Articles

https://doi.org/10.1038/s41567-021-01215-1

Embryonic tissues as active foams

Sangwoo Kim 1
, Marie Pochitaloff1, Georgina A. Stooke-Vaughan1 and Otger Campàs 1,2,3,4,5 ✉

The physical state of embryonic tissues emerges from non-equilibrium, collective interactions among constituent cells. Cellular
jamming, rigidity transitions and characteristics of glassy dynamics have all been observed in multicellular systems, but it is
unclear how cells control these emergent tissue states and transitions, including tissue fluidization. Combining computational
and experimental methods, here we show that tissue fluidization in posterior zebrafish tissues is controlled by the stochastic
dynamics of tensions at cell–cell contacts. We develop a computational framework that connects cell behaviour to embryonic
tissue dynamics, accounting for the presence of extracellular spaces, complex cell shapes and cortical tension dynamics. We
predict that tissues are maximally rigid at the structural transition between confluent and non-confluent states, with actively
generated tension fluctuations controlling stress relaxation and tissue fluidization. By directly measuring strain and stress
relaxation, as well as the dynamics of cell rearrangements, in elongating posterior zebrafish tissues, we show that tension fluc-
tuations drive active cell rearrangements that fluidize the tissue. These results highlight a key role of non-equilibrium tension
dynamics in developmental processes.

M
any essential processes in multicellular organisms, from These descriptions predict a density-independent rigidity tran-
organ formation to tissue homoeostasis, require a tight sition in confluent systems that depends on the balance between
control of the tissue physical state1,2. While tissue mechan- actomyosin-generated cortical tension T0 and cell adhesion W
ics and structure at supracellular scales emerge from the collective (treated as passive, effective tensions), with the resulting cell shape
physical interactions among the constituent cells, their control being the control parameter for solid/fluid states17. Since cell shape
occurs at cell and subcellular levels. Bridging these scales is essen- plays a central role in models of confluent systems, the ability to
tial to understand the physical nature of active multicellular systems accurately describe complex cell shapes, beyond polygonal shapes,
and to identify the processes that cells use to control the physical may be important to understand the physical state of the system, as
state of embryonic tissues. recently suggested24,25. Non-vertex models of deformable particles
In vitro experiments of cell monolayers on synthetic substrates have instead focused on configurations in non-confluent equilib-
have revealed characteristics of glassy dynamics3,4 and rigidity tran- rium systems, both in the presence26,27 and absence of cell adhe-
sitions5–7, which are thought to be linked to biological function and sion24,28. In contrast, self-propelled particle and Voronoi models that
multiple pathologies. In contrast, suspended epithelial monolayers focus on cell movements on synthetic substrates account for the
are largely solid-like in vitro8 and show evidence of fracture in vivo9. dynamics of the system and predict rigidity transitions that depend
Experiments with embryonic tissues have shown characteristics on cell density and self-propulsion29,30. All these descriptions cap-
of glassy dynamics in cell movements10, viscous behaviour at long ture some important aspects of the problem, but each neglects a
timescales11 and also structural signatures reminiscent of jamming subset of key cell behaviours and/or restricts the allowed configura-
transitions12, with cell divisions, cell shape and/or changes in cell tions. No current theoretical description accounts for all observed
adhesion suspected to play a role in the control of these emergent relevant cell behaviours in a common framework, hindering our
behaviours5,11,13,14. Recent in vivo experiments in developing zebraf- understanding of how cells control the emergent physical state of
ish embryos showed the existence of a rigidity transition underly- the tissue.
ing the formation of the vertebrate body axis, revealing a functional
role of rigidity transitions in embryonic development15 (Fig. 1a). Dynamic vertex model with extracellular spaces
Both the presence of adhesion-dependent spaces between cells To study the dynamics of embryonic tissues and their physical state,
(Fig. 1b–d) and the dynamics of cell–cell contacts (Fig. 1e,f) were we generalize two-dimensional (2D) vertex models by account-
shown to influence the physical state of the tissue15. However, the ing for (1) extracellular spaces (Fig. 1b,c,g and Methods), (2) the
relative roles of cell adhesion and cell–cell contact dynamics in the stochastic dynamics of active cortical tensions (Fig. 1e,f) and (3)
control of posterior tissue fluidization are still unclear. In general, complex cell shapes (Fig. 1a,g and Methods). Unlike previous vertex
little is known about how different cell behaviours control rigidity models, we do not assume the existence of a preferred cell perim-
transitions and tissue fluidization in embryonic tissues, and whether eter, as experimental evidence for this constraint is lacking. Instead,
all these observed emergent phenomena share a common since cells actively control adhesion, cortical tension and osmotic
physical origin. pressure1,2, we derive a physical description reminiscent of foams
The physical behaviour of multicellular systems has been stud- but with active tension dynamics. Tissue dynamics and structure,
ied theoretically using various approaches. Vertex models16–21 and as well as cell movements and their shapes, are all determined
cellular Potts models22,23 account for cell geometry and use equi- by the dynamics of vertices (Fig. 1g), which follow from force
librium formulations to describe the physical state of the system. balance, namely

Department of Mechanical Engineering, University of California, Santa Barbara, CA, USA. 2Center for Bioengineering, University of California,
1

Santa Barbara, CA, USA. 3Department of Molecular, Cellular and Developmental Biology, University of California, Santa Barbara, CA, USA. 4California
NanoSystems Institute, University of California, Santa Barbara, CA, USA. 5Cluster of Excellence Physics of Life, TU Dresden, Dresden, Germany.
✉e-mail: [email protected]

Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics 859

Articles NaTurE PhysiCs

a b e Tension fluctuations
0 I
(a.u.)
e Cell adhesion 2 200
Forc strength, W
le
dipo 4
6
0
0 100 200 300
Cortical tension, T Time (s)

d 4 f
c Extracellular spaces N=3

Junctional tension
(µm)
L0
∆T
Is
(a.u.) Cell
2 T0
200 0 100
Total area of system AT
Time (s)
N = 25 Time

Force
dipole
0

Til

Nik
Tjk
Nil
Tik

Nil Njk

Nij
Tij
Til

Fig. 1 | Characteristics of multicellular systems and simulation framework. a, Confocal section of embryonic tissue in a zebrafish embryo (membranes
labelled, yellow). b, Schematics of cortical tension and adhesion at cell–cell contacts in a multicellular system with spaces (red) between cells (grey).
c, Confocal section through zebrafish embryonic tissues showing the intensity (Is) of fluorescent reporter protein secreted extracellularly as well as cell
membranes (green), indicating the presence of extracellular spaces (red range). d, Schematics defining the cell size L0, cell number N and simulation
box area AT, which specify cell density. e, Top: kymograph of membrane signal intensity (I) along a tissue region (pink, left) containing a cell–cell contact,
showing contact length ℓ fluctuations (bottom). f, Top: simulated tension fluctuations causing cell–cell contact length fluctuations. Bottom: increasing
(decreasing) tension shortens (lengthens) cell–cell junctions. g, Schematics of the dynamic vertex model formulation. Triple vertices (physical vertices,
red) and non-physical intermediate vertices (blue) are shown. Scale bars, 10 μm.

dRα dTij
ηR Tij Θ(Tij ) + Nij , (1) τT = −(Tij − T0ij ) + ΔTξ , (2)
∑ ( )
dt
= dt
i,j∈F(α)
where ΔT is the amplitude of tension fluctuations and ξ is Gaussian
where t is time, Rα is the position of the vertex α, ηR is a friction white noise (Fig. 1f and Methods). The fixed point effective ten-
coefficient characterizing the dissipation associated with moving a sions depend on both the average cortical tension, T0, and average
vertex, Tij is the effective tension at the contact between cell i and j strength of cell–cell adhesion, W, which, like ΔT, are different at
(with F(α) representing the set of all cells sharing vertex α) and Nij is cell–cell contacts and free cell boundaries (Fig. 1b and Methods).
the normal forces acting on vertex α. Θ(⋅) represents the Heaviside Scaling all quantities, we obtain the relevant dimension-
step function and prevents unrealistic negative tensions15,31. Normal less parameters (Table 1 and Methods). Since P0L0/T0 and τT/τR
forces arise from osmotic pressure differences in adjacent cells and (Table 1) can be estimated from existing experimental data
are given by Nij = (ΔPi − ΔPj) Lij/2 (Methods), where Lij and ΔPi (Methods), we focus on the parameter space spanned by ΔT/T0,
(ΔPj) are, respectively, the contour length of the contact between W/T0 and the normalized cell density ρ (Table 1).
cells i and j and the osmotic pressure difference across cell i (j).
To capture the observed fluctuating nature and finite persis- Structural transitions and mechanics of equilibrium
tence of tension dynamics15, Tij, we describe them as an Ornstein– systems
Uhlenbeck process, with a tension that fluctuates around a fixed We first explore how the states and mechanics of the system change
point T0ij and has a persistence time τT (refs. 32,33), specifically if spaces between cells are allowed. In the absence of tension

860 Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics

NaTurE PhysiCs Articles
(Fig. 2a). Increasing cell density results in larger cellular volume frac-
Table 1 | Definition of the relevant dimensionless parameters in
tion ϕ (Fig. 2b) and cell contact number z (Fig. 2c), with the system
the problem
eventually becoming confluent at an adhesion-dependent critical
Dimensionless Description density, ρc(W/T0). For any fixed cell density, the system undergoes
parameters a non-confluent to confluent transition as W/T0 is increased, since
ΔT/T0 Magnitude of tension fluctuations (Fig. 1e,f). higher adhesion promotes stronger cell–cell contacts (Fig. 2a,b).
These results extend previous studies for purely repulsive deform-
W/T0 Relative strength of cell–cell adhesion W and average
able particles28 to arbitrary adhesion levels.
cortical tension T0 (Fig. 1b).
Changes in relative cell adhesion affect not only the volume frac-
ρ ≡ NL20 /AT Normalized system density (with AT and N being tion but also the structural characteristics of extracellular spaces.
the total area of the system and the number of cells, Equilibrium configurations sharply transition from a large num-
respectively; Fig. 1d). ber of small extracellular spaces to a few large extracellular spaces
P0L0/T0 Relative magnitude of normal to tensional forces. at W/T0 ≈ 0.23 (Fig. 2a,d), as small triangular extracellular spaces
τT/τR Ratio of persistence time of tension fluctuations and can only be stabilized below this value (Supplementary Section 1).
characteristic timescale τR ≡ ηTL0/T0 of dissipation at Concomitant to the presence of large extracellular holes, which are
vertices. reminiscent of epithelial fracture in vivo9, the spontaneous cluster-
ing of cells strongly resembles flocculation in sticky emulsions34,35
(Fig. 2a). Moreover, the system displays bistability between two pos-
fluctuations (ΔT/T0 = 0), the amount of extracellular spaces in equi- sible equilibrium configurations for W/T0 ≥ 0.23, namely a conflu-
librium configurations is determined by force balance and varies with ent state with stretched cells and a non-confluent state with sparse
both the cell density ρ and the relative cell adhesion strength W/T0 and large extracellular holes (Fig. 2e), with strong hysteresis in

a b d f Jamming transition (foams)

ϕ 1.0 ρ
1.4 1.0 1.44 (no adhesion, W = 0)
Confluent 1.00 6

NE /2N
ϕ=1 0.5 0.81
1.2 0.9
Non-confluent 5
z
0 1
ρ

z
2
ϕ= 0.8 3
1.0 0.9 30 4
ϕ= 5
0.8 20 4 6
AE /A0

0.7 0.9 1.0

10 ϕc ≃ 0.83 ϕ
c 0.8 z
1.4 6 0 g Density-independent transition
0 0.2 0.4 0.6 0
(vanishing tensions,T ij = 0)
5 W/T0
1.2 s
3.6 4.0 4.4 4.8
e 1.0 4.4
z=

4
ρ

6
Cell density, ρ

0.9 4.2
1.0 3
z=

0.8 s
z

4.0
=

2
4

0.8 0.7
0 0.2 0.4 0.6 0 0.25 0.50 0.75 1.00 3.8
Relative adhesion, W/T0
W/T0 W/T0 1.6 1.8 2.0
W/T0
h i j k σY/σ0
Shear (affine) ρ = 1 W/T 0.4 1.4
1.2 ρ=1
0.4
transformation ∆T/T0 = 0 0
1.5 ∆T/T0 = 0
0.9 1.0 0.3
1.2 0.3
σxy (t)/σ0

0.5
σY/σ0

0.10
0.25

0.2 0.2
0.6
ρ

0 0.2
1.0
0.25

1.5 1.5 0.3 0.1 0.1

Shear
stress
Strain

1.0 1.0
0.1

0.5 0.5 0 0 0.8 0

0 0 0 0.5 1.0 1.5 2.0 0 0.5 1.0 1.5
–10 0 10 –10 0 10 10–1 100 101 102
t/τT t/τT t/τR W/T0 W/T0

Fig. 2 | Equilibrium configurations and structural transitions. a–c, Representative equilibrium configurations (a), volume fraction ϕ (b) and mean
number of neighbours z (c) for varying values of relative adhesion W/T0 and cell density ρ. Green and purple boxes in a correspond to parameter values
specified as circle dots of the same colour in b and c. d, Number NE (top) and average area ĀE (bottom) of extracellular spaces for varying relative cell
adhesion and different cell densities, showing a sharp structural transition at W/T0 ≈ 0.23 (grey line) leading to the opening of large extracellular spaces.
e, Lowering (dashed line) and increasing (dash-dotted line) relative adhesion quasistatically shows bistable states and strong hysteresis in equilibrium
configurations. Equilibrium quenched states are also shown (solid line). f, Average neighbour number (cell contacts) z as the system volume fraction
changes at vanishing cell adhesion, showing the existence of a jamming transition at ϕc ≃ 0.83 and zc = 4 (configuration shown in inset). Power-law fits
z − zc = z0 (ϕ − ϕc )1/2 + z1 (ϕ − ϕc ) with coefficients z0 ≃ 1.45 and z1 ≃ 10.45 (red line) and with z0 ≃ 4.85 and z1 = 0 (grey line) are shown. g, Average
shape factor ( s̄ ) for varying relative adhesion showing a sharp increase at W/T0 = 2 (vanishing tensions), leading to anisotropic cell shapes (inset),
recapitulating density-independent transitions. h, Schematics of a simple shear deformation imposing a large strain step (ϵxy = 1.5), with associated
temporal evolution of both strain and shear stress. i, Temporal relaxation of shear stress σxy (normalized to σ0 ≡ ρ T0 /L0) after the imposed strain step
√
for varying adhesion levels. j,k, Dependence of the yield stress σY on the relative adhesion strength (j) and on both cell density and relative adhesion (k),
showing a maximum at the structural transition between confluent and non-confluent states (green line). Error bands, s.d.; N = 20 (b–d,f,g) and N = 10
(e,i–k) independent simulations for each parameter set.

Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics 861

Articles NaTurE PhysiCs

a b
101 1.0 ∆T/T0 = 0.5
Magnitude
100 usion of tension
0.5 Diff

α
fluctuations
10–1
∆T/T0
0
MSD (t)/L20
10–2 0 0.5 1.0 1.5 1.50
∆T/T0
10 –3 1.25
Caged
1.00
10–4 ∆T/T0 = 1.5
0.75
10–5 0.50
ρ = 1, W/T0 = 1
10–6 0.25
10–2 10–1 100 101 102
t/τT

c d
100

3 Cell contacts T1 transitions

10–1

Cellular NE rate (τ–1

R
)
MSD(t = 102τT)/L20

10–2
2
10–3

10–4
1

10–5

0 10–6
0 0.4 0.8 1.2 0 0.4 0.8 1.2
W/T0 W/T0

Fig. 3 | Tissue dynamics with finite tension fluctuations. a, MSD for varying magnitudes of tension fluctuations ΔT/T0, showing subdiffusive (0 < α < 1)
and diffusive (α = 1) behaviours as tension fluctuations increase (inset). b, Snapshots of dynamic configurations with examples of cell trajectories over
t/τT = 102. c, MSD at long timescales (t = 102τT ≫ τT > τR), showing non-monotonous behaviour for varying relative adhesion strength and minimal values
at the structural transition. d, Cellular NE rate for varying relative adhesion. Distinct types of NE event: gain/loss of cell contacts (left) and T1 transitions
(right). Cells and junctions undergoing NE events are highlighted as yellow/blue and red colour, respectively. Error bands, s.d.; N = 10 independent
simulations for each set of parameters.

equilibrium configurations if adhesion or cortical tension are var- vanishes when effective tensions vanish (Fig. 2i and Supplementary
ied in a quasi-static manner (Fig. 2e, Supplementary Video 1 and Fig. 5). Subsequently, shear stress relaxes with a characteristic times-
Methods). Some of the predicted system configurations (Fig. 2a) cale τR towards a constant value at long timescales, namely the yield
are analogous to those previously obtained for equilibrium systems stress σY (ref. 39), which depends non-monotonically on cell adhe-
with open boundaries27, but the bistability and hysteresis of configu- sion (Fig. 2j and Supplementary Video 2). For low relative adhesion
rations reported here can only be observed in closed systems that W/T0, the system is non-confluent and the yield stress increases with
more closely resemble epithelial tissues. increasing adhesion as extracellular spaces close down. In contrast,
In the limit of vanishing cell adhesion, the system should behave for adhesion values leading to confluence, increasing W/T0 leads to
as foams/emulsions, which display a jamming transition at a criti- a decreasing yield stress due to lower effective tensions. Vanishing
cal value ϕc. The isostatic condition (zc = 4 in 2D) sets the critical yield stress indicates a fluid tissue state, which occurs only for
volume fraction ϕc ≈ 0.83 of the system36 (Fig. 2f). Both this ϕc value W/T0 = 0 and W/T0 = 2 and corresponds to the jamming transition
and the power-law dependence of z (Fig. 2f) are in agreement with (Fig. 2f) and the density-independent rigidity transition (Fig. 2g),
recent equilibrium simulations of deformable particles28, indicat- respectively, with the tissue being solid for all other values in the
ing that our description accurately describes the foam limit. In 0 < W/T0 < 2 range. These results show that equilibrium systems are
the confluent limit, previous works reported density-independent maximally rigid at the structural transition between confluent and
rigidity transitions,
√ which are controlled by the cell shape factor non-confluent states (Fig. 2j,k), with increasing adhesion render-
s ( s = P/ A, with P and A being the cell perimeter and the area, ing the system more rigid in non-confluent systems and doing the
respectively). The system switches from a solid to a fluid state at opposite in confluent states.
approximately sc ≃ 3.81, with cells transiting from isotropic to
anisotropic shapes17,37,38. The fluid state in these descriptions is Dynamics of active multicellular systems
characterized by vanishing effective tensions (T0ij = 0), allowing Unlike equilibrium systems, tension dynamics at cell–cell con-
neighbour exchanges (NEs) at no energetic cost19. Setting W/T0 = 2 tacts can drive cell movements, NEs and cell shape changes
in our framework leads to vanishing effective tensions and a sharp (Supplementary Videos 3 and 4). At timescales longer than all
increase in average shape factor due to the emergence of anisotropic characteristic timescales (t ≫ τT > τR), cell movements are caged for
cell shape (Fig. 2g), recapitulating density-independent transitions. small ΔT/T0, as indicated by the saturation of the mean squared
To directly assess the physical state of the system, we monitor displacement (MSD) and bounded cell trajectories (Fig. 3a,b and
shear-stress relaxation after imposing an affine deformation (Fig. 2h Methods). For increasing magnitudes of tension fluctuations, cell
and Methods). The initial stress jump is largest with no adhesion and uncaging starts to occur and the asymptotic behaviour of the MSD

862 Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics

NaTurE PhysiCs Articles
a b d f h
ρ = 1, W/T0 = 0.5 τSR ∆T/T0 = 1.5
1.5 τR
1.2 10 0 1.00
100
Non-confluent 6
>10

σxy (t )/σ0

Cellular NE rate (τ–1

R
)
10–1
0.9 0.95
10–1 105
σxy (t )/σ0

10–2

∆T/T0
1.0 4

ϕ
0.6 100 101 102 10
∆T/T0 0.90
t/τR 10–2 s = 4.08
1.5 103
1.0 ∆T/T0 = 1.0
0.3
0.5 Confluent 102
10–3 0.85
0
0.5
0 101
10–1 100 101 102 100 101 102 0 0.4 0.8 1.2 0 0.4 0.8 1.2
t/τR t/τR W/T0 W/T0
e g s = 3.89
1.5
∆T/T0
τSR/τT = 10 2
4.2 ∆T/T0 = 0.5
c Onset of applied shear
1.5
1.25
0.75
0.50
Passive NE

1.00 0.25
Fluid 4.0

∆T/T0
Passive NE 1.0

s
Active NE
NE rate

Total NE Applied strain 3.8 s = 3.83

Active NE

Solid
3.6 3.8 4.0 4.2 4.4 >4.6
0.5
3.6 s
0 0 0.4 0.8 1.2 0 0.4 0.8 1.2
Time Tension fluctuations
W/T0 W/T0

Fig. 4 | Stress relaxation and structure in active multicellular systems. a,b, Stress relaxation (a) and temporal changes in cellular NE rate (b) after the
imposed strain step for varying magnitudes of tension fluctuations. The long timescale stress relaxation follows stretched exponentials (black dashed
lines) and eventually reaches the average value σA of active shear-stress fluctuations (horizontal lines in inset). The cellular NE rate quickly decays to zero
in the absence of activity (t ≈ τR), but remains finite in the presence of activity. c, Sketch showing the dynamics NE induced by the externally applied shear
strain (passive) and by tension fluctuations in cells (active). Active NE enables further stress relaxation and tissue fluidization after the initial passively
induced NE. d, Stress relaxation timescale τSR for varying magnitude of tension fluctuations and relative adhesion, showing a sharp increase as the structural
transition between non-confluent and confluent (green background) states. e, Phase diagram showing the transition between fluid and solid tissue states
for different activity values and relative adhesion strength. Solid states surround the structural transition and are found both in confluent and non-confluent
configurations for low enough activity. f–h, Dependence of the system volume fraction ϕ (f) and average shape factor s̄ (g; with red/blue squares indicating
fluid/solid states, respectively) on relative adhesion for different magnitudes of tension fluctuations. Representative snapshots of dynamic configurations for
fixed relative adhesion (W/T0 = 1; vertical dashed line in g) and increasing tension fluctuations are shown h, with red and blue outlines indicating fluid and
solid states, respectively. Error bands, s.d.; N = 10 independent simulations for each set of parameters.

for t ≫ τT becomes a power law (MSD ≈ tα), with an exponent α that displaying a slow stress decay rather than plateauing to a yield stress
increases with activity (Fig. 3a, inset). This evidences subdiffusive (Fig. 4a and Supplementary Video 5). This long timescale stress
(α < 1) cell movements for intermediate activities, and diffusive relaxation is driven by actively induced T1 transitions (Fig. 4b), and
(α = 1) behaviour for large enough tension fluctuations (Fig. 3a,b). can be accurately described by a stretched-exponential function
Comparing MSD values at long timescales (t = 102τT ≫ τT > τR) (Fig. 4a, inset), as previously done to explain the dynamics of systems
shows that increasing tension fluctuations always leads to higher with a large number of intrinsic timescales40,41. The stress relaxation
cell movements, regardless of cell adhesion strength, as recently timescale, τSR, at which shear stress reaches the level of active shear
reported for confluent systems33. However, for a fixed level of activ- stress in the unperturbed system (Fig. 4a, inset, and Supplementary
ity, MSD values vary non-monotonically as adhesion increases, Fig. 2), varies by over five orders of magnitude as the magnitude of
displaying very reduced cell movements at the transition between tension fluctuations or relative adhesion change slightly (Fig. 4d).
non-confluent and confluent states (Fig. 3c). Similar to the MSD, While larger activity values reduce τSR monotonically, increasing
the NE rate displays analogous non-monotonic behaviour, with the relative adhesion strength leads to non-monotonic changes in
minimal NE events at the structural transition. Close to confluence τSR, which rapidly increases in non-confluent states and displays the
for high adhesion levels, the NE rate is dominated by T1 transitions opposite behaviour in confluent states. The largest stress relaxation
and related to the MSD by MSD(t = 102 τ T ) ≈ k3/4 NE (Extended Data timescale occurs at the structural transition between non-confluent
Fig. 1), with the timescale for T1 transitions, 1/kNE, diverging as the and confluent states, indicating that tissues are minimally fluid at
activity vanishes, a signature of glassy dynamics. In contrast, at low long timescales close to this structural transition (Fig. 4d), consis-
adhesion levels leading to non-confluent states, loss and formation tent with the predicted low cell movements and NE rates close to the
of new cell contacts dominate NE events (Fig. 3d) and the MSD and transition (Fig. 3c,d).
NE behaviours differ close to jamming. The non-monotonic behav- In contrast to previous equilibrium vertex models, our results
iour of both MSD and NE events, with minimal values at the transi- show that a given tissue can behave as a fluid or a solid depend-
tion between non-confluent and confluent states, suggests that the ing on the time necessary to form embryonic structures. If the
non-equilibrium system is also maximally solid-like at the struc- characteristic timescale τd of developmental processes is larger
tural transition. (smaller) than the stress relaxation timescale τSR, namely τd/τSR ≫ 1
(τd/τSR ≪ 1), the tissue behaves as a fluid (solid). Using typical devel-
Physical state of active multicellular systems opmental timescales (τd ≈ 1–2 h; τd/τT ≈ 102), we obtain the tissue
To study the rigidity of active systems, we apply a shear-strain step phase diagram (Fig. 4e). Over a critical value of tension fluctuations,
and monitor stress relaxation, as described above (Fig. 2h). The the tissue is always fluid regardless of cell adhesion levels. Below
presence of finite tension fluctuations qualitatively changes stress that critical activity value, the tissue is fluid at both low and high
relaxation at long timescales compared with equilibrium systems, adhesion levels, but solid in between, in the region of the phase

Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics 863

Articles NaTurE PhysiCs

a e f j A P
6

Average
A P 0.20
A

Normalized

L
frequency
Posterior

Dorsal view
4 0.15
body

M
MPZ MPZ
elongation bleb

M
2 PSM MPZ
P
PSM MPZ

Simulations
V D 0
0 0.5 1.0
–1

b c g h

per cell per min)

On Droplet relaxation t/τT
actuation

R 2 4 0 5 10 15
Before

0.4

(b – b0)/R

NE rate

Experiments
–2
NE count (per cell)
b

(×10
0 2

Cumulative
–10 0 10 30 50 0 20 40
1
Magnetic
actuation

H Time (min) Time (min)

k
i

Normalized MSRD
d 10 0 MPZ

per cell per min)

1.0 4 PSM
2 –/–
PSM cdh2
(σA (t ) – σ0A)/σ0A

Cell
3 uncaging

NE rate
0.5 1
Caged cells
relaxation

0
Droplet

10–1 2
Stretched 0 20 40 0

–2
exponential fit 0 10 20 30

(×10
0 Time (min) 1
Time (min)

<15 m

>40 m
MPZ
MPZ
bleb
0 10 30 50 10
0
101 102

AD

AD
Time (min) Time (min)

Fig. 5 | Stress relaxation and tissue fluidization in posterior tissues during body axis elongation. a, Lateral view of a zebrafish embryo (scale bar,
200 μm) and sketches showing lateral and dorsal views of posterior tissues, indicating the fluid-like MPZ and solid-like PSM (ventral, V; dorsal, D; anterior,
A; posterior, P; medial, M; lateral, L). A confocal section showing a portion of the MPZ (cell membranes, yellow) with a magnetic droplet (magenta) is
highlighted. b, Schematic sketches and confocal snapshots of magnetic droplet actuation in the MPZ tissue. c, Time evolution of the strain, (b − b0)/R,
before, during and after magnetic actuation. d, Relaxation of anisotropic stress σA(t) after magnetic actuation in both linear and log-linear scales, with
σ0A = σA (t = 0). The fit (black line) corresponds to a stretched-exponential function. e, Example of NE events: cells in contact (blue line) before the
T1 transition (cyan; top) and not in contact after the transition (bottom). f, Normalized frequency of amplitude of junctional length fluctuations for the
MPZ, both in the absence and presence of blebbistatin (bleb; average amplitudes in inset). g, Cumulative NE events in the MPZ (away from the droplet,
cyan; dashed black line shows linear fit) and in the close neighbourhood of the droplet (around droplet, red). Inset: visual definition of droplet neighborhood
(red). h, Temporal evolution of NE rate in the MPZ (cyan) and in the region around droplet (red). Initial (t < 15 min) and final (t > 40 min) stages of droplet
relaxation are highlighted as green and pink shaded regions, respectively. i, Average NE rates in the region around droplet (AD, both during initial droplet
relaxation (green) and at its final stages (pink)) and in the MPZ, both in the absence (cyan) and presence (orange) of blebbistatin. j, Experimentally
measured and simulated dynamics of cell shapes in both the MPZ and the PSM, showing faster dynamics in the MPZ and largely static cell boundaries in
the PSM. Experimental data are an average intensity projection of a confocal section timelapse. k, MSRD shows uncaging behaviour for the MPZ but caged
for the PSM of both wild-type and cdh2−/− embryos. Error bands, s.d. (d,g) and s.e.m (i,k); N = 10 (d), N = 298 from 4 embryos (g–i, AD), N = 396 from
3 embryos (g–i, MPZ). Data in f, MPZ bleb in i and MSRD for MPZ and PSM in wild-type embryos in k were reanalysed from ref. 15. Scale bars, 10 μm,
unless otherwise stated.

diagram surrounding the structural transition between from vanishing junctional tensions in fluid states17, the large shape
non-confluent and confluent states. This non-monotonic behaviour factors reported here are due to spatiotemporal tension fluctua-
mirrors the behaviour of NE and cell movements, but care must be tions (Fig. 4h), with cell–cell contacts maintaining finite tensions
taken in inferring the physical state of the system solely from cel- (Supplementary Fig. 4 and Supplementary Video 6), as observed
lular movements (Extended Data Fig. 2). While cell density and cell experimentally31,42. In particular, we find the presence of both fluid
shape are the control parameters associated with jamming transi- and solid tissue states for adhesion values in the range 0 < W/T0 < 2
tions and density-independent transitions, respectively, our results that would correspond solely to solid states in previous equilibrium
show that tension fluctuations control a distinct rigidity transition descriptions. These results indicate that the tissue fluid or solid state
in both confluent and non-confluent states, as suggested recently cannot be inferred from static measurements of cell shape factor
for confluent states33. only, as the tissue physical state and its structure depend strongly on
To relate the mechanics of the system to its structure in the pres- the magnitude of tension fluctuations (Fig. 4g,h).
ence of tension dynamics, we study the configurations of the system.
Tension fluctuations generally promote transitions from confluent Fluidization of embryonic tissues
to non-confluent states by opening up extracellular spaces at weak Our theoretical results indicate that in the presence of active ten-
regions (Fig. 4f), implying that larger adhesion values are required sion fluctuations, active NE events control stress relaxation at long
to reach confluent states for increasing ΔT/T0. This effect becomes timescales both in confluent and non-confluent tissues. To experi-
negligible for large cell densities (ρ > 1) and volume fraction is mentally address the role of tension fluctuations in stress relaxation
then solely determined by cell density ρ and relative cell adhesion and tissue fluidization, we employed magnetically responsive oil
W/T0 (Supplementary Fig. 3). While cell shapes are only moder- droplets to directly measure strain and stress relaxation in posterior
ately affected by tension fluctuations in non-confluent regimes, tissues of developing zebrafish embryos (Fig. 5a), as these tissues
with low shape factors associated with rounder cells, increasing ten- have been previously shown to be in a fluid-like state15. After inject-
sion fluctuations in confluent regimes leads to substantially larger ing a single droplet in the mesodermal progenitor zone (MPZ),
shape factors (Fig. 4g,h). In contrast to density-independent transi- we induced large droplet deformations of multiple cell diameters
tions, where large shape factors (above approximately 3.81) result by applying a controlled magnetic field for a 15 min period15,43

864 Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics

NaTurE PhysiCs Articles
(Fig. 5b,c and Methods). We subsequently monitored the droplet author contributions and competing interests; and statements of
relaxation for 1 h after removing the magnetic field and measured data and code availability are available at https://doi.org/10.1038/
the decay of stress in the tissue (Fig. 5d and Methods). We observed s41567-021-01215-1.
an initial fast decay (~30 s) followed by a very slow relaxation at
long timescales that displays the stretched-exponential behaviour Received: 1 June 2020; Accepted: 25 February 2021;
predicted theoretically for a fluctuation-induced stress relaxation Published online: 12 April 2021
(Fig. 4a). This slow relaxation was previously interpreted as a
yield stress due to limitations in the measurement timescales15, a
well-known limitation in mechanical measurements of complex References
1. Heisenberg, C.-P. & Bellaiche, Y. Forces in tissue morphogenesis and
materials39. The present long timescale measurements reveal a very patterning. Cell 153, 948–962 (2013).
slow stress relaxation and indicate that posterior tissues completely 2. Guillot, C. & Lecuit, T. Mechanics of epithelial tissue homeostasis and
fluidize on timescales of approximately 1 h, enabling nearly com- morphogenesis. Science 340, 1185–1189 (2013).
plete remodelling of posterior tissues during axis elongation, as this 3. Angelini, T. E. et al. Glass-like dynamics of collective cell migration. Proc.
occurs at rates of approximately 45 μm h−1 (refs. 44,45). Natl Acad. Sci. USA 108, 4714–4719 (2011).
4. Malinverno, C. et al. Endocytic reawakening of motility in jammed epithelia.
To address the role of active NE (T1 transitions) in tissue flu- Nat. Mater. 16, 587–596 (2017).
idization (Fig. 4c), we quantified the characteristics of NE events 5. Park, J.-A. et al. Unjamming and cell shape in the asthmatic airway
(Fig. 5e) during droplet relaxation in the MPZ, as cells in this epithelium. Nat. Mater. 14, 1040–1048 (2015).
region display considerable cell–cell contact fluctuations (Fig. 5f 6. Palamidessi, A. et al. Unjamming overcomes kinetic and proliferation arrest
and Methods). Measurement of cumulative NE events in the MPZ in terminally differentiated cells and promotes collective motility of
carcinoma. Nat. Mater. 18, 1252–1263 (2019).
(away from the droplet) shows a linear increase over time (Fig. 5g 7. Sadati, M., Qazvini, N. T., Krishnan, R., Park, C. Y. & Fredberg, J. J.
and Methods), indicating an approximately constant NE rate in the Differentiation. Differentiation 86, 121–125 (2013).
MPZ with characteristic timescales of tens of minutes (Fig. 5h,i). 8. Harris, A. R. et al. Characterizing the mechanics of cultured cell monolayers.
Inhibiting myosin II activity with blebbistatin causes a reduction in Proc. Natl Acad. Sci. USA 109, 16449–16454 (2012).
the magnitude of cell–cell contact fluctuations (Fig. 5f and Methods) 9. Prakash, V. N., Bull, M. S. & Prakash, M. Motility-induced fracture reveals a
ductile-to-brittle crossover in a simple animal’s epithelia. Nat. Phys. https://
and a decrease of nearly 50% in the NE rate (Fig. 5i), indicating doi.org/10.1038/s41567-020-01134-7 (2021).
that the measured NE events in the MPZ are actively induced by 10. Schotz, E. M., Lanio, M., Talbot, J. A. & Manning, M. L. Glassy dynamics in
actomyosin-generated tension fluctuations. While NE events in the three-dimensional embryonic tissues. J. R. Soc. Interface 10, 20130726 (2013).
tissue directly adjacent to the droplet display the same NE rate as the 11. Petridou, N. I., Grigolon, S., Salbreux, G., Hannezo, E. & Heisenberg, C.-P.
rest of the MPZ in the final stages of droplet relaxation (40–55 min; Fluidization-mediated tissue spreading by mitotic cell rounding and
non-canonical Wnt signalling. Nat. Cell Biol. 21, 169–178 (2019).
Fig. 5g–i), the NE rate close to the droplet is higher than the rest 12. Atia, L. et al. Geometric constraints during epithelial jamming. Nat. Phys. 14,
of the MPZ during the beginning of droplet relaxation (0–15 min; 613–620 (2018).
Fig. 5g–i). This observed initial excess of NE events close to the 13. Firmino, J., Rocancourt, D., Saadaoui, M., Moreau, C. & Gros, J. Cell division
droplet indicates the presence of passive NE events caused by the drives epithelial cell rearrangements during gastrulation in chick. Dev. Cell
36, 249–261 (2016).
capillary stresses imposed by the droplet on the tissue during its
14. Ranft, J. et al. Fluidization of tissues by cell division and apoptosis. Proc. Natl
relaxation, as predicted theoretically (Fig. 4c). The measured low Acad. Sci. USA 107, 20863–20868 (2010).
and constant rate of active NE events in the MPZ is consistent with 15. Mongera, A. et al. A fluid-to-solid jamming transition underlies vertebrate
the measured stretched-exponential behaviour of the stress relax- body axis elongation. Nature 561, 401–405 (2018).
ation at long timescales (tens of minutes), and indicates that tissue 16. Noll, N., Mani, M., Heemskerk, I., Streichan, S. J. & Shraiman, B. I. Active
tension network model suggests an exotic mechanical state realized in
fluidization is largely caused by active NE events arising from ten- epithelial tissues. Nat. Phys. 13, 1221–1226 (2017).
sion fluctuations at cell–cell contacts. 17. Bi, D., Lopez, J. H., Schwarz, J. M. & Manning, M. L. A density-independent
To characterize the relative roles of tension fluctuations and rigidity transition in biological tissues. Nat. Phys. 11, 1074–1079 (2015).
cell–cell adhesion in rigidity transitions, we used N-cadherin 18. Kim, S. & Hilgenfeldt, S. Cell shapes and patterns as quantitative indicators of
(cdh2−/−) mutants lacking N-cadherin mediated cell–cell adhesion46 tissue stress in the plant epidermis. Soft Matter 11, 7270–7275 (2015).
19. Bi, D., Lopez, J. H., Schwarz, J. M. & Manning, M. L. Energy barriers and cell
(Methods). While the MPZ tissue in wild-type embryos is fluidized migration in densely packed tissues. Soft Matter 10, 1885–1890 (2014).
by active NE events (Fig. 5d–i) and displays cellular movements 20. Staple, D. et al. Mechanics and remodelling of cell packings in epithelia. Eur.
and mean squared relative displacements (MSRD) compatible with Phys. J. E 33, 117–127 (2010).
cell uncaging (Fig. 5j,k), the presomitic mesoderm (PSM) has been 21. Farhadifar, R., Röper, J.-C., Aigouy, B., Eaton, S. & Jülicher, F. The influence
shown to be in a solid-like state15, with cells caged by their neigh- of cell mechanics, cell–cell interactions, and proliferation on epithelial
packing. Curr. Biol. 17, 2095–2104 (2007).
bours (MSRD ≪ 1 at 30 min; Fig. 5j,k). Our measurements indicate 22. Graner, Fmc & Glazier, J. A. Simulation of biological cell sorting
that cell movements in the PSM of N-cadherin (cdh2−/−) mutants, a using a two-dimensional extended Potts model. Phys. Rev. Lett. 69,
tissue that has been shown to be characterized by reduced cell vol- 2013–2016 (1992).
ume fraction and lower yield stress but similar cell–cell contact fluc- 23. Chiang, M. & Marenduzzo, D. Glass transitions in the cellular Potts model.
tuations as wild-type embryos15, display the same MSRD behaviour Europhys. Lett. 116, 28009 (2016).
24. Boromand, A., Signoriello, A., Ye, F., O’Hern, C. S. & Shattuck, M. D.
as wild-type embryos (Fig. 5k). These results indicate that lowering Jamming of deformable polygons. Phys. Rev. Lett. 121, 248003 (2018).
cell–cell adhesion while maintaining low cell–cell contact fluctua- 25. Perrone, M. C., Veldhuis, J. H. & Brodland, G. W. Non-straight cell edges are
tions does not allow cells to uncage themselves to remodel the tissue. important to invasion and engulfment as demonstrated by cell mechanics
Altogether, both our computational and experimental results model. Biomech. Model. Mechanobiol. 15, 405–418 (2016).
indicate that actomyosin-generated tension fluctuations actively 26. Graner, F. & Sawada, Y. Can surface adhesion drive cell rearrangement? Part
II: a geometrical model. J. Theor. Biol. 164, 477–506 (1993).
drive structural rearrangements that cause stress relaxation in the 27. Teomy, E., Kessler, D. A. & Levine, H. Confluent and nonconfluent phases in
tissue, thereby controlling tissue fluidization and enabling tissue a model of cell tissue. Phys. Rev. E 98, 042418 (2018).
remodelling. 28. Boromand, A. et al. The role of deformability in determining the structural
and mechanical properties of bubbles and emulsions. Soft Matter 15,
5854–5865 (2019).
Online content 29. Henkes, S., Fily, Y. & Marchetti, M. C. Active jamming: self-propelled soft
Any methods, additional references, Nature Research report- particles at high density. Phys. Rev. E 84, 040301 (2011).
ing summaries, source data, extended data, supplementary infor- 30. Bi, D., Yang, X., Marchetti, M. C. & Manning, M. L. Motility-driven glass and
mation, acknowledgements, peer review information; details of jamming transitions in biological tissues. Phys. Rev. X 6, 021011 (2016).

Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics 865

Articles NaTurE PhysiCs
31. Sumi, A. et al. Adherens junction length during tissue contraction is 40. Abou, B., Bonn, D. & Meunier, J. Aging dynamics in a colloidal glass. Phys.
controlled by the mechanosensitive activity of actomyosin and junctional Rev. E 64, 243–246 (2001).
recycling. Dev. Cell 47, 453–463.e3 (2018). 41. Phillips, J. C. Stretched exponential relaxation in molecular and electronic
32. Curran, S. et al. Myosin II controls junction fluctuations to guide epithelial glasses. Rep. Prog. Phys. 59, 1133–1207 (1996).
tissue ordering. Dev. Cell 43, 480–492.e6 (2017). 42. Bambardekar, K., Clément, R., Blanc, O., Chardès, C. & Lenne, P.-F. Direct
33. Krajnc, M. Solid–fluid transition and cell sorting in epithelia with junctional laser manipulation reveals the mechanics of cell contacts in vivo. Proc. Natl
tension fluctuations. Soft Matter 16, 3209–3215 (2020). Acad. Sci. USA 112, 1416–1421 (2015).
34. Aveyard, R. et al. Flocculation transitions of weakly charged oil- 43. Serwane, F. et al. In vivo quantification of spatially varying mechanical
in-water emulsions stabilized by different surfactants. Langmuir 18, properties in developing tissues. Nat. Methods 14, 181–186 (2017).
3487–3494 (2002). 44. Lawton, A. K. et al. Regulated tissue fluidity steers zebrafish body elongation.
35. Trappe, V., Prasad, V., Cipelletti, L., Segre, P. N. & Weitz, D. A. Jamming Development 140, 573–582 (2013).
phase diagram for attractive particles. Nature 411, 772–775 (2001). 45. Banavar, S. P. et al. Mechanical control of tissue shape and morphogenetic
36. O’Hern, C. S., Silbert, L. E., Liu, A. J. & Nagel, S. R. Jamming at zero flows during vertebrate body axis elongation. Preprint at bioRxiv https://doi.
temperature and zero applied stress: the epitome of disorder. Phys. Rev. E 68, org/10.1101/2020.06.17.157586 (2020).
011306 (2003). 46. Lele, Z. et al. parachute/n-cadherin is required for morphogenesis and
37. Wang, X. et al. Anisotropy links cell shapes to tissue flow during convergent maintained integrity of the zebrafish neural tube. Development 129,
extension. Proc. Natl Acad. Sci. USA 117, 13541–13551 (2020). 3281–3294 (2002).
38. Kim, S., Wang, Y. & Hilgenfeldt, S. Universal features of metastable state
energies in cellular matter. Phys. Rev. Lett. 120, 248001 (2018). Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
39. Bonn, D., Denn, M. M., Berthier, L., Divoux, T. & Manneville, S. Yield stress published maps and institutional affiliations.
materials in soft condensed matter. Rev. Mod. Phys. 89, 035005 (2017). © The Author(s), under exclusive licence to Springer Nature Limited 2021

866 Nature Physics | VOL 17 | July 2021 | 859–866 | www.nature.com/naturephysics

NaTurE PhysiCs Articles
Methods changes (ΔW/T0 = −0.02). Similarly, the other branch was found by initializing
Pressure and volume relation. The (osmotic) pressure P of a cell has been the system at zero adhesion strength (W/T0 = 0) and progressively increasing it to
experimentally shown to vary with its volume according to P = K/V (ref. 47), W/T0 = 1 by small increments (ΔW/T0 = 0.02). After each adhesion adjustment, the
meaning that changes in the osmotic pressure difference between the inside (P) and system is relaxed to local equilibrium states.
outside (P0) of the cell lead to changes in the cell volume V, with K characterizing
the cell compressibility (we assumed for simplicity that cells are never compressed Introducing spaces between cells. Extracellular spaces are first introduced to
close to their dry mass limit). In our 2D description, the cell area A plays the role initial configurations of confluent states as cells with different properties. Each
of the volume, so that P = K/A, with the characteristic vertex is replaced by a small triangular extracellular space centred around the
√ cell size L0 (or preferred area
A0 = L20) being set by this relation, namely L0 ≡ K/P0 . However, while we use original vertex position, and with the three new vertices located on the each of
the experimentally measured relation between osmotic pressure and cell volume47 the original edges and of 1% their original length. These extracellular spaces
(area in 2D), the specific P–V (or P–A) functional form does not change our results then behave like ‘cells’ with different properties (see above) and their size and
qualitatively as long as the cell volume decreases with increasing applied osmotic geometrical features are determined by force balance at the vertices, as is the case
pressure, namely P − P0 = −Γ(A − A0) (with Γ being a coefficient characterizing the for cells too. When two extracellular space ‘cells’ become neighbours, they are
cell’s compressibility), and the deviations from the preferred cell area in 2D (A0) merged. This implementation allows non-confluent states in vertex models and
are mild. is different from previous descriptions of non-confluent states, which used centre
particles26,27 or deformable particles models28.
Fixed point tension and tension dynamics. The fixed point tension,
T0ij, corresponds to an average effective tension at cell–cell contacts and has Intermediate non-physical vertices. Intermediate vertices are introduced
contributions from the average actomyosin-generated cortical tension T0 in each for both cell–cell contacts and free cell boundaries to allow for more realistic cell
cell and the average cell–cell adhesion strength W, so that T0ij = 2T0 − W shapes. With intermediate vertices, individual edges consist of linear segments
(Fig. 1b). At free cell boundaries (cell boundaries contacting the extracellular joined together to form a piecewise linear edge. The desired segment length is
space), the average effective tension is just the average cortical tension of the cell, introduced as a parameter and the number of intermediate vertices for each
namely T0ij = T0. Here we assume the average tension T0 not to change with the edge is equal to the closest lower integer given by the ratio of instantaneous edge
cell perimeter, as no current experimental observations suggest that dependence. length to the segment length criterion. When the edge length increases (decreases),
In situations where the cell shape becomes very anisotropic or contorted an intermediate vertex can be added (deleted) following the criterion just
enough to have very large perimeters, a dependence of the average effective described. As the number of intermediate vertices changes, intermediate
tension on the perimeter could potentially arise from limitations in plasma vertex positions are reassigned uniformly along the edge. If the longest segment
membrane availability. is longer than twice the shortest segment in a given edge, intermediate vertex
The stochastic dynamics of tensions depends on the magnitude of tension positions are also reassigned uniformly along the piecewise linear line. This
fluctuations, which is ΔT at cell–cell contacts and ΔT/2 at free cell boundaries (cell implementation of intermediate vertices is similar to the one used for confluent
boundaries contacting the extracellular space). equilibrium systems25.
While we considered all vertices to have the same drag coefficient, physical
Parameter estimation. Since the relaxation timescale τR has been measured to be and non-physical vertices may, in general, be characterized by different drag
much smaller than the persistence timescale of tension variations (τR ≲ 20 s coefficients. This approximation is not due to technical reasons, as the simulation
(refs. 42,43); τT ≃ 90 s (ref. 15)), we set the ratio to τT/τR = 10. While the values of framework presented here can simulate different drag coefficients at different
osmotic pressure are unknown in vivo, they are expected to be larger than cortical vertices different drag coefficients. Previous studies have shown that the measured
stresses47, so that P0L0/T0 ≫ 1. Consequently, we fix P0L0/T0 = 10 for all simulations. stress relaxation timescales for movement of vertices is much smaller (~1–20 s
This ensures relatively mild cell size variations, as observed experimentally15. (refs. 42,43)) than the persistence timescale of active tension fluctuations, which
Fixing these parameters reduces the parameter space to the normalized amplitude is approximately 100 s (ref. 15). Since we are interested in the behaviour of the
of tension fluctuations ΔT/T0, the ratio W/T0 of average adhesion strength and system at long timescales (longer than the persistence timescale of active tension
average cortical tension, and the ratio NL20 /AT of the cell’s total preferred area and fluctuations (t ≫ τT)) and the timescales associated with viscous relaxation at
the total available area. vertices are much smaller than the persistence timescale of active stresses (τT ≫ τR),
we neglected the differences in drag coefficient at the different vertices, as they are
Numerical integration. The dimensionless version of the governing equations irrelevant in this limit.
(equations (1) and (2)) was integrated numerically using the Euler–Maruyama
method with a time step, Δt. We used Δt = 0.005τR for all simulations to have a Treatment of topological transitions. T1 transitions occurs in our description
time resolution much smaller than the smallest characteristic timescale in the when a given edge length is shorter than a critical length ℓc = 0.01 × 2πL0.
system, τR. When a T1 transition leads to the formation of a new vertex between
three cells, we introduce a small triangular extracellular space ‘cell’ and let it evolve
Initial configuration generation protocol. A polygonal tiling of confluent states in time, as described above. Free cell boundaries (a boundary of a cell and the
is first generated by random Poisson Voronoi tessellation in a square periodic extracellular space) occasionally intersect each other due to the system
box of total area AT. Before the introduction of extracellular spaces, the initial dynamics. This event corresponds to the formation of a new contact between
confluent configuration is first annealed to a local equilibrium state to prevent two cells. Therefore, when an intersection between any two free boundary
sudden adjustment of cell shape with extracellular spaces. Extracellular spaces are edges is detected, a new cell–cell contact is introduced, splitting the extracellular
introduced by replacing each vertex in the confluent configurations with a small space in two.
triangular extracellular space ‘cell’ centred around the original vertex position, While not simulated here, cell divisions and cell death, which
and with the three new vertices located on the each of the original edges and of would also introduce topological transitions, are straightforward to simulate
1% their original length. Small tension fluctuations (ΔT/T0 = 0.5) are applied for in this framework. Since these events would also induce cell
a duration τT. The resulting configuration is used as an initial configuration for rearrangements, they are also likely to enable tissue fluidization, as
simulations of both equilibrium configurations and dynamics. previously suggested5,11,13,14.
As for any model of multicellular systems, some of the observed
configurations depend on the initialization protocol. Regular hexagonal packings Cell trajectories and MSD. To obtain the MSD, we first computed the cell centres
can be obtained by choosing regularly arranged initial cellular configurations. (centroid of polygon), rC,i, for each time point, t, based on vertex positions. Then,
However, since most embryonic tissues typically show topologically disordered the cell trajectories were obtained by monitoring the changes of cell centres. Using
configurations, we used random Poisson Voronoi tessellation as an initial the cell trajectories rC,i(t), we obtained the MSD according to
configuration. 1 ∑ 2
MSD(t) = (rC,i (t0 + t) − rC,i (t)) . (3)
N i
Equilibrium configurations and quasi-static changes. Equilibrium states were
obtained by quenching the system from an initial state with tension fluctuations
(ΔT/T0 = 0.5) to an equilibrium state (ΔT/T0 = 0) for each parameter set. While we MSD values were averaged over all cells and also all initial timepoints t0 for any
observe small changes in the resulting configurations if the magnitude of the initial given set of model parameters.
tension fluctuations is varied in the annealing protocol, our results, including the
existence of the structural transition, remain qualitatively unchanged. Quasi-static Application of a step strain. Since tissue fluidization in biological tissues is
changes of the relative adhesion at equilibrium where performed by first quenching associated with their nonlinear mechanical response15, we impose large strains ϵxy
the system to equilibrium at a given parameter value, and then performing a small of 150%. The large step strain was applied by deforming the simulation box from
change in the parameter (W/T0) and letting the system relax. Specifically, the a square to a parallelogram of the target shear strain, namely 150%. An affine
system was initialized at a large value of the relative adhesion strength that ensured deformation is applied to all vertex positions when imposing the step strain. Lees–
a confluent state (W/T0 = 1) and then quenched to a local equilibrium state. The Edwards periodic boundary conditions were imposed throughout the simulation to
relative adhesion strength was then progressively reduced to W/T0 = 0 by small avoid mismatch of cell geometry across system boundaries.

Nature Physics | www.nature.com/naturephysics

Articles NaTurE PhysiCs
Stress calculation. The non-dimensional stress tensor can be computed from diluted in filtered 3M Novec 7300 fluorocarbon oil. To prevent non-specific
the transient tissue geometry with knowledge of junctional tensions and cell adhesion between cells and droplets, a fluorinated Krytox-PEG(600) surfactant
pressures48,49 (lowercase indicates normalized quantities), namely (008-fluorosurfactant, RAN Biotechnologies53) was diluted in the ferrofluid at a
  2.5% (w/w) concentration. The ferrofluid was calibrated before each
∑ ∑ lij,m lij,n experiment as previously described43, so that the applied magnetic stresses are
σ mn = ρ − Δpi ai δ mn + tij  , (4) known. Once prepared and calibrated, the ferrofluid was injected in the
|lij |
i:cell ij:edge MPZ tissue of phenotypically wild-type zebrafish embryos at the 4-6 somite stage
to form droplets of 30–40 μm diameter, as previously described15,43. Imaging
where Δpi ≡ pi − p0, m and n are indices indicating the spatial direction (m = x, y of droplets started at least 1.5 h after the injection to let the tissue fully recover
and n = x, y), ai is the dimensionless cell area, lij is the vector form of the edge length from it.
between cells i and j, and δmn is the Kronecker delta. The shear-stress term can be
written as Magnetic actuation of ferrofluid microdroplets. Actuation of ferrofluid droplets
  was performed following the previously described protocol43. Briefly, ferrofluid
∑ lij,x lij,y droplets were actuated by a uniform and constant magnetic field43 applied along
σ xy = ρ  tij  . (5) the direction of smaller droplet deformation, as given by the smallest semi-axis
|lij |
ij:edge of elliptical deformation before the magnetic field was applied. The droplet shape
was monitored before (15 min) and during magnetic actuation (15 min), as well as
during droplet relaxation (1 h).
Active shear stress. Due to tension fluctuations, the macroscopic shear stress
shows fluctuations around zero. To quantify the magnitude of these active Measurements of strain and stress relaxation. On application of a uniform,
shear-stress fluctuations, the shear-stress values are monitored over a time interval constant magnetic field, ferrofluid droplets acquire an ellipsoidal shape,
of 10τR and their standard deviation is computed for each parameter set (no elongated along the direction of the applied magnetic field and symmetrical
macroscopic imposed strain). The level of active shear stress corresponds to the about it43,54. The observed confocal section through the middle of the droplet was
computed standard deviation of shear stress, and increases approximately linearly fitted to an ellipse of short and long semi-axes a and b, respectively, as previously
with the magnitude of tension fluctuations for a given relative cell adhesion described43. The droplet strain, (b − b0)/R, is defined as the ratio of its elongation
strength (Supplementary Section 2). along the direction of applied magnetic field, namely (b − b0) (with b0 being the
value of b just before magnetic actuation), and the droplet radius R, and can be
Zebrafish husbandry, lines and experimental manipulations. Zebrafish (Danio monitored over time by measuring the change in the droplet long semi-axis b(t).
rerio) were maintained under standard conditions50. The cdh2tm101 mutant line46 was The stresses at supracellular (tissue) scales associated with a particular droplet
used to disrupt adhesions between cells, otherwise phenotypically wild-type lines deformation were obtained from the elliptical droplet deformation, as previously
were used. Animal husbandry and experiments were done according to protocols described15, namely σ A = 2γ (Hb − Ha ), where γ is the droplet interfacial tension
approved by the Institutional Animal Care and Use Committee (IACUC) at the and Hb and Ha are the mean curvatures of the droplet ellipsoidal shape at the
University of California Santa Barbara. Transgenic lines Tg(hsp70:secP-mCherry)p1 intersection of the principal axis with the fitted ellipsoid. Since the droplet shape
(ref. 51) and Tg(h2afva:eGFP)kca6 were used to visualize extracellular spaces is that of a prolate spheroid, Hb and Ha read Hb = b/a2 and Ha = 1/2a + a/(2b2). By
and nuclei, respectively. Both Tg(bact2:mem-neonGreen-neonGreen)hm40 and monitoring the changes in droplet shape over time, we obtained the time evolution
Tg(actb2:memCherry2)hm29 transgenic lines were used to visualize cell membranes of both the mean curvatures, namely Hb(t) and Ha(t). The droplet interfacial
in different experiments. In some cases, membranes were labelled by injection with tension γ was measured in each experiment in situ and in vivo, as previously
80–100 pg membrane-GFP messenger RNA at the 1-2 cell stage. described15,43. Using the measured value of the interfacial tension for each droplet
and the time evolution of the mean curvatures Hb(t) and Ha(t), we obtained the
Imaging. In all cases, 8–10 somite stage zebrafish embryos were mounted in 1% time evolution of the stresses during relaxation using σ A (t) = 2γ (Hb (t) − Ha (t)).
low-melting point agarose in a glass bottom Petri dish (MatTek Corporation) for
a dorsal view of the tailbud and imaged at 25 °C using a ×40 water immersion NE analysis. One-hour-long confocal timelapses of the MPZ region of
objective (LD C-Apochromat 1.1 W, Carl Zeiss) on an inverted Zeiss Laser membrane-labelled embryos containing a previously inserted droplet were
Scanning Confocal (LSM 710, Carl Zeiss). acquired at 1 frame every 5 s. A region of the MPZ away from the droplet and a
For kymograph of cell–cell contacts, confocal timelapse data of the MPZ region adjacent to the droplet were cropped and the time at which T1 transitions
and the PSM regions in Tg(h2afva:eGFP)kca6 × Tg(actb2:memCherry2)hm29 occurred in each region was recorded. T1 transitions were detected manually by
double transgenic embryos were acquired (2 s time intervals for 30 min). The inspecting cell configurations between frames. To improve the accuracy of the
intensity profile of a single junction was tracked over 300 s and measured over a analysis, smaller sections of each region containing approximately 10 to 15 cells
5-pixel-width segment line to minimize the error induced by the estimation of the were further cropped and analysed separately. The cumulative NE were obtained
junction location. by adding up the NE events occurring over time in each region. The NE rate was
Images of extracellular spaces were obtained using an outcross of obtained by taking the derivative of the cumulative NE function.
Tg(hsp70:secP-mCherry)p1 (refs. 51) and Tg(bact2:mem-neonGreen-neonGree
n)hm40, to label the extracellular spaces and cell membranes, respectively. To Reporting Summary. Further information on research design is available in the
trigger the expression of mCherry in the extracellular space, a 1 h heat shock Nature Research Reporting Summary linked to this article.
in a water bath at 39 °C was performed when the embryos were at 75% epiboly
stage. Confocal sections through the PSM and the MPZ tissues were subsequently
acquired. Data availability
Imaging of ferrofluid droplets was done as previously reported15,43. Source data are provided with this paper.
Briefly, images of a droplet previously inserted in the MPZ region of the
embryo were taken every 5 s. Ferrofluid droplets were fluorescently
labelled using a custom-synthesized fluorous cyanine dye FCy5 dissolved in the Code availability
fluorocarbon-based ferrofluid oil at a final concentration of 25 μM (ref. 52). The code developed for this paper is available in Supplementary Software 1.

Tracking of cell movements in the PSM of N-cadherin mutants. Both mutant

(cdh2−/−) and sibling (cdh2+/+;+/−) embryos were injected with 80–100 pg each of References
H2B-mRFP mRNA and membrane-GFP mRNA at the 1-2 cell stage to label nuclei 47. Guo, M. et al. Cell volume change through water efflux impacts
and membranes, respectively. Confocal stacks through the PSM were acquired cell stiffness and stem cell fate. Proc. Natl Acad. Sci. USA 114,
with a z-step size of 1 μm and time interval of 1 min for 30 min, and processed E8618–E8627 (2017).
using Imaris (Bitplane). Data were smoothed using a 1 pixel Gaussian filter, to 48. Ishihara, S. & Sugimura, K. Bayesian inference of force dynamics during
correct for photobleaching over time the normalize timepoints function was morphogenesis. J. Theor. Biol. 313, 201–211 (2012).
used, then attenuation correction was applied to correct for z attenuation. After 49. Yang, X. et al. Correlating cell shape and cellular stress in motile confluent
processing, data were cropped to the tissue of interest, then nuclei were detected tissues. Proc. Natl Acad. Sci. USA 114, 12663–12668 (2017).
using the spots function, and tracked using the Brownian motion algorithm. 50. Nüsslein-Volhard, C. & Dahm, R. Zebrafish: A Practical Approach (Oxford
Nuclei positions were output for further analyses. The MSRD is calculated from Univ. Press, 2002).
the cell trajectories and normalized using the cell diameter d. The normalized 51. Wang, J. et al. Anosmin1 shuttles Fgf to facilitate its diffusion, increase
MSRD is MSRD/d2 (ref. 15). its local concentration, and induce sensory organs. Dev. Cell 46,
751–766.e12 (2018).
Generation and injection of ferrofluid droplets. Ferrofluid droplets were 52. Lim, I. et al. Fluorous soluble cyanine dyes for visualizing perfluorocarbons
prepared as previously described15,43. Briefly, DFF1 ferrofluid (Ferrotec) was in living systems. J. Am. Chem. Soc. 142, 16072–16081 (2020).

Nature Physics | www.nature.com/naturephysics

NaTurE PhysiCs Articles
53. Holtze, C. et al. Biocompatible surfactants for water-in-fluorocarbon Foundation) under Germany’s Excellence Strategy - EXC 2068 - 390729961 - Cluster of
emulsions. Lab Chip 8, 1632–1639 (2008). Excellence Physics of Life of TU Dresden.
54. Rowghanian, P., Meinhart, C. D. & Campas, O. Dynamics of ferrofluid drop
deformations under spatially uniform magnetic fields. J. Fluid Mech. 802, Author contributions
245–262 (2016). S.K. and O.C. designed research; S.K. implemented and performed the simulations; M.P.
and G.A.S.-V. performed experiments; S.K., M.P. and G.A.S.-V. analysed data; S.K. and
O.C. wrote the paper; O.C. supervised the project.
Acknowledgements
We thank all members of the Campàs group for their comments and help, P. Rowghanian Competing interests
for help with cell segmentation, D. Kealhofer and E. Shelton for technical help, B. Shelby The authors declare no competing interests.
and the UCSB Animal Research Center for support with zebrafish, I. Lim and E. Sletten
(University of California, Los Angeles) for sharing custom-made fluorinated dyes,
and H. Knaut (New York University) and S. Megason (Harvard University) for kindly Additional information
providing the Tg(hsp70:secP-mCherry)p1 and Tg(actb2:memCherry2)hm29 transgenic Extended data is available for this paper at https://doi.org/10.1038/s41567-021-01215-1.
lines, respectively. The Tg(actb2:mem-neonGreen-neonGreen)hm40 line was generously Supplementary information The online version contains supplementary material
provided before publication by T. Kawanishi and I. Swinburne in S. Megason’s lab available at https://doi.org/10.1038/s41567-021-01215-1.
(Harvard University). This work was supported by the Eunice Kennedy Shriver National
Correspondence and requests for materials should be addressed to O.C.
Institute of Child Health and Human Development of the National Institutes of Health
(R01HD095797 to O.C.). We acknowledge support from the Center for Scientific Peer review information Nature Physics thanks the anonymous reviewers for their
Computing from the CNSI, MRL: an NSF MRSEC (DMR-1720256) and NSF CNS- contribution to the peer review of this work.
1725797, as well as from the Deutsche Forschungsgemeinschaft (DFG, German Research Reprints and permissions information is available at www.nature.com/reprints.

Nature Physics | www.nature.com/naturephysics

Articles NaTurE PhysiCs

Extended Data Fig. 1 | Power law relation between NE rate and MSD at long timescales. Power law relation between long time MSD values and NE rate
when the systems are close to confluence for high adhesion levels. NE rate and longtime MSD show a power law relation with an exponent of 0.75.

Nature Physics | www.nature.com/naturephysics

NaTurE PhysiCs Articles

Extended Data Fig. 2 | Comparison of solid/fluid phase diagrams obtained from stress relaxation and from cell movements. Solid/fluid phase diagrams
determined by mechanical measurement of stress relaxation (left) and cell movements, MSD=1/2 (middle) and MSD=1/4 (right). Green region indicates
confluent states.

Nature Physics | www.nature.com/naturephysics

 nature research | reporting summary
Corresponding author(s): Otger Campas
Last updated by author(s): Feb 15, 2021

Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see our Editorial Policies and the Editorial Policy Checklist.

Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.

A description of all covariates tested

A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)

For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.

For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.

Software and code

Policy information about availability of computer code
Data collection Commercial Zeiss Zen software was used to perform confocal imaging. Computer simulations were done in Matlab 9.6 (MathWorks).

Data analysis Commercial software used to analyze data: Imaris 9.3 (Bitplane) and Matlab 9.6 (MathWorks). ImageJ was also used to analyze data.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and
reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.

Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
- A description of any restrictions on data availability
April 2020

Source data supporting these findings are available online as Supplementary Data.

1
 nature research | reporting summary
Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

Life sciences study design

All studies must disclose on these points even when the disclosure is negative.
Sample size Sample size was chosen so that statistical variation of the data did not change considerably upon addition of more data points.

Data exclusions No data was excluded.

Replication Independent experiments were performed and statistical analysis done independently of these data sets. The results obtained were the same
within error bars in different, independent data sets.

Randomization No specific experimental groups were defined and all data was considered.

Blinding Analysis was done by automated software which was blind to data collection.

Reporting for specific materials, systems and methods

We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.

Materials & experimental systems Methods

n/a Involved in the study n/a Involved in the study
Antibodies ChIP-seq
Eukaryotic cell lines Flow cytometry
Palaeontology and archaeology MRI-based neuroimaging
Animals and other organisms
Human research participants
Clinical data
Dual use research of concern

Animals and other organisms

Policy information about studies involving animals; ARRIVE guidelines recommended for reporting animal research
Laboratory animals Zebrafish (Danio rerio) were used in this study. Since only embryos were studied, sex-specific experiments were not necessary, as
zebrafish embryos at the studies stage have not yet undergone sex determination.

Wild animals None.

Field-collected samples None.

Ethics oversight Animal husbandry and experiments were done according to protocols approved by the Institutional Animal Care and Use Committee
(IACUC) at the University of California Santa Barbara.
Note that full information on the approval of the study protocol must also be provided in the manuscript.
April 2020