1 2 3 4 5 6 7 8 9 10 11 12

Pseudopodial and -arrestin-interacting proteomes from migrating breast cancer cells upon PAR2 activation. Nikolaos Parisis1,2,3, Gergana Metodieva1, Metodi V. Metodiev1, *
1

Department of Biological Sciences/Proteomics Unit, University of Essex, Wivenhoe

Park, Colchester, Essex CO4 3SQ, United Kingdom;

2

Institute of Molecular Genetics (IGMM), CNRS, UMR5535, University of Montpellier I

and II, 34293 Montpellier, France; (current address)

3

Laboratory of Functional Proteomics, INRA, 34060 Montpellier, France (current

address) *Author for correspondence (mmetod@essex.ac.uk) Running title: Pseudopodial and -arrestin proteomes Key words: PAR-2; pseudopodium; beta-arrestin; cell migration; proteomics

13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Summary Metastatic cancer cells form pseudopodia (PD) to facilitate their migration. The proteinase-activated receptor-2 (PAR-2) transduces migratory signals from proteases, and it forms protein complexes with -arrestin and other signalling molecules that are enriched in pseudopodia. More generally, however, pseudopodial regulation is poorly understood. Here, we purified the pseudopodial proteomes of breast cancer cells after activation of the endogenous PAR-2 and we combined gel-based approaches with labelfree high-resolution mass spectrometry to identify proteins that accumulate at the pseudopodia upon PAR-2-mediated migration. We identified >410 proteins in the cell body and >380 in the pseudopodia upon PAR2 activation, of which 93 were enriched in the pseudopodia. One of the pathways strongly enriched in the PD was the clathrinmediated endocytosis signalling pathway, highlighting the importance of the scaffolding function of -arrestin in PAR-2 signalling via its endocytosis. We therefore immunoprecipitated -arrestins, and with mass spectrometry we identified 418 novel putative interactors. These data revealed novel -arrestin functions that specifically control PAR-2-regulated signalling in migrating breast cancer cells but also showed that some -arrestin functions are universal between GPCRs and cell types. In conclusion, this study reveals novel proteins and signalling pathways potentially important for migration of breast cancer cells. Introduction Cell migration is important in embryonic development, tissue homeostasis and in the operation of the immune system. When it malfunctions, it can contribute to several pathological conditions, including cancer metastasis [1]. Several chemotactic signals are transduced into directional cell migration (chemotaxis) via numerous membrane receptors and intracellular signalling pathways. The main characteristic of a chemotactic cell is the formation of a leading edge, termed pseudopodium (PD), as a result of spatiotemporally regulated accumulation of proteins towards the gradient [2]. The identification of proteins important for gradient sensing, pseudopodia extension, attachment and retraction is of vital importance for the understanding of chemotaxis.

32 33 34 35 36 37 38 39 40 41

42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70

Numerous proteins are involved in pseudopodia formation, such as plasma membrane receptors, integrins, kinases and phosphatases, GTPases, scaffolding proteins and others, important for actin dynamics, binding of the cytoskeleton to the extra-cellular matrix (ECM) as well as the control of further signalling events [3-5]. However, the PDenriched proteins, which are dependent on the type of the chemoattractant and the targeted receptor, as well as the cell type, are largely unknown. In addition to their role in degrading the extracellular matrix, trypsin-like proteases secreted from cancer cells have been implicated in the promotion of migration via the proteolytic activation of the protease-activated receptor 2 (PAR-2) [6]. PAR-2 is a ubiquitously but unevenly expressed GPCR receptor, which is activated by trypsin and trypsin-like proteases, including tryptase, clotting factors VIIa and Xa, and has important functions in immune system responses [7, 8]. PAR-2 has also been found overexpressed in cancer cell lines, human tumours and in cells of the tumour microenvironment in situ [9]. Ge et al. reported the secretion of biologically active trypsin-like proteases by the MDA-MB-231 breast cancer cell line that can promote cell migration [10]. PAR-2mediated migratory signals have also been reported in pancreatic cells [11], human melanoma and prostate carcinoma cells [12] and breast cancer cells [13, 14]. Similarly, increase on the migration and ERK signalling of BT549 cells showed the activation of PAR-2 by the clotting factors VIIa and Xa [15], suggesting that chronic inflammation contributes to tumour metastasis. However, little is known about the downstream signalling pathways regulated by this receptor. Following receptor activation, -arrestins antagonize with G-proteins, bind to the receptor, uncouple it from G-proteins [16] and transfer it via clathrin-coated pits to early endosomes [17]. Interestingly, during this arrestin-dependent endocytosis of PAR-2, a multiprotein signalling complex is formed containing the receptor, -arrestin, raf-1 and ERK [18]. In this complex, -arrestin works as a scaffold and retains ERK1/2 in the cytoplasm, thus ERK activity is predominantly cytosolic. Notably, this protein complex is enriched in the pseudopodia, where it promotes actin assembly along with other unidentified proteins [10, 19]. However, the proteins that interact with -arrestin upon PAR-2 activation remain unknown.

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

In this study, we used a breast cancer cell line endogenously expressing the PAR-2 receptor and we applied gel-based, label-free proteomics approaches [20, 21] in order to identify proteins and signalling pathways enriched in the pseudopodia. In addition, we immunoprecipitated -arrestin and identified PAR-2-regulated -arrestin partners in breast cancer cells (Fig. 1). Materials and Methods Materials MCF-7 and MDA-MB-231 were a kind gift from Dr. Elena Klenova and HEK293S were from Dr. Phillip Reeves. Vectors encoding FLAG-tagged -arrestin-1 or FLAG-tagged arrestin-2 (arr+FLAG) were kindly provided by Dr. Robert Lefkowitz and were described elsewhere [22, 23]. PAR-2-activating synthetic peptide (P2AP) SLIGRL-NH2, which is a more potent activator of human PAR-2 compared to SLIGKV-NH2 [24, 25], was from Sigma. Anti-FLAG antibody was from Sigma and anti-phospho p44/p42 MAPK from Cell Signaling. All other antibodies were from Santa Cruz Biotechnology and were used in 1/1000 dilution. Cell culture and plasmid preparations All cells, except HEK293, were cultured in Roswell Park Memorial Institute (RPMI; Lonza) medium supplemented with Ultraglutamine 1, 10% foetal calf serum (FCS; Biosera) and 1% antibiotic mixture (penicillin and streptomycin; Cambrex) at 37oC in a humidified environment (5% CO2). Serum deprivation was performed in the same medium but without FCS and for 12-16 hours. HEK293 were cultured in Dulbecos Modified Eagles Medium (DMEM; Cambrex) with L-glutamine, supplemented with 10% FCS and 1% antibiotic mixture in 37oC and 5% CO2. All cell lines were used at a low passage number. Bacterial (XL1-blue) cultures were grown in Luria Broth (LB) medium and/or LB agar supplemented with ampicillin (100 g/ml; Sigma) or kanamycin (50 g/ml; Sigma) with agitation at 250 rpm at 37oC. Bacterial transformations were conducted according to Inoue et al., [26]. Plasmids were prepared by using commercial kits (Qiagen or Fisher Scientific). Western Blotting

100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128

Protein samples were analysed by SDS-PAGE in 10% home-cast polyacrylamide gels with 4% stacking gels, according to Laemmli [27], and proteins were transferred onto PVDF membranes (Millipore). The membrane was blocked with 5% milk or Odyssey blocking buffer for 1 h. at room temperature (RT) and incubated with the appropriate primary antibodies for 12-16 hours at 4oC. After 3 washes with a PBS solution supplemented with Tween 20 (1%), the membrane was incubated with Alexa580- or Alexa680- conjugated antibodies for 45 min. at RT. Finally, the membrane was scanned by the Odyssey Infrared Imaging system (LI-COR). Band intensity was determined by the Odyssey software and normalized to the loading control. Bar charts were created in Microsoft Excel 2007. Transfections HEK293 cells were transfected with 10 g arr+FLAG by the calcium-phosphate method in 90 mm tissue culture dishes. MCF-7 cells were transfected by using Lipofectamine LTX following the manufacturer's instructions (Invitrogen). Fresh medium was added after 24 hours and the cells were used for experiments or collected 48 hours posttransfection. With these techniques, 90-100% transfection efficiency of HEK293 and 7080% transfection efficiency of MCF-7 was routinely achieved. Under-agarose migration assay The wells in the agarose gels were created by a special glass mould kindly provided by Dr. Y. Mousseau [28]. Briefly, the glass mould was placed in the middle of a 60 mm tissue culture dish and a 10 mL solution of 1% (w/v) agarose in PBS and serum-free RPMI (1:1 ratio) was poured in the dish around the glass and left to polymerize for 8-12 min in a Class II tissue culture hood. The glass mould was removed and the 10 mL of medium were added to equilibrate the gel. Before the experiment, 400 L of serum-free RPMI were added in the central well and left in the incubator for at least one hour to verify that there is not any leakage from the wells. Only agarose gels that did not leak were subsequently used in the agarose migration assay. MCF-7 (4-8x104) cells were seeded in the middle well, left to attach and serum-starved. After 16-20 hours, 50 M P2AP was placed in a side well and serum-free medium alone was placed in the opposite

129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158

side well. Cell migration was monitored on an Olympus IMT-2 Phase contrast inverted microscope and images were captured with an Olympus OM-2N camera. Pseudopodia purification For the separation of the pseudopodia from the cell bodies, both sides of several 24 mm polycarbonate membranes with 3 m pore size were coated with collagen IV (6 g/cm2; Sigma) and serum-starved MCF-7 cells (1.5x104) were seeded on the upper side. In the lower chamber, 50 M P2AP was added for 0-120 minutes. Formed pseudopodia or cell bodies were collected in SDS lysis buffer according to published protocols [4]. Lysates were cleared by centrifugation and stored at -80oC. Phosphopeptide enrichment Equal amounts of CB and PD samples were loaded on an SDS-PAGE gel and the electrophoresis was stopped as soon as the proteins migrated 0.5 mm in the resolving gel. The gel was stained with Zinc sulfate, the band was sliced and subjected to in-gel trypsin digestion. Peptide solutions were enriched for phosphorylated peptides with the Magnetic Phosphopeptide Enrichment Kit (Clontech) on a magnetic separator, following manufacturer's instructions. Immunoprecipitations MCF-7 or HEK293 cells were harvested with a cell scraper in 1% Triton X-100, 50 mM Tris-HCl pH 6.8 supplemented with protease (PMSF, pepstatin, aprotinin) and phosphatase (NaF, Na2VO3, Na4P2O7) inhibitors after rinsing with ice-cold PBS. Cells were left on ice for 30 min. with periodical vortexing and sonicated for complete lysis before being cleared by centrifugation at 14.000 rpm, 4oC for 20 min. Lysates were mixed with agarose beads anti-FLAG M2 (Sigma) or protein A/G beads and incubated for 12-16 hours at 4oC on a rotator. Unbound proteins were removed with 3-5 washes and bound proteins were eluted twice with 3x FLAG peptide, according to manufacturer's instructions, and subsequently with Laemmli buffer. For Western blot analysis, proteins were seperated by 10% SDS-PAGE and were processed as described above. For MS/MS analysis, the eluted proteins were separated by SDS-PAGE, the lanes were cut in 10 slices and in-gel digestion was performed with 20 g/mL trypsin (Trypsin Gold, Mass Spectrometry Grade, Promega) according to Shevchenko et al., [29].

159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189

Mass spectrometric analysis Tryptic peptides were injected onto an Ultimate II nano-LC (Dionex) coupled to Esquire HCT quadrupole ion trap mass spectrometer (Bruker Daltonics) via online nano electrospray ion source. Peptides were separated in a Pepmap 75, 75 m i.d., 150 mm long reverse-phase column packed with 3 m C18 silica particles (Dionex). The elution from the column and the data acquisition have been previously described [20]. The conditions for MS/MS analysis with the Orbitrap Velos have been recently described [21]. Bioinformatic analysis The generated data were converted to mzXML files by using the compassXport software (Bruker Daltonics) and MS/MS data were uploaded on and analysed by the Computational Proteomics Analysis System (CPAS) pipeline [30] with Mascot or X! Tandem as search engines. Peptide Prophet [31] and Protein Prophet [32] scores were generated to provide statistical analysis of protein identification by the Transproteomic Pipeline (TPP) from the Seattle Institute of Systems Biology, which was integrated into the CPAS. The following criteria were used to establish the confidence of the identified proteins: protein prophet 0.9, high peptide prophet scores (0.7), low Expect value, assignment of most prominent peaks as b- or y- ions, experimental sequence-theoretical sequence alignment, K or R as the most C-terminal residue, aspartic acid- or prolinedirected fragmentation. To avoid false-positive hits, only proteins matched with at least 2 peptides or 1 unique peptide were considered as true hits while all MS/MS spectra were manually inspected. Relative abundance was estimated by the spectral counting method [33]. For the pseudopodial proteins, pathway analysis, molecular function and biological processes of proteins were obtained from GeneOntology (www.geneontology.org) and pie charts were created in PANTHER classification system [34]. The trial version of the Ingenuity Pathway Analysis program was used for functional annotations, signalling pathway analysis and creation of bar charts [35]. For the -arrestin-interacting proteins, we performed two biological replicates and two technical replicates. The PANTHER classification system was used to classify proteins based on their molecular function and PANTHER protein class. In the analyses, we used only proteins that were not identified in the control experiment (transfection with an empty vector or untransfected cells) or

190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220

were identified with significantly higher (ten-fold) spectral counts in the transfected conditions. Results Endogenous expression of PAR-2 in MCF-7 cells and activation of ERK cascade by uniform concentrations of P2AP In order to study PAR-2 signalling in its physiological context, we used MCF7 cells, a breast cancer cell line which expresses PAR-2 endogenously (Fig. 2A). Signalling downstream of PAR-2 is strongly associated with ERKs and has important roles in migration [18]. The effect of chemotactic agents at uniform or gradient concentrations on the ERK pathway is cell-type dependent [36]. In order to test whether uniform concentrations of the PAR-2-activating peptide SLIGRL (P2AP) induce ERK phosphorylation, MCF-7 (low invasive) and MDA-MB-231 (highly invasive) cells were deprived of serum and then treated with P2AP or FCS for 0-120 min. As expected, the addition of P2AP for 30 min in culture significantly induced the ERK pathway in both cell lines and the pERK levels remained higher than basal over a time course of at least 2 hours (Fig. 2B). P2AP did not induce any change in the low levels of basal pERK in the non-invasive CAMA and ZR-75 cells (data not shown). These results confirmed that PAR-2 activation by uniform concentrations of P2AP induced the ERK pathway in MCF7 cells. Migratory effect of P2AP on MCF-7 cells Since MCF-7 cells responded well to uniform concentrations of P2AP with respect to ERK activity upon PAR-2 activation and as this cell line is considered as low invasive [37], we tested whether PAR-2 could sense gradient concentration of its ligand and trigger the ERK pathway as well as cell migration. PAR-2 activation is known to enhance migration of other breast cancer cells via the ERK pathway, including MDA-MB-231 and MDA-MB-468, as well as of mouse embryonic fibroblast cells NIH3T3 [10, 18] and there is one study that tested this in MCF-7 [14]. We studied the effect of PAR-2 activation on cell migration with the agarose migration assay. Indeed, even though the cells did not migrate under the gel in the agarose migration assay, probably due to their large size or their slow migration rate and their need for longer incubation times in order to continue migrating under the agarose, they accumulated at the side of the well towards

221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250

P2AP (Fig. 3A). These results are consistent with Kamath et al. [14] and suggest that MCF-7 cells sensed the gradient of P2AP, and the subsequent activation of PAR-2 promoted their migration. Proteins and signalling pathways enriched in the pseudopodia of breast cancer cells. The migration of cells is characterized by the extension of a leading edge and the formation of a pseudopodium [3]. Since activation of PAR-2 promoted the migration of MCF-7, we aimed to identify the proteins that accumulate at the pseudopodia of migrating MCF-7 cells upon activation of PAR-2, by the method developed by Klemke and colleagues [4]. First, we found the optimum timepoint to harvest the dynamic structure of pseudopodia at the time they were formed. The P2AP was left in the lower compartment for 0, 30, 60, 90 or 120 min and the PD were harvested. The protein amounts in each time point were: low at 30 and 90 min. and almost minimal at 0 and 120 min. The highest amount of proteins was harvested at the 60 min suggesting that this is the timepoint of pseudopodial formation (data not shown). Therefore, for the rest of the study we stimulated the cells for 60 min before PD purification. The efficiency of the purification was controlled with Western blots against the poly[ADP-ribose]polymerase 1 (PARP-1). Accordingly, PARP-1 was enriched in the cell body fractions and only trace amounts were detected in the pseudopodia fraction (Fig. 3C upper panel). Similarly, the band that corresponds to histones was absent from the PD fraction (Fig. 3B). These results showed the high efficiency of pseudopodia purification from migratory MCF-7 cells. Next, the ERK levels in each compartment were measured. Phospho-ERK1/2 levels were increased in both CB and PD upon PAR-2 activation by 2.4- and 2.8-fold, respectively (Fig. 3C lower panel and Fig. 3D). More detailed, pERK1 levels showed an increase of similar magnitude (2.95 and 3.05 times, respectively) while pERK2 levels showed higher increase in the PD (2.16-fold) than in the CB (1.57-fold). Two complementary workflows were used in order to identify PAR-2-regulated pseudopodial proteins and pathways during MCF-7 cell migration: (i) the pools of proteins from the pseudopodia and cell bodies were analysed by SDS-PAGE and the entire lanes were sliced into 10 regions (Fig. 3B); (ii) the pools of proteins from the pseudopodia and cell bodies were digested by trypsin and phosphopeptides were

251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280

enriched. Relative quantitation was performed with the spectral counting method; proteins with PD/CB ratio 1.5 were considered enriched in the pseudopodium, whereas those with a ratio 0.66 were considered to be rather excluded from the PD [4, 20, 33]. The sample fractionation at the protein level in workflow 1 was crucial for the identification of 3.2 times more proteins compared to workflow 2. In total, 250 nonredundant proteins were identified, of which 207 with workflow 1 and 63 with workflow 2 while 20 proteins were identified in both (that is 32% of the total proteins identified in workflow 2). Comparative analysis of CB and PD fractions showed a total of 93 proteins enriched in the PD, 117 proteins were rather depleted from PD, while 60 did not show any significant change. To increase the confidence of the identification, of the 20 proteins that overlapped in workflows 1 and 2, only those that showed identical enrichment profiles in both workflows were considered as PD enriched proteins (Fig. 4 and Table 1). Protein pathway analysis by the PANTHER Classification System [34] revealed the pathways that are most represented by the enriched proteins. In the PD, some of those were the apoptosis, interleukin or Huntington disease signalling pathways. Interestingly, several other receptor signalling pathways, such as dopamine receptor, EGF, FGF and PDGF, were also enriched in the PAR-2 mediated pseudopodia. In the CB, DNA replication-related pathways as well as biosynthetic and metabolic pathways (arginine, Oantigen, pyruvate and others) were enriched (Fig. 4). Furthermore, the breakdown of biological processes of the PD-enriched proteins was as follows: cellular processes (34.6%), transport (23.5%), cell communication (12.3%), cellular component organization (21%), metabolic process (49.4%), developmental processes (22.2%) and others (SI Fig. 3). Identification of -arrestin-interacting proteins Among the proteins found enriched in the pseudopodia of MCF-7 cells were the two large adaptins of the clathrin-associated adaptor protein (AP) complex (AP-2 complex subunit alpha-1 and subunit beta) as well as the clathrin heavy chain and coatomer subunits, indicating the significance of clathrin-mediated endocytosis to pseudopodia formation. The AP complex regulates the formation of the clathrin-coated vesicles, which are sequestered by -arrestins, the important mediators of 7-TMR endocytosis. As PAR-2

281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309

is dependent on the scaffolding ability of -arrestins for PAR-2 endocytosis as well as for downstream signalling, and the PAR-2-regulated -arrestin interactome is far from complete, we aimed to identify the proteins that are sequestered by -arrestins and hence to reveal novel -arrestin-dependent functions triggered by PAR-2 activation. In order to achieve this, we transiently transfected MCF-7 cells with FLAG-tagged arrestin-1 or -arrestin-2 in MCF-7 cells with endogenous levels of PAR-2, and after treatment with either P2AP, FCS or serum-free medium, the -arrestins and their interacting partners were immunoprecipitated by using the FLAG purification system. As controls, we performed parallel experiments with cells transfected with an empty vector or untransfected. The proteins identified in these negative controls were considered as background and excluded from further analyses (SI table 11). Furthermore, proteins that are usual contaminants in proteomics experiments, such as keratins, albumin, glyceraldehyde-3-phosphate dehydrogenase, heat shock protein, desmoplakin, or others that interact unspecifically with the anti-FLAG beads, were excluded from our analyses. With the above stringent criteria, we identified 418 proteins that co-immunoprecipitate with -arrestins in all conditions. Of these, 223 interacted with -arrestin-1, 325 with arrestin-2 while several of them interacted with both. More specifically, 115 proteins interacted with -arrestin-1 in uninduced conditions, and 35 of those interactions were lost after treatment; 28 interacted only in FCS-treatment; 57 only in PAR-2 activating conditions, while 23 interacted after both treatments (table 2 and SI table 9). Similarly, with -arrestin-2, 186 proteins interacted in uninduced conditions and 84 of these were no longer bound after treatment; 55 bound to -arrestin-2 only upon FCS treatment; 65 only in PAR-2 activating conditions, while 19 interacted after both treatments (table 3 and SI table 9). Finally, a number of proteins were found associated with one -arrestin in untreated conditions, but upon treatment were found associated with the other arrestin, and vice versa. In order to increase the confidence of our proteomics screen, we show by immunoblotting that two of the identified protein (nucleophosmin and heat shock protein 70) coimmunoprecipitate specifically with -arrestin-2 upon PAR-2 activation.

310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338

Canonical pathway analysis of -arrestin interactors Using Ingenuity Pathway Analysis, we generated and compared the canonical pathways of the proteins interacting with either -arrestin (SI Fig. 3). Interestingly, after PAR-2 activation, -arrestin-1 interactions are more abundant than -arrestin-2 in pathways of cell junction signalling, gap junction signalling, intrinsic prothrombin activation pathway, dendritic cell maturation. On the other hand, -arrestin-2 appears to be more involved in interleukin signalling, actin cytoskeleton signalling, clathrin-mediated endocytosis or other endocytic pathways, regulation of actin based motility by Rho, RhoA signalling, RhoGDI signalling, protein kinase A signalling, granzyme A signalling, PI3K/AKT signalling and others. We also used PANTHER to categorize proteins based on their GO Molecular Function and Biological Process (SI Fig. 5A-I) as well as PANTHER Protein Class (Tables 2, 3). In terms of GO Molecular Function, -arrestin-1-interacting proteins in untreated cells have binding, catalytic, enzyme regulator, transcription factor and structural molecule activities and when treated with P2AP, they are additionally involved in receptor, motor and transporter activities. For -arrestin-2-interacting proteins in untreated cells, they are involved in the same activities as for -arrestin-1 in addition to receptor and transporter activities and in PAR-2-activating conditions they are additionally engaged with ion channel and translation regulator activities. Discussion Protease inhibitors have been used in the treatment of several pathologies, such as asthma, emphysema, and others [8, 38], and over the past 10 years the possibility of using protease inhibitors or PAR-2 antagonists as therapeutics in tumour metastasis has been investigated [8]. Taking into consideration the recent finding that actively migrating cancer cells secrete trypsin proteases [10] as well as the heterogeneous nature of cancer, it is conceivable that final stage tumour cells, with an acquired ability to invade, secrete proteases for ECM degradation as well as for promoting migration of other earlier stage tumour cells. Therefore, identifying the molecular mechanisms that govern the PAR-2mediated metastasis of breast cancer cells is critical for the complete understanding of this process.

339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368

By identifying proteins enriched in the pseudopodia of MCF-7 breast cancer cells, which endogenously express the PAR-2 receptor, we cataloged the most important signalling pathways represented in this cell line (Table 1 and Fig. 4). In addition, we identified a number of putative -arrestin interactors downstream of the PAR-2 receptor (Tables 2 and 3). Several of these identified proteins interact with -arrestin upon angiotensin II type 1a receptor (AT1A-R) activation in HEK293 cells, but the majority are novel suggesting that they might be cancer-specific interactions during cancer metastasis. It is generally accepted that a migrating cell requires spatiotemporal re-arrangement of proteins that lead to the formation of a leading edge and a trailing back [2, 3]. It is shown here that, in MCF-7 breast cancer cells, the ERK pathway is activated by uniform concentrations of PAR-2-activating peptide (P2AP) and the levels remain high over a course of at least 2 hours (Fig. 2B). This is also the case with the highly invasive MDAMB-231, consistent with the findings of Ge et al., [10] (Fig. 2B). Upon gradient concentrations of P2AP, an increase in the pERK1/2 levels is observed in the entire cell but there is no spatial difference in either total pERK1/2 or pERK1 activity in cell bodies and pseudopodia, while pERK2 is slightly enriched in the PD and thus preferred in the regulation of the pseudopodia formation in MCF-7 cells (Fig. 3C, D). This suggests that ERK1 and ERK2 have diverse functions and target distinct substrates leading to cytoskeletal re-arrangments and subsequent formation of the pseudopodium. Whether these functions are interchangeable is something that has to be studied further. The relatively weak increase of pERK2 in PD could possibly explain the low ability of MCF7 to migrate and invade; perhaps a bigger increase would be observed in highly metastatic cells. Proteomic screens of the pseudopodial proteome have been published from experiments in COS-7 cells in response to lipopolysaccharide (LPS) [4, 5] and in human astrocytoma cells [39]. In the present study we used a cell line with low metastatic ability and we induced migration by activating the endogenous PAR-2 receptor (Fig. 3A). We analysed the proteins involved in the formation and extension of the pseudopodium after activation of PAR-2. Subsequent analyses were focused on the proteins found enriched in the PD fraction as these proteins are more likely to contribute to the formation of the

369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399

pseudopodia. Other proteins that are equally distributed in the PD and CB might also play a role and they can be the subject of further studies, especially of phosphoproteomic studies [40]. All the proteins confidently found enriched in the pseudopodium are presented in Table 1 and none of them have been previously described as PAR-2 regulated. Despite our phosphopeptide enrichment step, we did not detect phosphorylation sites, possibly due to the small size of pseudopodia formed by the MCF7 cells producing low protein abundant pseudopodial samples as starting material. Two of the proteins that are candidates for further studies are nidogen-1 and nidogen-2. After careful inspection of our mass spectrometry data we found that the peptides corresponding to these proteins were not proteotypic. Therefore, it was not possible to distinguish whether it is one of the two specific isoforms or both that were identified. In the tables we have included only nidogen-1 but it is highly likely that both are involved in pseudopodia formation as they have around 50% sequence similarity and they perform similar or perhaps identical functions [57]. Kubota et al., proposed a possible function for nidogen-1 in migration of C. elegans, during the interaction of integrins with the basement membrane, and that this function is evolutionarily conserved in mammals [41]. Furthermore, nidogen-1 knockout mice presented impaired wound healing [42]. What makes our finding more intriguing is the recent establishment of nidogen-2 as a biomarker for ovarian cancer, since it was overexpressed in the serum of women with ovarian cancer [43]. The PAR-2-regulated signalling pathways represented by the CB- and PD-enriched proteins are listed in Fig. 4. The results are consistent with the findings of Wang et al. [4] study, in which EGF, PDGF and Huntington's disease signalling were also enriched in the PD. In addition, here we show that clathrin-mediated endocytosis as well as heme biosynthesis, interleukin and PI3 kinase pathways signalling together with dopamine receptor and JAK/STAT signalling are also important pathways in cancer cell migration. These new findings are likely to be attributed to the migration of cancer cells specifically. With respect to the CB-enriched pathways, we found cell cycle-related and various metabolic pathways, similarly to the Wang et al study [4]. The AP-2 complex large subunits alpha-1 and beta were found only in PD. The AP-2 complex is localized at the plasma membrane and is associated with -arrestin and clathrin-mediated endocytosis but

400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428

here we show that upon PAR-2-mediated pseudopium formation AP-2 complex and clathrin signalling are enriched in the PD. The latter strongly suggests that receptor internalization is a critical process for the promotion of migratory signals. This is in agreement with previous studies [10, 19] and raises the question of whether blocking receptor internalization could be used in therapies against metastasis. Furthermore, although the actin cytoskeleton is heavily regulated in both compartments, the Rho signalling pathway was strongly enriched in PD showing the importance of this pathway in cell polarization and invasion, in agreement with others [44]. It has been reported by several groups that -arrestin is an important scaffold protein initiating signalling pathways during receptor internalization via clathrin-coated pits. Hence, we pulled-down -arrestins after activating the PAR-2 receptor and we identified -arrestin-interacting proteins. The -arrestin interactome has been studied by Lefkowitz and colleagues using proteomics downstream of the angiotensin II type 1a receptor in human embryonics kidney cells (HEK293) [45] and by other groups using other methods (reviewed in [46-48]), therefore, here we report the first interactome in cancer cells and upon stimulation of PAR-2. Several proteins identified in our study were previously known as -arrestin interactors, such as 14-3-3, serine/threonine kinase 38, clathrin heavy chain 1 and 2, copine, annexin, fatty acid synthase, ATP synthase and , phosphatases, actins, tubulins, tubulin -3, 8, F-actin capping protein, myosin-9, filamin, kinesin, dynein, titin, calpain-1 catalytic subunit, dehydrogenase, phospholipase, tropomyosin, histone H1, heat shock protein 70, 70.1, 90, T-complex protein 1, carboxypeptidase, treacle, nucleophosmin, nucleolin, hnRNP, ribosomal proteins S3, S7, L7, L19, L22, P0, P1, P2 and others (SI table 4) [45]. In addition, several isoforms of proteins that are known to interact with -arrestins, such as phosphatases and ribosomal proteins, were also found here. We also excluded proteins that were found in other proteomic screens as -arrestin-interactors but in our strategy were considered as background due to their identification in the negative controls (SI table 11). Also note that in this study, we did not over-express the receptor, reducing the possibility of unspecific binding due to exogenous protein expression.

429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458

Importantly, we identify new interactions, such as TRIM29, glutaredoxin, periplakin and others, and these interactions might play a role in PAR-2 signalling in cancer. The finding of alpha-actinin as well as several other actin isoforms in our screens suggests that actin cytoskeleton and -arrestins interact to manage receptor endocytosis [49] and leading edge formation [50]. Similarly, their interaction with S100A7 increases the role of arrestins in metastatic pathways [51]. It is also shown here that PAR-2 regulates the activity of kinases such as cyclin-dependent kinase (cdk) 13, N-acetyl-D-glucosamine kinase and microtubule-associated serine/threonine-protein kinase 2, in a -arrestindependent manner. More specifically, cdk13 and N-acetyl-D-glucosamine kinase are pulled down with both -arrestins in uninduced conditions but in induced conditions this association is lost. The opposite seems to be the case for microtubule-associated serine/threonine-protein kinase 2 (Tables 2, 3 and SI Table 9). Another two proteins we showed here by mass spectrometry and independent co-IPs that are genuine -arrestininteracting proteins downstream of PAR-2 are nucleophosmin B23 (NPM1) and heat shock protein 70 (HSP70) (Fig. 5). This is specifically interesting in the case of NPM1, as it is a multifunctional protein that is over-expressed in cancer cells, compared to normal cells, and it is considered a putative therapeutic target for hematological diseases [52-54]. Another protein found in our proteomic screen, the eukaryotic translation initiation factor 3 (eIF3), has been recently found as a -arrestin-2 interactor by another study [55]. This interaction, which is mediated by epidermal growth factor (EGF) stimulation, promotes binding of eIF3a to SHC and Raf-1 and subsequent suppression of the ERK pathway [55]. Whether all these interactions take place in other types of cancer or are also regulated by other GPCRs is a question that requires further study. Furthermore, we compared the overlap between the -arrestin-interactors and the PDenriched protein screens in order to identify other proteins of special interest (Table 4). These proteins are very likely to spatiotemporally translocate to the pseudopodium or retained away from it in a -arrestin-dependent manner. For example, copine, T-plastin, coatomer subunit beta and LDH-A were found associated with -arrestin-2 in untreated conditions but when PAR-2 is activated this interaction is lost and the proteins are found in pseudopodia. Alpha-actinin, AHNAK and elongation factor 1-alpha 1 are likely to be

459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484

recruited to the pseudopodia by -arrestins when PAR-2 is activated. Very importantly, RNA helicase DDX5 has also been found in this overlap. DDX5 is a nuclear DEAD-box containing protein with roles in development, miRNA pathway, in regulation of transcription and RNA processing and ribosome biogenesis. It is mainly nuclear but it was reported to translocate to the cytoplasm in order to recruit beta-catenin back to the nucleus and control transcription [56]. In our screens, it was found as -arrestin-2 partner only when cells were treated with FCS. Therefore, as the pseudopodia in our study were formed upon PAR-2 activation, this shows that DDX5 is enriched at the PD when not sequestered by -arrestin-2. As an additional validation of our screens, AP-2 complex subunits alpha-1 and beta were found in both screens. Conclusion In conclusion, this study presents the proteins and the signalling pathways enriched in the pseudopodia of migrating breast cancer cells by using pseudopodia purification and mass spectrometry-based proteomics. Furthermore, by using immunoprecipitation and mass spectrometry, several PAR-2-regulated -arrestin-1 and -arrestin-2 interactors were identified and revealed novel functions. This study therefore expanded the possible functions of -arrestins, especially during breast cancer cell migration, and provides directions for future research. Acknowledgment We acknowledge all members of the Metodiev laboratory for support and useful discussions. We are grateful to Robert Lefkowitz, for kindly providing the -arrestinexpressing vectors used in the study, Yoanne Mousseau for technical help and critical advice in setting up the agarose migration assay. NP is grateful to Daniel Fisher for his continuous support. We also acknowledge Daniel Fisher, Liliana Krasinska and James Hutchins for proofreading and for critical recommendations while writing the first version of the manuscript.

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

Figure legends. Figure 1. Schematic representation of the experimental strategies performed in this study. Figure 2. Endogenous expression of PAR-2 in MCF7 cells and ERK activity in untreated, FCS (10%) or PAR-2 (50 M) treated conditions. (A) Detection of PAR-2 mRNA in MCF7 cells by reverse transcription PCR. (B) Levels of ERK activity upon treatment with P2AP (50 M) or FCS. Cells were serum-starved and then treated or not with the indicated agents for the indicated times and pERK was detected by anti-phospho p44/p42 MAPK (Cell Signaling) in clear lysates by western blotting. Tubulin was used as loading control Figure 3. Effect of gradient concentrations of P2AP on chemotaxis and on ERK activity in MCF-7 cells. (A) Under agarose migration assay: Three wells were generated essentially as described in [28]. cells were seeded in the middle well while 50 M P2AP or RPMI only were added in the side wells. The chemoattractant diffused through the agarose creating a gradient to which cells responded by migrating towards the side with the chemoattractant. The image is representative of 3 individual experiments. Bar: 100 m; (B) The pools of proteins from cell bodies (CB) or pseudopodia (PD) separated by the pseudopodia purification assay were analysed by 10% SDS-PAGE. Arrow indicates the histone proteins. Lines indicate the sliced gel pieces from which proteins were digested and analysed by mass spectrometry. (C) Immunoblots of PARP-1 for validation of PD purification (upper panel) and pERK (lower panel) in PD and CB. (D) Levels of pERK1/2, pERK1 and pERK2 in PD and CB. Figure 4. Comparison of the protein identifications from the 2 different proteomic workflows used to identify PD and CB enriched proteins (see text for details). In parallel, proteins are categorized based on the PD/CB ratio, with the PD enriched proteins to the right-hand side and CB-enriched to the left-hand side. The most enriched signalling pathways represented by these proteins in each fraction are listed; the complete PANTHER pathway analysis is shown in SI figure 2.

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

Figure 5. Validation of nucleophosmin (A) and heat shock protein 70 (HSP70; B) as -arrestin interacting proteins by co-immunoprecipitation (co-IP). Control experiments were co-IPs with -FLAG beads from untransfected cells or with unreactive protein A/G beads from FLAG-tagged -arrestin-transfected cells. SI Figure 1. PANTHER pathway analysis of proteins enriched in the pseudopodia (left-hand panel) or cell bodies (right-hand panel) [34]. Numbers in square brackets indicate the percentage of the number of proteins corresponding to this pathway. SI Figure 2 Gene Ontology (GO) terms of PD- and CD-enriched proteins in pie charts generated by PANTHER Classification System (http://www.pantherdb.org/). SI Figure 3 Ingenuity Pathway Analysis of proteins that co-immunoprecipitate with -arrestin-1 (dark blue bars) or -arrestin-2 (light blue bars). SI Figure 4 A-I. Gene Ontology (GO) terms Molecular Function and Biological Process of -arrestininteracting proteins, upon treatment with FCS, P2AP or not treated, in pie charts generated by PANTHER Classification System (http://www.pantherdb.org/).

References
[1] Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, et al. Cell migration: integrating signals from front to back. Science. 2003;302:1704-9. [2] Devreotes P, Janetopoulos C. Eukaryotic chemotaxis: distinctions between directional sensing and polarization. Journal of Biological Chemistry. 2003;278:20445-8. [3] Chodniewicz D, Klemke RL. Guiding cell migration through directed extension and stabilization of pseudopodia. Experimental Cell Research. 2004;301:31-7. [4] Wang Y, Ding SJ, Wang W, Jacobs JM, Qian WJ, Moore RJ, et al. Profiling signaling polarity in chemotactic cells. Proceedings of the National Academy of Sciences of the United States of America. 2007;104:8328-33.

545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592

[5] Wang Y, Kelber JA, Tran Cao HS, Cantin GT, Lin R, Wang W, et al. Pseudopodium-enriched atypical kinase 1 regulates the cytoskeleton and cancer progression [corrected]. Proc Natl Acad Sci U S A. 2010;107:10920-5. [6] Nystedt S, Emilsson K, Larsson AK, Strombeck B, Sundelin J. Molecular cloning and functional expression of the gene encoding the human proteinase-activated receptor 2. European Journal of Biochemistry. 1995;232:84-9. [7] Dery O, Corvera CU, Steinhoff M, Bunnett NW. Proteinase-activated receptors: novel mechanisms of signaling by serine proteases. American Journal of Physiology. 1998;274:C142952. [8] Ramachandran R, Noorbakhsh F, Defea K, Hollenberg MD. Targeting proteinase-activated receptors: therapeutic potential and challenges. Nat Rev Drug Discov. 2012;11:69-86. [9] D'Andrea MR, Derian CK, Santulli RJ, Andrade-Gordon P. Differential expression of proteaseactivated receptors-1 and -2 in stromal fibroblasts of normal, benign, and malignant human tissues. American Journal of Pathology. 2001;158:2031-41. [10] Ge L, Shenoy SK, Lefkowitz RJ, DeFea K. Constitutive protease-activated receptor-2mediated migration of MDA MB-231 breast cancer cells requires both beta-arrestin-1 and -2. Journal of Biological Chemistry. 2004;279:55419-24. [11] Ikeda O, Egami H, Ishiko T, Ishikawa S, Kamohara H, Hidaka H, et al. Signal of proteinaseactivated receptor-2 contributes to highly malignant potential of human pancreatic cancer by up-regulation of interleukin-8 release. International Journal of Oncology. 2006;28:939-46. [12] Shi X, Gangadharan B, Brass LF, Ruf W, Mueller BM. Protease-activated receptors (PAR1 and PAR2) contribute to tumor cell motility and metastasis. Mol Cancer Res. 2004;2:395-402. [13] Hjortoe GM, Petersen LC, Albrektsen T, Sorensen BB, Norby PL, Mandal SK, et al. Tissue factor-factor VIIa-specific up-regulation of IL-8 expression in MDA-MB-231 cells is mediated by PAR-2 and results in increased cell migration. Blood. 2004;103:3029-37. [14] Kamath L, Meydani A, Foss F, Kuliopulos A. Signaling from protease-activated receptor-1 inhibits migration and invasion of breast cancer cells. Cancer Research. 2001;61:5933-40. [15] Morris DR, Ding Y, Ricks TK, Gullapalli A, Wolfe BL, Trejo J. Protease-activated receptor-2 is essential for factor VIIa and Xa-induced signaling, migration, and invasion of breast cancer cells. Cancer Research. 2006;66:307-14. [16] Krupnick JG, Benovic JL. The role of receptor kinases and arrestins in G protein-coupled receptor regulation. Annual Review of Pharmacology and Toxicology. 1998;38:289-319. [17] Ossovskaya VS, Bunnett NW. Protease-activated receptors: contribution to physiology and disease. Physiological Reviews. 2004;84:579-621. [18] DeFea KA, Zalevsky J, Thoma MS, Dery O, Mullins RD, Bunnett NW. beta-arrestin-dependent endocytosis of proteinase-activated receptor 2 is required for intracellular targeting of activated ERK1/2. Journal of Cell Biology. 2000;148:1267-81. [19] Ge L, Ly Y, Hollenberg M, DeFea K. A beta-arrestin-dependent scaffold is associated with prolonged MAPK activation in pseudopodia during protease-activated receptor-2-induced chemotaxis. Journal of Biological Chemistry. 2003;278:34418-26. [20] Alldridge L, Metodieva G, Greenwood C, Al-Janabi K, Thwaites L, Sauven P, et al. Proteome profiling of breast tumors by gel electrophoresis and nanoscale electrospray ionization mass spectrometry. J Proteome Res. 2008;7:1458-69. [21] Greenwood C, Metodieva G, Al-Janabi K, Lausen B, Alldridge L, Leng L, et al. Stat1 and CD74 overexpression is co-dependent and linked to increased invasion and lymph node metastasis in triple-negative breast cancer. J Proteomics. in print. [22] Laporte SA, Oakley RH, Zhang J, Holt JA, Ferguson SS, Caron MG, et al. The beta2-adrenergic receptor/betaarrestin complex recruits the clathrin adaptor AP-2 during endocytosis.

593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640

Proceedings of the National Academy of Sciences of the United States of America. 1999;96:3712-7. [23] Luttrell LM, Ferguson SS, Daaka Y, Miller WE, Maudsley S, Della Rocca GJ, et al. Betaarrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science. 1999;283:655-61. [24] Blackhart BD, Emilsson K, Nguyen D, Teng W, Martelli AJ, Nystedt S, et al. Ligand crossreactivity within the protease-activated receptor family. Journal of Biological Chemistry. 1996;271:16466-71. [25] Compton SJ, Cairns JA, Palmer KJ, Al-Ani B, Hollenberg MD, Walls AF. A polymorphic protease-activated receptor 2 (PAR2) displaying reduced sensitivity to trypsin and differential responses to PAR agonists. Journal of Biological Chemistry. 2000;275:39207-12. [26] Inoue H, Nojima H, Okayama H. High efficiency transformation of Escherichia coli with plasmids. Gene. 1990;96:23-8. [27] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227:680-5. [28] Mousseau Y, Leclers D, Faucher-Durand K, Cook-Moreau J, Lia-Baldini AS, Rigaud M, et al. Improved agarose gel assay for quantification of growth factor-induced cell motility. Biotechniques. 2007;43:509-16. [29] Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nat Protoc. 2006;1:2856-60. [30] Rauch A, Bellew M, Eng J, Fitzgibbon M, Holzman T, Hussey P, et al. Computational Proteomics Analysis System (CPAS): an extensible, open-source analytic system for evaluating and publishing proteomic data and high throughput biological experiments. J Proteome Res. 2006;5:112-21. [31] Keller A, Eng J, Zhang N, Li XJ, Aebersold R. A uniform proteomics MS/MS analysis platform utilizing open XML file formats. Mol Syst Biol. 2005;1:2005 0017. [32] Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A statistical model for identifying proteins by tandem mass spectrometry. Anal Chem. 2003;75:4646-58. [33] Liu H, Sadygov RG, Yates JR, 3rd. A model for random sampling and estimation of relative protein abundance in shotgun proteomics. Analytical Chemistry. 2004;76:4193-201. [34] Thomas PD, Kejariwal A, Campbell MJ, Mi H, Diemer K, Guo N, et al. PANTHER: a browsable database of gene products organized by biological function, using curated protein family and subfamily classification. Nucleic Acids Res. 2003;31:334-41. [35] Calvano SE, Xiao W, Richards DR, Felciano RM, Baker HV, Cho RJ, et al. A network-based analysis of systemic inflammation in humans. Nature. 2005;437:1032-7. [36] Brahmbhatt AA, Klemke RL. ERK and RhoA differentially regulate pseudopodia growth and retraction during chemotaxis. Journal of Biological Chemistry. 2003;278:13016-25. [37] Thompson EW, Paik S, Brunner N, Sommers CL, Zugmaier G, Clarke R, et al. Association of increased basement membrane invasiveness with absence of estrogen receptor and expression of vimentin in human breast cancer cell lines. Journal of Cellular Physiology. 1992;150:534-44. [38] Eden E, Hammel J, Rouhani FN, Brantly ML, Barker AF, Buist AS, et al. Asthma features in severe alpha1-antitrypsin deficiency: experience of the National Heart, Lung, and Blood Institute Registry. Chest. 2003;123:765-71. [39] Beckner ME, Chen X, An J, Day BW, Pollack IF. Proteomic characterization of harvested pseudopodia with differential gel electrophoresis and specific antibodies. Lab Invest. 2005;85:316-27. [40] Thingholm TE, Jensen ON, Larsen MR. Analytical strategies for phosphoproteomics. Proteomics. 2009;9:1451-68.

641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680

[41] Kubota Y, Ohkura K, Tamai KK, Nagata K, Nishiwaki K. MIG-17/ADAMTS controls cell migration by recruiting nidogen to the basement membrane in C. elegans. Proceedings of the National Academy of Sciences of the United States of America. 2008;105:20804-9. [42] Baranowsky A, Mokkapati S, Bechtel M, Krugel J, Miosge N, Wickenhauser C, et al. Impaired wound healing in mice lacking the basement membrane protein nidogen 1. Matrix Biology. 2008;29:15-21. [43] Kuk C, Gunawardana CG, Soosaipillai A, Kobayashi H, Li L, Zheng Y, et al. Nidogen-2: a new serum biomarker for ovarian cancer. Clinical Biochemistry. 2010;43:355-61. [44] Vishnubhotla R, Sun S, Huq J, Bulic M, Ramesh A, Guzman G, et al. ROCK-II mediates colon cancer invasion via regulation of MMP-2 and MMP-13 at the site of invadopodia as revealed by multiphoton imaging. Lab Invest. 2007;87:1149-58. [45] Xiao K, McClatchy DB, Shukla AK, Zhao Y, Chen M, Shenoy SK, et al. Functional specialization of beta-arrestin interactions revealed by proteomic analysis. Proc Natl Acad Sci U S A. 2007;104:12011-6. [46] Lefkowitz RJ, Rajagopal K, Whalen EJ. New roles for beta-arrestins in cell signaling: not just for seven-transmembrane receptors. Mol Cell. 2006;24:643-52. [47] Lefkowitz RJ, Shenoy SK. Transduction of receptor signals by beta-arrestins. Science. 2005;308:512-7. [48] Min J, Defea K. beta-arrestin-dependent actin reorganization: bringing the right players together at the leading edge. Mol Pharmacol.80:760-8. [49] Lamaze C, Fujimoto LM, Yin HL, Schmid SL. The actin cytoskeleton is required for receptormediated endocytosis in mammalian cells. J Biol Chem. 1997;272:20332-5. [50] Min J, Defea K. beta-arrestin-dependent actin reorganization: bringing the right players together at the leading edge. Mol Pharmacol. 2011;80:760-8. [51] Nasser MW, Qamri Z, Deol YS, Ravi J, Powell CA, Trikha P, et al. S100A7 enhances mammary tumorigenesis through upregulation of inflammatory pathways. Cancer Res.72:604-15. [52] Colombo E, Alcalay M, Pelicci PG. Nucleophosmin and its complex network: a possible therapeutic target in hematological diseases. Oncogene. 2011;30:2595-609. [53] Okuwaki M. The structure and functions of NPM1/Nucleophsmin/B23, a multifunctional nucleolar acidic protein. J Biochem. 2008;143:441-8. [54] Yung BY. Oncogenic role of nucleophosmin/B23. Chang Gung Med J. 2007;30:285-93. [55] Xu TR, Lu RF, Romano D, Pitt A, Houslay MD, Milligan G, et al. Eukaryotic translation initiation factor 3, subunit a, regulates the extracellular signal-regulated kinase pathway. Mol Cell Biol.32:88-95. [56] Yang L, Lin C, Liu ZR. P68 RNA helicase mediates PDGF-induced epithelial mesenchymal transition by displacing Axin from beta-catenin. Cell. 2006;127:139-55. [57] Ulazzi L, Sabbioni S, Miotto E, Veronese A, Angusti A, Gafa R, et al. Nidogen 1 and 2 gene promoters are abberantly methylated in human gastointestinal cancer. Mol Cancer. 2007;6:17