ARF1 regulates adhesion of MDA-MB-231 invasive breast cancer cells through formation of focal adhesions

CLS-08343; No of Pages 13 Cellular Signalling xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Cellular Signalling journal homepage: www.elsevier.com/locate/cellsig

3Q4

Sabrina Schlienger, Rodrigo Alain Migueles Ramirez, Audrey Claing ⁎

4

Department of Pharmacology, Faculty of Medicine, Université de Montréal, Montreal, QC H3C 3J7, Canada

5

a r t i c l e

6 7 8 9 10

Article history: Received 28 October 2014 Received in revised form 25 November 2014 Accepted 27 November 2014 Available online xxxx

11 12 13 14 15 16

Keywords: Epidermal growth factor receptor ADP-ribosylation factor-1 Focal adhesion kinase Breast cancer Focal adhesion

a b s t r a c t

T

E

D

P

Adhesion complex formation and disassembly is crucial for maintaining efficient cell movement. During migration, several proteins act in concert to promote remodeling of the actin cytoskeleton and we have previously shown that in highly invasive breast cancer cells, this process is highly regulated by small GTP-binding proteins of the ADPribosylation factor (ARF) family. These are overexpressed and highly activated in these cells. Here, we report that one mechanism by which ARF1 regulates migration is by controlling assembly of focal adhesions. In cells depleted of ARF1, paxillin is no longer colocalized with actin at focal adhesion sites. In addition, we demonstrate that this occurs through the ability of ARF1 to regulate the recruitment of key proteins such as paxillin, talin and FAK to ß1-integrin. Furthermore, we show that the interactions between paxillin and talin together and with FAK are significantly impaired in ARF1 knocked down cells. Our findings also indicate that ARF1 is essential for EGF-mediated phosphorylation of FAK and Src. Finally, we report that ARF1 can be found in complex with key focal adhesion proteins such as ß1-integrin, paxillin, talin and FAK. Together our findings uncover a new mechanism by which ARF1 regulates cell migration and provide this GTPase as a target for the development of new therapeutics in triple negative breast cancer. © 2014 Elsevier Inc. All rights reserved.

34 32 31

E

33

1. Introduction

36 37

Progression of cancer from primary tumors to metastatic carcinomas requires that cancer cells detach from the initial tumor to invade their surrounding. Through the lymphatic or circulatory systems, cells can attain distant organs and tissues to initiate the formation of metastasis. The ability of a cell to migrate is therefore essential for that process. Cell migration requires the dynamic interaction between the cell and the extracellular matrix onto which it is attached and migrates. Adhesion is mediated by membrane-proximal protein complexes associated with integrins. These are transmembrane proteins that function to link the intracellular cytoskeletal components of the cells to the substratum on which cells are anchored. The intracellular domain of integrins acts as a scaffold for focal adhesion proteins, which formation

46 47

R

N C O

44 45

U

42 43

R

35

40 41

17 18 19 20 21 22 23 24 25 26 27 28 29

C

30

38 39

O

R O

i n f o

F

2

ARF1 regulates adhesion of MDA-MB-231 invasive breast cancer cells through formation of focal adhesions

1Q3

Abbreviations: ARF, ADP-ribosylation factor; DMEM, Dulbecco's Modified Eagle Medium; DMSO, dimethyl sulfoxide; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; FAK, focal adhesion kinase; GAP, GTP activating protein; HCC70, Human Breast Cancer Cell Line 70; MDA-MB-231, MD Anderson-Mammary Breast cancer cell line-231; MDCK, Madin-Darby Canine Kidney; PBS, phosphate buffer saline; PI3K, phosphatidylinositol 3 kinase; PTK, protein-tyrosine kinase; RTK, receptor tyrosine kinase; SKBR3, Sloan Kettering human breast cancer cell line 3; siRNA, small interfering ribonucleic acids ⁎ Corresponding author at: P.O. Box 6128, Downtown station, Montreal, QC H3C 3J7, Canada. Tel.: +1 514 343 6352; fax: +1 514 343 2291. E-mail address: [email protected] (A. Claing).

and disassembly is controlled by more than 180 protein–protein interactions. These include cytoskeletal proteins such as paxillin and talin as well as enzymes such as FAK (Focal adhesion kinase) and small GTP-binding proteins [1,2]. Cell migration and invasiveness can be enhanced by numerous growth factors. For example, high expression of the epidermal growth factor receptor (EGFR) is a major cause of cancer development, progression, and is associated with a poor prognosis [3]. Basal or triple negative breast cancer cells, like MD AndersonMammary Breast cancer cell line-231 (MDA-MB-231), often express high EGFR levels, which is associated with a loss of sensitivity to hormonal therapies and/or a high probability of metastasis [4]. Increased activation of this receptor tyrosine kinase (RTK) has been correlated with, namely, enhanced survival, migration and invasiveness [5]. These three processes are mediated by intracellular signaling proteins of the Ras GTPase family [6]. Others and we have demonstrated that ADP-ribosylation factors (ARFs) are involved in the regulation of the biological effects induced by epidermal growth factor (EGF) [7–9]. Like all GTPases, ARFs are considered active when bound to GTP and inactive when bound to GDP. ARF1 and ARF6 are the best characterized from the six identified isoforms. ARF6 has been shown to localize to the plasma membrane, while ARF1 is classically associated with the Golgi and in some cell types can also be found associated with the plasma membrane [7,10,11]. These small GTP-binding proteins act as molecular switches to turn on signaling cascades associated with remodeling of the actin cytoskeleton, transformation of membrane lipids and vesicle formation [7,10,12,13].

http://dx.doi.org/10.1016/j.cellsig.2014.11.032 0898-6568/© 2014 Elsevier Inc. All rights reserved.

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73

114

126

Anti-paxillin, anti-FAK pY397 and anti-CD29 were purchased from BD Bioscience (Bedford, MA, USA). Anti-FAK pY861 polyclonal antibody, Lipofectamine 2000, Alexa Fluor 488-phalloidin and Rhodaminephalloidin were purchased from Invitrogen (Burlington, ON, Canada). Anti-ARF1 was obtained from ProteinTech (Chicago, IL, USA). AntiPan-Actin, anti-EGFR, anti-Src pY416 and anti-Giantin were from Cell Signaling Technology (Danvers, MA, USA). Anti-FAK (A-17), anti-ARF6 (3A1) and anti-talin were purchased from Santa Cruz Biotech Inc. (Santa Cruz, CA, USA). All others products were from Sigma Aldrich Company (Oakville, ON, Canada). LM11 compound was synthesized by Dr Pierre Lavallée (Combinatorial Chemistry Laboratory, Université de Montréal, QC, Canada). GST-GGA3 was a gift from Dr. Jean-Luc Parent (Université de Sherbrooke, Sherbrooke, QC, Canada).

127

2.2. Cell culture and transfection

128 129

MDA-MB-231, Human Breast Cancer Cell Line 70 (HCC70) and Sloan Kettering human breast cancer cell line 3 (SKBR3) cells were obtained from Dr. Sylvie Mader (Université de Montréal, Montreal, QC, Canada). Cells were maintained at 37 °C, 5% CO2, in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum and 10% penicillin/streptomycin. All cell culture reagents were purchased from Wisent Bioproducts (St-Bruno, QC, Canada). Transfection of small

99 100 101 102 103 104 105 106 107 108 109 110

115 116 117 118 119 120 121 122 123 124 Q5 125

130 131 132 133 134

C

97 98

E

95 96

R

93 94

R

91 92

O

89 90

C

87 88

N

85 86

U

83 84

2.3. Immunofluorescence

146

F

2.1. Reagents and antibodies

81 82

137 138 139 140 141 142 143 144 145

The cells were fixed with a phosphate buffer saline (PBS) solution containing 4% paraformaldehyde for 15 min at room temperature and then permeabilized with a solution of DMEM containing 0.05% saponin. The slides were incubated for 1 h with a primary antibody. After several washes, the plates were incubated for 1 h in the dark in the presence of an anti-mouse or anti-rabbit Alexa Fluor 488-antibody or Alexa-Fluor 568 (Invitrogen, Burlington, ON, Canada). They were then mounted onto slides using a solution of Aqua-mount (Fisher Scientific, Ottawa, ON, Canada) and observed using a confocal microscope (LSM510META, Carl Zeiss).

147

2.4. Migration assay

157

O

113

80

R O

2. Material and methods

78 79

135 136

P

112

76 77

interfering ribonucleic acids (siRNAs), (72 h) was conducted using Lipofectamine 2000 from Invitrogen (Burlington, ON, Canada) according to the manufacturer's instructions. siRNA corresponding to human ARF1 (-GAACCAGAAGUGAACGCGA- was used in all experiments and -CGGCCGAGATCACAGACAAG- a second sequence, was used to confirm specificity of the effects of the siRNA), to a scrambled ARF1 siRNA sequence and human FAK (-GGACAUUAUUGGCCACUGU-) were synthesized by Thermo Science Dharmacon (Lafayette, CO, USA). ARF1-HA was a generous gift from Nicolas Vitale (Université de Strasbourg, Strasbourg, France). For rescue experiments, ARF1-HA was transfected 24 h after introduction of the ARF1 siRNA sequence #1.

148 149 150 151 152 153 154 155 156

MDA-MB-231 cells were seeded onto collagen-coated (SigmaAldrich, Oakville, ON, Canada) Boyden Chambers (8 μm) (Corning, NY, USA) and treated with different doses of LM11 or vehicle. One hour later, cells were stimulated with 10 ng/ml EGF and incubated for 6 h. Cells were fixed using paraformaldehyde (4%) and stained with crystal violet (0.1% in 20% methanol, 16 h). Cells present in the upper chamber were removed with a cotton swab and cells that migrated were quantified in the lower chamber by counting.

158

2.5. Co-immunoprecipitation experiments

166

MDA-MB-231 cells were serum starved overnight and stimulated with EGF (100 ng/ml, 37 °C) for the indicated times. Coimmunoprecipitation experiments were described previously [24]. Briefly, cells were lysed into ice-cold TGH buffer (pH 7.3, 50 mM HEPES, 50 mM NaCl, 5 mM EDTA, 10% glycerol, 1% triton and protease inhibitors) and β1-integrin, talin, paxillin, ARF1 or EGFR were immunoprecipitated using the anti-β1-integrin, anti-talin, anti-paxillin, anti-ARF1 or antiEGFR antibodies. Interacting proteins were assessed by Western blot analysis.

167 168

2.6. GTPase activation assay

176

MDA-MB-231 cells transfected with scrambled siRNA or an siRNA targeting FAK (50 nM) were stimulated with EGF (10 ng/ml) for the indicated times and then washed with TBS and lysed in 200 μl of ice-cold MLB lysis buffer (pH 7.5, 25 mM HEPES, 150 mM NaCl, 10 mM MgCl2, 1 mM EDTA, 10% glycerol, 1% nonidet P-40, 0.3 mg/ml PMSF, 1.0 mM Na3VO4 and protease inhibitors). Samples were vortexed for 10 s and spun for 10 min at 10,000 g. GST-GGA3 coupled to glutathione sepharose 4B were added to each tube and samples were rotated at 4 °C for 45 min. Beads were washed and proteins were eluted in 25 μl SDS sample buffer by heating at 65 °C for 15 min. Detection of ARF1-GTP was performed by immunoblot analysis using a specific anti-ARF1 antibody. A similar methodology was used for the detection of activated ARF1 in conditions where cells were treated with the ARF1 biochemical inhibitor LM11. For those experiments, MDA-MB-231 cells were simply first treated with vehicle (dimethyl sulfoxide (DMSO) 1%) or LM11 (indicated concentration) before being stimulated.

177

T

111

During migration, dynamic adhesion involves the coordinated action of various classes of molecules. The protein tyrosine kinase, focal adhesion kinase (FAK) is a non-receptor protein–tyrosine kinase (PTK), ubiquitously expressed in cytoplasm and involved in the formation of the adhesions complex [14]. FAK plays a role in the assembly and disassembly of the focal adhesion point [15]. It is involved in signal transduction and in the regulation of cell survival and motility triggered by the binding of integrins to the extracellular matrix [16,17]. Following the activation of EGFR and/or integrins, FAK is recruited to these receptors indirectly through C-terminal domain-mediated interactions with integrin-associated proteins such as paxillin and talin [18]. FAK activation leads to the autophosphorylation of Tyr397, which binds to the Src family kinases. In turn, Src phosphorylates FAK on Tyr 861 [19]. FAK then becomes the main anchoring point from which focal adhesions complex are formed [18]. Certain ARF GAPs, such as GIT and ASAP1 have been shown to interact with FAK and paxillin [20,21]. Futhermore, in Swiss 3T3 Fibroblasts, ARF1 has been reported to mediate the recruitment of paxillin to focal adhesion [22]. In addition, we showed that another ARF isoform, ARF6, is required for remodeling of the focal adhesion complex in endothelial cells [23]. Despite the studies conducted, the detailed mechanism through which ARF1 regulates adhesion in highly invasive breast cancer cells remains unknown. In this study, we report that ARF1 controls formation of focal adhesions complex. Using RNAi and a chemical inhibitor targeted to inhibit the expression or activation of this ARF isoform, we provide evidence that ARF1 acts to regulate key protein–protein interactions between ß1-integrin, paxillin, talin, and FAK. Furthermore, ARF1 expression and activation impacts the ability of FAK, as well as Src, to become phosphorylated. Finally, we show that ARF1 is an integral part of focal adhesion complex by demonstrating that this GTPase can also be found interacting with these key molecules. By defining the key signaling proteins regulating adhesion, we will gain further insights into the mechanism by which tumor cells become motile and contribute to the progression and aggressiveness of breast cancer. The identification of proteins that orchestrate fundamental cellular responses such as migration will help us identify new molecular therapeutic targets for the treatment of highly invasive breast cancers for which therapeutical options remain limited.

D

74 75

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

E

2

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

159 160 161 162 163 164 165

169 170 171 172 173 174 175

178 179 180 181 182 183 184 185 186 187 188 189 190 191 192

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

2.9. Statistical analysis

207

212

Quantification of the digital images obtained by Western blot analysis was performed using ImageJ 1.46o software (National Institutes of Health, USA). Statistical analyses were calculated using a one-way or two-way analysis of variance followed by a Tukey's or Bonferroni's multiple comparison test, using GraphPad Prism Software (ver. 5.02; San Diego, CA, USA).

213

3. Results

214 215

3.1. ARF1 expression and activation controls formation of focal adhesion complex in different invasive breast cancer cell lines

216

To further define how ARF1 regulates cell migration, we examined the ability of invasive breast cancer cells to dynamically interact with substratum. We first examined focal adhesions by staining for paxillin and F-actin. Paxillin is a key focal adhesion-associated phosphotyrosine-containing protein and is required for regulating cell spreading and motility [25]. Because of its numerous protein–protein binding domains, it acts as a scaffold for cytoskeletal proteins, tyrosine and serine/threonine kinases, GTPases regulatory factors as well as other adaptors [25]. In this first series of experiments, we examined

C

E

R

224

merge

R

222 223

scrambled siRNA

221

N C O

219 220

ARF1 siRNA 1

217 218

paxillin

actin

IB: ARF1 IB: actin

ARF1 siRNA 2

210 211

U

208 209

F

206

O

205

MDA-MB-231 cells were treated with various concentrations of LM11 or DMSO and after 16 h cell viability was determined by Trypan blue assay.

R O

203 204

199 200

ra AR mbl F1 ed AR si siR F1 RN NA si A 1 R N A 2

2.8. Survival assay

198

sc

202

196 197

P

201

Cells were harvested and total soluble proteins were run on polyacrylamide gels and transferred onto nitrocellulose membranes. The membranes were blotted for relevant proteins using specific primary antibodies (as described for each experiment). Secondary antibodies were HRP-conjugated (R&D Systems, Minneapolis, MN, USA) and fluorescence was detected using ECL detection reagent. Quantification of the digital images obtained was performed using ImageQuant 5.2 software.

D

194 195

the role of ARF1 in 3 types of invasive breast cancer cell lines: MDA- 225 MB-231, HCC 70 and SKBR3. Here, control and ARF1 knocked down 226 cells were plated onto collagen for 15 min to first assess kinetics. As 227 illustrated in Fig. 1, paxillin was displayed as a punctate peripheral 228 staining in MDA-MB-231 cells transfected with a scrambled siRNA, 229 colocalizing with F-actin. In conditions where cells were depleted of 230 ARF1 (with two different siRNA), paxillin remained in the cytosol, 231 while actin was mainly organized in ruffles at the cell periphery (Fig. 1). 232 In order to further define the role played by the activation of ARF1 in 233 focal complex, we used a biochemical inhibitor, LM11 [26]. This com- 234 pound has been shown to block migration of Madin-Darby Canine 235 Kidney (MDCK) cells [26]. First, we examined the integrity of the Golgi 236 following LM11 treatment of our cells by assessing localization of 237 Giantin, a conserved membrane protein of the Golgi [27]. As illustrated 238 in Fig. 2A and B, Golgi organization remained unaffected when 239 cells were treated with 10 μM of LM11 for 8 h or 100 μM for 30 min. 240 However, defaults in Golgi were observed when cells were treated for 241 longer periods of time with a low concentration (18 h; 10 μM) or with 242 a high concentration (1 h; 100 μM) (Fig. 2A, B). LM11 treatment was ef- 243 fective in inhibiting migration of MDA-MB-231 cells in a dose- 244 dependent fashion (Fig. 2C). However, when used for 16 h, cell viability 245 was compromised (Fig. 2D). To avoid any effect on the integrity of the 246 Golgi, we therefore treated the cell with LM11 for 30 min and examined 247 the effect of a low and a high dose (10 μM and 100 μM,respectively). In 248 Q6 control conditions and when cells were allowed to attach to the matrix 249 for 15 min, paxillin was observed in punctate, at the cell periphery 250 (Fig. 3). However, when cells were pretreated with LM11 (10 μM), 251 those focal points of adhesion assessed by paxillin distribution, were 252 partially inhibited. Interestingly, using higher concentrations of this bio- 253 chemical agent (100 μM), paxillin remained cytosolic (Fig. 3). We next 254 investigated the effect of LM11 on cell migration. Upon treatment of 255 the MDA-MB-231 cells with this biochemical inhibitor (10 μM; 256 30 min), EGF-dependent migration was completely inhibited (Fig. 4A). 257 Furthermore, in those conditions, activation of ARF1 was blocked by 258 LM11 (Fig. 4B). To confirm these data, we examined the importance of 259 ARF1 in two other invasive breast cancer cell lines. In HCC70 and 260 SKBR3 cells, knock down of this ARF isoform also prevented localization 261 of paxillin to focal adhesions (Fig. 5). 262 Together, these results demonstrate that ARF1 is a key regulator of 263 focal adhesion assembly and that it is the ability of this GTPase to 264

E

2.7. Western blotting

T

193

3

Fig. 1. Depletion of ARF1 blocks the recruitment of paxillin to focal adhesions. MDA-MB-231 cells were transfected with a scrambled, or two different sequences of ARF1 siRNA. Cells were seeded onto collagen coverslips. After 15 min, cells were fixed, permeabilized and incubated with specific anti-paxillin antibody. Labeling of paxillin was performed using a secondary antibody coupled to Alexa-Fluor 488 and actin stained using Alexa Fluor 568-phalloidin. Images are representative of at least 20 cells observed in three independent experiments. Scale bar, 10 μm.

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

4

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

A

B giantin

actin

merge

giantin

actin

merge

giantin

actin

merge

giantin

actin

merge

No giantin

actin

giantin

actin

merge

giantin

actin

merge

giantin

R O

O

actin

giantin

actin

No giantin

actin

C

D

C R

0.6 *** 0.4

R

0.2

***

60

40

20

***

10 µM 10 0µ M

1µ M

M

LM11

C

LM11

0. 1µ

ic le

O M

M

10 µM

1µ

e cl

0. 1µ

hi ve

0

ve h

***

0.0

10 0µ M

cell migration (fold/basal)

E

**

% cell death in 16h

1.0 0.8

80

T

E

D

P

60 min vehicle

merge

vehicle

F

60 min 30 min LM11 100 μM

8 hours

merge

18 hours

LM11 10 μM

merge 30 min vehicule

16 hours

vehicule

8 hours

merge

U

N

Fig. 2. Concentration and time-dependent effects of the ARF1 inhibitor LM11 in MDA-MB-231. (A, B) MDA-MB-231 cells were treated with vehicle or LM11 (10 or 100 μM) for the indicated times. Cells were fixed, permeabilized, and labeled using an anti-Giantin antibody, a secondary antibody coupled to Alexa-Fluor 633, and Alexa Fluor 488-phalloidin. Golgi distribution was assessed by confocal microscopy. These images are representative of at least 100 cells observed in three independent experiments. Scale bar, 10 μm. (C) MDA-MB-231 cells were reseeded into Boyden chambers, treated with different concentrations of LM11 or DMSO. Cell migration was assessed after 6 h of stimulation with EGF. Results are the mean ± S.E.M. of three independent experiments. **P b 0.01 E and *** P b 0.001 are values compared to the vehicle condition. Statistical analyses were calculated using a one-way analysis of variance followed by a Bonferroni's multiple comparison test. (D) MDA-MB-231 cells were treated with various concentrations of LM11 or DMSO and after 16 h cell viability was assessed by Trypan blue assay. Results are the mean ± S.E.M. of three independent experiments. ***P b 0.001 are values compared to the vehicle condition.

265

268 269

become activated that regulates, namely, localization of paxillin to the plasma membrane. Because the localization of proteins to focal adhesion sites is consistent with the ability of key proteins to physically interact, we next investigated whether ARF could act to modulate interactions of proteins present in this complex.

270

3.2. ARF1 controls the association of ß1-integrin with its binding partners

271 272

Integrins constitute the major transmembrane receptors that mediate dynamic interactions between the extracellular matrix and the cell cytoskeleton during motility [2]. Studies in human breast cancer cells

266 267

273

have revealed key roles for ß1-integrin in the regulation of cancer cell functions that promote malignant progression and metastasis [28]. Interestingly, recycling of ß1-integrin, needed to form new adhesion sites, has been shown to be regulated by two GTPases, ARF6 and Rab 11 [29]. We therefore focused on this specific integrin. We first examined whether EGF stimulation of MDA-MB-231 cells, our prototypical cellular model, led to the association of endogenously expressed β1integrin and paxillin. As illustrated in Fig. 6A, EGF stimulation promoted a rapid association of ß1-integrin and paxillin that was maximal at 1 min and remained sustained after 15 min of stimulation. Knock down of ARF1 expression totally inhibited the ability of ß1-integrin to

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

274 275 276 277 278 279 280 281 282 283 284

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

paxillin

A

actin

***

2

Migration (Fold/basal)

vehicle

merge

5

1 ***

***

EGF

LM11 100 μm

LM11 10 μM

vehicle ST

0

G

F

B

1

0

1

30

LM11 100 μM 0

1

30

min

R O

IB: ARF1-GTP

30

O

LM11 10 μm

ve ve hi cl hi e cl e +E G F LM LM 11 11 +E G F

0

IB: ARF6-GTP

R

R

N C O

300 301

3.3. Depletion of ARF1 controls the association of proteins present in focal adhesion complex

302

We next investigated the role of ARF1 in regulating some of the interactions between other components of the adhesion complex.

287 288 289 290 291 292 293 294 295 296 297

303

U

298 299

interact with its partner. ß1-integrins have also been shown to form a complex with talin, a cytoskeletal protein highly localized to cellsubstratum contact sites [30]. In MDA-MB-231 cells, the interaction between the two proteins was rapid and transient where maximal effects were observed after 1 min of EGF treatment (Fig. 6B). Similar to what we previously observed, depletion of ARF1 markedly reduced the ability of ß1-integrin to interact with talin, this other key partner. We next examined the ability of ß1-integrin to interact with a regulator of focal adhesions, the Focal Adhesion Kinase (FAK). FAK is a cytoplasmic tyrosine kinase that has long been recognized as a regulator of cell migration [31]. EGF stimulation rapidly and transiently promoted the recruitment of FAK to ß1-integrin. Again, this interaction was completely abolished in cells that expressed very low levels of ARF1 (Fig. 6C). Altogether, these data suggest that ARF1 is required for EGFdependent β1-integrin interaction with focal adhesion proteins.

IB: ARF1

IB: ARF6

IB: actin vehicle LM11(10 μM) LM11 (100 μM)

2.5 ARF1 activation (folb/basal)

E

C

T

No secondary

Fig. 3. Inhibition of ARF1 activation blocks the recruitment of paxillin to focal adhesions. MDA-MB-231 cells were treated with vehicle or LM11 (10 or 100 μM) for 30 min. Cells were seeded onto collagen coverslips. After 15 min, they were fixed and labeled as in Fig. 1. Top right, magnifications showing the presence or absence of focal adhesions. Images are representative of at least 20 cells observed in three independent experiments. Scale bar, 10 μm.

285 286

P

E

vehicle

D

total cell lysate

No primary

2.0 1.5 1.0

** ****

****

0.5 0.0 0

10 20 30 EGF stimulation (min)

Fig. 4. Effect of ARF1 inhibitor LM11 in MDA-MB-231 cells. (A) Cells were treated with DMSO or LM11 (10 μM) and stimulated or not with EGF. Migration was assessed after 6 h using the Boyden chamber assay. Results are the mean ± S.E.M. of two other independent experiments. ***P b 0.001 are values compared to the vehicle condition. Statistical analyses were calculated using a one-way analysis of variance followed by a Bonferroni's multiple comparison test. (B) MDA-MB-231 cells treated with DMSO or LM11 (10 or 100 μM) for 30 min, were stimulated with EGF (100 ng/ml) for the indicated time. The active forms of ARF1 and ARF6 as well as endogenous level of ARF1, ARF6 and actin were analyzed by Western blot. Data are representative of four independent experiments. Statistical analyses were calculated using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. ** P b 0.01 and **** P b 0.0001 are values compared to each vehicle condition.

More than 150 proteins have been shown to associate with adhesion sites [2]. We first examined the ability of paxillin to form a complex with talin [32]. As illustrated in Fig. 7A, EGF stimulation promoted the association of paxillin with talin where maximal interaction was observed after 5 min. In cells depleted of ARF1, the ability of these two proteins to form a complex was abrogated. Similar results were observed when we examined the ability of paxillin and talin to interact with FAK. This tyrosine kinase is a central component of focal adhesion sites and acts as a structural and signaling protein. Through its Cterminal domain, FAK can directly interact with paxillin and talin [33].

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

304 305 306 307 308 309 310 311 312 313

6

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

paxillin

actin

ra AR mbl F1 ed si siR R N N A A 1

merge

sc

scrambled siRNA

A HCC 70

ARF1 siRNA

IB: ARF1

F

IB: actin

O R O

scrambled siRNA

No primary

actin

E R

sc

ra

T

paxillin

C

merge

IB: ARF1 IB: actin

No primary

scrambled siRNA

U

N

C

O

R

ARF1 siRNA

scrambled siRNA

B SKBR3

AR mbl F1 ed si siR R N N A A 1

E

D

P

No secondary

No secondary

Fig. 5. Depletion of ARF1 blocks the recruitment of paxillin to focal adhesions in two other invasive breast cancer cell lines. HCC 70 (A) or SKBR3 (B) cells were transfected with a scrambled, or ARF1 siRNA. Cells were seeded onto collagen coverslips. After 20 and 30 min respectively, cells were fixed, permeabilized and incubated with specific anti-paxillin antibody. Labeling of paxillin was performed using a secondary antibody coupled to Alexa-Fluor 488 and actin stained using Alexa Fluor 568-phalloidin. Top right, magnifications showing the presence or absence of focal adhesions. Images are representative of at least 20 cells observed in three independent experiments. Scale bar, 10 μm.

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

A EGF IP: β1-integrin

ctl

scrambled siRNA 0 1 2 5

15

0

ARF1 siRNA 1 2 5 15 min scrambled siRNA ARF1 siRNA

IB: paxillin

2

β1/paxillin

IB: β1-integrin

whole cell lysate

7

IB: paxillin IB: β1-integrin

*

1

0 0

5 10 EGF stimulation (min)

scrambled siRNA 0 1 2 5

15

0

ARF1 siRNA 1 2 5 15 min

R O

ctl

O

IB: actin

EGF IP: β1-integrin

scrambled siRNA ARF1 siRNA

IB: talin

5 4

P β1/talin

D

IB: talin IB: β1-integrin

E

scrambled siRNA 0 1 2 5

15

0

ARF1 siRNA 1 2 5 15

N C O

R

IB: β1-integrin

whole cell lysate

0 0

5 10 EGF stimulation (min)

15

min scrambled siRNA ARF1 siRNA

IB: FAK IB: β1-integrin

2

β1/FAK

IB: FAK

1

R

ctl

C

IB: actin

C

2

T

IB: ARF1

**** **

3

E

whole cell lysate

IB: β1-integrin

EGF IP: β1-integrin

15

F

IB: ARF1

B

*

**

1

0 0

5 10 EGF stimulation (min)

15

IB: ARF1

Q1

314 315 316 317 318 319 320

U

IB: actin

Fig. 6. ARF1 controls β1-integrin interaction with focal adhesion partners. MDA-MB-231 cells transfected with a scrambled or ARF1 siRNA were stimulated with EGF (100 ng/ml) for the indicated time. β1-integrin was immunoprecipitated and associated with paxillin (A), talin (B) or FAK (C) were detected by Western blotting. Data are the mean ± SEM of three to four experiments. Statistical analyses were calculated using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. *P b 0.05, **P b 0.01, and ****P b 0.0001 are values compared with the scrambled siRNA condition at each time point.

In MDA-MB-231 cells, increased association of paxillin and FAK, as well as talin and FAK, was observed in EGF stimulated control conditions. These interactions were rapid and transient and maximal association was observed after 5 min (Fig. 7B and C). Knock down of ARF1 blocked the ability of FAK to interact with both paxillin and talin. These evidences therefore provide a molecular basis for the role of ARF1 in the formation of focal complex.

3.4. Phosphorylation of FAK is regulated by the expression and activation of 321 ARF1 322 FAK was reported to have over 30 phosphorylation sites [34]. Phosphorylation of specific tyrosine residues has been demonstrated to be a key step regulating cell motility, proliferation and apoptosis [34]. The inability of FAK to interact with one of its focal adhesion

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

323 324 325 326

8

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

A ctl

EGF

scrambled siRNA 0 1 2 5

15

0

ARF1 siRNA 1 2 5 15

min

IP: talin IB: paxillin

scrambled siRNA ARF1 siRNA

*

IB: talin talin/paxillin

2

IB: talin

*

1

0

IB: ARF1

0

5 10 EGF stimulation (min)

scrambled siRNA 0 1 2 5

15

0

ARF1 siRNA 1 2 5 15

min

IB: FAK

scrambled siRNA ARF1 siRNA

4

D E

IB: paxillin

IB: ARF1

scrambled siRNA 0 1 2 5

15

E

IB: FAK

ARF1 siRNA 1 2 5 15

R

IB: talin

O IB: talin

C

whole cell lysate

IB: FAK

1

5 10 EGF stimulation (min)

IB: ARF1

15

min

scrambled siRNA ARF1 siRNA

4

****

3

*

2 1 0 0

5 10 EGF stimulation (min)

15

IB: actin

U

N

2

0

R

IP: talin

0

**

0

talin/FAK

ctl

C

IB: actin

EGF

*

3

T

whole cell lysate

IB: FAK

paxillin/FAK

P

IB: paxillin

C

R O

ctl

EGF IP: Paxillin

O

IB: actin

B

15

F

whole cell lysate

IB: paxillin

*

Q2

327 328 329 330 331 332 333

Fig. 7. ARF1 controls the EGF-mediated interaction of proteins present in focal adhesion complex. MDA-MB-231 cells transfected with a scrambled or ARF1 siRNA were stimulated with EGF (100 ng/ml) for the indicated time. Talin was immunoprecipitated and associated with paxillin (A) or FAK (C) was detected by Western blotting. Data are the mean ± SEM of four experiments. Statistical analyses were calculated using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. *P b 0.05 and ****P b 0.0001 are values compared with the scrambled siRNA condition at each time point. (B) Cells were used as in (A), paxillin was immunoprecipitated and associated with FAK was detected by Western blotting. Data are the mean ± SEM of three experiments. Statistical analyses were calculated using a two way analysis of variance followed by a Bonferroni's multiple comparison test. *P b 0.05 and **P b 0.01 are values compared with the scrambled siRNA condition at each time point.

binding partner could result from a change in FAK phosphorylation at the autophosphorylation site Y397, or at the major Src phosphorylation site Y861. Phosphorylation of both tyrosines have been reported to be required for cell motility induced by integrin [18]. Here, we therefore next investigated whether ARF1 could control FAK phosphorylation. As illustrated in Fig. 8, EGF stimulation promoted an increase in FAK phosphorylation of Y397 and Y861. For both tyrosine residues,

phosphorylation was rapid and transient. In ARF1 depleted cells, EGFdependent phosphorylation at both residues was abolished (Fig. 8A). This effect was reversed by re-expressing ARF1-HA (Fig. 8B). Next, to further examine whether the effects observed in ARF1 depleted cells depended on the lack of ARF1 activity, we used the ARF1 inhibitor LM11. MDA-MB-231 cells were pretreated for 30 min with the vehicle or the inhibitor (10 μM) before being stimulated with

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

334 335 336 337 338 339 340

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

A EGF

scrambled siRNA 0 1 2 5 15

0

ARF1 siRNA 1 2 5 15 min

scrambled siRNA

9

ARF1 siRNA

2

4

IB: P-FAK Y861

IB: FAK

IB: ARF1

* **

2 1 0

0 5 10 EGF stimulation (min)

0

15

R O

IB: P-FAK Y861

P

IB: FAK

IB: ARF1

vehicle 1 2 5 15

0

1

LM11 2 5

15

min

E

0

D

IB: actin

EGF

vehicle

T

E

IB: FAK

N C O

0

0

2 1

5 10 EGF stimulation (min)

15

0

5 10 EGF stimulation (min)

15

ARF1 siRNA 1 2 5 15 min

6

*

P-Src

IB: P-Src

IB: Src

*

0

0

R

IB: actin

scrambled siRNA 0 1 2 5 15

1

R

IB: ARF1

*

3

C

IB: P-FAK Y861

4

P-FAK Y861

P-FAK Y397

2

EGF

LM11

**

IB: P-FAK Y397

D

15

O

ARF1 siRNA + ARF1-HA 0 1 2 5 15 min

IB: P-FAK Y397

C

5 10 EGF stimulation (min)

F

IB: actin scrambled siRNA + empty vector EGF 0 1 2 5 15

*

1

0

B

*

3 P-FAK Y861

P-FAK Y397

IB: P-FAK Y397

*

scrambled siRNA ARF1 siRNA

4 2

IB: ARF1

U

IB: actin

0 0

5 10 EGF stimulation (min)

15

Fig. 8. ARF1 expression and activation control FAK phosphorylation. (A) MDA-MB-231 cells transfected with a scrambled or ARF1 siRNA were stimulated with EGF (100 ng/ml) for the indicated time. Endogenously expressed FAK (phosphorylated and total), actin and ARF1 were detected by Western immunoblotting using specific antibodies. Quantifications are the mean ± SEM of five independent experiments. Statistical analyses were calculated using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. *P b 0.05 and **P b 0.01 are values compared with the scrambled siRNA condition at each time point. (B) ARF1-HA was transfected 24 h after introduction of the ARF1 siRNA 1. MDA-MB-231 cells were stimulated with EGF (100 ng/ml) for the indicated time. Endogenously expressed FAK (phosphorylated and total), actin and ARF1 were detected by Western immunoblotting using specific antibodies. This result is representative of four independent experiments. (C) MDA-MB-231 cells treated vehicle or LM11 (10 μM) for 30 min, and were then stimulated with EGF (100 ng/ml) for the indicated time. Endogenously expressed FAK (phosphorylated and total), actin and ARF1 were detected by Western immunoblotting using specific antibodies. Quantifications are the mean ± SEM of four independent experiments. Statistical analyses were calculated using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. *P b 0.05, are values compared with the control condition at each time points. (D) Cells were used in the same conditions as in (A). Endogenously expressed Src (phosphorylated and total), actin and ARF1 were detected by Western immunoblotting using specific antibodies. Quantifications are the mean ± SEM of six independent experiments. Statistical analyses were calculated using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. *P b 0.05 are values compared with the scrambled siRNA condition at each time points.

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

Because FAK regulates the activation of so many key proteins, in focal adhesions but also in other parts of the cells, we next assessed whether it could modulate ARF1 activity. As reported previously, EGF stimulation of MDA-MB-231 cells led to the activation of this ARF isoform (Fig. 9) [7]. In conditions where FAK expression was knocked down, activation of ARF1 was markedly impaired. These results demonstrate that FAK has the potential to regulate activation of this small GTPase. To further investigate how FAK acts to control ARF1 activation,

3.6. ARF1 is a component of the focal adhesion complex

373

Because ARF1 appears to be such an important regulator of focal adhesion protein interactions and activity, we finally examined the possibility that ARF1 could be an integral component of the focal adhesion complex by interacting with key proteins. We therefore assessed whether ARF1 could be found in complex with β1-integrin, paxillin, talin, and FAK in MDA-MB-231 cells. As shown in Fig. 10, EGF stimulation of the cells promoted the association of ARF1 with all the above mentioned proteins. The most rapid association we found was with talin and FAK, which peaked at 1–2 min. Although the association of ARF1 with ß1-integrin and FAK was also significantly enhanced after such short times of EGF stimulation, it continued to increase and maximal interactions were observed at 5 min.

374 375

scrambled siRNA T

GS

0

1

30

FAK siRNA 0

C E

IB: ARF1

R

scrambled siRNA

EGF

T GS

scrambled siRNA FAK siRNA

2

0

1

2

0 1 30 EGF stimulation (min)

FAK siRNA 0

1

2

min

scrambled siRNA FAK siRNA

**

8

IB: phospho-Y

IB: FAK

****

0

IP: EGFR

IB: EGFR

***

1

R

IB: FAK

B

min

30

T

IB: ARF1-GTP

IB: actin

1

E

EGF

D

A

O

363

total cell lysate

361 362

C

359 360

total cell lysate

357 358

N

350 351

U

348 349 Q7

F

356

347

O

3.5. EGF mediated activation of ARF1 is dependent upon FAK and the ability of this enzyme to control EGF receptor phosphorylation

345 346

364 365

R O

354 355

343 344

we examined whether FAK could modulate EGFR phosphorylation. As previously reported and illustrated in Fig. 9B, exposure of the cells to EGF rapidly promoted phosphorylation of its receptor. Interestingly, knock down of FAK expression had a major impact on this event. Because it is generally recognized that ligand binding promotes direct autophosphorylation of the intracellular tyrosine kinase domains, we propose that FAK may be part of a feedback mechanism controlling levels of phosphorylated receptors. This would results in the subsequent impairment of ARF1 activation.

P

352 353

EGF. As presented in Fig. 8C, LM11 was highly effective in preventing auto-phosphorylation of FAK at Y397 and Src-mediated phosphorylation of this key enzyme on residue Y861. To further investigate the role of ARF1 in regulating phosphorylation of Y861, we next examined the effect of ARF1 depletion on Src activation by assessing the ability of this enzyme to become phosphorylated. We focused on assessing phosphorylation of Y416 present in the kinase domain, recognized as the major autophosphorylation site, and associated with Src activation. In MDA-MB-231 cells, EGF induced a rapid phosphorylation of Src, peaking at 2 min (Fig. 8D). Depletion of ARF1 drastically impaired Src phosphorylation upon EGF stimulation. Altogether, these findings suggest that ARF1 controls EGF-dependent Src activation, which in turn, regulates FAK phosphorylation.

EGFR activation (folb/basal)

341 342

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

ARF1 activation (folb/basal)

10

6 4

*

2 0

0 1 2 EGF stimulation (min)

IB: EGFR

IB: actin Fig. 9. FAK controls ARF1 activation through EGFR phosphorylation. (A) MDA-MB-231 cells were transfected with a scrambled, or FAK siRNA and were stimulated with EGF (100 ng/ml) for the indicated time. The active forms of ARF1 as well as endogenous level of ARF1, FAK and actin were analyzed by Western blot. Data are the mean ± SEM of six experiments. Statistical analyses were calculated using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. ***P b 0.001 and ****P b 0.0001 are values compared with the scrambled siRNA condition at each time point. (B) MDA-MB-231 cells were transfected and stimulated as in (A). EGFR was immunoprecipitated and associated phospho-tyrosines were detected by Western blotting. Data are the mean ± SEM of four experiments respectively. Statistical analyses were performed using a two-way analysis of variance followed by a Bonferroni's multiple comparison test. *P b 0.05 and **P b 0.01 are values compared with the scrambled siRNA condition for each time point.

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

366 367 368 369 370 371 372

376 377 378 379 380 381 382 383 384 385

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

B 0

1

2

5

15

min

5 ARF1-β1 integrin

IB: paxillin

IB: talin

3 2 1 0 0

IB: FAK

C ARF1-paxillin

IB: ARF1 IB: β1-integrin

5 10 EGF stimulation (min)

4 3 2 1 0

D

2

15

15

*

T

E

IB: ARF1

ARF1-talin

D

IB: FAK

5 10 EGF stimulation (min)

P

0

IB: talin

*

**

IB: paxillin

N C O

R

R

E

C

IB: actin

1

0 0

E

5 10 EGF stimulation (min)

15

**

4 3

ARF1-FAK

whole cell lysate

**

*

4

IB: β1-integrin

F

ctl

O

EGF IP: ARF1

R O

A

11

2 1 0 0

5 10 EGF stimulation (min)

15

U

Fig. 10. EGF stimulation promotes the association of ARF1 with focal adhesion proteins. MDA-MB-231 cells were stimulated with EGF (100 ng/ml) for the indicated time. ARF1 was immunoprecipitated and associated with talin, FAK, paxillin, and β1-integrin were detected by Western blotting. Data are the mean ± SEM of four to seven experiments. Statistical analyses were performed using a one-way Anova analysis of variance followed by a Tukey's multiple comparison test. *P b 0.05 and **P b 0.01 are values compared with the basal level of interaction.

386 387

Altogether, these data show that ARF1 is a regulator of focal adhesion signaling and is an integral part of these structures.

388

4. Discussion

389 390

We have previously shown that ARF1 plays a major role in breast cancer tumorigenesis by controlling proliferation, migration and invasion of MDA-MB-231 cells [7,35,36]. We have partially delineated the mechanism by which ARF1 regulates migration by demonstrating that this ARF isoform overexpressed in highly invasive cells control the ability of the EGFR to signal through the phosphatidylinositol 3 kinase (PI3K) pathway, but also that this GTPase regulates Rac1, a member of

391 392 393 394 395

the Rho family of small GTP-binding proteins, that actively regulate lamellipodia formation [7,37]. In the present study, we demonstrate a role for ARF1 in the regulation of focal adhesion. Using the siRNA strategy, we show that the presence of ARF1 is required for EGF-dependent focal adhesion formation. When ARF1 is depleted, no signal integration from growth factor receptors is made. FAK cannot bind the cytoskeletal proteins paxillin and talin, which normally mediate FAK localization to integrin-enriched focal adhesions. These observations may explain why in ARF1-depleted conditions, FAK no longer interacts with β1integrin. Furthermore, we report that ARF1 plays a key role in regulating the processes mediating FAK phosphorylation. Phosphorylation of Tyr397 serves to create a high-affinity binding site for Src-family PTKs

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

396 397 398 399 400 401 402 403 404 405 406 407

431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 Q8 462 463 464 465 466 467 468 469 470 471 472 473

Acknowledgments

481

This work was supported by the Canadian Institutes of Health Research (grant number: MOP-106596 to AC). AC is the recipient of a Senior Scientist Award from the Fonds de Recherche du Québec-Santé.

482

O

F

475 476

References

R O

429 430

In conclusion, we have shown that ARF1 regulates cellular migration of highly invasive triple-negative breast cancer cells, through the regulation of the focal adhesions complex. Gaining an understanding of the molecular mechanisms by which ARF1 controls key features of cancer cells will provide news insights into the development of new therapeutic tools.

[1] R. Zaidel-Bar, S. Itzkovitz, A. Ma'ayan, R. Iyengar, B. Geiger, Nat. Cell Biol. 9 (2007) 858–867. [2] A. Huttenlocher, A.R. Horwitz, Cold Spring Harb. Perspect. Biol. 3 (2011) a005074. [3] J.G. Klijn, P.M. Berns, P.I. Schmitz, J.A. Foekens, Endocr. Rev. 13 (1992) 3–17. [4] R.I. Nicholson, R.A. McClelland, J.M. Gee, D.L. Manning, P. Cannon, J.F. Robertson, I.O. Ellis, R.W. Blamey, Breast Cancer Res. Treat. 29 (1994) 117–125. [5] J.T. Price, T. Tiganis, A. Agarwal, D. Djakiew, E.W. Thompson, Cancer Res. 59 (1999) 5475–5478. [6] S. Malaney, R.J. Daly, J. Mammary Gland Biol. Neoplasia 6 (2001) 101–113. [7] P.L. Boulay, M. Cotton, P. Melancon, A. Claing, J. Biol. Chem. 283 (2008) 36425–36434. [8] M. Morishige, S. Hashimoto, E. Ogawa, Y. Toda, H. Kotani, M. Hirose, S. Wei, A. Hashimoto, A. Yamada, H. Yano, Y. Mazaki, H. Kodama, Y. Nio, T. Manabe, H. Wada, H. Kobayashi, H. Sabe, Nat. Cell Biol. 10 (2008) 85–92. [9] S.W. Kim, M. Hayashi, J.F. Lo, Y. Yang, J.S. Yoo, J.D. Lee, J. Biol. Chem. 278 (2003) 2661–2668. [10] L.A. Cohen, A. Honda, P. Varnai, F.D. Brown, T. Balla, J.G. Donaldson, Mol. Biol. Cell 18 (2007) 2244–2253. [11] J.G. Donaldson, A. Honda, Biochem. Soc. Trans. 33 (2005) 639–642. [12] M.C. Galas, J.B. Helms, N. Vitale, D. Thierse, D. Aunis, M.F. Bader, J. Biol. Chem. 272 (1997) 2788–2793. [13] J.E. Casanova, Traffic 8 (2007) 1476–1485. [14] J. Cornillon, L. Campos, D. Guyotat, Med. Sci.: M/S 19 (2003) 743–752. [15] D. Ilic, Y. Furuta, T. Suda, T. Atsumi, J. Fujimoto, Y. Ikawa, T. Yamamoto, S. Aizawa, Biochem. Biophys. Res. Commun. 209 (1995) 300–309. [16] M.B. Calalb, T.R. Polte, S.K. Hanks, Mol. Cell. Biol. 15 (1995) 954–963. [17] M.D. Schaller, Biochim. Biophys. Acta 1540 (2001) 1–21. [18] D.J. Sieg, C.R. Hauck, D. Ilic, C.K. Klingbeil, E. Schaefer, C.H. Damsky, D.D. Schlaepfer, Nat. Cell Biol. 2 (2000) 249–256. [19] D.J. Sieg, C.R. Hauck, D.D. Schlaepfer, J. Cell Sci. 112 (Pt 16) (1999) 2677–2691. [20] Z.S. Zhao, E. Manser, T.H. Loo, L. Lim, Mol. Cell. Biol. 20 (2000) 6354–6363. [21] Y. Liu, J.C. Loijens, K.H. Martin, A.V. Karginov, J.T. Parsons, Mol. Biol. Cell 13 (2002) 2147–2156. [22] J.C. Norman, D. Jones, S.T. Barry, M.R. Holt, S. Cockcroft, D.R. Critchley, J. Cell Biol. 143 (1998) 1981–1995. [23] Z. Daher, J. Noel, A. Claing, Cell. Signal. 20 (2008) 2256–2265. [24] M. Cotton, P.L. Boulay, T. Houndolo, N. Vitale, J.A. Pitcher, A. Claing, Mol. Biol. Cell 18 (2007) 501–511. [25] M.D. Schaller, Oncogene 20 (2001) 6459–6472. [26] J. Viaud, M. Zeghouf, H. Barelli, J.C. Zeeh, A. Padilla, B. Guibert, P. Chardin, C.A. Royer, J. Cherfils, A. Chavanieu, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 10370–10375. [27] A.D. Linstedt, H.P. Hauri, Mol. Biol. Cell 4 (1993) 679–693. [28] J.S. Desgrosellier, D.A. Cheresh, Nat. Rev. Cancer 10 (2010) 9–22. [29] A.M. Powelka, J. Sun, J. Li, M. Gao, L.M. Shaw, A. Sonnenberg, V.W. Hsu, Traffic 5 (2004) 20–36. [30] K. Burridge, L. Connell, J. Cell Biol. 97 (1983) 359–367. [31] S.K. Mitra, D.A. Hanson, D.D. Schlaepfer, Nat. Rev. Mol. Cell Biol. 6 (2005) 56–68. [32] C.E. Turner, Nat. Cell Biol. 2 (2000) E231–E236. [33] C.T. Mierke, Phys. Biol. 10 (2013) 065005. [34] P.R. Grigera, E.D. Jeffery, K.H. Martin, J. Shabanowitz, D.F. Hunt, J.T. Parsons, J. Cell Sci. 118 (2005) 4931–4935. [35] P.L. Boulay, S. Schlienger, S. Lewis-Saravalli, N. Vitale, G. Ferbeyre, A. Claing, Oncogene 30 (2011) 3846–3861. [36] S. Schlienger, S. Campbell, A. Claing, Mol. Biol. Cell 25 (2014) 17–29. [37] S. Lewis-Saravalli, S. Campbell, A. Claing, Cell. Signal. 25 (2013) 1813–1819. [38] E. Haines, C. Saucier, A. Claing, J. Biol. Chem. 289 (2014) 5687–5703. [39] P.A. Randazzo, J. Andrade, K. Miura, M.T. Brown, Y.Q. Long, S. Stauffer, P. Roller, J.A. Cooper, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 4011–4016. [40] C.R. Hauck, D.J. Sieg, D.A. Hsia, J.C. Loftus, W.A. Gaarde, B.P. Monia, D.D. Schlaepfer, Cancer Res. 61 (2001) 7079–7090.

P

427 428

474

D

425 426

5. Conclusion

T

423 424

C

421 422

E

419 420

R

417 418

R

415 416

O

414

C

412 413

N

410 411

and the adaptor protein Shc, whereas Tyr-925 phosphorylation mediates binding of Grb2, another adaptor protein. We have previously shown that p66Shc and Grb2 can be found interacting with ARF1, thereby playing an adaptor role for the recruitment of this ARF isoform to the activated EGFR [38]. As partners, we can speculate that depletion of ARF1 can ultimately affect signalization through both adaptors and therefore FAK activation. In addition, Src binding to FAK promotes increased Src kinase activity and in turn, Src-mediated phosphorylation of FAK C-terminal domain residues Tyr-861. This site of FAK phosphorylation acts to modulate FAK localization within cells and promotes the association of FAK with integrin. Our data show that, in absence or inactivation of ARF1, phosphorylation of both tyrosine residues is affected, and FAK cannot interact with β1-integrin to promote adhesion. Overall, our data indicate that ARF1 is a central regulator of FAK activation. Conversely, we found a role for FAK on ARF1 activation. It was proposed that activated FAK recruits ARF GTPase-Activating Proteins (ARF GAPs), like ASAP1. By modulating the activation state of ARFs, ARF GAPs could reduce adhesion and the formation of adhesions complex [39]. In MDA-MB-231 cells, FAK may therefore coordinate processes leading to ARF inactivation. Furthermore, it was shown that inhibition of FAK expression or activity, in human carcinoma cells, disrupted EGF-stimulated signaling [40], which may involve the activation of ARF GTPases. All these data suggest a mechanism, by which FAK expression can regulate indirectly ARF1 activation by affecting EGFR activation. The demonstration that ARF1 regulates focal adhesion complements previous reports that highlighted the role of another ARF isoform, ARF6, in this process [23]. This ARF isoform was reported to regulate both FAK phosphorylation and its association with Src. ARF6 is a central regulator of focal adhesion turnover. Depletion of this GTPase led to increased FAK phosphorylation and constitutive association with Src. In these conditions, the ARF GAP GIT1 can no longer be recruited to FAK and contribute to focal adhesion disassembly, which ultimately impairs endothelial cells motility. We previously demonstrated that in MDA-MB-231 cells, EGF stimulation activates both ARF1 and ARF6 to modulate, respectively, the PI3K and Erk pathways essential for migration [7,41]. Although the physiological endpoint might be similar; in that case, cell migration via focal adhesion formation, the mechanism may be different. Indeed, both depletion of ARF1 and ARF6 block cell migration. ARF1 depletion results in decreased FAK activation, while ARF6 depletion locked FAK in a constitutive activation state. Although ARF1 has been reported to influence the recruitment of paxillin to focal adhesion on Swiss 3T3 Fibroblasts, we demonstrate for the first time its physiological role in invasive breast cancer cells [22]. We show by using RNAi-based approaches, that knock down of ARF1 significantly attenuates actin dynamics of MDA-MB-231 cells by activating focal adhesion formation. This observation was further supported by the use of two other cell lines HCC70, another triplenegative breast cancer cell line, and SKBR3, a HER2 + breast cancer cell line, suggesting that the effect of ARF1 on focal adhesion is not cell type specific, but could represent a general mechanism. Interestingly, it has been previously reported that depletion of ARF1 in MDA-MB231 cells led to cell invasion arrest by controlling the RhoA/RhoC pathway [36]. These two other GTPases are also known to be important in the assembly of focal adhesions in other cell types [42,43]. All this data support the mechanism through which ARF1 acts as a key mediator of breast cancer progression. We demonstrated that phosphorylation of the membrane-associated non-receptor tyrosine kinase Src was regulated by ARF1. Src has been shown to induce the migration, proliferation, adhesion, and growth of cancer cells and has been reported to be overexpressed or activated in many cancers including breast cancer [44,45]. There are a number of anti-Src therapeutics developed or in development for use as cancer therapies [46]. Together, our findings demonstrate that ARF1 is a key regulator of breast cancer cell progression and could also be a potential target for the development of new therapies.

U

408 409

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

E

12

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032

477 478 479 480

483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548

S. Schlienger et al. / Cellular Signalling xxx (2014) xxx–xxx

[44] M.C. Frame, Biochim. Biophys. Acta 1602 (2002) 114–130. 554 [45] D.L. Wheeler, M. Iida, E.F. Dunn, Oncologist 14 (2009) 667–678. [46] L. Morgan, R.I. Nicholson, S. Hiscox, Endocr. Metab. Immune Disord. Drug Targets 8 (2008) 273–278.

N C O

R

R

E

C

T

E

D

P

R O

O

F

[41] S.E. Tague, V. Muralidharan, C. D'Souza-Schorey, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 9671–9676. [42] S.T. Barry, D.R. Critchley, J. Cell Sci. 107 (Pt 7) (1994) 2033–2045. [43] S.T. Barry, H.M. Flinn, M.J. Humphries, D.R. Critchley, A.J. Ridley, Cell Adhes. Commun. 4 (1997) 387–398.

U

549 550 551 552 553

13

Please cite this article as: S. Schlienger, et al., Cell. Signal. (2014), http://dx.doi.org/10.1016/j.cellsig.2014.11.032