During cell spreading, mammalian cells migrate using lamellipodia formed from a large dense branched actin network which produces the protrusive force required for leading edge advancement. The formation of lamellipodia is a dynamic process and is dependent on a variety of protein cofactors that mediate their local regulation, structural characteristics and dynamics. In the present study, we show that mRNAs encoding some structural and regulatory components of the WAVE [WASP (Wiskott–Aldrich syndrome protein) verprolin homologous] complex are localized to the leading edge of the cell and associated with sites of active translation. Furthermore, we demonstrate that steady-state levels of ArpC2 and Rac1 proteins increase at the leading edge during cell spreading, suggesting that localized protein synthesis has a pivotal role in controlling cell spreading and migration.
Fibroblasts are highly migratory mesenchymal cells that mediate formation and remodelling of epithelial tissues and wound healing. As such, fibroblasts are key determinants in the outcomes of tumour progression and therefore a significant target in the development of new cancer therapies . The directional chemotaxis of mesenchymal cells can be initiated by various extracellular biochemical stimuli which include growth factors and chemokines; however, cell migration and direction can also be influenced by mechanical forces and adhesion to ECM (extracellular matrix) proteins in what still remains a poorly understood process .
The mechanism of cell migration towards external attractants comprises a cycle in which cells are stimulated to polarize by the actions of Rho GTPases  and extend protrusions formed of large broad lamellipodia, or narrow elongate filopodia, both of which are driven and formed of filamentous actin networks [3,4]. These protrusions are then stabilized by focal adhesions that link the actin cytoskeleton to the ECM . Such lamellipodia are large flattened structures formed from the large dense branched actin networks, and cells migrating on to a 2D (two-dimensional) substrate often form a spatulate or fan-like morphology with the broad lamellipodium on the leading edge tapering to the trailing edge of the cell [4,5]. Elongation of the barbed ends of actin filaments in lamellipodia and filopodia are hypothesized to produce the protrusive force required for the leading edge advancement observed during cell migration . Formation of lamellipodia is therefore dependent on a variety of protein cofactors that mediate their local regulation, structural characteristics and dynamics . These include the Rho family GTPase Rac1, which stimulates the recruitment of the WAVE complex to the plasma membrane , the constituent factors of the WAVE complex itself , and the actin nucleation Arp2/3 complex (actin-related protein 2/3 complex), including Arp2/3 subunits that promote branched actin nucleation such as ArpC2 . Arp2/3 consists of a stable complex comprising seven polypeptides; the Arp2 and Arp3 subunits (that share homology with actin) form a pseudo actin dimer that nucleates actin when in a conformation that is dependent on the presence of further ArpC2/ArpC3 subunits as well as ArpC4 and ArpC5 subunits . However, Arp2/3 is not sufficient to nucleate branched actin alone . Formation of actin structures is regulated by WASP (Wiskott–Aldrich syndrome protein) family proteins, which consist of two principal classes, WASP and WAVE (WASP verprolin homologous) proteins, both of which activate Arp2/3 to nucleate new actin filaments . WAVE proteins form the scaffold for the WAVE complex, which is composed of five subunits and is required for lamellipodium formation, activated downstream of the small GTPase Rac1. Cell signalling cascades and actinomyosin-dependent contractility promote the disassembly of adhesion at the trailing edge of the cell, and this allows the cell to progress forward .
It is known that the localization of mRNA into specific subcellular compartments is required for the establishment and maintenance of cell polarity , and the localization of actin mRNA itself has been fairly well characterized [15–17]. Actin mRNA is transported along filamentous actin within the cell by ZBP1 (Zipcode-binding protein-1) which binds to a ‘Zipcode’ comprising bipartite RNA elements in the 3′ UTR (untranslated region) of β-actin mRNA [15,16]. This localization has been shown to affect the density of dendritic filopodia and filopodia synapses . Moreover, it has been demonstrated previously that all seven constituent mRNAs of the Arp2/3 complex are compartmentalized to protrusions in fibroblasts, although it is unclear as to the nature of the protrusions, or whether they involved active protein synthesis .
Protein synthesis is carried out in three stages (initiation, elongation and termination), with the initiation stage of translation generally accepted as a major site of regulation of gene expression [19–24]. This pivotal role reflects the regulated binding of mRNA to the ribosome, facilitated by the assembly of eIFs (eukaryotic initiation factors) into a multiprotein complex known as eIF4F (eIF4E, eIF4A and eIF4G) which is often unregulated in tumour cells. In turn, the activity of this complex is regulated by both phosphorylation and the inherent structural properties of the recruited mRNA [19,20,22]. The formation of the eIF4F complex reflects the regulated availability of eIF4E to participate in initiation, a process controlled by a number of general and mRNA-specific regulatory proteins. Using a conserved motif, 4E-BPs (eIF4E-binding proteins) compete with eIF4G for a common surface on eIF4E and inhibit eIF4F assembly. The localized association of 4E-BPs with eIF4E is acutely modulated by multisite phosphorylation dependent on mTORC1 [mTOR (mammalian target of rapamycin) complex 1] signalling [21,23], effectively integrating signals from mitogens and nutrients with the translational apparatus [23,24]. mTORC1 inhibition has been demonstrated to inhibit adhesion and invasiveness of prostate cancer cells, and fibronectin-induced migration of a human arterial smooth muscle line, among others [25,26].
We have demonstrated previously that translation initiation factors localize to the leading edge of migrating fibroblasts in loci enriched with actively translating ribosomes . Furthermore, we also demonstrated that ribosomes co-localize with nascent focal adhesions in spreading cells in a cell-matrix-dependent manner . In the present study, using human lung fibroblasts, we extend the work to examine the localization of mRNA encoding both structural and regulatory components of the WAVE complex, key components in the formation of branched actin during lamellipodia formation and cytoskeleton remodelling. We have also used puromycilation  to demonstrate that these mRNAs co-localize to lamellipodial-located foci where active protein synthesis is taking place in polarized and migrating cells.
MATERIALS AND METHODS
SV40 (simian virus 40) transformed MRC5 human lung fibroblasts, were cultured in 10-cm plates in MEM (minimum essential medium)-GlutaMAX™ (Gibco®) supplemented with 15% fetal bovine serum (PAA Laboratories) at 37°C in a humidified atmosphere containing 5% CO2 as described previously .
Preparation of cell extracts
SDS/PAGE and immunoblotting
Cells at specific densities were seeded on to 15-cm plates coated with either 10 μg/ml collagen I or 20 μg/ml fibronectin and cultivated as described in the Table 1 legend. The polysome extraction was performed as follows: 100 μg/ml cycloheximide (Sigma) was added to the cells directly into the medium, and then incubated for a further 3 min at 37°C. Plates were placed on ice and washed twice with ice-cold PBS containing 100 μg/ml cycloheximide, and the cells were lysed on the plate with 400 μl of polysome lysis buffer [10 mM Tris/HCl (pH 7.8), 10 mM KCl, 15 mM MgCl2, 0.5% deoxycorticosterone, 1% (v/v) Nonidet P40, 40 mM 2-glycerophosphate, 2 mM DTT (dithiothreitol), 100 μg/ml cycloheximide and 1× EDTA-free protease inhibitor cocktail (Roche)] for 10 min and the lysate was pre-cleared by centrifugation at 10000 g for 3 min at 4°C. The supernatant was loaded on to a non-linear 10–60% sucrose gradient [50 mM Tris (pH 7.4), 150 mM KCl, 15 mM MgCl2, 2 mM DTT, 100 μg/ml cycloheximide and 1× EDTA-free protease inhibitor cocktail] and centrifuged at 38000 rev./min (256136 g) for 150 min at 4°C in a SW40 Ti rotor (Beckman). Visualization of polysome profiles and collection of 1 ml fractions from the gradients were performed using a UV–visible detector (Isco) at 254 nm.
|mRNA .||Fold change .||S.D. .||P value*** .|
|mRNA .||Fold change .||S.D. .||P value*** .|
RNA extraction from polysome fractions
To extract RNA from polysome fractions (1 ml), RNA was precipitated by adding one volume of propan-2-ol, then 150 μl of 2 M NaCl and 20 μg of glycogen (Roche) per ml of fraction. Fractions were vortex-mixed, incubated for 10 min at room temperature (20°C) and then at −20°C overnight. Precipitates were collected at 13000 g in a bench-top Microfuge for 45 min at 4°C. Pellets were washed twice in 1 ml of ice-cold 70% ethanol followed by centrifugation for 15 min at 4°C. The supernatant was removed and RNA was extracted from the pellets using Nucleospin RNA II kits (Macherey-Nagel). Quality and quantity of RNA was verified using a Nanodrop spectrophotometer.
First-strand DNA synthesis
Total cDNA libraries were prepared from RNA samples using random primers with an Improm II kit (Promega) as per the manufacturer's instructions.
QPCR (quantitative PCR)
QPCR experiments were undertaken using 40 cycles in an Mx3005P QPCR cycler (Agilent). The amount of target cDNA was quantified using KAPA SYBR FAST 2X Universal mastermix (Anachem). Template equivalent to 5 ng of RNA in the cDNA library per reaction was added to each 20 μl reaction mixture with a final primer concentration of 200 nM per reaction. Crossing thresholds were determined using MxPro software (Agilent), and the fold-difference in RNA quantity was calculated using the relative quantification method (2−ΔΔCT).
IF (immunofluorescence) analysis
Coverslips were coated with 10 μg/ml collagen I (Sigma) in 30% ethanol and incubated at 37°C until dry. For cell-spreading assays, cells were detached from their substrate with trypsin, washed and held in suspension in a shaker in complete tissue culture medium for a recovery time of 45 min. Cells (5×104) were then seeded on to each 22-mm coverslip, and the cells were recovered for fixation and staining at predetermined time points, as described previously [27,28]. Single-stain, bleed-through controls and antibody cross-reaction controls were prepared for each sample (results not shown). Costes’ approach Pearson's correlation coefficients and P values were calculated using ImageJ  (http://www.uhnresearch.ca/facilities/wcif/imagej/).
Puromycilation assay by IF
The puromycilation assay was performed as described in .
In situ hybridization
mRNA FISH (fluorescence in situ hybridization) was performed according to  with the following modifications: the 3′ ends of DNA oligomer probes were conjugated with DIG (digoxygenin) using a DIG Oligonucleotide 3′-labelling kit (Roche). Probe sequences were as follows: Rac1 probe, 5′-gggacagtggtgccgcacctcaggataccactttgcacggacattttcaaatgatgcag-3′; scrambled Rac1 probe, 5′-gtgttcgcgtacctgtgtacacaactagcggagcatgctaaactactcgagatcaggcg-3′; β-actin probe, 5′-caggactccatgcccaggaaggaaggctggaagagtgcctcagggcagcggaaccgctc-3′; scrambled β-actin probe, 5′-ctaagacaagtgcatcgggtgacgacgagccagccgagccaccgatgtgggcgagccag-3′; ArpC2 probe, 5′-ttccgggactacctgcactaccacatcaagtgctctaaggcctatattcacacacgtatg-3′; and scrambled ArpC2 probe, 5′-acgactctactacctactaccgccgaatgaactcacaattgtcttcggtgtctggcacaa-3′. WAVE1 was a mixture of 48 single-labelled probes (20-mers) conjugated to CAL Fluor Orange 560 [as per the manufacturer's instructions (Biosearch Technologies)].
The probe (1 μl containing 50 ng) was mixed with 1 μl of salmon sperm DNA (10 μg/μl; Boehringer) and 1 μl of yeast tRNA (10 μg/μl; Sigma) per sample. Two volumes of 100% ethanol were then added and the mixture was freeze-dried to dryness. Probes were then resuspended in 2× hybridization buffer [4× SSC, 20% (w/v) dextran sulfate, 2% (w/v) BSA and RNase inhibitor (2 units/μl; Fermentas)] (1× SSC is 0.15 M NaCl/0.015 M sodium citrate), and formamide to a final volume of 10 μl/sample with the percentage of formamide being calculated to adjust the probe annealing temperature to 65°C [1% (v/v) formamide=Δtemperature −0.62°C] and denatured at 74°C for 5 min. Cells were seeded at a density of 15×105 on to slides coated with collagen I as described above, and cultured overnight as described above. Unless stated otherwise, a slide rack and staining jar were used for washes and incubations. Slides were washed once in 200 ml of PBS and then fixed for 20 min in 4% paraformaldehyde/PBS. Fixed slides were washed three times in 200 ml of PBS for 5 min and then immersed in 200 ml of 0.1% Triton X-100/PBS for 5 min to permeabilize the cells, followed by a further three washes in PBS. Slides were then dehydrated with sequential 5 min incubations of 80, 95 and 100% ethanol, and then desiccated completely on a hot-block at 42°C.
The probe/hybridization mixture per sample was pipetted (10 μl) on to dehydrated slides, and each sample was overlaid with a 22-mm2 glass coverslip edge-coated with silicone sealant, and then incubated at 64°C in a humidified environment for 1 h. Coverslips were removed with a razor blade and the slides were washed three times at 64°C for 5 min in 1× SSC, containing the same percentage formamide as the hybridization buffer. Slides were then washed three times in 2× SSC at 64°C for 5 min. Slides were left in the final wash to cool to room temperature, and then placed in 4× SSC at room temperature. Slides were blocked in 3% BSA/4× SSC/0.1% Triton X-100 for 20 min, then washed in 4× SSC. Slides were then incubated with 0.3% H2O2/4× SSC to inhibit endogenous peroxidases then washed three times in 4× SSC/0.1% Triton X-100. Samples were then incubated with sheep anti-DIG-POD (Roche) diluted to 1:100 in 3% BSA/4× SSC/0.1% Triton X-100 for 1 h at room temperature, then washed three times in 4× SSC. Samples were incubated with tyromide-FITC diluted to 1:50 for 5 min using the TSA-Plus Fluorescein System (PerkinElmer) as per the manufacturer's instructions, then washed three times in 4× SSC/0.1% Triton X-100. Cells were then incubated with DAPI (4′,6-diamidino-2-phenylindole) as described above to stain nuclei and then mounted in Mowiol as described above.
Live cell imaging
Cells were cultured at 37°C in CO2-independent medium (Gibco®) supplemented with 15% (v/v) fetal bovine serum (PAA Laboratories) and 1× Glutamine (Gibco®). Video microscopy images were captured using Motic Live Imaging Module software (Motic) via a Nikon Eclipse TS-100 inverted microscope equipped with a Moticam 2000 video camera placed in an LMS tissue culture incubator at 37°C. For cell wounding experiments, cells were cultured in multi-well chamber slides (Lab-Tek) until 100% confluent. Wounds were introduced on the surface of the confluent cells using a micropipette tip. Images were captured at the rate of one frame every 10 min for 20 h. Images were converted into stacks, and cells were tracked using ImageJ (http://www.uhnresearch.ca/facilities/wcif/imagej/) equipped with the Manual Tracking plug-in. Cell tracking data were exported to Chemotaxis and Migration Tool version 2.0 (http://www.Ibidi.com) for statistical analysis. For chemotaxis experiments, cells were seeded on to Ibidi μ-slide chemotaxis slides (Thistle) as per the manufacturer's protocol, with a mean gradient concentration of 15% used to induce chemotaxis. Images were captured at the rate of one frame every 10 min for 20 h. Images were converted into stacks, and cells were tracked using ImageJ (http://www.uhnresearch.ca/facilities/wcif/imagej/) equipped with the Manual Tracking plug-in. Cell tracking data were exported to Chemotaxis and Migration tool version 2.0 (www.Ibidi.com) for statistical analysis.
Inhibition of protein synthesis or vesicle transport decreases lamellipodia formation in fibroblasts
In fibroblastic cells, Src family kinases play an essential role in integrin adhesive function and focal adhesion structures . This serves to establish and maintain the mechano-tension necessary for the stabilization and recruitment of new subunits to assembled focal adhesions . Previously, we have shown the association of translation initiation factors and ribosomes with subcellular membranes, with fibroblastic cell spreading sensitive to inhibition of intracellular vesicle transport by BFA (Brefeldin A) . As focal adhesion integrity is essential for the maintenance of cell protrusions such as lamellipodia , we have used the Src inhibitor Src1-I to disassemble these structures in fully spread polarized cells (Figure 1). To monitor the assembly of focal adhesions, cells were subsequently released from Src inhibition by replenishment with replacement growth media, in the absence or presence of emetine (an inhibitor of translation elongation) or BFA and visualized by IF. As shown in Figure 1, treatment of cells with Src1-I resulted in the withdrawal of lamellipodium-like structures (Figure 1B compared with Figure 1A). However, release of cells into complete growth medium resulted in the rapid formation of new focal adhesions and lamellipodia (Figure 1C compared with Figure 1B). Using this model system of cell release into either BFA which interferes with vesicle transport (Figure 1E) or emetine at a level which inhibits protein synthesis (Figure 1D), resulted in a failure to assemble new lamellipodial structures.
Inhibition of protein synthesis prevents lamellipodia formation in MRC5 fibroblasts
Inhibition of mTOR signalling affects cell migration velocity and distance
The association of 4E-BPs with eIF4E is acutely modulated by multisite phosphorylation dependent on mTORC1 signalling [19–24], with mTORC1 inhibition shown previously to inhibit protein synthesis [33–35], adhesion, invasiveness and migration of human cells [25,26]. The rapamycin analogue RAD001 (everolimus) inhibits the mTORC1 complex with a high degree of specificity, thus leading to translation inhibition via 4E-BP1 dephosphorylation . The mTOR catalytic site inhibitor, Ku-0063794, inhibits translation more efficiently than RAD001 , which might reflect that Ku-0063794 also affects mTORC2 activity [23,24].
To investigate the effect of inhibition of mTORC1 and mTORC1/2 on migrating cells, cells were seeded on to Boyden chambers, starved of growth factors for 2 h, and then allowed to migrate in the absence or presence of a growth factor gradient supplied by fetal bovine serum. Individual cells were also spatially tracked over 20 h in the presence of 10 μM RAD001 or Ku-0063794, as indicated, concentrations that inhibited mTORC1  or mTORC1/2 signalling  respectively, and impaired protein synthesis (results not shown). The resulting data were analysed statistically (n=25) (Figure 2A) showing that inhibition of mTORC1 signalling (and protein synthesis) had a profound effect on cell movement. The mean square displacement of the data (Figure 2B) indicate that the directional movement of cells incubated with Ku-0063794 (×) or RAD001 (▲) were equally inhibited when compared with cells incubated with chemo-attractant only (■). This inhibition of movement was observed from the earliest times investigated, was consistent throughout the time course (i.e. not due to a general long-term inhibition of mTORC2), and similar to that observed following release of arrested cells into emetine (Figure 1, and see Figure 5C later). Relative to control cells, there was a significant impairment of migration velocity (distance travelled/time) for both the Ku-0063794- and RAD001-treated cells (Figure 2C). These data are consistent with the approximately equal reduction observed in both accumulated (Figure 2D) and Euclidean migration distances (Figure 2E) following inhibition of mTOR signalling and protein synthesis. Interestingly, migration straightness (accumulated distance/Euclidean distance) was significantly higher for the inhibitor-treated cells, suggesting a role for mTOR/protein synthesis in steering/directionality during fibroblast migration (Figure 2F). As RAD001-inhibited cell migration to the same extent as Ku-0063794-treated cells, these data indicate mTORC1, but not mTORC2, as a central player in controlling cell migration and steering. Although mTORC1 is well known to regulate protein synthesis, given its multiple roles in cell function, other processes downstream of mTORC1 may also be involved here.
Inhibition of protein synthesis using mTOR inhibitors results in decreased migration velocity and distance in MRC5 fibroblasts
Inhibition of mTORC1 results in an impairment of wound healing and lamellipodium morphology
To further investigate the effect of inhibition of mTOR signalling on fibroblast migration and lamellipodium morphology, cells were grown to a confluent monolayer which was then wounded by scratching with a micropipette tip. The cells were allowed to migrate into the wound without (Figure 3A, upper panels) or with inhibition of mTORC1/2 signalling (and protein synthesis ) using either Ku-0063794 (Figure 3A, lower panels), or with inhibition of mTORC1 using RAD001 (Figure 3B). Individual cell migration was tracked, and the area of the wounds covered after 20 h of cell migration was quantified using ROIs (region of interest) created with MBF ImageJ. In cells where protein synthesis was inhibited (in the presence of Ku-0063794 ), they failed to form lamellipodia when migrating into the wound; instead, the cells exhibited an irregular morphology (Figure 3A, lower panels), and only achieved 40% of the wound coverage of the untreated cells (Figure 3B). In agreement with the data shown in Figure 2, less migration was observed with mTORC1/2 (Figure 3C) or mTORC1 (RAD001; Figure 3B) inhibition than observed with control cells.
mTOR inhibition during wound healing confers an irregular morphology consistent with defective lamellipodium formation
Recruitment of ribosomes to Rho family mRNAs during cell spreading
To induce assembly of new lamellipodia-like structures in bulk culture, we enzymatically detached cells from their substrate, kept them in suspension until their cytoskeletons disassembled  and then allowed them to spread on a collagen-coated substrate for 30 min . To monitor global rates of protein synthesis, cells were pulse-labelled with [35S]methionine for 30 min while in suspension, confluent on collagen I or spreading on collagen I. As shown in Supplementary Figure S1 (at http://www.biochemj.org/bj/452/bj4520045add.htm), there was no significant change in the global rates of protein synthesis under such conditions. As reported previously, confocal microscopy confirmed that ribosomes active in translation were found in lamellopodia  (Supplementary Figure S2 at http://www.biochemj.org/bj/452/bj4520045add.htm). Extracts were prepared after 30 min of cell spreading and polysomes were isolated by sucrose density-gradient centrifugation as described in the Materials and methods section. Polysome-enriched fractions were collected and the mRNA was purified and used for first-strand synthesis, as described in the Material and methods section. Using QPCR, the resulting cDNA libraries were compared with those collected from confluent cells, which lack extensive lamellipodia (Supplementary Figure S2). As summarized in Table 1, there was a modest 2-fold induction in the recruitment of Cdc42 (cell division cycle 42) mRNA to ribosomes during lamellipodia formation, with a smaller change in the total levels of mRNA (Supplementary Table S1 at http://www.biochemj.org/bj/452/bj4520045add.htm). This marginally exceeds the 1.3-fold increase in total cellular Cdc42 mRNA. A similar result was observed with the preferential recruitment of Rac1 mRNA (Table 1 and Supplementary Table S1). Such a small increase in both total and polysome-associated mRNAs suggests that the latter probably reflects changes in the cellular level of the mRNA as a result of increased transcription and/or nuclear export during lamellipodia formation. However, with RhoA, there was a 5.5-fold increase in RhoA mRNA recruited to polysomes (Table 1), with smaller changes in levels of total mRNA (Supplementary Table S1).
Selective ribosomal recruitment of mRNA encoding WAVE regulatory protein during cell spreading
To investigate a role for selective localized protein synthesis during cell spreading, we have also used these conditions to analyse the association of a number of mRNAs encoding WAVE complex recruitment factors to ribosomes (Table 1 and Supplementary Table S1). PIP5K1a (PtdIns4P 5-kinase type-1 α) catalyses the phosphorylation of PtdIns4P to form the second messenger, PtdIns(4,5)P2, the precursor to PtdIns(3,4,5)P3. PtdIns(4,5)P2 is required for the formation of membrane ruffles, actin re-organization and focal adhesion formation during directional cell migration in response to integrin activation . IRSp53 (insulin receptor tyrosine kinase substrate p53) is implicated in the recruitment and clustering of WAVE complexes to the plasma membrane and therefore the formation of lamellipodia [39,40]. Moreover, PtdIns(3,4,5)P3 and IRSp53 combine to synergize with Rac1 to activate the WAVE complex at the plasma membrane . Our QPCR analysis (Table 1) showed a 6.4-fold increase in the recruitment of PIP5K1a mRNA to ribosomes. In addition, there was a 2.6-fold increase in the recruitment of IRSp53 mRNA to polysomes.
Brk1 has been reported to exist in a free cytoplasmic pool in the form of a homotrimeric precursor. Following the induction of WAVE complex formation, these trimers dissociate and single Brk1 subunits associate with newly synthesized WAVE subunits to facilitate the assembly of a functional WAVE complex [41,42]. Interestingly, QPCR analysis of polysome-derived mRNA from spreading cells showed a large (58.7-fold) stimulation of Brk1 mRNA recruitment of ribosomes following induction of lamellipodium assembly (Table 1). The large apparent induction of Brk1 mRNA translation could be hypothesized as a mechanism to support the maintenance of a large local critical concentration of Brk1 trimers to facilitate the assembly of active WAVE complexes.
The WCA domain of activated WAVE complex presents ATP-loaded G-actin (globular actin) monomers to the Arp2/3 complex, thus facilitating polymerization of the actin filament . The Arp2/3 complex initiates the formation of actin daughter filaments [44,45]; the ArpC2 subunit as part of the structural core of the complex is thought to make substantial contact with the mother filament. The WAVE complex also contains Abl interactor Abi1, Nck-associated protein Nap1, and specifically Rac1-associated protein Sra-1, where Abi1 plays a central role in holding the complex together, and Sra provides a binding site for active Rac. In addition, interaction of the Abi1 protein with the p85 regulatory subunit of PI3K (phosphoinositide 3-kinase) represents the link between growth receptor signalling and actin cytoskeleton remodelling . This highly regulated WAVE complex is maintained in its inactive state in the cytosol due to Nap1, which blocks access to the WCA domain, thus inhibiting Arp2/3 activity . This inhibition is released by an allosteric mechanism following the interaction of the WAVE complex with activated Rac1 . Sra-1 (also known as CYFIP), functions to protect the WAVE protein from degradation . As the proteins are central to regulating the WAVE complex, we have also investigated a role for their selective localized protein synthesis during cell spreading. Table 1 shows that in spreading cells, Nap1 mRNA shows a 31-fold increase in association with polysomes, WAVE mRNA shows a 12.4-fold increase, Sra-1 shows a 13-fold increase and Abi1 shows a 2.6-fold increase. Similarly, we see a 31-fold increase in association of ArpC2 mRNA with ribosomes during lamellipodia assembly. Further studies showed that during cell spreading, β-actin mRNA (Table 1 and Supplementary Table S1), actinin and filamin mRNAs, and a large number of other mRNAs encoding proteins related to cytoskeletal re-organization were also recruited on to polysomes during cell spreading. These data are consistent with the localized translation of WAVE and regulatory proteins during lamellipodium assembly following the induction of cell spreading.
Rac1, ArpC2 and β-actin mRNAs co-localize with actively translating ribosomes on lamellipodia
To analyse selective localized translation directly, we have used a combination of mRNA FISH, puromycilation and IF . We chose to investigate Rac1 as the primary signalling molecule that stimulates the formation of branched actin, the ArpC2 subunit of the Arp2/3 complex and β-actin. Using a combination of puromycilation  and IF microscopy using antisera raised against the 40S ribosomal subunit protein S6, rpS6, we have shown that the lamellipodia of migrating fibroblasts are enriched for foci active in protein synthesis (Figure 4A). Furthermore, using FISH, we have shown that ArpC2 mRNA co-localizes with both 40S ribosomes (Figure 4B) and translationally active foci (Figure 4C). Statistical analysis of these data using Pearson co-localization coefficients (Robs) indicate these events to be highly significant (n=200). Moreover, when we used FISH in combination with IF microscopy, we found that ArpC2 protein also co-localized with ArpC2 mRNA in lamellipodia with a significant Pearson's co-localization coefficient (Figure 4D). In addition to WAVE1 mRNA (Figure 4G), these translationally active foci are also enriched with Rac1 and β-actin mRNA (Figures 4E and 4F respectively).
Rac1, ArpC2 and β-actin mRNAs co-localize with actively translating ribosomes on lamellipodia
Steady-state levels of ArpC2 and Rac1 protein increase during lamellipodium formation
To investigate the role of localized synthesis of ArpC2 in lamellipodium formation, we used RNAi (RNA interference) to deplete the level of cellular ArpC2 protein (Figure 5A). ArpC2 siRNA (small interfering DNA)-treated cells exhibited an increased incidence in membrane blebbing compared with the untreated control, consistent with a decrease in the formation of lamellipodium structures (Figure 5D compared with Figure 5B). More extensive membrane blebbing was observed following incubation of MRC5 cells with emetine (Figure 5C compared with Figure 5B). Preventing the recruitment of translation factors to the lamellipodia using short-term treatment with BFA [27,28] revealed a highly blebbed membrane morphology alongside a complete abrogation of lamellipodium formation (Figure 5E), similar to that observed with RNAi depletion of ArpC2. These data might suggest that the localized de novo synthesis of Arp2/3 complex plays a role in regulating lamellipodial structures observed on the leading edge of migrating cells.
Depletion of ArpC2 results in reduced levels of lamellopodia and an increase in membrane blebbing
We have used SDS/PAGE (15% gel) and Western blotting to investigate whether the steady-state levels of ArpC2 and Rac1 protein increase during cell spreading. Attempts to monitor the synthesis of these proteins by pulse-labelling with radioactive methionine and immunoprecipitation did not yield conclusive data owing to limitations with the commercial antisera (results not shown). Cells were detached from their substrate and kept in suspension for 45 min to allow disassembly of their cytoskeleton. The cells were allowed to spread on a collagen I or fibronectin-coated substrate for 1 h, allowing the formation of extensive lamellipodia . As shown in Figure 6(A) (and quantified in Figure 6B), cell spreading led to a 1.8-fold increase in total Rac1 and, to a lesser extent, ArpC2 protein when there was an increase in recruitment of ArpC2 mRNA to polysomes (Table 1). Similarly, there was a 3-fold increase in total levels of Rac1 protein with cells spread on collagen I. The relatively lower fold-change in the steady-state levels of ArpC2 protein compared with the recruitment of ArpC2 mRNA to ribosomes (31-fold) may reflect a large background amount of pre-existing ArpC2 protein compared with the steady-state ribosome loading of ArpC2 mRNA.
Steady-state levels of ArpC2 and Rac1 proteins increase during cell spreading
Cell migration comprises a cycle in which cells are stimulated to polarize by the actions of Rho GTPases  and extend protrusions formed of large broad lamellipodia, driven and formed of filamentous actin networks [3,4,43–45,47]. Formation of lamellipodia are dependent on a variety of protein cofactors that mediate their local regulation, structural characteristics and dynamics [7,43–45,47]. The WAVE–Arp2/3-mediated assembly of a functional lamellipodium is dependent on the activity of adaptor proteins such as Nck, which mediate not only the membrane-associated assembly and activation of WAVE–Arp2/3, but also the clustering of the complexes to achieve the required actin network .
The role for localized protein synthesis in regulating lamellopodia formation and cell migration has not been resolved. The most characterized example of regulated localized mRNA translation is provided by β-actin mRNA. Localization of actin mRNA is mediated by ZBP1 which binds to a Zipcode comprising bipartite RNA elements in the 3′ UTR of β-actin mRNA and is removed following localized src signalling to allow translation at focal adhesions [15–17]. Recent work has shown that binding of ZBP1 to the mRNAs encoding E-cadherin, β-actin, α-actinin and the Arp2/3 complex facilitates localization of the mRNAs, which stabilizes cell–cell connections and focal adhesions . In agreement with such a model, we have demonstrated previously an ECM-sensitive recruitment of ribosomes to focal adhesions during cell spreading , suggesting that the localized synthesis of focal adhesion components during cell spreading. In addition to src, mTOR signalling, which impinges on the regulation of initiation factor activity [19–23] and localization [25,26], has also been shown to modulate actin dynamics. The latter occurs via p70RSK (p70 ribosomal S6 kinase), an effector kinase downstream of mTORC1, possibly mediated through activation of Rac1. Moreover, local mTOR activation has been shown to modulate spreading [25,26] and lamellipodium formation in platelets through activation of Rac1 . Our data, which shows that inhibition of mTOR signalling and protein synthesis affects cell migration velocity and distance (Figure 2), as well as impairing wound-healing and lamellipodium morphology (Figure 3), are consistent with a pivotal role for mTOR signalling in regulating actin dynamics.
In addition to localized signalling, the recruitment of mRNA into specific subcellular compartments is required for the establishment and maintenance of cell polarity . Localization of actin mRNA using ZBP1 in neuronal cells has been shown to affect the density of dendritic filopodia and filopodia synapses [15–17]. In the present study with MRC5 cells, we show that β-actin mRNA is similarly localized to active foci on the leading edge of the cell and is furthermore recruited to polysomes. This might reflect a top-up mechanism as there is a large reservoir of available soluble actin monomers within the cytoplasm. In addition, we show that mRNA encoding proteins involved in actin branching and bundling, such as filamin and α-actinin (Table 1), are also recruited to polysomes during cell spreading. These data suggest that localized protein synthesis may help modulate actin levels and polymerized structures. Consistent with this is the finding that all seven constituent mRNAs of the Arp2/3 complex are compartmentalized to protrusions in fibroblasts, although it is unclear as to the nature of the protrusions, or whether they are involved active protein synthesis . However, our observations show that lamellipodia structures rapidly collapse following inhibition of protein synthesis (Figure 5). Our data have shown that mRNA encoding WAVE–Arp2/3-associated proteins is co-localized with foci enriched with active protein synthesis at the leading edge during cell migration (Figure 4). In addition, we show that the steady-state levels of ArpC2 and Rac1 protein increase during lamellipodium formation in MRC5 cells (Figure 6). This could provide an additional layer of regulation on the assembly of these structures where local de novo synthesis in response to key signalling events could act as a trigger for WAVE activation and complex clustering. Bio-informatic analysis of ArpC2 mRNA alone reveals a potentially complex regulatory structure. The 5′ UTR contains a number of interesting motifs which may impinge on translation: a putative G-quadruplex structure; an FMRP (fragile X mental retardation protein)-binding motif implicated in the negative regulation of translation ; four large stem-loop structures reminiscent of a cellular IRES (internal ribosome entry site) ; a putative GAIT (interferon-γ-activated inhibitor of translation) domain involved in reversible selective translation control ; a putative SECIS (selenocysteine insertion sequence) domain involved in the recoding of the translational stop codon in response to selenium availability [55,56]; and a short (270 nt) upstream open reading frame which is located immediately on the 3′ end of the fourth putative stem-loop/SECIS domain. Further work is required to determine which, if any, of these motifs contribute to the localized translation of ArpC2 mRNA.
mTOR signalling is also likely to play a key role in modulating actin dynamics, cell spreading and lamellipodium formation in MRC5 cells (Figure 2). In platelets, this is mediated by p70RSK, an effector kinase downstream of mTORC1, possible through the activation of Rac1 . Interestingly, Sra-1, which is a subunit of the WAVE and protects WAVE protein from degradation , has been shown to be a 4E-BP in the brain where it represses the translation of FMRP mRNA until it reaches the synapse . This discovery of Sra-1 as an mRNP (messenger ribonucleoprotein particle)-binding partner presents an attractive hypothesis where it might also be involved in the recruitment and translational repression of WAVE complex or related mRNA to lamellipodia.
Cell spreading is also associated with an increase in the association of PIP5K1a mRNA with ribosomes (Table 1), suggesting that localized protein synthesis could modulate PtdIns(3,4,5)P3 levels, facilitating mTOR activation and WAVE complex clustering. Brk1, a modulator of the WAVE complex [41,42], has also been investigated in the present study. Our data show that there is 58-fold stimulation in Brk1 mRNA recruitment of ribosomes on induction of lamellipodium assembly, possibly to support the maintenance of Brk1 to facilitate the assembly of active WAVE complexes. In spreading cells, we also see a 31-fold increase and a 12.4-fold increase in the association of NAP-1 mRNA and WAVE1 mRNA with ribosomes respectively. Similarly, we see a 31.3-fold increase in association of ArpC2 mRNA with ribosomes during lamellipodia assembly (Table 1), with little change in total mRNA levels (Supplementary Table S1). Furthermore, studies have shown that depletion of Scar/WAVE3 in MDA-MB-231 cells results in larger and less dynamic lamellipodia, with cells moving less rapidly and typically only showing one lamellipod . These data are consistent with localized translation of WAVE protein and regulatory proteins during lamellipodium assembly following the induction of cell spreading. One possible scenario that still needs to be addressed is that membrane protrusions are maintained by de novo protein synthesis, and that WAVE–Arp2/3 is stabilized by the arrival of Sra-1-associated mRNAs.
complex, actin-related protein 2/3 complex
cell division cycle 42
eukaryotic initiation factor
fluorescence in situ hybridization
fragile X mental retardation protein
insulin receptor tyrosine kinase substrate p53
mammalian target of rapamycin
mTOR complex 1
p70 ribosomal S6 kinase
PtdIns4P 5-kinase type-1 α
ribosomal protein S6
small interfering RNA
Wiskott–Aldrich syndrome protein
WASP verprolin homologous
Mark Willett and Michele Brocard carried out the experimental work and provided expert input into the design and interpretation of data; Hilary Pollard carried out the QPCR analysis; Simon Morley supervised the work, generated the paper and secured funding for the work.
We thank Jonathan W. Yewdell [Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH), Bethesda, MD 20892, U.S.A.] for the reagents used in this work.
This work was supported by the Biotechnology and Biological Sciences Research Council (BBSRC) U.K. [grant number BB/H018956].