Cells Respond to Mechanical Stress by Rapid Dis Assembly of Caveolae

8/6/2019 Cells Respond to Mechanical Stress by Rapid Dis Assembly of Caveolae

1/12

Cells Respond to Mechanical Stress

by Rapid Disassembly of CaveolaeBidisha Sinha, 1 , 2 , 14 Darius Ko ster, 1 ,2 ,14 Richard Ruez, 3 ,4 Pauline Gonnord, 3 ,4 Michele Bastiani, 8 ,9 Daniel Abankwa, 8 ,9Radu V. Stan, 10 Gillian Butler-Browne, 11 Benoit Vedie, 12 Ludger Johannes, 3 ,4 Nobuhiro Morone, 13 Robert G. Parton, 8 ,9Grac a Raposo, 3 ,5 ,6 Pierre Sens, 7 Christophe Lamaze, 3 , 4 , 15 ,* and Pierre Nassoy 1 ,2 ,15 ,*1 Universite P. et M. Curie/CNRS UMR1682 Institut Curie, Centre de Recherche, Laboratoire Physico-Chimie3 CNRS UMR1444 Institut Curie, Centre de Recherche, Laboratoire Trac, Signalisation et Ciblage Intracellulaires5 PICT IBiSA Institut Curie6 Centre de Recherche, Laboratoire Structure et Compartiments Membranaires, Institut Curie26 rue dUlm, 75248 Paris Cedex 05, France7 ESPCI, CNRS-UMR 7083, Physico-Chimie The orique, 10 rue Vauquelin, 75231 Paris Cedex 05, France8 The University of Queensland, Institute for Molecular Bioscience9 Center for Microscopy and Microanalysis

Brisbane, Queensland 4072, Australia10 Dartmouth Medical School, Borwell 502W, HB7600, One Medical Center Drive, 03756 Lebanon, NH, USA 11 Institut de Myologie, Ho pital Pitie -Salpe trie `re, UM76 UPMC, U974 Inserm, UMR7215, CNRS-AIM, 47, bld de lho pital, 75651 Paris Cedex13, France12 Laboratoire de Biochimie, Ho pital Europe en Georges Pompidou, 20 rue Leblanc, 75015 Paris, France13 National Center of Neurology and Psychiatry, National Institute of Neuroscience, Department of Ultrastructural Research, 4-1-1Ogawa-Higashi, Kodaira, Tokyo 187-8502, Japan14 These authors contributed equally to this work15 These authors contributed equally to this work*Correspondence: [email protected] (C.L.), [email protected] (P.N.)DOI 10.1016/j.cell.2010.12.031

SUMMARY

The functions of caveolae, the characteristic plasmamembrane invaginations, remain debated. Their abundance in cells experiencing mechanical stressled us to investigate their role in membrane-mediatedmechanical response. Acute mechanical stressinduced by osmotic swelling or by uniaxial stretchingresults in a rapid disappearance of caveolae, in areduced caveolin/Cavin1 interaction, and in anincrease of free caveolins at the plasma membrane.Tether-pulling force measurements in cells and inplasmamembranespheres demonstrate thatcaveolaattening and disassembly is the primary actin- and ATP-independent cell response that buffers mem-brane tension surges during mechanical stress.Conversely, stress release leads to complete caveolareassembly in an actin- and ATP-dependent process.The absence of a functional caveola reservoir in myo-tubes from muscular dystrophic patients enhancedmembrane fragility under mechanical stress. Our ndings support a new role for caveolae as a physio-logical membrane reservoir that quickly accommo-dates sudden and acute mechanical stresses.

INTRODUCTION

Caveolae were rst described in the early 1950s through theseminal electron microscopy studies of Palade and Yamada( Palade, 1953; Yamada, 1955 ). These characteristic 6080 nmcup-shaped uncoated invaginations are highly enriched incholesterol and sphingolipids ( Richter et al., 2008 ). Present atthe plasma membrane of many cells with the exception of neurons and lymphocytes, they are particularly abundant inmuscle cells, adipocytes, and endothelial cells. The identicationof caveolin-1 (Cav1) ( Rothberg et al., 1992; Kurzchalia et al.,1992 ) and caveolin-2 ( Scherer et al., 1996 ) as the main constitu-ents of the caveolar structure was instrumental to gain insight

into the cell biology, structural, and genetic features of caveolae( Stan, 2005 ). They have been associated with endocytosis, cellsignaling, lipid metabolism, and other functions in physiologicalas well as in pathological conditions. Nevertheless, the role of these specialized membrane domains remains debated, andlittle is known about the molecular mechanisms involved in theirformation and proposed functions ( Parton and Simons, 2007 ).

Recent studies have suggested that the distribution of Cav1and caveolae-mediated signaling can be affected by externalmechanical cues. In endothelial cells, chronic shear exposureactivates the ERK pathway in a caveolae-dependent manner( Boyd et al., 2003; Park et al., 2000; Rizzo et al., 2003 ). Insmooth-muscle cells, cyclic stretch can cause association of

402 Cell 144 , 402413, February 4, 2011 2011 Elsevier Inc.
mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.cell.2010.12.031http://dx.doi.org/10.1016/j.cell.2010.12.031mailto:[email protected]:[email protected]


2/12

some kinases with Cav1 ( Sedding et al., 2005 ). To date, the roleof Cav1/caveolae in mechanotransduction is mainly viewed asa downstream signaling platform, whereas their function inprimary mechanosensing has not been directly addressed.

A recent theoretical study has proposed that budded membranedomains like caveolae could play the role of membrane-medi-ated sensors and regulators of the plasma membrane tension( Sens and Turner, 2006 ). Endowed with a high membrane andlipid storage capacity, owing to the invaginated structure andhigh lipid packing, caveolae are well equipped to play sucha role.

We have challenged the homeostasis of the plasma mem-brane tension with different types of controlled mechanicalstresses and analyzed the role of caveolae in the cell short-term response. We show in endothelial cells and muscle cellsthat functional caveolae are required to buffer the variations of membrane tension induced by sudden and transient mechanicalstress via a two-step process of rapid caveola disassembly and

slower reassembly.

RESULTS

Mechanical Stress Leads to the Partial Disappearanceof Caveolae from the Plasma MembraneWe examined the response of caveolae when cells wereexposed to acute mechanical stresses. Osmotic swelling causesan increase of the membrane tension of cells unless someadditional membrane is delivered to the cell surface ( Dai andSheetz, 1995; Dai et al., 1998; Morris and Homann, 2001 ).Cav1-EGFP-transfected HeLa cells were exposed to hypo-osmotic medium (30 mOsm). We observed a 35% increase of the cell volume within the rst 5 min and a slow decrease there-after ( Figures 1 A and 1B). Upon reversing back to iso-osmolarity(300 mOsm) after 30 min of hypotonic shock, the volumedecreased below the initial cell volume. These observationssupport the existence of a compensatory mechanism knownas regulatory volume decrease, which restores the osmoticbalance by activating ion channels ( DAlessandro et al., 2002 ).Our data, however, suggest that this process is not dominantduring the rst 5 min following hypo-osmotic shock. To distin-guish caveolae at the plasma membrane from the internal Golgipool of Cav1, we used total internal reection uorescence(TIRF) microscopy ( Figure 1 C and Figures S1 A andS1B availableonline). Upon hypo-osmotic shock, we observed that thenumber of caveolae signicantly decreased by $ 30% at the

cell surface ( Figures 1 C and 1D) and that the loss correlatedwith the magnitude of the shock ( Figure 1 E). Importantly, thecell footprint and the adhesion between the cell and the glasssurface were unaltered, as shown by reection interferencecontrast microscopy (RICM) ( Figure S1 C). Because caveolaeexhibit different types of dynamics at the plasma membrane( Pelkmans and Zerial, 2005 ), we also checked whether anyparticular pool was selectively affected. Within minutes of hypo-osmotic shock, slow-moving caveolae reduced theirmobility ( Figure S1 D), and fast dynamics were abolished ( MovieS1 and Movie S2 ), whereas caveolae displaying all kinds of mobility were reduced in number ( Figures S1 E and S1F). Similarresults were obtained in mouse lung endothelial cells (MLEC)

( Figure S1 B). Although osmotic shocks have been extensivelyused to mimic the osmolarity changes that cells experience( Lang et al., 1998 ), we sought to rule out any indirect inuenceof cell swelling on caveolae. We developed a stretching device

based on thin transparent silicone substrates to challenge thecell membrane with a different mechanical stress. It allowedimaging of caveolae by TIRF before and after stretch and wascombined with micropatterning ( Chen et al., 1997 ) to controlthe cell adhesion area and its orientation with the stretchingaxis. The number of caveolae present at the basal footprint of Cav1-EGFP HeLa cells decreased upon stretching ( Figures 1 Fand 1G), and the loss correlated with the extent of stretch ( Fig-ure 1 H). Therefore, acute mechanical stress induced either byhypo-osmotic shock or membrane stretching leads to a rapidand signicant loss of caveolae from the cell surface. We nextperformed electron microscopy (EM) on MLEC. These endothe-lial cells experience chronic cyclesof shear stressfrom thebloodow in lungs vessels in vivo. MLEC immunostaining shows

multiple subcellular Cav1 positive structures, which are localizedpredominantly at the plasma membrane and at the Golgi appa-ratus ( Figure S3 B; Murata et al., 2007 ). In contrast, Cav1

/

MLEC derived from cells knocked out for the CAV1 gene donot present any Cav1staining. However, Cav1-EGFPexpressioncan be induced by transfection in WT and Cav1

/ MLEC ( Fig-ure S3 C). EM analysis showed a signicant decrease of thenumber of caveolae (50%) upon a 5 min exposure of WTMLEC to a hypo-osmotic shock ( Figures 2 A and 2B). Thesedata conrm the results obtained by TIRF imaging and extendour conclusions to endogenous caveolae present on the entiresurface of the cell.

Flattening and Disassembly of Caveolaeupon Hypo-Osmotic Shock The contribution of caveolae to the general endocytic activity of the cell is believed to be minimal ( Nabi and Le, 2003 ), and endo-cytosis is disfavored at high membrane tension ( Dai et al., 1997 ).We still tested whether the loss of caveolae upon hypo-osmoticwas due to increased caveola endocytosis. We used dynasore,an inhibitor of the dynamin GTPase that is involved in caveolainternalization ( Henley et al., 1998; Macia et al., 2006 ). Indeed,dynasore signicantly increased the caveolae density at theplasma membrane, reecting the efcient inhibition of caveolaendocytosis ( Figure 2 C). However, upon hypo-osmotic shock,a similar loss of caveolae was measured ( Figures 2 C and 2D).We next examined Cavin1, which is part of the caveolar complex

and is required to maintain caveola invagination ( Hill et al., 2008;Hansen et al., 2009 ). Cavin1 does not bind to free Cav1 oligo-mers or to Cav1 present on the Golgi apparatus. We founda high level (64%) of Cav1-EGFP colocalization with Cavin1-mCherry ( Figure 2 E). Upon hypo-osmotic shock, there wasa similar or even higher loss of Cavin1-labeled structures, con-rming the partial loss of caveolae ( Figure 2 F). We also observeda decreased colocalization with Cavin1 (35%) for the remainingcaveolae, suggesting a loss of their invaginated structure. Wequantied the interaction between Cav1 and Cavin1 using uo-rescence lifetime imaging microscopy (FLIM). When Cavin1 ispresent in caveola, the close proximity of mRFP-Cav3 andCavin1-EGFP results in FRET and a decrease in the EGFP

Cell 144 , 402413, February 4, 2011 2011 Elsevier Inc. 403


3/12

uorescence lifetime ( Abankwa et al., 2008; Hill et al., 2008 ).There was a signicant increase in the uorescence lifetimeupon hypo-osmotic shock, indicating the dissociation of Cavin1from Cav1 ( Figure 2 G). We also performed Cav1 immuno-EMon MLEC before and after hypo-osmotic shock ( Figure 3 A andFigure S2 A). We found a 10-fold increase in the number of goldparticles associated with Cav1 in noncaveolar membranes after5 min of hypo-osmotic shock ( Figure 3 B). Upon returning to iso-osmolarity, cells recovered the initial number of caveolae, and

Cav1 was mainly associated with caveola ( Figures 3 B and 3C).Under iso-osmotic conditions, deep-etched EM showeda majority of budded caveolae with characteristic tight striatedcoats ( Morone et al., 2006 ). After hypo-osmotic shock, severalat structures with loose striated coats reminiscent of formerlybudded caveolaewere observed. Upon iso-osmolarity recovery,all caveolae were budded ( Figure 3 D and Figure S2 B forthree-dimensional view). These ndings clearly indicate thatcellsrespond to acute mechanical membrane stress by the rapid

BA

1.2 ( a u

)Iso

0 10 20 30

0.71.0

V o

l u m e

Time (min)

Hypo

Rec

C

Iso Hypo

ED

0.8

1.2

m a

l i z e

d

c a v e o

l a e

0 8

1.0

1.2

HypoIsoIso Hypo

0.0

0.4 N o r m

n o .

o f

300 200 100 0

0.6

0.8

Osmolarity (mOsm)

F G

0 6

0.8

1.0

1.2

0.4

0.8

1.2

N o r m a

l i z e

d

o .

o f c a v e o

l a e

H

0.0 0.2 0.4

0.6

(L-L0)/L

0

- +0.0

n o

Stretch

Figure 1. Mechanical Stress Induces Partial Disappearance of Caveolae

(A) YZ maximum-intensity projection of confocal stacks of Cav1-EGFP HeLa cells. Projection of four cells under iso-osmotic conditions (Iso), hypo-osmoticconditions (Hypo, 5 min), and 3 min after returning to iso-osmolarity (Rec). Scale bar, 5 mm. Dashed lines mark out the initial cell boundary.(B) Volume of Cav1-EGFP HeLa cells tracked from (and normalized to) iso-osmotic conditions through hypo-osmotic shock (onset: t = 0 min) and upon returningto iso-osmolarity (t $ 29 min). Arrow indicates return to iso-osmolarity. Data were derived from multiple measurements (n = 5) in three independent experiments.Error bars represent standard deviations (SD).(C) TIRF images of Cav1-EGFPHeLa cells underiso-osmotic conditions (Iso) and after 4 min hypo-osmotic shock (Hypo). Dotted linemarks out the cellfootprint.Scale bar, 5 mm.(D) Change in the number of caveolaefor single Cav1-EGFPHeLa cells after hypo-osmoticshock (Hypo) normalized to the number counted before hypo-osmoticshock (Iso) (n = 18). Error bars represent SD (p = 4 3 10

11 ).(E) Evolutionof the lossof caveolae per cellwith decreasing osmolarity. The sameCav1-EGFP HeLacellswere exposedto decreasing osmolaritiesduring $ 1 minfor eachosmolarity.From correlation analysis, theloss of caveolaeis positively correlated withthe decreasein externalosmolarity(r 2 = 0.85). Error barsrepresentSD (n = 3).(F) TIRF images of a Cav1-EGFP HeLa cell on the stretching device at 0% (left) and 20% stretch (right). Dotted lines mark out cell boundaries before and afterstretch. Scale bar, 5 mm.(G) Change in thenumber of caveolae for single Cav1-EGFPHeLa cells after stretching (15% 1%) normalized to the number counted before stretching. Data arederived from multiple measurements (n = 7; p = 0.00033) in seven independent experiments. Error bars represent SD.

(H) Evolution of the number of caveolaefor single Cav1-EGFP HeLacells stretched to different lengths characterized by (L L0 )/L0 wherein L 0 andL arethe initialand nal lengths of the cell footprint in the stretching direction. Each point is measured on a single cell. The number of caveolae is found to be negativelycorrelated to the extent of stretch (n = 7; r 2 = 0.85) as measured in seven independent experiments.



4/12

atteningof a fraction of the caveola.We also measured the levelof Cav1-EGFP diffusing freely at the membrane using uores-

cence recovery after photobleaching (FRAP). At steady state,the fraction of freely diffusing Cav1 in the plasma membranewas low (10%; Figure 3 E), as reported ( Pelkmans et al., 2004;Hill et al., 2008 ). However, we measured a higher mobile fraction(30%) in cells exposed to hypo-osmoticshockfor at least 10 min.This increase is likely to reectthe release of Cav1 from attenedcaveolae.

Caveolae Are Selectively Required for BufferingMembrane TensionCaveola attening is likely to release the amount of membranestored within the caveolar invagination and thereby to providethe additional membrane required to maintain membrane

tension homeostasis during mechanical stress ( Sens andTurner,2006 ). We tested this hypothesis with the tether pulling

technique ( Dai and Sheetz, 1995 ), which measures the cellmembrane tension. Optically trapped beads adhering to theplasma membrane served as handles to extract membranetethers ( Figure 4 A, Figure S3 A and Movie S3 ). The restoringtether force f , which was derived from the bead displacement,is an indicator of the effective membrane tension ( Sheetz,2001 ). Its value is proportional to the square root of the effectivetension ~s , which corresponds to the sum of the lipid bilayertension s and the cytoskeleton-to-membrane adhesion energyW0 . The latter term represents $ 75% of ~s ( Dai and Sheetz,1999 ) and arises from all molecular interactions betweenmembrane and cytoskeleton. The presence of exogenousproteins is thus likely to have an intricate inuence on W 0 , as

A

0.8

v e o

l a e

m

Iso

B

Iso Hypo0.0

.

N o .

o f c a

p e r

Hypo

C

0.8

1.2

r m a

l i z e

d

f c a v e o

l a e

D

0.8

1.0

1.2

CtrlDyn

a l i z e

d

e r o

f c a v

Hypo

FE Iso H o

Iso Ctrl Dyn0.0

. N o

n o .

Hypo0 10 20 30 40 50

0.6 N o r

n u m

b

Time (min)

a v 1 - E

G F P

90

120

150

o f c a v e o

l a e

Cav1-EGFPCavin1-mCherry

G - m

C h e r r y

Iso Hypo0

30

N o .

C a v

i n 1

1.9

2.2

m e

( n s

)

av n

Cavin1 + Cav3*

M e r g e

Iso Hypo

1.6 L i f e - t

i

Figure 2. Caveolae Morphology and Cav1-Cavin1 Interaction Are Lost upon Hypo-Osmotic Shock (A) Ultrathin cryosections of WT MLEC before (Iso)and 5 min after (Hypo) switch to hypo-osmotic

medium examined by EM. Arrows mark outcaveolae. Scale bar, 150 nm.(B) Quantication of caveolae detected per mmof plasma membrane on ultrathin cryosectionsof WT MLEC before (Iso) and 5 min after switchto hypo-osmotic medium (Hypo) reveals a signi-cant decrease (p = 0.047) in the number of caveolae after hypo-osmotic shock. Total mem-brane used for quantication was 76 and 67 mmfor iso- and hypo-osmotic conditions, respec-tively, imaged from different sections of multiplerandomly selected cells (>10). Data representmean SD.(C) Time evolution of caveolae number ( SD)detected by TIRF in control (Ctrl) and dynasore-treated cells (Dyn) normalized to the caveolaenumber before addition of dynasore (t = 5 min).Dynasore was added at t = 0, and hypo-osmoticshockwasapplied att $ 45 min(30 mOsm,shadedregion). Error bars represent SD; n = 4.(D) Change in number of caveolae for single cellsin control (Ctrl) and in dynasore-treated cells(Dyn) after hypo-osmotic shock normalized to thenumber before shock (Iso) (n = 9). Error barsrepresent SD (p = 1 3 10

4 ).(E) TIRF images of HeLa cells expressing Cavin1-mCherry and Cav1-EGFP before (Iso) and 5 minafter switch to hypo-osmotic conditions (Hypo).Scale bar, 10 mm.(F) Change in number of Cav1 and Cavin-1 struc-tures per cellbefore and after hypo-osmotic shockof 5 min. Error bars represent SD (n = 3).(G)HeLa cells transientlyexpressingCavin1-EGFPwith or without Cav3-mRFP were exposed toiso-osmotic (Iso) or hypo-osmotic (Hypo) mediafor 15 min and analyzed by FLIM. Data representmean EGFP uorescence lifetime standarderrors (SE). n = 4070 cells; *p < 4 3 10

12 .



5/12

shown later. Tether force measurements do not enablea priori toseparate s and W 0 . Therefore, we used the tether pulling tech-nique as a differential assay to probe the relative changes in

tether forces under conditions that keep W 0 unaltered. In partic-ular, osmotic shocks have been assumed to mostly affect s withminor changes on adhesion ( Dai et al., 1998 ). By quantifying the

DA

I s o

HypoIso

MembraneB H y p o

C

50

100

r c e n

t a g e o

f

C a v

1 l a b e

l i n g

***

*** 0.8

f c a v e o

l a e

p e r m

Iso Hypo Rec0

P e r

s u r f a c

1.00

o v e r y

Iso

E IsoRec0.0

N o .

R e

0.25

0.50

0.75

F l u o r e s c e n c e

R e c o Hypo

Rec

0 100 200 300. F

Time (sec)

Rec

Figure 3. Caveolae Flatten Out and Disassemble upon Hypo-Osmotic Shock and Reassemble upon Recovering Iso-Osmolarity (A) Immuno-EM images of ultrathin cryosections with gold-tagged Cav1 antibody of WT MLEC under iso-osmotic (Iso), hypo-osmotic (Hypo, 5 min), andrecovered iso-osmotic (Rec, 5 min) conditions. Scale bar, 150 nm. See also Figure S2 A.(B) Percentage of gold particles found in caveolae and endosomal structures close to the plasma membrane versus those found in noncaveolar membranes

(*p = 43

10 3

; **p = 43

102

; ***p = 13

103

). Total membrane used for quantication was 23, 24, and 35 mm, respectively, for iso-osmotic, hypo-osmotic, andrecovered iso-osmotic conditions, imaged from different sections of multiple cells. Error bars represent SE.(C) Comparison between the number of caveolae per mm of membrane before hypo-osmotic shock (Iso) and after return to iso-osmotic medium (Rec) analyzedfrom ultrathin cryosections of WT MLEC using 60 mm ofmembrane imagedfrom different sections (n = 8) of multiple cells. Data represent mean SE(p = 1 3 10

2 ).(D) Deep-etched EM images of MLECs under iso-osmotic (Iso), hypo-osmotic (Hypo, 5 min), and recovered iso-osmotic (Rec, 5 min) conditions. Scale bar,200 nm. Leftinsets depict representative images of clathrin-coated pits. Rightimagesdepict representative images of caveolae. Scale bars(insets), 100 nm. Seealso Figure S2 B.(E) Fluorescence recovery after photobleaching (FRAP) of Cav1-EGFP in HeLa cells in iso-osmotic (Iso; n = 8), hypo-osmotic (Hypo; n = 8), and recovered iso-osmotic conditions (Rec; n = 8). Lines show t for the curves to standard recovery equation. Hypo-osmotic shock results in a statistically higher uorescencerecovery (p = 2 3 10

4 ) than in iso-osmotic conditions. Data represent mean SE.



6/12

actin bundles architecture, we found that the cortical actin cyto-skeleton was unaltered within the rst 5 min of osmotic shock( Figures S3 DS3F). At these timescales, variations in f thusdirectly mirror changes in s . We measured the tether force f 0 inisotonic conditions and recorded the variations of the tetherforce f while MLEC were exposed to hypo-osmotic shock. Fig-ure 4 A shows representative temporal traces of the relativetether force changes, (f-f 0 )/f 0 , obtained in WT and Cav1

/

MLEC. We found that, upon hypo-osmotic shock (150 mOsm),f remained almost identical to f 0 in WT MLEC. In contrast, thetether force increased by 200% in Cav1

/ MLEC. Our dataimply that the membrane tension increase is buffered in WTMLEC by the presence of caveolae. Importantly, the expressionof functional caveolae by transfection of Cav1-EGFP in Cav1

/

MLEC restored membrane tension buffering. Therefore, the lack

of membrane tension buffering is due to theabsence of caveolaeand not to other cellular structures that may have been altered inCav1

/ MLEC ( Figure 4 A). M-b -cyclodextrin, which attenscaveolae through cholesterol depletion ( Rothberg et al., 1992 ),led also to membrane tension increase upon hypo-osmoticshock in WTMLEC( Figure 4 A), conrming that caveola atteningis required for buffering the membrane tension surge. Finally, wehave calculated that the number of lost caveolae per cellobserved by TIRF and EM is in excellent agreement with theamount of released area ( $ 0.3%) required to buffer s (seeSupplemental Results ).

We also tested whether clathrin-coated pits (CCP), anothertype of plasma membrane invagination, could buffer membrane

A B

C D

Figure 4. Caveolae Buffer the MembraneTension Rise during Hypo-Osmotic Shock (A) Representative force curves for tethers ex-tracted from WT MLEC, Cav1

/ MLEC, Cav1 /

MLEC transfected with Cav1-EGFP, and WT

MLEC treated with m bCD exposed to hypo-osmotic shock (150 mOsm). Hypo-osmotic shockis indicated by an arrow (break from 1.34 to2.7 min).(B) Relative change of the mean tether force afterhypo-osmotic shock (5 min) for WT (n = 9), andCav1

/ MLEC (n = 9; p = 0.01502), WT (n = 3) andCav1

/ MEFs(n=4;p=8 3 104 ),and HeLa cells

(n = 4). Data represent mean SE. See text fordetails.(C) Cav3 immunostaining in differentiated WT andCav3-P28L human myotubes. Scale bar, 5 mm.(D) Relative change of the mean tether force afterhypo-osmotic shock (5 min) for WT (n = 11) andP28L myotubes (n = 12; p = 5 3 10

8 ). Datarepresent mean SE.

tension. Thus, we quantitatively analyzedthe fate of CCPs upon hypo-osmoticshock in MLEC and HeLa cells. Weobserved that the number of CCPs lostat the membrane was, at most, one-tenthof the number of lost caveolae, bothin WT and Cav1

/ MLEC ( FiguresS4 AS4C). Accordingly, deep-etch EMshowed that the structure of CCPs was

not affected ( Figure 3 D). Additionally, under hypo-osmoticshock, membrane tension was buffered to the same extentwhether clathrin was expressed or knocked down in WT MLEC( Figures S4 D and S4E). In contrast, Cav1

/ MLEC havingCCPs could not buffer the membrane tension surge. Theseresults rule out a contribution of CCPs in membrane tensionregulation and establish caveolae as the primary stress-respon-sive membrane structure.

Finally, we obtained similar results in Cav1-EGFP HeLa and inembryonic broblasts (MEF) lacking or expressing Cav1 ( Fig-ure 4 B). We could further extend the physiological signicanceof these ndings by measuring the stress reactivity of humanmuscle cells. Several human muscular dystrophies have beenassociated with mutations in Cav3, the muscular isoform of caveolin ( Woodman et al., 2004 ), and more recently with muta-

tions in Cavin1 ( Hayashi et al., 2009 ). Most Cav3 mutationsprevent caveolae assembly at theplasma membrane by seques-tration of Cav3 in the Golgi apparatus. We studied differentiatedhuman myotubes bearing the P28L Cav3 mutation described infamilial hyperCKaemia (FHCK), a hereditary form of musculardystrophy ( Woodman et al., 2004 ). Muscle bers that were iso-lated from these patients show a strong decrease of Cav3expression and reduced Cav3 staining at the cell surface ( Merliniet al., 2002 ). Accordingly, we found that Cav3 immunostainingwas restricted to the Golgi apparatus in P28L Cav3 myotubes,whereas WT Cav3 was mainly found at the plasma membrane( Figure 4 C). Under hypo-osmotic shock, membrane tensionwas increased in P28L Cav3 myotubes and buffered in WT



7/12

Cav3 myotubes ( Figure 4 D). P28L myotubes and Cav1 / MLEC

also showed an increased tendency to membrane rupture underhypo-osmotic shock ( Figure S5 ). This mayexplain the high blood

level of creatine kinase found in these patients in the absence of a functional caveolar reservoir during the repeated extension-relaxation cycles.

Membrane Tension Surge Buffering by CaveolaFlattening Occurs in an ATP- and Actin-IndependentProcess Although the actin cytoskeleton was unaffected by hypo-osmotic shocks at early times, we still investigated its potentialrole in caveola attening. Under hypo-osmotic shock (30mOsm), none of the actin-perturbing drugs prevented the lossof caveolae from the membrane ( Figure 5 A). This was furtherconrmed by cell ATP depletion before hypo-osmotic shock.

ATP depletion abolished the mobility of the internal pool of cav-eolae, conrming the inhibition of active cellular processes( Movie S4 and Movie S5 ). However, a similar loss of caveolae

occurred in ATP-depleted cells ( Figure 5 A). Likewise, upon cellstretching, caveolae still disappeared in ATP-depleted and cyto-chalasin D (CD)-treated cells ( Figures 5 B and Figure S6 A).Furthermore, membrane tension measurements under hypo-osmotic shock (150 mOsm) showed that the tether force of Cav1

/ MLEC still increased by 200% in cells treated with CDor depleted in ATP ( Figures 5 C and Figure S6 B). In contrast, nosignicant tether force variation was measured for WT MLEC,indicating that the buffering is not dependent on ATP and actindynamics. Importantly, the invaginated shape of caveola wasunaffected by CD or ATP depletion ( Figure S7 ). Finally, we couldunambiguously establish that membrane tension buffering is anintrinsic mechanical property of caveolae by using plasma

0.8

1.0

e

BA

0.8

1.0

e

0.2

0.4

0.6

N o r m a l i z e

d

n o .

o f c a v e o

l a

0.2

0.4

0.6

N o r m a l i z e

d

n o .

o f c a v e o

l a

Ctrl Crtl CD no ATP0.0

+++Stretch -

4 D

Iso Ctrl CD LatA Jas no ATP0.0

Hypo

C Cav1-EGFPCavin1-mCherry

2

3

( f - f

0 ) / f 0

Ctrl CD no ATP Ctrl CD no ATP0

1

Wt Cav1-/-

Cav1-EGFP

E F

0.4

Cavin1-mCherry

5 Pa22 23 24

0.6

Wt - transfectedCav1 -/-

0.40.0

0.2

30 Pa l i z e

d i n t e n s

i t y

0.3

( f - f 0 ) / f 0

0 10 20 30 400.0

0.2

Distance (m)

N o r m

5 10 15 20 25 30 35

.

Aspiration pressure (Pa)

Figure 5. Membrane Tension Surge Buff-ering by Caveolae Flattening Occurs in an ATP- and Actin-Independent Process(A) Normalized number of caveolae per cell afterhypo-osmotic shock (Hypo). Reference is the

number before hypo-osmotic shock (Iso) forcontrol cells(Ctrl; n = 18). Referenceis thenumberafter the drug treatment and before hypo-osmoticshock for cytochalasin D (CD; n = 10, p = 2 3

105 ), latrunculin A (Lat; n = 21, p = 7 3 10

11 ), jasplakinolide (Jas, n = 11, p = 6 3 10

8 ) treatedcells, and ATP-depleted cells (no ATP; n = 10, p =2 3 10

5 ). Data represent mean SD.(B) Change in number of caveolae for single HeLaCav1-EGFP cells after stretching (15% 1%).Same normalization asin (A)for control cells(n = 7;p = 3 3 10

4 ),and forcellstreatedwith cytochalasinD (CD; n = 5; p = 0.01) and ATP-depleted cells (no ATP; n = 5; p = 0.04). Data represent mean SD.(C) Relative change of the tether force after hypo-osmotic shock (5 min) for WT and Cav1 / MLECfor control (Ctrl;n = 9 for WTand n = 5 for Cav1 / ;p = 0.015), cytochalasin D treated (CD; n = 9 forWTandn=10for Cav1

/ ; p = 3 3 105 ), and ATP

depleted (no ATP; n = 6 for WT n = 5 for Cav1 / ;

p = 2 3 104 ) cells. Data represent mean SE.

(D) Confocal image of a WT MLEC transfectedwith Cav1-EGFP (green) and Cavin1-mCherry(red) after incubation for 6 hr in PMS buffer. Scalebar, 10 mm.(E) (Top) Confocal image of a PMS positive forCavin1-mCherry (red) and Cav1-EGFP (green)after micropipette aspiration (white lines) andformation of a membrane tether with an opticallytrapped bead (white disk). Scale bar, 10 mm.(Bottom) Line scans of cavin1-mCherry (red) andCav1-EGFP (green) normalized intensity along thecircumference of the PMS shown above for twoaspiration pressures. Arrows indicate regions of colocalization.(F) Relative change of the tether force as a func-tion of the micropipette aspiration pressure inPMS obtained from Cav1-GFP + Cavin1-mCherryMLEC (black squares) and from Cav1

/ MLEC( Cav1

/ , red circles). Data were obtained in eightindependent experiments and represent the meanvalue of 60 s measurements SD.



8/12

membrane spheres (PMS). PMS are composed of plasmamembrane and cytosol, whereas the subcellular compartments( Lingwood et al., 2008 ) and lamentous actin ( Figure S6 C) areexcluded. PMS were prepared from MLEC transfected with

Cav1-EGFPand Cavin1-mCherry. As expected, a major colocal-ization of Cavin1 and Cav1 was observed at the plasmamembrane of the donor cell ( Figure 5 D). Confocal imagingrevealed that Cavin1 was present inside PMS and along themembrane, where it colocalized with Cav1 as a punctuatedpattern likely to reect the incorporation of caveolar invagina-tions in these vesicles ( Figure 5 E). PMS were then mechanicallystressed by micropipette aspiration ( Dimova et al., 2006 ).Membrane tethers were pulled with optical tweezers from PMSaspirated under minimal pressure ( Figure 5 E). Upon increasingaspiration pressure,PMS from Cav1

/ MLECexhibiteda steadytether force increase, whereas f remained constant over a signif-icant range of aspiration pressures in PMS from WT MLEC ( Fig-ure 5 F). Concomitantly, over the aspiration range corresponding

to the force plateau, the number of Cavin1/Cav1 colocalizationsalso decreased to noise level ( Figure 5 E). When the aspirationwas increased further, the force plateau was followed by anincrease in f , which is in agreement with the complete atteningof caveolae and thus depletion of all residual membrane reser-voir ( Figure 5 F). We also measured identical amounts of freecholesterol in WT and Cav1

/ PMS ( $ 75 20 nmol/mg of cellprotein), indicating that the lack of buffering effect in Cav1

/

PMS is not related to changes in lipid composition. These resultson PMSreinforce our ndings obtained with actin drugs and ATPdepletion in cells and denitely establish that membrane tensionbuffering by caveolae attening is a purely passive mechanismsolely driven by membrane mechanics.

Caveolae Reassemble in an Actin- and ATP-DependentProcess upon Stress RelaxationWe next examined the behavior of caveolae when cells were leftin hypo-osmotic medium (30 mOsm) for 5 min and returned toiso-osmotic medium (300 mOsm). Within 25 min, caveolaereappeared at the plasma membrane ( Figures 3 A3D and Fig-ure 6 A). Accordingly, there was a decrease of both the fractionof mobile Cav1 ( Figure 3 E) and Cav1 immunogold labeling innoncaveolar membranes ( Figures 3 A and 3B and Figure S2 A).We also measured a decrease in EGFP uorescence lifetime,indicating the reassociation of Cav1 and Cavin1 and thereforethe reassembly of functional caveolae at the cell surface( Figure 6 B).

On average, the initial rate of reassembly was about 10 caveo-lae per min per cell and required ATP ( Figure 6 C). In contrast tocaveolae disassembly, reassembly was enhanced when actindynamics were blocked by CD. Finally, we tested whethercaveolae reassembled directly from the plasma membranepool of free Cav1 or from other compartments. The disruptionof the Golgi apparatus by Brefeldin A (BFA) did not prevent reas-sembly excluding a major contribution of the Cav1 Golgi pool( Figure 6 D). Furthermore, we found no signicant transfer of the photoconverted Golgi pool of Cav1 to the plasma membraneon returning to iso-osmolarity. However, the nonphotoconvertedCav1 present throughout the cell showed an increased punctu-ated localization at the plasma membrane ( Figure 6 E and 6F).

DISCUSSION

Since their discovery in 1953, the precise role of caveolae hasremained a matter of considerable debate. Freeze-fracture anal-

ysis of the surface membrane of smooth and striated musclecells in the early 1970s led to the rst hypothesis that caveolaecould atten out under stretching conditions ( Dulhunty andFranzini-Armstrong, 1975; Prescott and Brightman, 1976 ). Laterstudies have also associated caveolae with the mechanosens-ing response of the cell ( Boyd et al., 2003; Park et al., 2000;Rizzo et al., 2003; Sedding et al., 2005; Kawamura et al.,2003; Kozera et al., 2009 ). Whether caveolae are directlyinvolved in the cell response to mechanical stress, and by whichmechanisms, still remain unknown. In this study, we establishthat the primary cell response to an acute mechanical stressoccurs through the rapid attening of caveolae into the plasmamembrane. Upon cell stretching or osmotic swelling, caveolaeattening provides the additional amount of membrane that

inhibits any surge of membrane tension. This response occursalso in ATP-depleted cells and in membrane-derived vesiclesdevoid of actin, indicating that the ability to buffer membranetension is intrinsic to the caveola structure. These results arein line with a theoretical model proposing that invaginateddomains can serve as a membrane reservoir to regulatethe membrane tension ( Sens and Turner, 2006 ). However, incontrast to this model, which predicted that attening of existingcaveolae or budding of new caveolae was driven by the devia-tion of the membrane tension with respect to the resting tension,our data indicate that the regulatory mechanism is asymmetric.Whereas stress-induced attening of caveolae is truly passive,the reassembly of new caveolae is assisted by ATP and actindynamics. The energy landscape for at and invaginated cav-eola is thus characterized by an activation energy. The transitionbetween the two forms thus requires an energy barrier to beovercome or lowered. In this model, the at conguration wouldbe stabilized upon application of a mechanical force, whereasthe formation of invaginated caveola is energetically favoredthrough interactions with cortical actin in the presence of ATP.This is consistent with the known role of actin dynamics andvarious kinases in caveola function ( Pelkmans et al., 2005;Pelkmans and Zerial, 2005 ). Our results further indicate thatclathrin-coated pits, the other abundant invaginations presentat the plasma membrane, are not involved in the fast responseto membrane strain. Whereas CCPs exchange coat proteinswithin seconds, most caveolae are rather stable at the plasma

membrane. The stability of caveolae combined with their lowendocytic activity allows this membrane reservoir to be readilyavailable to respond to sudden mechanical strains. The stabilityof the caveolae reservoir is a striking feature of cells thatexperience pulsating mechanical stresses in their lifetime(muscle cells, cardio-myocytes, and endothelial cells) becausethey all express very high numbers of caveolae. Indeed, wecould associate the lack of Cav3 expression at the surface of myotubes with impaired membrane tension buffering in patientsbearing the Cav3 P28L mutation found in a form of humanmuscular dystrophy. Interestingly, P28L myotubes appear tobe more fragile than WT myotubes when exposed to acuteosmotic shock.



9/12

We have shown that the timescales of caveolae and actincortex responses to osmotic shock are well separated (5 min, respectively). Though the immediate response tomechanical stress relies primarily on caveola attening, it is

likely that other processes such as endocytosis, exocytosis,and actin dynamics may prolong or complete the initialresponse at longer times. Whether and how caveolae alsocontribute to long timescale regulation remain to be investi-gated. In endothelial cells, the application of chronic and repet-itive shear stress tensions results in a several-fold increase inthe number of caveolae at the plasma membrane through themobilization of the Cav1 pool associated with the Golgi complex( Boyd et al., 2003; Park et al., 1998 ). In agreement with the slowdelivery rate of Cav1 from the Golgi complex ( Tagawa et al.,2005 ), we show that caveolae are reassembled independentlyfrom the Golgi complex when the mechanical stress is relaxed.It is thus important to distinguish between the short-term

mechanical function of caveolae and the long-term adaptationof the cell to chronic stress. In this context, we also analyzedthe contribution of caveolae to the setting of the membranetension under resting conditions. At steady state, we measured

a lower membrane tension in Cav1 /

than in WT MLEC;however, it was identical in WT and Cav1 / MEF, albeit toa lower level ( Figure S8 A), raising the possibility of a peculiarityof MLEC. Furthermore, the resting tension was drasticallyaffected by m- b -cyclodextrin, cytochalasin D, or ATP depletiontreatments in the different cell types, independently from theexpression of caveolae ( Figure S8 B). Although caveolae playa key role in cell tension homeostasis, their direct contributionto the resting tension remains intricate. As previouslymentioned, the dynamic membrane-to-cytoskeleton adhesion,which is the main contribution to the apparent membranetension, is likely to be regulated by cell line-dependent compen-satory mechanisms.

A B

C D

E F

Figure 6. Caveolae Reassembly at thePlasma Membrane upon Recovering Iso-Osmolarity (A) TIRF imaging of caveolae reassembly aftershifting from hypo-osmotic (Hypo) to iso-osmotic

(Rec) conditions. Dotted line marks out the cellboundary. Scale bar, 5 mm.(B) HeLacellstransientlyexpressingCavin1-EGFPwith or without Cav3-mRFP were exposed tohypo-osmotic medium (Hypo) for 15 min followedby iso-osmotic medium (Rec) for 10 min andwere analyzed by FLIM. Data represent meanEGFP uorescence lifetime SE (n = 4060 cells;*p = 8 3 10

7 ).(C) After hypo-osmotic shock (15 min), cells werereturned to iso-osmolarity (t = 0 min). Data repre-sent the number of caveolae per cell for cytocha-lasin D-treated (CD), ATP-depleted (no ATP), andcontrol cells (Ctrl). Error bars represent SD.(D) (Left) TIRF images show a typical BFA-treatedcell after 15 min hypo-osmotic shock (Hypo) andafter 3 min of recovering iso-osmolarity.Scale bar,10 mm. (Right) Quantication of the number of caveolae before hypo-osmotic shock, after 15 minhypo-osmotic shock, and after 3 min of recoveryto iso-osmotic conditions in BFA-treated (n = 3)and control cells (n > 10). *p = 8 3 10

29 ; **p = 1 3

1022 . Error bars represent SD.

(E) Cav1-Dendra2 photoconversion from greento red uorescence at the Golgi apparatus(rectangle) of HeLa cells followed through hypo-osmotic shock and recovery to iso-osmoticconditions. Arrow marks out the section of theplasma membrane used for plotting intensityproles.(F) Intensity proles of green (Cav1-Dendra2) andred (photoconverted Cav1-Dendra2) uorescenceat the plasma membrane before (top) and after(bottom) switch from hypo-osmotic to iso-osmoticconditions.



10/12

The well-conserved scaffolding domain of caveolin (CSD) hasbeen involved in both the assembly of caveolae and the interac-tion with several signaling effectors in vitro ( Parton et al., 2006;Parton and Simons, 2007 ). Because several of these effectorshave been associated with mechanotransduction ( Vogel andSheetz, 2006 ), stress-induced disassembly of caveolae maygenerate mechanosensitive signals mediating the short- andlong-term cell response to mechanical challenges. It is temptingto speculate that the mechanical release of free Cav1 oligomerscould favor the interaction between signaling effectors and CSD,which is otherwise hidden in the caveolar structure ( Kirkhamet al., 2008; Parton et al., 2006 ). This mechanosensitive signalingwould be terminated through the reassembly of free Cav1 oligo-mers into caveolae when the mechanical stress is relaxed.Endocytosis, which is favored for the retrieval of the excess of

membrane during tension relaxation, may also contribute tosignaling termination through the internalization of free Cav1and its degradation in the endolysosomal pathway. The recentlycharacterized Cavin1 protein, which was rst described asa transcription factor ( Jansa et al., 2001 ), may also contributeto mechanosignaling regulation through the release from at-tened caveola. Indeed, the redistribution of Cav1 to noncaveolarportions of the plasma membrane, the increased mobility of Cav1, and the decreased association of Cav1 and Cavin1 uponhypo-osmotic treatment are all consistent with the effect of Cavin1 knockdown on these parameters ( Hill et al., 2008 ), sug-gesting that dissociation of the Cav1-cavin module may becrucial in the caveolar response.

Our study establishes a new physiological mechanism bywhich cells can respond immediately to sudden variations inmembrane tension induced by acute mechanical stress ( Fig-ure 7 ). The different proposed roles of caveolae should therefore

be reconsidered through this unique ability to respond tomechanical stress, especially in situations in which cells experi-ence physiological or pathological membrane strains such asosmotic swelling, shear stress, or mechanical stretching.

EXPERIMENTAL PROCEDURES

A list of chemicals and materials can be found in the Extended ExperimentalProcedures .

Cell Culture and TreatmentsMLEC ( Murata et al., 2007 ) were maintained in EGM-2/20% FBS medium.They were transfected with AMAXA HUVEC nucleofector kit and used forexperiments after 2472 hr. HeLa cells stably transfected with Cav1-EGFP( Pinaud et al., 2009 ) and MEFs were maintained in DMEM/10% FBS. HeLa

cells were transfected with FuGENE or Lipofectamine 2000 and used after16 hr. Human muscle cells were maintained in X medium (64% DMEM, 16%199 Medium, 20% FBS, 2.5 ng/ml HGF, 50 mg/ml gentamycin + 10

7 M dexa-methasone). For differentiation, cells were grown on collagen type I coatedsurface for 710 days in DMEM supplemented with 50 mg/ml gentamycin,10 mg/ml of bovine insulin, and 100 mg/ml of human apotransferrin. For dyna-sore treatment, cells were washed thrice with PBS 2+ (Phosphate BufferSaline + 1.5 mM Ca 2+ + 1.5 mM Mg 2+ ) and overlaid with 80 mM dynasore inPBS 2+ . Cytochalasin D (CD), Latrunculin A (Lat), and Jasplakinolide (Jas)were used for 15 min at 37 C at 5 mg/ml, 1 mM, or 1 mM, respectively in normalmedium. ATP depletion was performed by incubating the cells for 30 min at37 C in PBS 2+ with 10 mM deoxy-D-glucose and 10 mM NaN 3 . For disruptingthe Golgi apparatus, cells were pretreated for 50 min with 10 mg/ml BFA. Cellswere incubated for 50 min in PBS 2+ supplemented with 5 mMm bCDto extractcholesterol. Hypo-osmotic shock was performed on pretreated cells by usinggrowth medium diluted appropriately in deionized water (1:9 dilution for

30 mOsm hypo-osmotic shock and 1:1 for 150 mOsm hypo-osmotic shock).The concentration of all drugs was maintained during the hypo-osmotic shockas well as during recovery.

Membrane Tether Extraction and Force MeasurementsPlasma membrane tethers were pulled out from cells by a concanavalin A-coated bead trapped in optical tweezers ( Cuvelier et al., 2005 ). After extrac-tion, thetetherwasheldat a constantlength between 5 and150 mm,and tetherforces were measured from the detected position of the bead after calibrationof the optical trap. Samples were maintained at 37 C throughout.

PMS Formation and Micropipette AspirationPlasma membranespheres(PMS) weregenerated by a protocol adapted fromLingwood et al. (2008) . Cells were grown on glass coverslips and were incu-bated for 68 hr in PBS 2+ supplemented with 10 mM MG132. Individual PMSwereselected for micropipette aspiration experimentsas described previouslyfor lipid vesicles( Sorreet al.,2009 ).In brief,PMS wereheld with a micropipette(diameter $ 3 mm) under slight aspiration. By partially entering the pipette, themembrane is strained. Tether forces were measured as explained above, andthe aspiration of the PMS was gradually increased.

Fluorescence Imaging and AnalysisTIRF microscopy was performed using a 100 3 , 1.45 NA objective and anEMCCD camera (Hamamatsu Photonics,Japan)in a Zeiss Axiovert200 micro-scope. Confocal imaging was performed on a Nikon A1R microscope witha 100 3 , 1.4 NA objective at 1 airy unit pinhole aperture. TIRF-FRAP experi-ments (5 3 5 mm bleaching region) were performed on a Nikon Eclipse 2000microscope equipped with an EMCCD camera (Roper Scientic, Tucson, AZ). Cells were maintained at 37 C during imaging. Confocal images of PMSwere taken in a Z section of width 0.4 mm around the equatorial plane. Image

Figure 7. Cells Respond to Acute Mechanical Stresses by RapidDisassembly and Reassembly of CaveolaeIn resting conditions, caveolae present at the plasma membrane are mostlybudded. Magnication shows oligomerized Cav1 and Cavin1 in the caveolarstructure. Upon acute mechanical stress (hypo-osmotic shock or stretching),caveolae atten out in the plasma membrane to provide additional membraneand buffer membranetension.Magnication shows disassembly and diffusionof Cav1 in the plasma membrane and loss of interaction between Cav1 andCavin1. Return to resting conditions allows the reassembly of the caveolarstructure together with Cavin1 interaction. This cycle represents the primarycell response to an acute mechanical stress.



11/12

analysis(caveolae detection, volume measurement, FRAP,and colocalization)was done using Labview 8.5 and Vision 8.5 (National Instruments, Austin, TX)as detailed in the Extended Experimental Procedures . FLIM microscopy wasperformed as described previously ( Hill et al., 2008 ).

Cell StretchingCells were grown on a rectangular PDMS sheet (thickness $ 100 mm, dimen-sions $ 12 3 7 mm) decorated with adhesive micropatterns and stretcheduniaxially using a custom-built device equipped with a motorized linearactuator (PI, Karlsruhe, Germany) and a temperature controller. Imaging wasdone on the TIRF microscope. Strains were applied in $ 5 s, and the numberof caveolae was measured from images taken within 13 min of stretching.

Immunouorescence StudiesCells were xed, immunolabeled, and imaged on an inverted microscope(Leica, Wetzlar, Germany).

Deep-Etched EM of the Cytoplasmic Surface of MLECCells grownon coverslips wereprepared by rst unroong their apical surface,and then xing and freezing the basal surface. The cytoplasmic surface was

deeply etched and rotary shadowed with platinum/carbon at an angle of 22 C from the surface and with carbon from the top. The replica was mountedon the sample grid and observed using a transmission electron microscope.

Electron MicroscopyFor conventional EM, cells were grown on coverslips, xed, dehydrated, andembedded in epon. For ultrathin cryosectioning and immunogold labeling,cells were xed and processed for ultracryomicrotomy and were immunogoldlabeled. Quantication of caveolae was done by counting only the number of neck-open caveolae at the plasma membrane in EM images that had at leastone caveola.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Extended Results, Extended ExperimentalProcedures, eight gures, and ve movies and can be found online at doi:10.1016/j.cell.2010.12.031 .

ACKNOWLEDGMENTS

This work was supported by the Institut Curie (PIC Division Cellulaire, Polarite et Cancer), Agence Nationale pour la Recherche, Fondation de France (pro-gramme tumeurs), Association Francaise contre les Myopathies, the NationalHealth and Medical ResearchCouncil of Australia and the Australian ResearchCouncil (R.G.P. and D.A.), and NIH grants HL83249 and HL92085 (R.V.S.).Microscopy (M.B.) was performed at the ACRF/IMB Dynamic Imaging CentreforCancer Biology. D.K. wasfunded bya doctorate fellowship from theInstitutCurie, and B.S. was funded by postdoctoral fellowships from the Institut Curieand the Association pour la Recherche sur le Cancer. D.A. is a fellow of theSwiss National Science Foundation (PA00A-111446) and is indebted toK. Alexandrov for support. The authors gratefully acknowledge the NikonImaging Centre at Institut Curie and the CNRS consortium CellTis. We thankF. Perez for the pDendra2-C plasmid, A. Helenius for the Cav1-EGFP plasmidand MEF cells, and F. Pinaud for Cav1-EGFP HeLa. We are also grateful toM. Piel and N. Carpi for help in micropatterning and P. Bassereau for accessto the confocal microscope coupled to optical tweezers. We thank C. Viarisand C. Blouin for carefully reading the manuscript, and we are indebted to A. Bigotfor providinghuman muscle cells. We would like to thankEurobiobankand Pr. C. Minettefor giving access to primary cells to isolate the immortalizedCav3 cell line. C.L. wishes to dedicate this work to the memory of his father,Robert Lamaze, and to all patients suffering from pancreatic cancer.

Received: January 7, 2010Revised: October 27, 2010 Accepted: December 23, 2010Published: February 3, 2011

REFERENCES

Abankwa,D., Hanzal-Bayer, M.,Ariotti,N., Plowman,S.J., Gorfe, A.A., Parton,R.G., McCammon, J.A., and Hancock, J.F. (2008). A novel switch region regu-lates H-ras membrane orientation and signal output. EMBO J. 27 , 727735.

Boyd, N.L., Park, H., Yi, H., Boo, Y.C., Sorescu, G.P., Sykes, M., and Jo, H.(2003). Chronic shear induces caveolae formation and alters ERK and Aktresponses in endothelial cells. Am. J. Physiol. Heart Circ. Physiol. 285 ,H1113H1122.

Chen,C.S., Mrksich, M., Huang, S., Whitesides, G.M., and Ingber, D.E.(1997).Geometric control of cell life and death. Science 276 , 14251428.

Cuvelier, D., Dere nyi, I., Bassereau, P., and Nassoy, P. (2005). Coalescence of membrane tethers: experiments, theory, and applications. Biophys. J. 88 ,27142726.

Dai, J., and Sheetz, M.P. (1995). Regulation of endocytosis, exocytosis, andshape by membrane tension. Cold Spring Harb. Symp. Quant. Biol. 60 ,567571.

Dai, J., and Sheetz, M.P. (1999). Membrane tether formation from blebbingcells. Biophys. J. 77 , 33633370.

Dai, J.,Ting-Beall, H.P., andSheetz,M.P. (1997). Thesecretion-coupledendo-cytosis correlates with membrane tension changes in RBL 2H3 cells. J. Gen.Physiol. 110 , 110.

Dai, J., Sheetz, M.P., Wan, X., and Morris, C.E. (1998). Membrane tension inswelling and shrinking molluscan neurons. J. Neurosci. 18 , 66816692.

DAlessandro, M., Russell, D., Morley, S.M., Davies, A.M., and Lane, E.B.(2002). Keratin mutations of epidermolysis bullosa simplex alter the kineticsof stress response to osmotic shock. J. Cell Sci. 115 , 43414351.

Dimova, R., Aranda, S., Bezlyepkina, N., Nikolov, V., Riske, K.A., andLipowsky,R. (2006). A practical guideto giantvesicles.Probing themembranenanoregime via optical microcopy. J. Phys. Condens. Matter 18 , S1151S1176.

Dulhunty, A.F.,and Franzini-Armstrong, C. (1975). The relativecontributions of the folds and caveolae to the surface membrane of f rog skeletal muscle bresat different sarcomere lengths. J. Physiol. 250 , 513539.

Hansen, C.G., Bright, N.A., Howard, G., and Nichols, B.J. (2009). SDPRinduces membrane curvature and functions in the formation of caveolae.Nat. Cell Biol. 11 , 807814.

Hayashi, Y.K., Matsuda, C., Ogawa, M., Goto, K., Tominaga, K., Mitsuhashi,S., Park, Y.E., Nonaka, I., Hino-Fukuyo, N., Haginoya, K., et al. (2009). HumanPTRFmutations cause secondarydeciencyof caveolinsresulting in musculardystrophy with generalized lipodystrophy. J. Clin. Invest. 119 , 26232633.

Henley, J.R., Krueger, E.W., Oswald, B.J., and McNiven, M.A. (1998). Dyna-min-mediated internalization of caveolae. J. Cell Biol. 141 , 8599.

Hill, M.M., Bastiani, M., Luetterforst, R., Kirkham, M., Kirkham, A., Nixon, S.J.,Walser, P., Abankwa, D., Oorschot,V.M., Martin, S., et al. (2008). PTRF-Cavin,a conserved cytoplasmic protein required for caveola formation and function.Cell 132 , 113124.

Jansa, P., Burek, C., Sander, E.E., and Grummt, I. (2001). The transcriptrelease factor PTRF augments ribosomal gene transcription by facilitatingreinitiation of RNA polymerase I. Nucleic Acids Res. 29 , 423429.

Kawamura, S., Miyamoto, S., and Brown, J.H. (2003). Initiation and transduc-tion of stretch-induced RhoA and Rac1 activation through caveolae: cytoskel-etal regulation of ERK translocation. J. Biol. Chem. 278 , 3111131117.

Kirkham, M., Nixon, S.J., Howes, M.T., Abi-Rached, L., Wakeham, D.E.,Hanzal-Bayer, M., Ferguson, C., Hill, M.M., Fernandez-Rojo, M., Brown,D.A., et al. (2008). Evolutionary analysis and molecular dissection of caveolabiogenesis. J. Cell Sci. 121 , 20752086.

Kozera, L., White, E., and Calaghan, S. (2009). Caveolae act as membranereserves which limit mechanosensitive I( Cl,swell ) channel activation duringswelling in the rat ventricular myocyte. PLoS ONE 4 , e8312.

Kurzchalia, T.V., Dupree, P., Parton, R.G., Kellner, R., Virta, H., Lehnert, M.,and Simons, K. (1992). VIP21, a 21-kD membrane protein is an integral

http://dx.doi.org/doi:10.1016/j.cell.2010.12.031http://dx.doi.org/doi:10.1016/j.cell.2010.12.031http://dx.doi.org/doi:10.1016/j.cell.2010.12.031http://dx.doi.org/doi:10.1016/j.cell.2010.12.031


12/12

component of trans-Golgi-network-derived transport vesicles. J. Cell Biol.118 , 10031014.

Lang, F., Busch, G.L., Ritter, M., Vo lkl, H., Waldegger, S., Gulbins, E., andHa ussinger, D. (1998). Functional signicance of cell volume regulatory mech-anisms. Physiol. Rev. 78 , 247306.

Lingwood, D., Ries, J., Schwille, P., and Simons, K. (2008). Plasmamembranes are poised for activation of raft phase coalescence at physiolog-ical temperature. Proc. Natl. Acad. Sci. USA 105 , 1000510010.

Macia, E., Ehrlich, M., Massol, R., Boucrot, E., Brunner, C., and Kirchhausen,T. (2006). Dynasore, a cell-permeable inhibitor of dynamin. Dev. Cell 10 ,839850.

Merlini, L., Carbone, I., Capanni, C., Sabatelli, P., Tortorelli, S., Sotgia, F.,Lisanti,M.P., Bruno, C., and Minetti,C. (2002). Familialisolated hyperCKaemiaassociated with a new mutation in the caveolin-3 (CAV-3) gene. J. Neurol.Neurosurg. Psychiatry 73 , 6567.

Morone, N., Fujiwara, T., Murase, K., Kasai, R.S., Ike, H., Yuasa, S., Usukura,J., and Kusumi, A. (2006). Three-dimensional reconstruction of the membraneskeleton at the plasma membrane interface by electron tomography. J. CellBiol. 174 , 851862.

Morris, C.E., and Homann, U. (2001). Cell surface area regulation andmembrane tension. J. Membr. Biol. 179 , 79102.

Murata, T., Lin, M.I., Stan, R.V., Bauer, P.M., Yu, J., and Sessa, W.C. (2007).Genetic evidence supporting caveolae microdomain regulation of calciumentry in endothelial cells. J. Biol. Chem. 282 , 1663116643.

Nabi, I.R., and Le, P.U. (2003). Caveolae/raft-dependent endocytosis. J. CellBiol. 161 , 673677.

Palade, G.E. (1953). An electron microscope study of the mitochondrial struc-ture. J. Histochem. Cytochem. 1 , 188211.

Park, H., Go, Y.M., St John, P.L., Maland, M.C., Lisanti, M.P., Abrahamson,D.R., and Jo, H. (1998). Plasma membrane cholesterol is a key molecule inshear stress-dependent activation of extracellular signal-regulated kinase.J. Biol. Chem. 273 , 3230432311.

Park, H.,Go, Y.M., Darji, R.,Choi,J.W., Lisanti, M.P.,Maland,M.C., andJo, H.(2000). Caveolin-1 regulatesshear stress-dependent activation of extracellularsignal-regulated kinase. Am. J. Physiol. Heart Circ. Physiol. 278 , H1285H1293.

Parton, R.G., and Simons, K. (2007). The multiple faces of caveolae. Nat. Rev.Mol. Cell Biol. 8 , 185194.

Parton, R.G., Hanzal-Bayer, M., and Hancock, J.F. (2006). Biogenesis of caveolae: a structural model for caveolin-induced domain formation. J. CellSci. 119 , 787796.

Pelkmans, L., and Zerial, M. (2005). Kinase-regulated quantal assemblies andkiss-and-run recycling of caveolae. Nature 436 , 128133.

Pelkmans, L., Bu rli, T., Zerial, M., and Helenius, A. (2004). Caveolin-stabilizedmembrane domains as multifunctional transport and sorting devices inendocytic membrane trafc. Cell 118 , 767780.

Pelkmans, L., Fava, E., Grabner, H., Hannus, M., Habermann, B., Krausz, E.,and Zerial, M. (2005). Genome-wide analysis of human kinases in clathrin-and caveolae/raft-mediated endocytosis. Nature 436 , 7886.

Pinaud, F., Michalet, X., Iyer, G., Margeat, E., Moore, H.P., and Weiss, S.(2009). Dynamic partitioning of a glycosyl-phosphatidylinositol-anchoredprotein in glycosphingolipid-rich microdomains imaged by single-quantumdot tracking. Trafc 10 , 691712.

Prescott, L., and Brightman, M.W. (1976). The sarcolemma of Aplysia smoothmuscle in freeze-fracture preparations. Tissue Cell 8 , 241258.

Richter, T., Floetenmeyer, M., Ferguson, C., Galea, J., Goh, J., Lindsay, M.R.,Morgan, G.P., Marsh, B.J., and Parton, R.G. (2008). High-resolution 3D quan-titative analysis of caveolar ultrastructure and caveola-cytoskeleton interac-tions. Trafc 9 , 893909.

Rizzo, V., Morton, C., DePaola, N., Schnitzer, J.E., and Davies, P.F. (2003).Recruitment of endothelial caveolae into mechanotransduction pathways byow conditioning in vitro. Am. J. Physiol. Heart Circ. Physiol. 285 , H1720H1729.

Rothberg, K.G., Heuser, J.E., Donzell, W.C., Ying, Y.S., Glenney, J.R., and Anderson, R.G. (1992). Caveolin, a protein component of caveolae membranecoats. Cell 68 , 673682.

Scherer, P.E., Okamoto, T., Chun, M., Nishimoto, I., Lodish, H.F., and Lisanti,M.P. (1996). Identication, sequence, and expression of caveolin-2 denesa caveolin gene family. Proc. Natl. Acad. Sci. USA 93 , 131135.

Sedding, D.G., Hermsen, J., Seay, U., Eickelberg, O., Kummer, W.,Schwencke, C., Strasser, R.H., Tillmanns, H., and Braun-Dullaeus, R.C.(2005). Caveolin-1facilitates mechanosensitiveproteinkinase B (Akt) signalingin vitro and in vivo. Circ. Res. 96 , 635642.

Sens, P., and Turner, M.S. (2006). Budded membrane microdomains astension regulators. Phys. Rev. E Stat. Nonlin. Soft Matter Physiol. 73 , 031918.

Sheetz, M.P. (2001). Cell control by membrane-cytoskeleton adhesion. Nat.Rev. Mol. Cell Biol. 2, 392396.

Sorre, B., Callan-Jones, A., Manneville, J.-B., Nassoy, P., Joanny, J.-F., Prost,J., Goud, B., and Bassereau, P. (2009). Curvature-driven lipid sorting needsproximity to a demixing point and is aided by proteins. Proc. Natl. Acad. Sci.USA 106 , 56225626.

Stan, R.V. (2005). Structure of caveolae. Biochim. Biophys. Acta 1746 ,334348.

Tagawa, A., Mezzacasa, A., Hayer, A., Longatti, A., Pelkmans, L., andHelenius, A. (2005). Assembly and trafcking of caveolar domains in the cell:caveolae as stable, cargo-triggered, vesicular transporters. J. Cell Biol. 170 ,769779.

Vogel, V., and Sheetz, M. (2006). Local force and geometry sensing regulatecell functions. Nat. Rev. Mol. Cell Biol. 7 , 265275.

Woodman, S.E., Sotgia, F., Galbiati, F., Minetti, C., and Lisanti, M.P. (2004).Caveolinopathies: mutations in caveolin-3 cause four distinct autosomaldominant muscle diseases. Neurology 62 , 538543.

Yamada, E. (1955). The ne structure of the gall bladder epithelium of themouse. J. Biophys. Biochem. Cytol. 1 , 445458.

Documents

Cells Respond to Mechanical Stress by Rapid Dis Assembly of Caveolae