organism and tissue growth - collège de france · 2019-12-11 · thomas lecuit 2018-2019 8 tissue...

Organism and Tissue Growth

Thomas Lecuitchaire: Dynamiques du vivant

Course 5: Growth control and mechanics

1

• Control of growth arrest at target size

Thomas LECUIT 2019-2020

— What is measured? Intrinsic « ruler » of growth: « size-meter » Organ intrinsic feedback of target organ size on cell growth/proliferation

— Organ and organism size may be regulated by: growth duration and/or growth rate.— Need to regulate growth arrest at the appropriate target size: process vs endpoint regulation.

arrest

time

mass/size

target size

2

1. Energy and mass/volume2. Pattern and mass/volume3. Mechanics and mass/volume

—Need to consider scaling between factors that promote growth and size itself

—Need to consider feedback mechanisms operating across scales (organ/tissue to cell)


• Conclusions

— Cell growth is controlled intrinsically— What are the mechanisms of organ specific tissue size sensing/measurement?

• Organ-size dependent negative feedback on cell growth/proliferation (chalone)• « Size-meter »: scaling of chalone concentration and tissue size

distance between activator and feedback inhibitor mechanical signal that scales with size

patterning cue: morphogen gradient

• Organ-size dependent growth inducer: Morphogen and growth arrest: Spatial model: gradient scaling and gradient slope. Temporal model: gradient scaling and rate of signal increase

— What is measured? Intrinsic « ruler » of growth: « size-meter », scaling.

• Importance of considering robustness of mechanisms: e.g. integral feedback

3

• Energy delivery/demand unbalance.

Programmed vs Self-organised regulation of Growth

• hierarchical• modular• deterministic rules (ie. genetically encoded)


• no hierarchy• feedbacks• statistical rules

4


Mechanical regulation of tissue size?

5

• Organs grow until they reach a fixed size characteristic of each organ/species• Growth is often uniform in the face of non uniform growth factor distribution

(morphogens)• Tissues can compensate for growth heterogeneities and reach a normal final

size arguing that cells assess their environment

—How do cells compare their growth rates?—How is growth uniform on average in a tissue?—How is tissue size determined?

• Cell-cell mechanical interactions

6Thomas LECUIT 2018-2019

Tissue Growth and Homeostasis

Thomas LECUIT 2018-2019 7


A B

�-tubulin Myosin II Phosphorylated Moesin

Cytokinetic ring

Supracellularcable

Myosin II

Supracellular ring

• cell division

• cell death/extrusion

Tissue growth



—Tissue growth depends on the balance between:• Rate of cell division• Rate of cell death and cell extrusion

—Tissue growth leads to homeostasis at a fixed size where cell growth/division and death/extrusion balance each other (unless growth is arrested): eg. gut.

• How is such a balance achieved?• Feedback interactions• Does growth impact on extrusion?• Does extrusion lead to compensatory growth?

—What are the consequence of differential tissue growth rates?• Cell and tissue competition (eg. tumours)• Tissue buckling (morphogenesis)

Biomechanical signalling


Biochemistry Mechanics

Information(mechano-chemistry)

9

driving tissue growth

Thomas LECUIT 2018-2019 10ASG. Curtis and GM Seehar, Nature 274, 52-53 (1978)

Cells in culture are subjected to low frequency stretching (60μm) for 1 hour

Control of cell division by mechanics: tension


Control of cell division by mechanics: geometry

11

• Increased cell-substrate adhesion causes reduction of cell height and increased cell diameter

• Cell proliferation increases as cells spread/adhere more to the substrate

• Is it adhesion per se or cell shape per se?

• Density dependence of cell proliferation and contact inhibition are characteristic of cells in culture

J. M. Folkman and A. Moscona, Nature 273, 345 (1978)


Geometric Control of Cell Life and DeathChristopher S. Chen, Milan Mrksich, Sui Huang,

George M. Whitesides, Donald E. Ingber*

12

Control of cell division by mechanics: geometry or adhesion?

—The role of surface of adhesion:• Cell division increases as the adhesive

island surface increases• Cell death by apoptosis decreases in

parallel

C. Chen et al., and D. Ingber. Science (1997):Vol. 276, Issue 5317, pp. 1425-1428DOI: 10.1126/science.276.5317.1425

—Disentangling cell geometry and substrate adhesion using a controlled micro fabricated substrate• The projected surface area, but not the

surface of adhesion to the ECM impacts on cell proliferation and cell death.

• Does this reflect cell shape per se or cell stresses?

ECM contact area


Control of cell growth by mechanics

13

—Disentangling the role of cell stress and cell shape • using micro fabricated substrates on colony island

of multiple cells (rather than single cells)• Cell division patterns depend on the geometry

colony islands

endothelial cells on fibronectin BrdU incorporation density maps

BrdU Proliferation map

C. Nelson et al. C. Chen. (2005) P.N.A.S. 102: 11594–11599 www.pnas.orgcgidoi10.1073pnas.0502575102

nN

traction force microscopy

• Traction Force microscopy confirms the deduced mechanical stress pattern.

Stress from FEM

• Use of finite element modelling (FEM) to deduce stress patterns in colony islands.

• Cell division patterns correlate with stress patterns that emerge from cell contractility and connectedness between cells and tissue geometry.

—Mapping stress pattern


—Disentangling the role of cell stress and cell shape

• Mechanical forces generated by the actomyosin cytoskeleton control cell proliferation patterns

Rho1 GDP Rho1 GTPGEF

GAP

MLCP

ROCKMRLC MRLC PDia

F-actin MyoIIminifilament

actinfilaments

passivecrosslinker

myosinIImotors

cell cortex

BrdUdensity

map

Y-27632

RhoV14*

dominant negative VE-cadherin (blocks mechanical coupling at cell junctions)

C. Nelson et al. C. Chen. (2005) P.N.A.S. 102: 11594–11599 www.pnas.orgcgidoi10.1073pnas.0502575102




15S. Dupont et al., and S. Piccolo. Nature (2011). 474::179-183

—Identification of transcription factors that mediate mechanically induced cell proliferation

a b

100

50

0si

Co.si

YZ1si

YZ240

kPa0.7kPa

mR

NA

exp

ress

ion Luciferase activity

90

60

30

0

CTGF ANKRD1 4xGTIIC-lux

10050

040 0.7

40 kPa 0.7 kPaNuclear

YAP/TAZ (%)

YAP

/TA

ZTO

TO3** ** **

β

β

β

• YAP/TAZ are known components of theHippo/LATS growth pathway that translocates in the nucleus once activated.• Substrate stiffness induces YAP activation

transcriptional targets of YAP

c

Micropillars

d

10,000 300 μm2

10,0

002,

025

1,02

430

0

Unpat

t.

100755025

0

Nuclear YAP/TAZ (%)

10,000 2,025 1,024 300 μm2Unpatt.

e

6,0003,000

0

300

μm2

Unpat

t.M

icrop

illars

Cell area(μm2)

10050

0

ECM area(μm2)

**

6,0003,000

0

YAP

/TA

ZYA

P/T

AZ

YAP

/TA

Z

NuclearYAP/TAZ (%)

• Cell substrate adhesion/geometry affects YAP/TAZ activation

• Cell projected area, but not adhesion per se activates YAP//TAZ

**



16

YAP

/TA

Z

Control C3

Noco

Lat.A

NSC23766

mR

NA

exp

ress

ion

CTGF

100806040200

C3La

t.ANSCNoc

o.Co.

1007550250

Nuc

lear

YA

P/T

AZ

(%)a b

C3La

t.ACo.

YAP

/TA

Z ** **

ANKRD1

S. Dupont et al., and S. Piccolo. Nature (2011). 474::179-183


• The actomyosin cytoskeleton (but not the microtubule network) is required for YAP/TAZ activation.


GAP

MLCP

ROCKMRLC MRLC PDia


C3 Transferase

Y-27632

BlebbistatinLatrunculin A no actin polymerisation

c

125100

755025

0

Luci

fera

se a

ctiv

ity

4xGTIIC-lux

Co.

Lat.A

Actin R62D

d

Luci

fera

se a

ctiv

ity

4xGTIIC-lux

e

YAP

/TA

Z

Rigidpillars

Elasticpillars

Elastic

Rigid

100

50

0

Nuc

lear

YAP

/TA

Z (%

)

f Y27

632

Blebbist

.

Co.

Y27632 Blebbist. Control

YAP

/TA

Z

Nuc

lear

YAP

/TA

Z (%

)

100

50

0

Y2763

2Bleb

bist.

Co.

403020100

actinfilaments

passivecrosslinker

myosinIImotors

cell cortex



17

30

20

10

0

30

20

10

0

60

40

20

0

60

40

20

0Pro

lifer

atio

n (%

) A

pop

tosi

s (%

)

ss s

d e

fAdhesive island area (μm2)

10,0

002,

025

1,02

430

010

,000

2,02

51,

024

300

10,0

002,

025

1,02

430

0

10,0

002,

025

1,02

430

010

,000

2,02

51,

024

300

Adhesive island area (μm2)

Pro

lifer

atio

n (%

) A

pop

tosi

s (%

)

Control +5SA-YAP siControl siYAP/TAZ siCo. + C3

n.s. n.s.

S. Dupont et al., and S. Piccolo. Nature (2011). 474::179-183


• Cell stretching induces activation of cell proliferation and inhibition of apoptosis

• YAP/TAZ are both required for this process

• Constitutively active YAP expression induces cell proliferation and inhibits apoptosis

• This effect is independent of upstream Hippo/LATS pathway activity (not shown)

YAP/TAZ — YAP/TAZ * (C.A.)

a

c b

40 kPa 40 kPa

siYZ1siCo. + C3

40 kPa1 kPa40 kPa40 kPa

siYZ2

Ad

ipog

enes

is(A

.U.)

siCo.

siYZ1

1 40 40 (kPa)

60

40

20

0Ost

eoge

nesi

s(A

.U.)

40302010

0

siCo.

siYZ1

40 40 40 (kPa)40 1 40

siYZ2

+C3

siCo.

[see also A. Engler et al, D. Discher. (2006) Cell 126, 677–689]

• YAP/TAZ similarly mediates the cell differentiation of stem cells dependent on substrate stiffness

• On soft substrates adipocytes, on stiff substrates osteocytes.

substrate stiffness



18

—Cell spreading affects cell volume via water efflux

M. Guo et al. J. Lippincott-Schwartz and D. Weitz. PNAS. (2017) E8618–E8627 www.pnas.org/cgi/doi/10.1073/pnas.1705179114

• Cells spread more on stiffer substrates and reduce concomitantly their volume

• Restricting cell area using smaller adhesive islands, increases cell volume

• Dynamic cell spreading is associated with dynamic change in cell volume

• The volume dependence on spread area is the same irrespective of how area is attained (substrate stiffness, adhesion area, dynamics of spreading)

substrate stiffness

size of adhesive

• Volume could be changed by protein concentration changes or water efflux: no [protein] change

• Volume change caused by water efflux



19

—Volume changes induced by cell spreading or osmotic stress affect nuclear volume and molecular crowding in the nucleus.


• Hyper-osmotic stress (with high PEG concentration in the medium) causes a reduction in cell volume through water efflux

irrespective of substrate stiffness.• Blocking Cl ion channels (NPPB) blocks

volume change as a function of substrate stiffness so volume adjustment involves changes in osmolytes.

• Blocking cell contractility also affects stiffness dependent volume change.

• Histine H2::GFP mean square displacement is lower in compressed cells consistent with increased molecular crowding

reviewed in R. Phillips, T. Ursell, P. Wiggins and P. Sens (2009) Nature 459: 379-385

• Consistent with tension dependent regulation of mechanosensitive ion channels, affecting osmolyte concentration and water flux.

Hyper-osmotic stress

Blocking Cl ion channels (

g yBlocking cell contractility

• In all conditions nuclear volume adjusts to cell volume



20

—Volume changed induced by osmotic stress affects differentiation of stem cells


• Differentiation of stem cells into osteocytes on a stiff substrate

• Hypertonic medium rescues differentiation on soft substrate.

• This is associated with volume reduction of cell and nucleus and increased cell stiffness

# YAP forms liquid like nuclear condensates in response to molecular crowding induced by hyperosmotic stressD. Cai et al. and J. Lippincott-Schwartz. bioRxiv. http://dx.doi.org/10.1101/438416. (2019)

ie. stiff substrate

osteogenesis

now in press in Nature Cell Biology

• On soft substrates, conversely cells differentiate into adipocytes

• Hypotonic medium rescues adipogenesis if on stiff substrate

Nat Cell Biol. 2019 Dec;21(12):1578-1589. doi: 10.1038/s41556-019-0433-z.


Control of cell growth/division & death by mechanics

21

Geometry: spreading, volume reduction (but not adhesion per se)

Mechanics: tension

Mechanochemical signalling: ion channels, YAP, etc.

Cell division No cell division Apoptosis

compression

• Summary


Mechanics of Tissue Growth and Homeostasis

—Tissue growth depends on the balance between:• Rate of cell division• Rate of cell death and cell extrusion

—Tissue growth leads to homeostasis at a fixed size where cell growth/division and death/extrusion balance each other (unless growth is arrested): eg. gut.

• How is such a balance achieved?• Feedback interactions?• Does growth impact on extrusion?• Does extrusion lead to compensatory growth?

—What are the consequence of differential tissue growth rates?• Cell and tissue competition (eg. tumours)• Tissue buckling (morphogenesis)


Control of cell growth/division by mechanics

23

A B

C

—Changes in cell behaviour during collective tissue growth in vitro: contact inhibition

A. Puliafito et al. and B. Shraiman. (2012) PNAS 109:739–744

• Growth of isolated colonies to disentangle cell contact from mechanics

• « free growth » regime is associated with exponential growth while cells are in contact

cell density

A BC

• Colony growth slows down, limited by constrained velocity of edge cells (ie. friction).

• This is associated with a transition to increased cell density and slower motility which are mechanically induced

cell density and size cell motilitycell division

• Increased cell density is associated with gradual increase of division time

Thomas LECUIT 2018-2019 24A. Puliafito et al. and B. Shraiman. (2012) PNAS 109:739–744

—1D vertex mechanical model of self-limiting growth • Short time elastic response of cells• Friction coupling to the substrate (elastic

restoring force)• Active random cell motion (Langevin driving

force) averages to zero in colony bulk but not at colony edges

• Cell growth is modelled by increasing the rest length of a cell edge when neighbouring cells stretch it.

• Cell division halves rest length• Cell division rate depends on cell size

—Changes in cell behaviour during collective tissue growth in vitro: contact inhibition

—Simulation results• 2 phases: exponential and linear growth• Cell division in the bulk and gradual arrests due to

increased cell density• Cell division persists at colony edges as cells are stretched.

time

cell size

time

time

trac

tion

forc

e

position

Control of cell growth/division by mechanics


Balancing cell growth and extrusion by mechanics

14 h

AP

26 h

AP

a

OM

OM

M

OM

OM

M

14–26 h AP

Cells that delaminate before division

Cells that delaminate after dividing

Cells that divide and only one daughter cell delaminates

Daughter cell that delaminates

b

Marinari, E., et al, and Baum, B. (2012). Nature 484, 542–545

—Compression induced cell extrusion

• Cells in the midline of a growing tissue delaminate.

–40 min –15 min –10 min

E-cad–GFP

0 min

0

0.4

0.2

0.6

0.8

1.0

0246

Nor

mal

ized

are

a

Junction number

• Cell delamination is associated with apical area reduction and junction remodelling (loss).

junction number

PTEN

control overgrowth undergrowth

Del

amin

atio

n (%

)

0

20

40

60

80

pnr-G

AL4pnr

-GAL4

> p11

0pnr

-GAL4

> PTE

N iRpnr

-GAL4

> Tsc1

> Ts

c2pnr

-GAL4

> PI(3

)K iR

pnr-G

AL4pnr

-GAL4

> p11

0pnr

-GAL4

> PTE

N iRpnr

-GAL4

> Tsc1

> Ts

c2pnr

-GAL4

> PI(3

)K iR

Midline

Outside midline

n = 3

P < 0.01P < 0.05

P < 0.05 P < 0.01 P < 0.05P < 0.05 P < 0.05

pnr-GAL4 > p110 pnr-GAL4 > Tsc1,Tsc2

a

bpnr-GAL4

Insulin

Pi3K

mTOR TSC

Growth/Cell division

• Tissue overgrowth (induced by activation of the Pi3K pathway) increases cell delamination

• Reducing tissue growth reduces cell delamination/extrusion

• Tissue crowding induces cell delamination


a

Compressibility Line tension Contractility

lij

ji

Aα

W = (Aα – A0)2K2

+ +lij 22lij

cells α junctions lj junctions lj

b c

d e

Initial state Final state

0

10

20

30

40

50

Time

Del

amin

atio

n (%

)

Time0

10

20

0

20

10

Del

amin

atio

n (%

)

High crowding

Wild type

No crowding

Anisotropic tissue

Isotropic tissue

M OM

T1 T2

Myosin II


Marinari, E., et al, and Baum, B. (2012). Nature 484, 542–545

• Mechanical vertex model of junction remodelling underlying compression induced cell delamination.



2 cm

2 cmStretch

28%

Release

Seed membrane with MDCK cells

Silicone membraneT0 1 h 2 h 3 h

Time After Release

Control

a

b c

jF-actin

FibronectinDAPI

d e2 h0.5 h 6 h

DAPIZO-1β-cat

hg Control 2 h0.5 h i******

***

Contro

l0.

5 1

1.5 2 3 4 6

Time (h)

f 150

140

130

120

110

100

Cel

ls p

er 1

00 μ

m2


Eisenhoffer, G.T., et al., and Rosenblatt, J. (2012). Nature 484, 546–549.

• A tissue monolayer under compression adjusts its cell density in two phases:

• Phase 1: Density first increases rapidly due to compression.

• Phase 2: Cell density then slowly decreases until reaching initial steady state.


b

Inhibitors against

***

***

**

**

* * ** *****

*

* *** * *

**

*Con

trol

0.5 1

1.5 2 3 4 6

Time post-overcrowding (h)

80

60

40

20

0Ext

rud

ing

cells

per

1,0

00 c

ells Non-apoptotic Apoptotic 80

60

40

20

0Ext

rud

ing

cells

per

1,0

00 c

ells

Contro

l

ROCK SK

S1P2

Non-apoptoticApoptoticApoptotic blocked

a

• Adjustment of cell density is associated with cell extrusion by apoptosis and independent of apoptosis.

• Cell extrusion requires ROCK, the sphingosine kinase SK and sphingosine phosphate (SIP2).


Dev

elop

ing

zeb

rafis

h ep

ider

mis

Non-apoptotic cell extrusion Apoptotic cell extrusion

Hum

an c

olon

tis

sue

Cel

l cul

ture

a b

DAPICasp3F-actin

d e

g h

c

Extruding cell

Proliferating cell

Actin ring in neighbouring cells

f

i

F-actin/DNAd

Bright-field (DIC)

Unt

reat

ed

hhh

Cell migrationc e

*37 h 49 h 56 h

6

4

2

0

8

10

*

**

***

***

No.

of e

xtru

din

g ce

lls p

er fi

sh

No.

of H

3P-p

ositi

ve c

ells

per

100

μm

2

g

Gd

3+p

iezo

1 p

hoto

-MO

hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhf

i j k

Control Gd3+ piezo1 photo-MO

6

4

No.

of e

xtru

din

g ce

lls p

er fi

sh

2

0*** ***

Stretch-activated signals

Sphingosine-1-phosphate

Extrusion

Gd3+

Homeostatic cell death

Rho

Anoikis

e

Piezo1


Eisenhoffer, G.T., et al., and Rosenblatt, J. (2012). Nature 484, 546–549.

• In whole organisms: gut villi and fins in zebrafish, cell crowding causes cell extrusion.

• This requires the mechanonosensitive receptors Piezo.




Gudipaty, S.A., et al., and Rosenblatt, J. (2017). Nature 543, 118–121.

—Stretch induced cell divisions

• Stretching of cell cultures (MDCK cells)induces cell division.• Wound healing in a cell monolayer causes

stretching of cells after closure and induces a burst of cell division after closure

Time after stretch (h)Ctrl 0.5 1 2 4 6

12

14

16

18

20

22

Cel

l len

gth

(μm

)

* *

b

Ctrl 0.5 1 2 4 6 0

50

100

150

H3P

+ c

ells

per

10,0

00 c

ells

Time after stretch (h)

***

**

a

c

12 (closure) 18 5

0

2

4

6

8

Time after wounding (h)

Div

idin

g ce

lls p

er fr

ame

5 10 15 20

closure

• Piezo activation gives rise to calcium signalling, ERK activation and Cyclin B induction

Ca2+

Piezo1 Microtubules Cell–cell junctions

ERK1/2 Cyclin B

Wound or stretch

Epithelia at steady state Stretch-induced mitosis

30 60 90 1200

200

400

600

Time after stretch (min)

Cyc

lin B

+ c

ells

per

10,

000

cells

Gd3+α-amanitin

No stretch

Control

Control Piezo1CRISPR

0

20

40

60

H3P

+ c

ells

per

fish

****

e f Piezo1 CRISPRControl

DNA H3P

Ctrl

Rosco

.Gd

3+

Piezo1

KD

0

50

100

150

H3P

+ c

ells

per

10,

000

cells

ControlStretch

***********

+Piezo1

aControl

0

1

2

3

4

5

6

4 8 12 16 20 24

Wound closed

Div

idin

g ce

lls p

er fr

ame


Piezo1 KD

0

1

2

3

4

5

6

Wound closed

4 8 12 16 20 24

Div

idin

g ce

lls p

er fr

ame


c

Pi

1A

ti

• The mechanosensitive receptor Piezo 1 is required for cell division in MDCK culture cells and in zebrafish fins




Regulation of cell cycle progression by cell–cell and cell–matrix forces

M. Uroz et al. and X. Trepat. (2018) Nature Cell Biology 20:646–654

g h iControl G1/S arrest

0 4 8 12 16 20 24 28 32 36

100200300400500600700

Time (h)

Num

ber

of n

ucle

i

Control G1/S arrest

0 4 8 12 16 20 24 28 32 36

8001,2001,6002,0002,4002,800

Time (h)

Cel

l are

a (μ

m2 )

Control G1/S arrest

Time (h)

0 4 8 12 16 20 24 28 32 36

0.00.10.20.30.40.50.60.70.8

Mon

olay

er a

rea

(mm

2 )

• MDCK cell culture expand at a constant cell density

• When cell division is blocked (G1/S arrest), cell area increases

• Cell division is sustained while the colony expands and thereby maintains a constant density.

Cdt1::RFP: G1-SGeminin::GFP: S-G2-M

G1/S arrest

control




M. Uroz et al. and X. Trepat. (2018) Nature Cell Biology 20:646–654

d e f

Time (h)

0 3 6 9 12 15 18 21 24 27 30

100

200

300

400

500

600Cdt1GemininG1 S–G2–M

0 3 6 9 12 15 18 21 24 27 30

200

300

400

500

600

700

800

Time (h)

0 2 4 6 8 10 12 14 16 18 20

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Time (h)

Tra

ctio

n ∣T

∣ (P

a)

Ten

sion

σ (

Pa)

Nuc

lear

inte

nsity

(no

rm.)

Traction Tx (on substrate)

Cdt1::RFP: G1-SGeminin::GFP: S-G2-M

Intercellular tension σ

σ

Bay

es fa

ctor

G1 104

103

102

101

100

A.

σ τc A τ

cσ τ

c A 0c

σ τc A τ

n

σ τc

τc

σ τc A τ

nT τ

c A τc

T τc A τ

nT τ

c A 0c

T τc A τ

cσ τ

c A 0n

Eτc (μJ)

G1

dura

tion

(h)

q

0.0 0.1 0.2 0.3

8

12

16

20

24

• Cell tension and surface energy E ( . A) are the best predictors of cell cycle arrest (over cell geometry).

• High surface energy correlates with a short G1 duration: cells with same surface should enter G1 earlier if tension is higher.

Large scale cross-correlation analysis

E ( . σ σ A)

A τc

σ τc A 0

n

Eτc (μJ)

Thomas LECUIT 2018-2019 32Gudipaty, S.A., et al., and Rosenblatt, J. (2017). Nature 543, 118–121.


• Cells sense compression and stretch via the same mechanosensitive receptor Piezo and adjust their density depending on their mechanical state.

• Cell density evolves until the population reaches a steady state X where cell death/extrusion and cell division balance each other

—in 2D, balance between Stretch induced cell divisions and compression induced cell extrusion

Control Stretch0

50

100

Eve

nts

per

10,

000

cells

a

0

20

40

60

80

Eve

nts

per

10,

000

cells

b

**** ****

Control Crowd

DivisionExtrusion/death

DivisionExtrusion/death

X: steady-state density where cell death/extrusion and cell division balance each other


Balancing cell growth and death by mechanics: tumours

—In 3D, homeostatic pressure: a tissue property where cell growth/division balances apoptosis

Markus Basan , Thomas Risler , Jean-François Joanny , Xavier Sastre-Garau & Jacques Prost (2009) HFSP Journal, 3:4, 265-272, DOI: 10.2976/1.3086732

Tissue grows against wall resisted by a spring. Growth proceeds until internal pressure p = ph where growth and apoptosis balance each other.

The spring position is stable, since further growth increases the pressure above this value ph and induces apoptosis, whereas recession of the piston decreases the pressure and causes cell division.

P = P h. loserP < P h. winner P = P h. winnerP > P h. loserP < P h. winner

Net growth > 0 Net growth > 0Net growth = 0 Net growth < 0 Net growth = 0Current Biology

A. Matamoro-Vidaland R. LevayerCurrent Biology (2019)29, R762–R774,

• Cell competition

(see 26 Nov 2019)

pH,hpT,h >

As cells divide, the pressure rises and first reaches the homeostatic pressure pH,h. At this point, tissue H stops growing while tissue T continues to proliferate and drives the pressure above pH,h. As a result, the apoptosis rate of H becomes larger than its division rate, resulting in its recession. The process continues until tissue H completely disappears

Tissue HTissue T• Tumour growth

• Homeostatic pressure


Levayer, R., Hauert, B., and Moreno, E. (2015). Nature 524, 476–480.

—Differential growth by cell competition requires cell intercalation

• Differential cortical tension between winner-winner cell interfaces and loser-loser cell interfaces causes cell intercalation at the winner/loser interface, cell isolation and cell death.

• Mechanics (differential cortical tension) drives cell competition and differential growth.

Balancing cell growth vs death by mechanics: competition

Winner–loser mixingWinner expansionIncreased loser

apoptosis

w–w > w–l , w–l = l-l

Filamentous actin and tension modification

Winner LoserDifferential PIP3

0.3

0.9

1.5

clone

n =

40

n =

40

n =

27

WT in tub-dmyc pi3kDN in WT

n =

39

n =

35

n =

36

P = 0.03

P = 0.013

P = 0.89

P = 0.03

P = 0.007

P = 0.48

clone

clone

f

Vm

ax (μm

s–1

)

E-cad::GFP

t = –2 s

t = 12.4 s

1.5

2

2.5

3

3.5

4

4.5

–2.4 0.0

2.4

4.8

7.1

9.5

11.9

Vert

ex d

ista

nce

(μm

)

Time (s)

Vmax

1

3

1

313

2 4 2 4 2 4

370 min335 min225 min

RFP (WT loser in tub-dmyc)E-cad::GFP

l-l l-lw-w w-w


Moreno, E. et al and Levayer, R. (2019). Current Biology 29, 23–34


—Differential growth by cell competition requires tissue compaction induced cell delamination


Mechanical cues affects cell competition in 3 ways:

• Modulating the geometry of cell signalling leading to loser cell elimination,

• Inducing cell death by compression,

• Stimulating the compensatory growth of winner cells around extruded/dying loser cells via cell stretching.

Pressure

Loca

tion

Zone of death

Zone of death

* *

*

*

*

Cell intercalation (T1 transition)

Cell sorting Cell mixing

Winner cell (WT)

Loser cell

Apoptotic cell

Junction shrinking

Newly formed junction

A

Reviewed in:

A. Matamoro-Vidal and R. Levayer

Current Biology (2019) 29, R762–R774, https://doi.org/10.1016/j.cub.2019.06.030


see B. Shraiman PNAS, (2005) 102:3318–3323


Mechanical regulation of tissue size

37

• Organs grow until they reach a fixed size characteristic of each organ/species• Growth is often uniform in the face of non uniform growth factor distribution

(morphogens)• Tissues can compensate for growth heterogeneities and reach a normal final

size arguing that cells assess their environment

—How do cells compare their growth rates?—How is growth uniform on average in a tissue?—How is tissue size determined?


—How do cells compare their growth rates?—How is growth uniform on average in a tissue?• Mechanical Feedback

>0, then p increases until p0 where =0 ��r, t� � ��t��

��r, t� � ��t�� <0, then p decreases until p0 where =0

>0��r, t� � ��t��

��r, t� � ��t�� <0

If

If

• Inhomogeneous growth is smoothened through the mechanical negative feedback

��r, t� � ��t��

��r, t� � ��t��r, t��r, t� or until cells die ( <0)

38


B. Shraiman PNAS, (2005) 102:3318–3323

• Mechanics of inhomogeneous growth dp�r, t�dt

��K

1 � ��r, t� � ��t��

1�

p�r, t�,

growth rate

pressure

shear-rigidity relaxation time(ie. cell rearrangements)

bulk modulus

00

00

1

Stress p

Gro

wth

rat

e

ΓM

M

0

stronger feedback• Negative feedback of pressure

on growth


— Regulation as an integral feedback:

• Compression or stretching of a group of cells is dependent on integral over the history of the clone of its growth rate with respect to average surrounding growth.

• Cell rearrangements limits the « memory » of this feedback so adaptation of growth rate is not perfect.

• At later stages, the rate of growth is reduced to match that of the background, resulting in a ratio of mutant to background clone sizes (the overgrowth ratio) independent of time

• Initial rate of growth is faster than that of a background clone.

Ln (#cells)

time

39


B. Shraiman PNAS, (2005) 102:3318–3323

dp�r, t�dt

��K

1 � ��r, t� � ��t��

1�

p�r, t�,

Initial rate of growth i

At later stages


— Non-autonomous effect: The mechanical feedback as a possible mechanisms underlying cell competition

40


B. Shraiman PNAS, (2005) 102:3318–3323

• Non-autonomous cell death around fast growing clone

• Dynamics of a fast-growing clone in a background of slow growing cells

• Fast growing cells increase their pressure p to p0 and reduce their growth rate until they reach the growth rate of the surrounding

• Neighbouring slow growing cells may die as a consequence of high pressure

p0

p


— Integration of morphogen and mechanical feedback

41


E�ri, � � �� a�V � V0�

2 � b ��

� � ��2 � c� � 1�2�perimeter

volumeheight

cost for cell heightmismatch deviation from

rest statedeviation from

rest stateline tension

00

0 1 20

1

Stress p

Gro

wth

rat

e

M

ΓM

M

0

��r, t� � ��M�r, t�, p�r, t��

1) growth rate goes down with increasing stressie. negative mechanical feedback.

2) growth rate has a maximum 3) the maximum depends on morphogen

concentration above a threshold value M0

M(r) � me�r/�M(r)A B C

D E F

Stress p is proportional to � 1 � 1 >0 � 1 <0

:compression:stretch:stretch:compression

Increased pressure reduces perimeter and increases height

cell divisions

L. Hufnagel et al. and B. Shraiman (2007). PNAS. 104:3835–3840 www.pnas.orgcgidoi10.1073pnas.0607134104

4) the morphogen M does not scale



L. Hufnagel et al. and B. Shraiman (2007). PNAS. 104:3835–3840 www.pnas.orgcgidoi10.1073pnas.0607134104


00

0 1 20

1

Stress p

Gro

wth

rat

e

M

ΓM

M

0

• Growth arrest: when M falls below threshold at edges, growth arrests there and pressure increases in the bulk thereby causing growth arrest.

0 4 810

1

102

103

0 50

50

t

N(t)J

Growth arrests at fixed size

0 2 40

10

20

30

d

λ• Tissue size depends on morphogen levels m, and length scale

but also on feedback strength q

size

length scaleM(r) � me�r/�

(m, q)(m, 2q)

(2m, q)

λ

p

γ = 0

γ = 0.01

γ = 0.08

γ = 0.5

M

p

γ = 0

γ = 0.01

γ = 0.08

γ = 0.5

M

p

γ = 0

γ = 0.01

γ = 0.08

γ = 0.5

0.7 0.9 1.1

1

0.1

0.01

M

IHG

0.7 0.9 1.1

1

0.1

0.01

0.7 0.9 1.1

1

0.1

0.01

D E F

Division follows iso-growth lines in the M, p phase plane

• Growth homogeneity: The mechanical feedback suppresses the effect of the high concentration of the morphogen at the center of the tissue and ensures uniform growth




• Two morphogens that scale with tissue size• Cell stretching induces growth above a threshold• No growth if both stretch and morphogen are too low

• Initially growth higher in center due to higher growth factors GF

• As a result cells are stretched at periphery, which promotes growth.

• Growth lowers stretching but cells remain stretched and compress the center, thereby reducing growth.

• The wider the periphery the greater the compression.• Growth arrests when compression overcomes the effect

of growth factors

Assumptions:

Results:

T. Aegerter-Wilmsen et al. Mechanisms of Development 124 (2007) 318–326


�

�

L. Le Goff et al . and T. Lecuit Development 140, 4051-4059 (2013) doi:10.1242/dev.090878

Y. Mao et al. and N. Tapon. The EMBO Journal (2013) 32, 2790–2803. doi:10.1038/ emboj.2013.197

— Pattern of tissue deformations in growing imaginal discs


• Gradient of cell area consistent with compression at the centre.

• Evidence of tissue anisotropy in a growing epithelium at tissue periphery

Δ Δ ΔΔ Δ Δ

Texture tensor

Graner, F., et al. (2008). Eur. Phys. J. E 25, 349-369.


�



Distal

Proximal

P/D junctions

Lateraljunctions

Time (s)

D –

D0

(μm

)

–5 0 5 10 15

0

0.4

0.8

1.2

D –

D0

(μm

)

0 5–5 10 15

0

0.4

0.8

1.2

Time (s)

P/D junctions Lateral junctions

Proximaledge

Proximaledge

Distal centre Distal centre

— Stress patterns in growing imaginal discs


• Cells are stretched tangentially in the growing tissue

-laser ablation experiments and relaxation kinetics (related to tension over friction)

• Planar polarised distribution of actomyosin contractile network


�

�

�

�



— Stress patterns in growing imaginal discs


G

Anisotropy alignment average

distance from clone edge (micron)

tang

entia

l alig

nmen

t S

color = anisotropy

Segmented ban-expressing clone

H

Y. Pan et al and B. Shraiman and K. Irvine. (2016) PNAS www.pnas.org/cgi/doi/10.1073/pnas.1615012113

• The induction of a fast-growing clone causes non-autonomous tissue deformation (anisotropy)

e.g. activation of the Hippo pathway (Yorkie++ or Expanded mutant) or of targets of Hippo pathway (mRNA bantam)


— Cytoskeletal actomyosin tension modulates tissue size

47


C. Rauskolb et al and K. Irvine. (2014) Cell 158, 143–156


GAP

MLCP

ROCKMRLC MRLC PDia


ExpandedBantam

• Activation of Rho1/Rok affects wing size

• Modulates Yorkie (YAP) pathway activity (Expanded expression)

• and Adjuba recruitment at cell junctions



C. Rauskolb et al and K. Irvine. (2014) Cell 158, 143–156

— Cytoskeletal actomyosin tension increases tissue size via activation of the Hippo growth pathway

— Tension at E-cadherin complexes recruits Adjuba and the kinase Warts at cell junctions and inhibits its activity via « sequestration » at cell cortex.—This releaves Yorkie repression by Warts and causes Yorkie nuclear translocation, hence activation of the pathway.


— Tension is reduced in fast growing clones

49



Growth

Compression

1

23

Transcription

F

Compression

2

1. overgrowth compresses cells; 2. compression decreases cytoskeletal tension (conversely, stretching increases actomyosin network and tension)3. decrease in cytoskeletal tension triggers reduction in the cortical recruitment of Jub and Wts, and in Yki activity hence on growth.

• Mechanical Feedback

I

Cell shape reflects balance between edge tension and internal pressure

— This emerges from the mechanical feedback model where mechanical cell response adapts to external mechanical stress pattern.


— Adjuba and Warts localization at cell junction are decreased in overgrowing clone of cells

50



J K

A

E

Growth

Compression

1

23

Transcription

3

and rescued when Myosin2 is constitutively active (and tension increased in these overgrowing cells)



A

G

Growth

Compression

1

23

Transcription Growth

1

Trranscription


— Yorkie induced overgrowth downregulates Yorkie activity (evidence of negative feedback)

— Blocking the feedback affects causes growth heterogeneity

D F

J

H

A

• Forcing Bantam expression independent of mechanical feedback• Removing Warts so it cannot be subjected to feedback regulation• Forcing Myosin2 constitutively


— Integration of chemical and mechanical signalling

52


Feedback mechanisms responsible for uniform growth:

ABuild-up of compression during growth

Center Pouch edge

Com

pres

sion

Center Pouch edgeCenter Pouch edge

0

Compression > threshold

Compression slope < threshold

1

2

Growth termination when:

Dpp Wg

Arm

Yki

Fj

Ds

Notch Brk

Growth

Vg

MechanicalCompression

Fj

D

Ds

Hypotheticalvg quadrant enhancervg boundary enhancer

A P

DV

B

DppWgNotch

A

Tinri Aegerter-Wilmsen et al. and C. Aegerter and K. Basler. Development 139, 3221-3231 (2012) doi:10.1242/dev.082800

• vertex model with cell shape governed by

• Growth rate

Center

YkiNotch Brk

Growth

Vg


YkiNotch Brk

Growth

Vg


1

Termination

-increased growth leads to increased compression, which decreases growth,

Compression slope read-out

Yki

Fj

Ds

Vg


Fj

D

Ds

Growth

D Yki Vg

Growth


Large compression difference:

2

-it leads to a compression gradient that induces growth in regions with lower growth rates



— Integration of chemical and mechanical signalling

Tinri Aegerter-Wilmsen et al. and C. Aegerter and K. Basler. Development 139, 3221-3231 (2012) doi:10.1242/dev.082800

Non-trivial model predictions:

L M L

D

L M L

wt uniform Dpp

M L M L0

5

10

15

Ave

rage

cel

l cyc

le p

rogr

essi

on (

%/h

r)

wt uniform DPP

**

0 20 40 60 80 100101

102

103

104

Time (hours)

Num

ber

of c

ells

tkvQD clone; BrdU staining

+ 38 hrs

A A’

�

�



Mechanical feedback: - corrects the consequence of non uniform growth rates, and

smoothens out heterogeneities: growth homogeneity.- may be involved in compensatory mechanisms (eg. cell

competition) in the context of tissue growth and homeostasis.

- part of growth arrest mechanisms at the scale of organs.

Yet, tissue growth can be highly heterogeneous. - normal development: growth zone (limb bud, beak

morphogenesis), tissue folding. - disease: tumour evolution.

Thomas LECUIT 2019-2020 T. Tallinen et al and L. Mahadevan. PNAS (2014) 111:12667

physical mimic of brain

GI: gyration index (total area/exposed area)

brain

simulations

physical model of brain like instability

swelling of outer layer (growth)

Differential growth and tissue folding

• Cortex convolutions


E6 E8 E10 E12 E13 E15 E16 E17

No muscle + Circ. muscle + Long. muscle + Long. muscle

up until E6Smooth

E8 - E12Ridges

E13 - E15Zigzags

E16 onwardVilli

αS

MA

DA

PI

αS

MA

DA

PI

• Epithelium and mesenchyme form annealed surfaces ensheathed by sequential layers of smooth muscles

• Formation of Ridges in epithelium and mesenchyme parallel formation of circumferential smooth muscles

• Formation of Zigzags parallels formation of longitudinal muscles

• Formation of Villi formation of longitudinal muscles

Ridges Zigzag Villi

A. Shyer et al, C. Tabin and L. Mahadevan. Science 342: 212-218 (2013)

Differential growth and tissue folding

• Gut vilification:


—Computation model

—Theoretical studies:

A. Shyer et al, C. Tabin and L. Mahadevan. Science 342: 212-218 (2013)

M. Ben Amar and F. Jia. PNAS 110: 10525–10530 (2013)

E. Hannezo J. Prost and J-F. Joanny PRL 107, 078104 (2011)

Growth induced mechanical instabilities: Gut vilification

• Gut vilification


Tissue growth and mechanical feedback: - corrects the consequence of non uniform growth rates, and smoothens

out heterogeneities: growth homogeneity.- may be involved in compensatory mechanisms (eg. cell competition) in

the context of tissue growth and homeostasis.- required for growth arrest at the scale of organs.

Summary

Mechanical regulation of cell division and cell death/extrusion

- Sensing: mechanosensitive receptors (Piezo), cortical tension and YAP pathway- Cellular response: mechanics of junction remodelling (see Courses 2018)

- Cells respond differently to compressive and tensile stresses: - Cell growth, cell division, cell extrusion or simple cell shape changes, cell

intercalation.- These processes underlie viscoelastic properties of tissues: energy

dissipation in homeostatic conditions- Cell mechanics underlies tissue homeostatic pressure.- Not well understood what factors determine these properties

organism and tissue growth - collège de france · 2019-12-11 · thomas lecuit 2018-2019 8 tissue...

Documents