membrane ruffles in cell migration: indicators of

www.elsevier.com/locate/yexcr

Experimental Cell Resear

Membrane ruffles in cell migration: indicators of inefficient lamellipodia

adhesion and compartments of actin filament reorganization

Bodo Borm, Robert P. Requardt, Volker Herzog, Gregor Kirfel*

Institute for Cell Biology, University of Bonn, 53121 Bonn, Germany

Received 14 June 2004, revised version received 25 August 2004

Available online 30 September 2004

Abstract

During epithelial cell migration, membrane ruffles can be visualized by phase contrast microscopy as dark waves arising at the leading

edge of lamellipodia that move centripetally toward the main cell body. Despite the common use of the term membrane ruffles, their

structure, molecular composition, and the mechanisms leading to their formation remained largely unknown. We show here that membrane

ruffles differ from the underlying cell lamella by more densely packed bundles of actin filaments that are enriched in the actin cross-linkers

filamin and ezrin, pointing to a specific bundling process based on these cross-linkers. The accumulation of phosphorylated, that is,

inactivated, cofilin in membrane ruffles suggests that they are compartments of inhibited actin filament turnover. High Rac1 and low RhoA

activities were found under conditions of suboptimal integrin–ligand interaction correlating with low lamellipodia persistence, inefficient

migration, and high ruffling rates. Based on these findings, we define membrane ruffles as distinct compartments of specific composition that

form as a consequence of inefficient lamellipodia adhesion.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Cell migration; Membrane ruffle; Lamellipodia; Rac; Rho; Integrin; Cell adhesion; Filamin; Motility assay

Introduction

Cell migration is central not only to many biological

processes, including embryogenesis, the inflammatory

response, tissue repair, and regeneration, but also to

pathological processes such as cancer and metastasis. The

onset of migration is accompanied by the acquisition of a

spatial asymmetry that is manifested in a polarized cell

morphology, that is, a clearly distinguishable front and rear

of the cell. Cell migration can be viewed as a multistep cycle

beginning with the extension of membrane protrusions

(lamellipodia, filopodia) at the cell lamella, which is the flat,

almost organelle free front region of motile cells. The cycle

is continued by the formation of cell substrate adhesions

near the tips of protrusions and the acto-myosin powered

0014-4827/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.yexcr.2004.08.034

* Corresponding author. Institute for Cell Biology, University of Bonn,

Ulrich-Haberlandstr. 61a, 53121 Bonn, Germany.

E-mail address: [email protected] (G. Kirfel).

contraction of the cell body to be finally completed by the

release of cell-substrate adhesions at the cell rear [1,2].

The formation of lamellipodia and filopodia at the cell

front is driven by the assembly of actin filaments from the

cytoplasmic pool of monomeric actin bound to sequestering

proteins such as profilin and thymosin h4 [3]. The signals

for actin assembly to start are provided by constituents of

the extracellular matrix (ECM) such as fibronectin (FN) and

by soluble motogens from the large families of growth

factors including the epidermal growth factor (EGF) [4,5].

Both kinds of signals are sensed by appropriate trans-

membrane receptors that, upon ligand binding, activate

different signaling pathways [6,7]. The most prominent

receptors for ECM components are the proteins of the

integrin gene family [6] that, upon activation by ligand

binding, can activate the small GTPases of the Rho family

such as RhoA, Rac1, and Cdc42 [5,8]. Rac1 and Cdc42

have been shown to act as regulators of actin assembly

essential for lamellipodia and filopodia protrusion [9]. Of

the many effector proteins that interact with activated Rac1

ch 302 (2005) 83–95

B. Borm et al. / Experimental Cell Research 302 (2005) 83–9584

and Cdc42, the Wiskott–Aldrich syndrome protein (WASP)

and its ubiquitous family member N-WASP provide a direct

link to actin assembly by activating the Arp2/3 complex

[10]. This complex acts as a nucleation point for the

assembly of the branched actin network that is the

predominant cytoskeletal element of growing lamellipodia.

According to their role as promoters of actin assembly,

Arp2/3 and WASP are concentrated at the tips of lamelli-

podia of migrating cells [11,12].

Since the cell is a bounded compartment, rapid assembly

of actin filaments cannot continue for long without being

balanced by rapid and highly regulated disassembly. The

ubiquitous proteins of the ADF/cofilin family have been

shown to bind to actin filaments and, thereby, promote their

disassembly [13]. The activity of ADF/cofilin depends on

LIM-kinase-controlled phosphorylation that blocks its inter-

actions with actin [14]. Since LIM-kinase is a downstream

effector of Rho family GTPases, actin assembly is promoted

in two ways: by activating the Arp2/3 nucleation complex

and by inhibiting disassembly via ADF/cofilin regulation.

Recent findings additionally point to a positive engagement

of cofilin in actin assembly by generating free barbed ends

that are a prerequisite for filament elongation by the addition

of actin monomers [15].

For sustained migration to occur, the newly formed

membrane protrusions have to be stabilized by the

formation of firm adhesions to the substrate [16]. The

initial step during the establishment of these adhesions has

been shown to be the aggregation of ligand-bound integrins

in the form of small focal contacts (FC) at the leading edge

of protrusions. This aggregation of integrins then triggers

the activation of signaling molecules such as the small

GTPase RhoA that is responsible for the recruitment of

cytoskeletal- and FC-associated proteins, resulting in the

formation of contractile acto-myosin bundles and the

generation of the larger and more organized focal adhesions

(FA) [8,9,16,17]. The small FCs at the cell front are thought

to be essential for cell locomotion and act as anchoring sites

for the transmission of acto-myosin-driven propulsive

forces, whereas the larger FAs seem to inhibit cell migration

by forming stable and less dynamic anchoring structures

[18,19].

Lamellipodia that fail to establish stable adhesions are

believed to become detached from the substrate and

retracted toward the cell body [6,16,17]. It has been

speculated that during this retraction process, membrane

ruffles are formed that are clearly visible in the light

microscope as waves of dark contrast on the frontal surface

of cells. These ruffles emerge at the cell edge, move

centripetally, and finally disappear at the border between the

cell lamella and the main cell body [20,21]. What still

remained largely unknown are their structure and their

molecular composition as well as the signals and the

mechanisms leading to ruffle formation. We have applied

high-resolution scanning and transmission electron micro-

scopy (EM) to reveal the structural organization of

membrane ruffles. To gain insight into the process of ruffle

formation and the potential reorganization of the actin

filament system, we have analyzed their protein composi-

tion, particularly in relation to that of lamellipodia. Our

observations show that ruffles contain densely packed arrays

of actin filaments endowed with a specific set of associated

proteins including the cross-linkers filamin and ezrin, and

the assembly regulator ADF/cofilin. To elucidate the func-

tional significance of ruffle formation during cell migration,

we have quantified simultaneously with highest resolution

in space and time cell migration velocity, the frequency of

ruffle formation, the persistence of lamellipodia, and their

dependence on the adhesive properties of the substrate

[20,22]. Our findings show that conditions of suboptimal

cell-substrate adhesion that lead to a significantly reduced

lamellipodia persistence and an inefficient cell migration

result in a dramatic increase in ruffle frequency. Our data

point to integrin-dependent and RhoGTPases-mediated

signaling pathways controlling lamellipodia protrusion and

ruffle dynamics. We conclude that ruffles are formed as a

consequence of inefficient integrin–ligand interaction at the

leading edge of lamellipodia and that they act as compart-

ments of actin reorganization.

Materials and methods

Cells and cell culture

Normal human epidermal keratinocytes from neonatal

foreskin were purchased from Cambrex Bio Science

(Verviers, Belgium). Subcultures were grown at 378C and

5% CO2 in complete keratinocyte basal medium (Cambrex

Bio Science). Cells were harvested for subculturing or for

subsequent experiments using 0.025% Trypsin and 0.01%

EDTA in HBSS (Cambrex Bio Science) only between

passages 2 and 4. For immunofluorescence staining and

scanning EM, cells were subconfluently plated on glass

cover slips or CELLocatek cover slips (Eppendorf,

Hamburg, Germany). For quantification of cell motility,

cells were plated into chambered cover slips (Nalge Nunc

International, Wiesbaden, Germany). For biochemical

measurements of Rac1 and Rho activity, the cells were

plated on 10-cm culture plates. All cells were plated on

uncoated or FN-coated surfaces (see below). One hour

before starting the quantification of cell motility, staining of

integrin h1, or performing Rac1 and RhoA activity assays,

the cells were stimulated with 50 nM human recombinant

EGF (Sigma-Aldrich, Deisenhofen, Germany) to induce cell

migration.

Cell adhesion substrates and anti-integrin antibodies

Coating with human FN (Cell-Systems, St. Katharinen,

Germany) at concentrations of 2.5, 5, 7.5, or 10 Ag/cm2

was done as described previously [23]. To inhibit the

B. Borm et al. / Experimental Cell Research 302 (2005) 83–95 85

interaction between integrins and their ligands, function-

blocking antibodies against integrin a5 (clone P1D6) and

h1 (clone JB1A) (Chemicon, Temecula, USA) were used at

10 and 40 Ag/ml, respectively, after saturation with 1%

BSA for 30 min.

Quantification of cell motility

Phase contrast image series of motile keratinocytes were

obtained using an inverted LSM 510 (Zeiss, Oberkochen,

Germany) equipped with a 63� 1.4 NA Ph3 plan

apochromat objective and an incubation chamber for

constant temperature, and were controlled by the LSM

510 standard software. Lamella dynamics and cell migra-

tion velocity were analyzed by the computer-assisted

stroboscopic analysis of cell dynamics (SACED) (Fig. 1)

that has been described before [20,22] and renamed

recently as kymographic assay [24]. For each substrate, at

least 15 individual cells were monitored over a 10-min

period by capturing digital images every 2 s. Subsequently,

four areas of interest were marked on each image by lines

that crossed the cell lamella (Fig. 1a). The resulting 1-pixel-

wide areas were cut (Fig. 1b) and lined up in time space

plots (Fig. 1c) that allowed the quantification of relevant

motility data. In these stroboscopic time space plots,

Fig. 1. The SACED motility assay. Cells are monitored over periods of up

to 10 min by digitally capturing phase contrast images every 2 s.

Subsequently, an area of interest is marked on each image by lines that

cross the cell lamella and the leading edge (a). The resulting 1-pixel-wide

areas were cut out (b) and lined up in time space plots (c). These plots allow

the simultaneous quantification of different motility parameters. In these

stroboscopic time space plots, membrane ruffles appeared as dark

descending lines (c, black line) whereas lamellipodia protrusions were

represented by linear ascending contours (c, white line). The slope of these

lines corresponds to the velocity of ruffle and lamellipodia movement (v =

dx (Am)/dt (min)). Projections of these lines along the x-axis (time) were

used to calculate the persistence of lamellipodia, that is, the period a

protrusion lasts before its retraction begins (c, dashed white line). The

number of lines per time was counted to determine the frequency of ruffle

formation. Cell migration velocity was determined by exploiting the clear

halo that represents the border between the main cell body and the lamella.

Marking this halo (c, dashed black line) allowed to calculate cell body

displacement in relation to the substrate, which is a common definition of

cell translocation [25].

membrane ruffles appeared as dark descending lines (Fig.

1c, black line), whereas lamellipodia protrusions were

represented by linear ascending contours (Fig. 1c, white

line). The slope of these lines corresponds to the velocity of

ruffle and lamellipodia movement (v = dx (Am)/dt (min)).

Projections of these lines along the x-axis (time) were used

to calculate the persistence of lamellipodia, that is, the

period a protrusion lasts before its retraction begins (Fig.

1c, dashed white line). The number of lines per time was

counted to determine the frequency of ruffle formation. Cell

migration velocity was determined by exploiting the clear

halo that represents the border between the main cell body

and the lamella. Marking this halo (Fig. 1c, dashed black

line) allowed to calculate the cell body displacement in

relation to the substrate, which is a common definition of

cell translocation [25].

Immunofluorescence staining

For immunofluorescence staining of integrin h1 (mouse

monoclonal antihuman integrin h1, clone P5D2, Chemi-

con), filamin (mouse monoclonal anti-filamin, clone PM6/

317, Chemicon), and ezrin (mouse monoclonal anti-ezrin,

clone 3C12, Sigma-Aldrich), the cells were fixed in 4%

freshly made paraformaldehyde (PFA) in cytoskeleton

buffer [26] (150 mM NaCl, 5 mM MgCl2 5 mM EGTA, 5

mM Glucose, 10 mM MES pH 6.1) for 20 min at 378C,washed in PBS containing 30 mM Glycin (G-PBS),

permeabilized in PBS/0.2% Triton X-100 for 10 min, rinsed

in PBS, and blocked in 3% BSA/PBS for 30 min. The

samples were incubated with primary antibodies diluted in

0.3% BSA in PBS for 45 min at 378C, washed three times

with PBS, and incubated with goat anti-mouse Cy3-

conjugated secondary antibodies (Jackson Laboratories,

West Grove, PA, USA) for 45 min at 378C together with

Alexa-488 Phallodin (Molecular Probes, Leiden, Nether-

lands) for correlative actin filament staining. Immunofluore-

scence staining of cofilin and phosphorylated ADF/cofilin

was modified from Ref. [27]. The cells were fixed with 3%

PFA together with 0.1% (vol/vol) glutaraldehyde in

cytoskeleton buffer for 1 h, washed extensively in PBS,

and permeabilized with 0.5% Triton X-100 in PBS for

15 min. After three times washing in PBS, once in TBS

(20 mM Tris, 145 mM NaCl, pH 8) and twice with 2 mg/ml

NaBH4 in TBS for 15 min, the samples were rinsed four

times in PBS and blocked in 3% BSA/PBS for 30 min. Cells

were further incubated for 45 min at 378C with anti-cofilin

antibody (rabbit polyclonal anti-cofilin, Cytoskeleton Inc.,

Denver, USA) and anti-phospho ADF/cofilin antibody

(rabbit antiserum raised against a phoshopeptide epitope

of ADF/cofilin [28], gift from J.R. Bamburg, Colorado State

University, USA), respectively, followed by incubation with

a Cy2-conjugated anti-rabbit antibody and Alexa546 Phal-

loidin for 1 h at 378C. All samples were mounted in anti-

fade reagent (Biomeda Corporation, Foster City, CA, USA)

and analyzed by using a LSM 510 equipped with a 63� NA


1.4 plan apochromat objective. All images were obtained as

confocal z stacks spanning the whole cell in its z range with

subsequent projection of the image series into a single

overlay using the LSM 510 standard software. For

fluorescence quantification using Image-Pro Plus 5 (Media

Cybernetics), staining conditions and microscope settings

were identical for all samples. All measurements were made

on unprocessed images. In brief, the total fluorescence of an

area of interest (sum of all gray levels) marking membrane

ruffles or lamellipodia was measured and normalized to the

size of the respective area. On this basis, the fluorescence

ratios of membrane ruffles versus lamellipodia from five

individual cells were calculated for actin/filamin and for

actin/ezrin.

Correlative light and electron microscopy

Cells grown on CELLocatek cover slips were labeled

against actin filaments using Alexa546-Phalloidin (Mole-

cular Probes) as described above. After light microscopic

analysis, the cells were treated as described below for

scanning EM.

Scanning electron microscopy

For scanning EM, cells were fixed in 2% glutaraldehyde

in 0.1 M sodium cacodylate, pH 7.3, for 20 min, transferred

to 0.1% aqueous tannic acid, and rinsed with distilled water.

For scanning EM of the cytoskeleton, the cell membrane

was removed by detergent lysis. In detail, cells were

incubated in PEG-GTX (10 mM PIPES, 50 mM EDTA,

10 mM KOH, 27 mM KCl, 0.1% (vol/vol) Triton X-100,

4% (vol/vol) polyethylenglycol (MW 6000), 10% (vol/vol)

glycerin, pH 7.2) for 5 min, washed briefly in PBS, and

fixed with 4% (vol/vol) glutaraldehyde in PBS. All speci-

mens were dehydrated through a graded series of ethanol

and critical point dried from CO2 in 10 cycles according

to Ref. [29] using a Balzers CPD 030 (BAL-TEC,

Schalksmqhlen, Germany). Dried specimens were mounted

on aluminum sample holders and coated with a 2-nm layer

of platinum/palladium in a HR 208 sputter coating device

(Cressington, Watford, UK). SEM was performed at an

acceleration voltage of 3 kV using a XL 30 SFEG (Philips,

Eindhoven, Netherlands) equipped with a through lens

secondary electron detector.

Rac1 and rhoA activity assays

Many Rho family effector proteins specifically recog-

nize the activated, GTP-bound form of RhoA and Rac1.

We used PAK-1- and Rhotekin-coupled agarose beads

(Cytoskeleton) to precipitate PAK-1/RacGTP- and Rhote-

kin/RhoGTP complexes that can be identified in Western

blots using anti-Rac and anti-Rho antibodies. In brief,

keratinocytes grown on 10-cm culture plates or FN-coated

10-cm culture plates were stimulated with 50 nM EGF for

1 h and lysed with lysis buffer (50 mM TRIS, 0.5 M NaCl,

1% Triton X-100, 10 mM MgCl2, pH 7.5). After addition

of agarose beads to the cell lysates and incubation at 48Cfor 60 min, the agarose beads were washed and sub-

sequently collected by centrifugation at 5000 g. The beads

were then resuspended in Laemmli-buffer, separated by

SDS-PAGE, and detected by Western blot using rabbit

polyclonal anti-Rac and mouse monoclonal anti-Rho anti-

bodies (Cytoskeleton) and appropriate secondary HRP-

conjugated antibodies. Staining was detected using

enhanced chemiluminescence and densitometry has been

done using Optiquant (Packard Instruments, Meridan,

USA).

Statistical analysis

Statistical evaluation of SACED-derived motility data

was performed with SPSS software (SPSS Inc., Chicago,

USA). Results were expressed as mean values with bars

representing 95% confidence interval of mean values.

Statistical significance between data groups was determined

by Whitney U test and considered to be significantly

different at values P b 0.01. Data were tested for linear

regression and rank correlation was determined according to

Spearman and considered to have statistical significance at

values P b 0.01.

Results

Ruffles are flat membrane folds on the lamella surface

In phase contrast microscopy of motile keratinocytes,

membrane ruffles were clearly visible as waves of dark

contrast that move centripetally from the leading edge of

the organelle free cell lamella toward the organelle

containing main cell body where they finally disappeared.

Cells grown on not precoated glass surfaces migrated

inefficiently (about 0.2 Am/min; see below) and exhibited

numerous ruffles evenly distributed over the cell lamella

(Fig. 2a, arrows), whereas cells on FN-coated surfaces

migrated efficiently (about 0.5 Am/min; see below) and

bore only few if any ruffles (Fig. 2b). Scanning EM

revealed a narrow, fold-bearing lamella for cells grown on

not precoated glass surfaces (Fig. 2c) and an extended,

smooth lamella for cells grown on FN (Fig. 2d). At higher

magnifications, ruffles appeared as flat membrane folds

that protruded from the lamella resembling waves on a

water surface (Fig. 2e, arrows). Underneath these mem-

brane folds newly formed membrane protrusions were

observed (Fig. 2e, arrowheads). Time-lapse studies of the

cell lamella allowed showing that lamellipodia detachment

and their subsequent retraction toward the cell body are

the main steps of ruffle formation (Figs. 3a and b).

Analysis of cells grown on microgrid cover slips revealed

that the dark waves on the cell lamella visualized by phase

Fig. 3. Time lapse sequences (25 s) of the cell lamella provided evidence

that ruffle formation results from lamellipodia detachment and their

subsequent retraction toward the cell body (a and b). Using microgrid

cover slips allowed to show that the wave-like regions of dark contrast that

were visible in phase contrast (c) correspond to the membrane folds

observed by scanning EM (d) and contain high amounts of actin filaments

as shown by phalloidin staining (e). Scanning EM of the cell lamella after

removal of the plasma membrane (f) revealed that ruffles contain densely

packed arrays of actin filaments that lie upon a more loosely arranged actin

filament network (inset in f). Arrowheads in c: the edges of microgrids

became visible due to light refraction. Scale bars: (a, b, and f) 1 Am; (c–e)

10 Am; (inset in f) 200 nm.

Fig. 2. Phase contrast (a and b) and scanning EM (c–e) of motile

keratinocytes. Keratinocytes grown on not precoated glass surfaces

showed numerous ruffles (a, arrows) evenly distributed over the cell

lamella, which appeared narrow and fold bearing (c). On 2.5 Ag/cm2 FN,

only few if any ruffles were observed (b) on a broad and smooth lamella

(d). Higher magnification revealed ruffles as flat membrane folds

protruding from the lamella surface (e, arrows). At the cell edge, newly

formed membrane protrusions were observed (e, arrowheads). Scale bars:

(a–d) 10 Am; (e) 1 Am.


contrast microscopy (Fig. 3c, arrows) indeed corresponded

to the extended membrane folds observed by scanning EM

(Fig. 3d, arrows).

Membrane ruffles contain densely packed actin filaments

with a characteristic set of associated proteins

Scanning EM of the keratinocyte lamella after removal

of the plasma membrane by detergents revealed that ruffles

contained densely packed arrays of thin filaments (Fig. 3f,

arrows), which were identified as actin filaments by

phalloidin staining (Fig. 3e, arrows) or immunogold

labeling (not shown). These actin filaments were located

upon a more loosely arranged network of actin filaments

that were regularly distributed across the cell lamella (Fig.

3f). Immunofluorescence studies on the composition of

membrane ruffles revealed a specific set of actin-associated

proteins that differed from the underlying lamella by a

clear enrichment of the actin cross-linkers filamin (Figs. 4a

and b, circle) and ezrin (Figs. 4c and d, circle), which were

only sparsely distributed in the underlying lamella.

Fluorescence quantification revealed a 50% higher density

of actin filaments within membrane ruffles as compared to

lamellipodia. This is consistent with the dense packaging

of actin filaments observed by SEM (Fig. 3f). An even

higher increase in the concentration of ezrin by 150% and

of filamin by 100% was measured. This corresponds to an

about 2-fold increase in the filamin/actin and an about 3-

fold increase in the ezrin/actin ratio in membrane ruffles as

compared to lamellipodia and points to a specific recruit-

ment of both proteins during membrane ruffle formation

and concomitant actin filament reorganization. Intriguingly,

the actin filament severing protein ADF/cofilin was found

in ruffles exclusively in its phosphorylated, that is,

inactive, form (Figs. 3e and f, circle), whereas the

Fig. 4. Immunocytochemical (a, c, e, and g) and phase contrast studies (b,

d, f, and h) on the composition of membrane ruffles revealed a specific

set of actin-associated proteins that differ from the underlying lamella by

increased filamin/actin (a, circle) and ezrin/actin ratios (c, circle). The

actin filament severing protein ADF/cofilin was found in ruffles

exclusively in its phosphorylated, that is, inactive form (e, circle),

whereas nonphospho cofilin was more evenly distributed over the lamella

(g). Scale bars: 10 Am.


activated form was evenly distributed across the lamella

region (Figs. 3g and h).

The SACED motility assay allows to quantify the dynamics

of ruffles and lamellipodia in relation to migration velocity

The SACED cell motility assay-derived stroboscopic

time space plots allowed to simultaneously visualize the

dynamics of membrane ruffles, which appeared as dark

descending lines (Figs. 5a and e, black lines), and of

lamellipodia, which represented linear ascending contours

(Figs. 5a and e, white lines). The slope of these lines

corresponded to the velocity of ruffle and lamellipodia

movement and projections of these lines along the x-axis

(time) allowed to calculate the persistence of lamellipodia,

that is, the period a protrusion lasts before its retraction

begins (Fig. 5a, dashed white line). The number of lines per

time was counted to determine the frequency of ruffle

formation. In addition, cell migration velocity was deter-

mined by exploiting the clear halo, which represents the

border between the main cell body and the lamella.

Marking this halo (Fig. 5a, dashed black line) allowed to

calculate the cell body displacement in relation to the

substrate that is a common definition of cell translocation

[25].

High ruffle frequency indicates low lamellipodia persistence

and inefficient cell migration

Evaluation of the SACED-derived motility data showed

that migration velocity reached a minimum of about 0.21

(F0.18) Am/min on not precoated glass surfaces (Fig. 5b),

whereas significantly higher velocities occurred on FN-

coated surfaces with a maximum of about 0.53 (F0.3) Am/

min at 2.5 Ag/cm2 FN. With increasing FN concentrations,

the migration velocity dropped in a linear fashion to only

0.24 Am/min at 10 Ag/cm2 FN. Lamellipodia persistence

reached lowest values on not precoated glass surfaces (Fig.

5c) with an average life time of about 0.37 (F0.15) min,

whereas at 2.5 Ag/cm2 FN, a nearly 10-fold higher

persistence of about 3.34 (F1.33) min was determined.

As for migration velocity, the persistence of lamellipodia

decreased in a linear relation with increasing FN concen-

trations dropping to about 0.6 min at 10 Ag/cm2 FN. Ruffle

frequencies reached maximum values on not precoated

glass surfaces with about 0.2 (F0.08) ruffles/min (Fig. 5d),

whereas lowest ruffle frequencies were measured with

about 0.06 (F0.04)/min at 2.5 Ag/cm2 FN. With increasing

FN concentrations, a linear rise in ruffle frequencies

occurred up to 0.14 (F0.06)/min at 10 Ag/cm2. All data

point to a clear correlation between migration velocity, that

is, the efficiency of migration, the persistence of lamelli-

podia, and the frequency of membrane ruffle formation.

Under substrate conditions that supported the formation of

highly persistent lamellipodia, efficient migration occurred

and ruffling became less frequent. Under conditions

favoring short-lived lamellipodia, migration was slow, that

is, inefficient, and ruffle formation was more frequent.

Hence, membrane ruffles might be considered as indicators

of inefficient migration due to reduced lamellipodia

persistence.

Lamellipodia persistence and ruffle formation depend on

the interaction of integrin a5b1 with its ligand FN

Newly formed lamellipodia are known to be stabilized

by the interaction of integrins with their ECM ligands [16].

After addition of antibodies blocking the interaction of the

integrins a5 and h1 with FN, we observed a significantly

reduced migration velocity of keratinocytes dropping by

75% to about 0.08 (F0.04) Am/min for integrin h1 (Fig.

5f). Simultaneously, lamellipodia persistence decreased to

0.037 (F0.15) min thus reaching only 1% of control

values (Fig. 5g), whereas ruffle formation became dramati-

cally more frequent reaching 5-fold values of controls (Fig.

Fig. 5. SACED-derived time space plots of motile keratinocytes visualized the dynamics of membrane ruffles (a, e, black lines) and of lamellipodia (a, e, white

lines). Time space plots from keratinocytes grown on not precoated glass surfaces showed short-lived lamellipodia and high ruffling frequencies (a, top). On 2.5

Ag/cm2 FN, the cells produced more persistent lamellipodia with reduced ruffling frequencies (a, middle). On 10 Ag/cm2 FN, lamellipodia persistence decreased

whereas ruffling frequency increased (a, bottom). Statistical evaluation of migration velocity (b), lamellipodia persistence (c), and ruffle frequency (d) showed a

clear correlation between migration, persistence of lamellipodia, and the frequency of ruffles formation depending on substrate adhesiveness. High lamellipodia

persistence due to appropriate adhesiveness is a prerequisite for efficient cell migration that is indicated by low ruffling frequencies. Keratinocytes grown on 2.5

Ag/cm2 FN after addition of antibodies inhibiting the interaction between FN and integrin a5 (e, top) and h1 (e, bottom) showed significantly reduced

lamellipodia persistence (g) that is correlated to decreased migration efficiency (f) and is marked by clearly increased ruffle frequencies (h). Error bars: 95%

confidence interval of mean values.


5h). Incubation with 1% BSA had no effect on cell

motility (not shown). Using function-blocking antibodies

against various other integrins expressed by migrating

keratinocytes caused no significant effects on the motility

parameters described above (not shown). Our data indicate

that integrin a5h1 interaction with FN is essential for the

attachment of lamellipodia leading to high lamellipodia

persistence and allowing efficient cell migration. Hence,

the formation of ruffles is apparently a consequence of a

suboptimal interaction between integrin a5h1 and its

ligand FN.

Efficient migration depends on low rac and high rho activity

The small GTPases Rac1 and RhoA belonging to the

Rho-family are known to be involved in the downstream


signaling of ligand-bound integrins. We applied Rac1 and

RhoA activity assays to determine the relationship between

GTPase activity and cell migration efficiency. Rac1 activity

was low for efficiently migrating keratinocytes exhibiting

low ruffle frequencies and highly persistent lamellipodia at

2.5 Ag/cm2 FN (Fig. 6a). Rac1 activity was found to be

significantly increased when cells were grown on high FN

concentrations or on not precoated glass surfaces, that is, on

substrates that allowed only reduced migration efficiency

accompanied by diminished lamellipodia persistence and

frequent ruffle formation. Highest Rac1 activity was found

at 10 Ag/cm2 reaching about 3-fold higher levels as

compared to 2.5 Ag/cm2 FN. About 2.5-fold higher levels

of Rac1 activity were also measured for cells grown on not

precoated glass surfaces. Contrary to Rac1, RhoA showed

highest activity levels in cells grown on 2.5 Ag/cm2

dropping significantly on higher FN concentrations to only

40% on 10 Ag/cm2 and to 60% for cells grown on not

precoated glass surfaces (Fig. 6b). Hence, high RhoA and

low Rac1 activities coincided with efficient cell migration,

that is, fast cell movement, high lamellipodia persistence,

and low ruffling frequencies.

Fig. 6. Rac1 (a) and RhoA (b) activities were determined in relation to

substrate adhesiveness by quantitative Western blot analysis of cell lysates

and displayed in relation to the amounts of total Rac1 (d, black columns)

and total RhoA (d, grey columns), respectively. Rac1 activity, stimulating

the formation of lamellipodia and ruffles, was found to be low for

efficiently migrating keratinocytes on 2.5 Ag/cm2, whereas high Rac1

activity was measured for inefficiently migrating keratinocytes on high FN

concentrations or on not precoated glass surfaces (a). RhoA activity

stimulating the formation of contractile acto-myosin complexes and focal

adhesions was high in efficiently migrating cells on 2.5 Ag/cm2 but was

reduced in inefficiently migrating cells on higher FN concentrations or on

not precoated glass surfaces (b). Actin blots of cell lysates are shown to

demonstrate equal loading of all samples (c).

Formation of stress fibers and focal adhesions are

characteristics of inefficiently migrating cells

Downstream effectors of the Rho GTPase are various

actin-associated proteins that modulate the formation of

contractile acto-myosin fibers and the clustering of integrin

h1 [17] in the form of FC at the front of lamellipodia that

are required to convert contraction into traction forces

necessary for cell migration [18]. Cells grown on not

precoated glass surfaces, which have been shown to contain

low levels of activated RhoA and high levels of Rac1, were

enriched with numerous densely packed bundles of actin

filaments (Figs. 7b and c, arrows) and showed a nearly even

distribution of integrin h1 but no obvious clustering at the

lamella region (Fig. 7a). Cells grown on 2.5 Ag/cm2 FN,

which exhibited the highest levels of RhoA activity and low

levels of Rac1 activity, contained only few and significantly

smaller bundles of actin filaments (Figs. 7e and f), and

integrin h1 was clustered at the cell lamella in the form of

distinct patches (Fig. 7d, circles) with a size of about 2 Amthat is typical for FCs [30]. Very fine and short bundles of

actin filaments were observed that were orientated along the

direction of cell migration (Fig. 7e, circles) and were linked

to the putative FC sites (Fig. 7f, circles). With increasing

FN concentrations, we observed significantly more and

thicker bundles of actin filaments representing stress fibers

(Fig. 7h, arrows) and larger but less distinct integrin h1patches (Fig. 7g).

Discussion

The protrusion of lamellipodia at the cell front is

regarded as the initial step of a cyclic process that underlies

cell migration [1,2]. For efficient migration to occur,

lamellipodia have to be stabilized by the formation of cell

substrate adhesions that represent molecular bridges by the

heterodimeric transmembrane receptors of the integrin

family connecting the cytoskeleton and the ECM [16].

Depending on the adhesiveness of the substrate, the

formation of adhesions can fail, and it has been speculated

that lamellipodia without appropriate anchorage become

detached and are retracted from the leading lamella

centripetally toward the main cell body [17], thereby

forming characteristic structures on the cell surface that

are generally described as membrane ruffles [20,21].

To validate the assumed relationship between lamellipo-

dia adhesion and ruffle formation, we applied the high-

resolution SACED cell motility assay [20,22,23] that allows

to acquire and quantify various parameters of cell migration

and motility simultaneously, and has been successfully used

to characterize the motogenic activity of growth factors on

epidermal keratinocytes and melanocytes [22,31]. To learn

more about the signals and mechanisms involved in ruffle

formation, we investigated the structural organization and

the protein composition of membrane ruffles, and analyzed

Fig. 7. Immunofluorescence labeling against Integrinh1 (red) and actin filaments (green) of keratinocytes grown on not precoated glass surfaces (a–c), 2.5 Ag/cm2

FN (d–f), or 10 Ag/cm2 FN (g–i). Cells on not precoated glass surfaces were enriched with massive bundles of actin filaments (b, arrows) showing the

characteristic appearance of stress fibers. Integrin h1 was almost evenly distributed across the cell lamella without any obvious clustering (a). Cells on

2.5 Ag/cm2 FN exhibited only few and significantly smaller bundles of actin filaments (e). Very fine and short bundles of actin filaments were observed (e,

f, circles) that were oriented along the direction of cell migration and inserted in the cell lamella at sites of integrin h1 clusters with a size of about 2.5 Am(d, circles). At 10 Ag/cm2 FN, the number of large actin filament bundles increased (h, arrows), and larger but less distinct integrin h1 patches were

observed (g). Scale bars: 20 Am.


the activities of the small GTPases Rac and Rho that are

known as downstream effectors of integrin signaling and as

mediators of cytoskeletal and adhesive structures formation.

Ruffling is a function of lamellipodia persistence and

directly related to migration efficiency

FN is one of the major constituents of the dermal ECM

and has been shown to represent the preferred migration

substrate for keratinocytes during the process of epidermal

wound healing [32]. Surfaces coated with FN have been

observed to be appropriate for efficient migration of

keratinocytes in vitro. In the absence of motogenic growth

factors, a saturable increase in migration velocity has been

shown in relation to FN concentration for keratinocytes

[23], whereas a biphasic relation between migration velocity

and the adhesiveness to substrates has been described for

CHO [33]. In contrast, this report shows that keratinocytes

exhibit this biphasic relation only after EGF stimulation

with maximal values occurring at intermediate FN levels,

which has also been reported for NR6 fibroblasts [34].

Lower FN concentrations failed to support efficient cell

migration, which might be due to insufficient cell–matrix

adhesion at the cell front that in turn is essential for the

generation of traction forces during cell migration [18]. On

FN concentrations above 5 Ag/cm2, we observed a

significant decrease of migration velocity, which is consi-

dered to be a consequence of insufficient detachment of

adhesion sites at the cell rear [35].

A clear relationship between the application of different

motogenic growth factors, the frequency of ruffle formation,

and cell migration velocity has been reported [20,23].

Accordingly, the formation of membrane ruffles is usually

considered as a sign of increased lamella dynamics and of

elevated migration levels [36]. To address the question

whether membrane ruffles result from lamellipodia that have

failed to form firm attachments, we measured the persis-

tence of lamellipodia, that is, the period a lamellipodium

exists before it is retracted, which we found to be a

sensitive, rate-limiting, and adhesion-dependent motility

parameter. Highly persistent lamellipodia were found to be

characteristic of efficiently, that is, fast moving, cells on

appropriate FN substrates and to be correlated in all cases

with minimal ruffle frequencies. Reciprocal proportions

were observed for inefficiently migrating cells that gener-

ated only short-lived lamellipodia accompanied by high

ruffling frequencies. A correlation between lamellipodia

persistence and cell migration efficiency has also been

reported recently for fibroblasts [24] without addressing

directly the relationship with ruffle formation. Our findings

provide for the first time clear evidence that ruffles originate

in lamellipodia that fail to properly attach to the substrate

and are retracted toward the main cell body. Thus,

membrane ruffle formation can be looked upon as a specific


but indirect criterion for the evaluation of lamellipodia

persistence, which depends on substrate adhesiveness and is

directly related to cell migration efficiency (Fig. 8).

Lamellipodia persistence depends on the specific

interaction between integrin a5b1 and FN

Cell-substrate adhesion is mainly accomplished by the

interaction of ECM receptors from the integrin family and is

rate limiting for efficient cell migration [1], and any

disturbance of the ECM–integrin interaction has inhibitory

effects on cell migration [37]. Keratinocytes migrating on

FN use integrin a5h1 for the interaction with the substrate,

whereas integrin a3h1 is responsible for dynamic inter-

actions with laminin 5 [38]. Antibodies that block the

interaction of the integrins a5 and h1 have been shown

recently to significantly inhibit keratinocyte migration

velocity on FN [37,39], whereas inhibition of integrin a3

had measurable effects on migration of mammary carci-

noma cells on laminin 5 [40]. Our SACED assay-derived

data on keratinocytes show that inhibition of the FN

receptor components integrin a5 and h1 causes a significant

decrease in lamellipodia persistence and migration velocity,

whereas ruffle formation becomes dramatically more

frequent. The FN-binding integrin a5h1 has been shown

previously to be essential for the formation of FC at the

front of newly generated lamellipodia [41], which, besides

anchoring the lamellipodia, are indispensable for the trans-

formation of cellular contraction forces into traction forces,

that is, cell body translocation during migration [18]. Our

findings show that small but very distinct integrin clusters

characteristic of FC are formed at the lamellipodia front of

efficiently migrating cells, but absent from the lamella of

inefficiently migrating cells. The interaction of integrin

a5h1 with FN can be considered the molecular anchor, the

Fig. 8. Schematic illustration summarizing the correlation between

adhesion-dependent lamellipodia persistence, migration efficiency, and

ruffling rates. High ruffling rates indicate high Rac-activity, suboptimal

adhesion, and low persistence of lamellipodia that results in inefficient

migration. Low ruffling rates point to high Rho activity, adequate adhesion,

and highly persistent lamellipodia that result in efficient cell migration.

stability of which decides whether lamellipodia are stabi-

lized and, therefore, contribute to efficient cell migration or

whether they are retracted as ruffles.

Ruffle formation is regulated by the small GTPases

rac and rho

Besides acting as molecular linkers between FN and the

actin filament system, integrin h1 can activate GTPases of

the Rho family following ligand binding [42,43]. Adhesion

of keratinocytes to laminin 5 via the integrins a3h1 and a6h4

leads to increased levels of activated Rho [44], and the

adhesion of untransformed NIH-3T3 cells on FN specifi-

cally enhances RhoA activity [45]. Reexpression of integrin

h1 in h1-deficient epithelial cells triggered the activation of

both RhoA and Rac1, which raises the possibility that

integrin h1 might regulate the motility of epithelial cells by

the use of this small GTPases [46]. In mice, loss of integrin

h1 caused a severe defect in epidermal wound healing and

h1-null keratinocytes showed impaired migration [47]. An

increase in RhoA activity occurred after prolonged contact

to FN, although an initial decrease was observed upon

plating NIH3T3-fibroblasts on FN [48]. RhoA activity,

which has been found to antagonize lamellipodia formation

[49,50] also, stimulates cellular contractility as a prere-

quisite for cell migration [51]. RhoA activity, therefore,

might be a prerequisite for sustained cell migration under

appropriate substrate conditions by regulating cellular

contractility and focal adhesions. Our findings on keratino-

cytes show that high Rho activity coincides with the

formation of highly persistent lamellipodia, efficient cell

migration, and low ruffling rates. Under these conditions,

we observed the formation of very faint bundles of actin

filaments terminating in FC sites. It has been demonstrated

that 3–10 actin filaments are sufficient to support the forces

applied to individual FC [52]. Hence, these bundles of actin

filaments might represent the contractile structures respon-

sible for cell body contraction during keratinocyte migration

(see Figs. 7d–f).

Upon activation by ligand-bound integrins and growth

factor receptors, the GTPase Rac1 triggers the activation of

the Arp2/3 complex, resulting in increased assembly of actin

filaments, thereby promoting lamellipodia protrusion

[10,16,17]. Expression of a constitutively active form of

Rac1 showed that activation of this GTPase induces the

formation of lamellipodia and ruffling but is not sufficient to

propagate cell movement [53,54]. Accordingly, Rac1-

deficient macrophages show normal migration and chemo-

taxis but reduced membrane ruffling in response to colony

stimulating factor-1 [55]. Our findings that highest Rac1

activity is observed under conditions of inefficient migra-

tion, low lamellipodia persistence, and high ruffling rates

support the view that Rac1 triggers the protrusion of

lamellipodia. Without appropriate adhesion these lamellipo-

dia are not able to support forward movement and are

quickly retracted as membrane ruffles. Membrane ruffles are


apparently an appropriate structural indicator for Rac1

activity pointing to suboptimal lamellipodia adhesion and

inefficient cell migration (Fig. 8).

Filamin and ezrin might act as organizers and stabilizers of

membrane ruffles

The actin-associated protein filamin is known as an actin

filament cross-linker that is able to form either three-

dimensional orthogonal networks when the filamin/actin

ratio is about 1:150 or parallel bundles when the ratio is

about 1:10 [56]. Our observations showing that the filamin/

actin ratio in ruffles is about 2-fold higher than in

lamellipodia point to filamin as the bundling protein

responsible for the dense packaging of actin filaments

observed in membrane ruffles. These results imply that

membrane ruffle formation is accompanied by compaction

of actin filaments. Besides its cross-linking function, filamin

has been shown to be mechanically linked to integrins [57],

thereby providing a link between the plasma membrane and

the actin filament system that might be necessary for the

stability of membrane ruffles during their retraction.

Accordingly, filamin-deficient human melanoma cells are

still capable to form protrusions upon serum stimulation, but

the protrusions form blebs and ruffling is not observed

[58,59]. Finally, filamin has been shown to be integrated in

cellular signaling pathways that use Rho GTPases as

mediators and lead to a phosphorylation-dependent modu-

lation of the binding capacity of filamin for actin [60]. The

activity levels of Rac1 and RhoA observed in migrating

keratinocytes might represent a direct switch for filamin to

preferentially stabilize actin filament bundles as found in

membrane ruffles. Ezrin is another prominent actin binding

protein known to tether actin filaments to the plasma

membrane by its interaction with the hyaluron receptor

CD 44 [61]. Functional ablation of ezrin has been shown to

block ruffling and migration of fibroblasts [62]. Our finding

that the ezrin/actin ratio in ruffles is 3-fold higher than in

lamellipodia provides further support for its suggested role

during the formation and stabilization of membrane ruffles.

The actin filament severing protein ADF/cofilin has been

shown to produce free actin monomers that are necessary

for the Arp2/3-dependent assembly of actin filaments,

which power lamellipodia protrusion [3]. Recent findings

additionally point to a positive engagement of cofilin in

actin assembly by generating free barbed ends that are a

prerequisite for filament elongation by the addition of actin

monomers [15].

Upon phosphorylation by the Rho effector kinase LIM,

ADF/cofilin becomes inactivated [63]. Our observation that

ruffles contain high amounts of phosphorylated cofilin

points to ruffles as a compartment where the Arp2/3- and

Cofilin-dependent actin filament turnover that powers

lamellipodia protrusion becomes inactivated during mem-

brane ruffle formation. The specific protein composition of

ruffles and their characteristic arrays of actin filaments

suggest that membrane ruffles might act as a compartment

involved in the highly regulated reorganization of actin

filaments that become necessary when lamellipodia are

retracted due to impaired adhesion.

Acknowledgments

The authors wish to thank Prof. J. R. Bamburg, Colorado

State University, USA, for providing the anti-phospho ADF/

cofilin antibody, Prof. W. Alt, University of Bonn,

Germany, for excellent discussions, and Dr. W. Neumqllerfor critical reading of the manuscript. We greatly appreciate

the technical assistance of C. Hennes and N. Florin.

References

[1] D.A. Lauffenburger, A.F. Horwitz, Cell migration: a physically

integrated molecular process, Cell 84 (1996) 359–369.

[2] A.J. Ridley, M.A. Schwartz, K. Burridge, R.A. Firtel, M.H. Ginsberg,

G. Borisy, J.T. Parsons, A.R. Horwitz, Cell migration: integrating

signals from front to back, Science 302 (2003) 1704–1709.

[3] T.D. Pollard, G.G. Borisy, Cellular motility driven by assembly and

disassembly of actin filaments, Cell 112 (2003) 453–465.

[4] A.Y. Chan, S. Raft, M. Bailly, J.B. Wyckoff, J.E. Segall, J.S.

Condeelis, EGF stimulates an increase in actin nucleation and filament

number at the leading edge of the lamellipod in mammary

adenocarcinoma cells, J. Cell Sci. 111 (1998) 199–211.

[5] C.H. Damsky, D. Ilic, Integrin signaling: it’s where the action is, Curr.

Opin. Cell Biol. 14 (2002) 594–602.

[6] R.O. Hynes, Integrins: bidirectional, allosteric signaling machines,

Cell 110 (2002) 673–687.

[7] A. Wells, EGF receptor, Int. J. Biochem. Cell Biol. 31 (1999)

637–643.

[8] K.A. DeMali, K. Wennerberg, K. Burridge, Integrin signaling to the

actin cytoskeleton, Curr. Opin. Cell Biol. 15 (2003) 572–582.

[9] P. Aspenstrom, Effectors for the Rho GTPases, Curr. Opin. Cell Biol.

11 (1999) 95–102.

[10] T. Takenawa, H. Miki, WASP and WAVE family proteins: key

molecules for rapid rearrangement of cortical actin filaments and cell

movement, J. Cell Sci. 114 (2001) 1801–1809.

[11] M.D. Welch, A.H. DePace, S. Verma, A. Iwamatsu, T.J. Mitchison,

The human Arp2/3 complex is composed of evolutionarily conserved

subunits and is localized to cellular regions of dynamic actin filament

assembly, J. Cell Biol. 138 (1997) 375–384.

[12] A.M. Weaver, M.E. Young, W.L. Lee, J.A. Cooper, Integration of

signals to the Arp2/3 complex, Curr. Opin. Cell Biol. 15 (2003)

23–30.

[13] J.R. Bamburg, A. McGough, S. Ono, Putting a new twist on actin:

ADF/cofilins modulate actin dynamics, Trends Cell Biol. 9 (1999)

364–370.

[14] N. Yang, O. Higuchi, K. Ohashi, K. Nagata, A. Wada, K. Kangawa,

E. Nishida, K. Mizuno, Cofilin phosphorylation by LIM-kinase 1 and

its role in Rac-mediated actin reorganization, Nature 393 (1998)

809–812.

[15] M. Ghosh, X. Song, G. Mouneimne, M. Sidani, D.S. Lawrence, J.S.

Condeelis, Cofilin promotes actin polymerization and defines the

direction of cell motility, Science 304 (2004) 743–746.

[16] K.A. DeMali, K. Burridge, Coupling membrane protrusion and cell

adhesion, J. Cell Sci. 116 (2003) 2389–2397.

[17] K. Burridge, K. Wennerberg, Rho and Rac take center stage, Cell 116

(2004) 167–179.


[18] K.A. Beningo, M. Dembo, I. Kaverina, J.V. Small, Y.L. Wang,

Nascent focal adhesions are responsible for the generation of strong

propulsive forces in migrating fibroblasts, J. Cell Biol. 153 (2001)

881–888.

[19] N.Q. Balaban, U.S. Schwarz, D. Riveline, P. Goichberg, G. Tzur, I.

Sabanay, D. Mahalu, S. Safran, A. Bershadsky, L. Addadi, B.

Geiger, Force and focal adhesion assembly: a close relationship

studied using elastic micropatterned substrates, Nat. Cell Biol. 3

(2001) 466–472.

[20] B. Hinz, W. Alt, C. Johnen, V. Herzog, H.W. Kaiser, Quantifying

lamella dynamics of cultured cells by SACED, a new computer-

assisted motion analysis, Exp. Cell Res. 251 (1999) 234–243.

[21] M. Abercrombie, The crawling movement of metazoan cells, Proc. R.

Soc. Lond., Ser. B 207 (1980) 129–147.

[22] G. Kirfel, B. Borm, A. Rigort, V. Herzog, The secretory beta-amyloid

precursor protein is a motogen for human epidermal keratinocytes,

Eur. J. Cell Biol. 81 (2002) 664–676.

[23] T. Libotte, H.W. Kaiser, W. Alt, T. Bretschneider, Polarity, protrusion-

retraction dynamics and their interplay during keratinocyte cell

migration, Exp. Cell Res. 270 (2001) 129–137.

[24] J.E. Bear, T.M. Svitkina, M. Krause, D.A. Schafer, J.J. Loureiro, G.A.

Strasser, I.V. Maly, O.Y. Chaga, J.A. Cooper, G.G. Borisy, F.B.

Gertler, Antagonism between Ena/VASP proteins and actin filament

capping regulates fibroblast motility, Cell 109 (2002) 509–521.

[25] D.R. Soll, The use of computers in understanding how animal cells

crawl, Int. Rev. Cytol. 163 (1995) 43–104.

[26] J.V. Small, Organization of actin in the leading edge of cultured cells:

influence of osmium tetroxide and dehydration on the ultrastructure of

actin meshworks, J. Cell Biol. 91 (1981) 695–705.

[27] A.Y. Chan, M. Bailly, N. Zebda, J.E. Segall, J.S. Condeelis, Role of

cofilin in epidermal growth factor-stimulated actin polymerization and

lamellipod protrusion, J. Cell Biol. 148 (2000) 531–542.

[28] P.J. Meberg, S. Ono, L.S. Minamide, M. Takahashi, J.R. Bamburg,

Actin depolymerizing factor and cofilin phosphorylation dynamics:

response to signals that regulate neurite extension, Cell Motil.

Cytoskeleton 39 (1998) 172–190.

[29] T.M. Svitkina, A.A. Shevelev, A.D. Bershadsky, V.I. Gelfand,

Cytoskeleton of mouse embryo fibroblasts. Electron microscopy of

platinum replicas, Eur. J. Cell Biol. 34 (1984) 64–74.

[30] B. Geiger, A. Bershadsky, R. Pankov, K.M. Yamada, Transmembrane

crosstalk between the extracellular matrix-cytoskeleton crosstalk, Nat.

Rev., Mol. Cell Biol. 2 (2001) 793–805.

[31] T. Quast, S. Wehner, G. Kirfel, K. Jaeger, M. De Luca, V. Herzog,

sAPP as a regulator of dendrite motility and melanin release in

epidermal melanocytes and melanoma cells, FASEB J. 17 (2003)

1739–1741.

[32] R.O. Hynes, Fibronectins, Springer-Verlag, New York, 1990.

[33] S.P. Palecek, J.C. Loftus, M.H. Ginsberg, D.A. Lauffenburger, A.F.

Horwitz, Integrin-ligand binding properties govern cell migration

speed through cell-substratum adhesiveness, Nature 385 (1997)

537–540.

[34] G. Maheshwari, A. Wells, L.G. Griffith, D.A. Lauffenburger,

Biophysical integration of effects of epidermal growth factor and

fibronectin on fibroblast migration, Biophys. J. 76 (1999) 2814–2823.

[35] S.P. Palecek, A. Huttenlocher, A.F. Horwitz, D.A. Lauffenburger,

Physical and biochemical regulation of integrin release during rear

detachment of migrating cells, J. Cell Sci. 111 (1998) 929–940.

[36] J.V. Small, T. Stradal, E. Vignal, K. Rottner, The lamellipodium:

where motility begins, Trends Cell Biol. 12 (2002) 112–120.

[37] F. Decline, P. Rousselle, Keratinocyte migration requires alpha2beta1

integrin-mediated interaction with the laminin 5 gamma2 chain, J. Cell

Sci. 114 (2001) 811–823.

[38] F.M. Watt, M.D. Hertle, Keratinocyte integrins, in: I. Leigh, E.B.

Lane, M. Watt (Eds.), The Keratinocyte Handbook, Cambridge Univ.

Press, Cambridge, England, 1994.

[39] J.P. Kim, K. Zhang, J.D. Chen, K.C. Wynn, R.H. Kramer, D.T.

Woodley, Mechanism of human keratinocyte migration on fibronectin:

unique roles of RGD site and integrins, J. Cell. Physiol. 151 (1992)

443–450.

[40] R.B. Lichtner, A.R. Howlett, M. Lerch, J.A. Xuan, J. Brink, B.

Langton-Webster, M.R. Schneider, Negative cooperativity between

alpha 3 beta 1 and alpha 2 beta 1 integrins in human mammary

carcinoma MDA MB 231 cells, Exp. Cell Res. 240 (1998) 368–376.

[41] B.J. Dzamba, H. Bultmann, S.K. Akiyama, D.M. Peters, Substrate-

specific binding of the amino terminus of fibronectin to an integrin

complex in focal adhesions, J. Biol. Chem. 269 (1994) 19646–19652.

[42] H. Miao, S. Li, Y.L. Hu, S. Yuan, Y. Zhao, B.P. Chen, W. Puzon-

McLaughlin, T. Tarui, J.Y. Shyy, Y. Takada, S. Usami, S. Chien,

Differential regulation of Rho GTPases by beta1 and beta3 integrins:

the role of an extracellular domain of integrin in intracellular

signaling, J. Cell Sci. 115 (2002) 2199–2206.

[43] A.L. Berrier, R. Martinez, G.M. Bokoch, S.E. LaFlamme, The integrin

beta tail is required and sufficient to regulate adhesion signaling to

Rac1, J. Cell Sci. 115 (2002) 4285–4291.

[44] B.P. Nguyen, X.D. Ren, M.A. Schwartz, W.G. Carter, Ligation of

integrin alpha 3beta 1 by laminin 5 at the wound edge activates Rho-

dependent adhesion of leading keratinocytes on collagen, J. Biol.

Chem. 276 (2001) 43860–43870.

[45] E.H. Danen, P. Sonneveld, A. Sonnenberg, K.M. Yamada, Dual

stimulation of Ras/mitogen-activated protein kinase and RhoA by cell

adhesion to fibronectin supports growth factor-stimulated cell cycle

progression, J. Cell. Biol. 151 (2000) 1413–1422.

[46] C. Gimond, A. van Der Flier, S. van Delft, C. Brakebusch, I.

Kuikman, J.G. Collard, R. Fassler, A. Sonnenberg, Induction of cell

scattering by expression of beta1 integrins in beta1-deficient epithelial

cells requires activation of members of the rho family of GTPases and

downregulation of cadherin and catenin function, J. Cell Biol. 147

(1999) 1325–1340.

[47] R. Grose, C. Hutter, W. Bloch, I. Thorey, F.M. Watt, R. Fassler, C.

Brakebusch, S. Werner, A crucial role of beta 1 integrins for

keratinocyte migration in vitro and during cutaneous wound repair,

Development 129 (2002) 2303–2315.

[48] X.D. Ren, W.B. Kiosses, M.A. Schwartz, Regulation of the small

GTP-binding protein Rho by cell adhesion and the cytoskeleton,

EMBO J. 18 (1999) 578–585.

[49] W.T. Arthur, K. Burridge, RhoA inactivation by p190RhoGAP

regulates cell spreading and migration by promoting membrane

protrusion and polarity, Mol. Biol. Cell 12 (2001) 2711–2720.

[50] R.A. Worthylake, K. Burridge, RhoA and ROCK promote migration

by limiting membrane protrusions, J. Biol. Chem. 278 (2003)

13578–13584.

[51] M. Chrzanowska-Wodnicka, K. Burridge, Rho-stimulated contractility

drives the formation of stress fibers and focal adhesions, J. Cell Biol.

133 (1996) 1403–1415.

[52] A. Kishino, T. Yanagida, Force measurements by micromanipulation

of a single actin filament by glass needles, Nature 334 (1988) 74–76.

[53] A.J. Ridley, H.F. Paterson, C.L. Johnston, D. Diekmann, A. Hall, The

small GTP-binding protein rac regulates growth factor-induced

membrane ruffling, Cell 70 (1992) 401–410.

[54] D.G. Stupack, S.Y. Cho, R.L. Klemke, Molecular signaling

mechanisms of cell migration and invasion, Immunol. Res. 21

(2000) 83–88.

[55] C.M. Wells, M. Walmsley, S. Ooi, V. Tybulewicz, A.J. Ridley, Rac1-

deficient macrophages exhibit defects in cell spreading and membrane

ruffling but not migration, J. Cell Sci. 117 (2004) 1259–1268.

[56] B.M. Jockusch, P. Bubeck, K. Giehl, M. Kroemker, J. Moschner, M.

Rothkegel, M. Rudiger, K. Schluter, G. Stanke, J. Winkler, The

molecular architecture of focal adhesions, Annu. Rev. Cell Dev. Biol.

11 (1995) 379–416.

[57] M. Pfaff, S. Liu, D.J. Erle, M.H. Ginsberg, Integrin beta cytoplasmic

domains differentially bind to cytoskeletal proteins, J. Biol. Chem.

273 (1998) 6104–6109.

[58] L.A. Flanagan, J. Chou, H. Falet, R. Neujahr, J.H. Hartwig, T.P.

Stossel, Filamin A, the Arp2/3 complex, and the morphology and


function of cortical actin filaments in human melanoma cells, J. Cell

Biol. 155 (2001) 511–517.

[59] R.K. Vadlamudi, F. Li, L. Adam, D. Nguyen, Y. Ohta, T.P. Stossel, R.

Kumar, Filamin is essential in actin cytoskeletal assembly mediated by

p21-activated kinase 1, Nat. Cell Biol. 4 (2002) 681–690.

[60] J.M. Bellanger, C. Astier, C. Sardet, Y. Ohta, T.P. Stossel, A. Debant,

The Rac1- and RhoG-specific GEF domain of Trio targets filamin to

remodel cytoskeletal actin, Nat. Cell Biol. 2 (2000) 888–892.

[61] S. Tsukita, K. Oishi, N. Sato, J. Sagara, A. Kawai, ERM family

members as molecular linkers between the cell surface glycopro-

tein CD44 and actin-based cytoskeletons, J. Cell Biol. 126 (1994)

391–401.

[62] R.F. Lamb, B.W. Ozanne, C. Roy, L. McGarry, C. Stipp, P. Mangeat,

D.G. Jay, Essential functions of ezrin in maintenance of cell shape and

lamellipodial extension in normal and transformed fibroblasts, Curr.

Biol. 7 (1997) 682–688.

[63] K. Ohashi, K. Nagata, M. Maekawa, T. Ishizaki, S. Narumiya, K.

Mizuno, Rho-associated kinase ROCK activates LIM-kinase 1 by

phosphorylation at threonine 508 within the activation loop, J. Biol.

Chem. 275 (2000) 3577–3582.

membrane ruffles in cell migration: indicators of

Documents