and Kenneth M. Yamada. Fibroblasts Lead the Way: A Unified View of 3D CellMotility.
Trends in Cell Biology , 25(11):666–674, 2015.[49] Masha Prager-Khoutorsky, Alexandra Lichtenstein, Ramaswamy Krishnan, Kavitha Rajendran, AviMayo, Zvi Kam, Benjamin Geiger, and Alexander D. Bershadsky. Fibroblast polarization is amatrix-rigidity-dependent process controlled by focal adhesion mechanosensing.
Nature Cell Biology ,13(12):1457–1465, 2011.[50] R. Rens, M. Vahabi, A. J. Licup, F. C. MacKintosh, and A. Sharma. Nonlinear mechanics of athermalbranched biopolymer networks.
Journal of Physical Chemistry B , 120(26):5831–5841, 2016.[51] Mehdi Saadaoui, Francis Corson, Didier Rocancourt, Julian Roussel, and Jerome Gros. A tensile ringdrives tissue flows to shape the gastrulating amniote embryo. bioRxiv , 33(0):1–25, 2018.[52] Monirosadat Sadati, Nader Taheri Qazvini, Ramaswamy Krishnan, Chan Young Park, and Jeffrey J.Fredberg. Collective migration and cell jamming.
Differentiation , 86(3):121–125, 2013.[53] Preeti Sahu, J M Schwarz, and M Lisa Manning. Signatures of interfacial tension between two tissuetypes in 3D models. arXiv , pages 1–14, 2021.[54] Preeti Sahu, Daniel M. Sussman, Matthias Rübsam, Aaron F. Mertz, Valerie Horsley, Eric R. Dufresne,Carien M. Niessen, M. Cristina Marchetti, M. Lisa Manning, and J. M. Schwarz. Small-scale demixingin confluent biological tissues.
Soft Matter , 16(13):3325–3337, 2020.[55] Michael F. Staddon, Kate E. Cavanaugh, Edwin M. Munro, Margaret L. Gardel, and Shiladitya Baner-jee. Mechanosensitive Junction Remodeling Promotes Robust Epithelial Morphogenesis.
BiophysicalJournal , 117(9):1739–1750, 2019.[56] Cornelis Storm, Jennifer J. Pastore, F. C. MacKintosh, T. C. Lubensky, and Paul A. Janmey. Nonlinearelasticity in biological gels.
Nature , 435(7039):191–194, 2005.[57] Daniel M. Sussman, M. Paoluzzi, M. Cristina Marchetti, and M. Lisa Manning. Anomalous glassydynamics in simple models of dense biological tissue.
Epl , 121(3):1–8, 2018.[58] Daniel M. Sussman, J. M. Schwarz, M. Cristina Marchetti, and M. Lisa Manning. Soft yet SharpInterfaces in a Vertex Model of Confluent Tissue.
Physical Review Letters , 120(5):58001, 2018.[59] Eial Teomy, David A Kessler, and Herbert Levine. Confluent and nonconfluent phases in a model ofcell tissue.
Physical Review E , 042418:1–14, 2018.[60] V. Trappe, V. Prasad, Luca Cipelletti, P. N. Segre, and D. A. Weitz. Jamming phase diagram forattractive particles.
Nature , 411(6839):772–775, 2001.[61] Shubham Tripathi, Priyanka Chakraborty, Herbert Levine, and Mohit Kumar Jolly. A mechanism forepithelial-mesenchymal heterogeneity in a population of cancer cells.
PLoS Computational Biology ,16(2):1–27, 2020.[62] Daniel J. Tschumperlin. Fibroblasts and the ground they walk on.
Physiology , 28(6):380–390, 2013.[63] Tamás Vicsek and Anna Zafeiris. Collective motion.
Physics Reports , 517(3-4):71–140, 2012.[64] Xun Wang, Matthias Merkel, Leo B. Sutter, Gonca Erdemci-Tandogan, M. Lisa Manning, and Karen E. Kasza. Anisotropy links cell shapes to tissue flow during convergent extension.
Proceedings of theNational Academy of Sciences of the United States of America , 117(24):13541–13551, 2020.[65] Soichiro Yamada and W. James Nelson. Localized zones of Rho and Rac activities drive initiation andexpansion of epithelial cell-cell adhesion.
Journal of Cell Biology , 178(3):517–527, 2007.[66] Takaki Yamamoto, Daniel M. Sussman, Tatsuo Shibata, and M. Lisa Manning. Non-monotonic flu-idization generated by fluctuating edge tensions in confluent tissues. arXiv , 2020.[67] Le Yan and Dapeng Bi. Multicellular Rosettes Drive Fluid-solid Transition in Epithelial Tissues.
Physical Review X , 9(1):11029, 2019.[68] Jing Yang, Parker Antin, Geert Berx, Cédric Blanpain, Thomas Brabletz, Marianne Bronner, KyraCampbell, Amparo Cano, Jordi Casanova, Gerhard Christofori, Shoukat Dedhar, Rik Derynck, Heide L.Ford, Jonas Fuxe, Antonio García de Herreros, Gregory J. Goodall, Anna Katerina Hadjantonakis,Ruby J.Y. Huang, Chaya Kalcheim, Raghu Kalluri, Yibin Kang, Yeesim Khew-Goodall, HerbertLevine, Jinsong Liu, Gregory D. Longmore, Sendurai A. Mani, Joan Massagué, Roberto Mayor, DavidMcClay, Keith E. Mostov, Donald F. Newgreen, M. Angela Nieto, Alain Puisieux, Raymond Runyan,Pierre Savagner, Ben Stanger, Marc P. Stemmler, Yoshiko Takahashi, Masatoshi Takeichi, Eric Theve-neau, Jean Paul Thiery, Erik W. Thompson, Robert A. Weinberg, Elizabeth D. Williams, Jianhua Xing,Binhua P. Zhou, and Guojun Sheng. Guidelines and definitions for research on epithelial–mesenchymaltransition.
Nature Reviews Molecular Cell Biology , 21(6):341–352, 2020.[69] Jennifer A Zallen and Eric Wieschaus. Patterned Gene Expression Directs Bipolar Planar Polarityin Drosophila that eliminate dorsal or ventral cell types (Irvine and Wieschaus, 1994). Instead, thepatterning of cell fates along the A-P axis, perpendicular to the direction of cell.
Developmental Cell ,6:343–355, 2004.[70] Juliane Zimmermann, Brian A. Camley, Wouter Jan Rappel, and Herbert Levine. Contact inhibitionof locomotion determines cell-cell and cell-Substrate forces in tissues.