Association of syntenin-1 with M-RIP polarizes Rac-1 activation during chemotaxis and immune interactions

T cell migration and activation in response to antigen are characterized by morphological changes dictated by the subcellular redistribution of a number of molecules, including chemokine receptors, adhesion molecules, signaling effectors and cytoskeletal components (Vicente-Manzanares and Sanchez-Madrid, 2004). During T cell migration, the activation of integrins and chemokine receptors triggers many signaling pathways, including the activation of Rho GTPases, which actively induce actin polymerization and reorganization. This results in the acquisition of a polarized morphology (Vicente-Manzanares and Sanchez-Madrid, 2004) with a leading edge and a trailing edge, which is characterized by a pseudopod-like projection, the uropod (Sanchez-Madrid and Serrador, 2009). Similarly, antigen-driven T cell activation involves the formation of a highly organized asymmetric complex of surface receptors and intracellular signaling molecules at the cell-to-cell contact, known as the immunological synapse (IS) (Dustin, 2009). The formation of the IS involves a rapid reorientation of the T cell cytoskeleton, including the formation of a tight structure of polymerized actin at the interface between the T cell and the antigen-presenting cell (APC), clustering of CD3, and relocalization of the microtubule-organizing center (MTOC) towards the contact area (Vicente-Manzanares and Sanchez-Madrid, 2004). This coordinated polarized reorganization is central to the maturation of the immune synapse and subsequent T cell activation (Martín-Cófreces et al., 2008; Dustin, 2009).

Chemotaxis and activation of T cells are both accompanied by polarization of the plasma membrane receptors, cytoskeletal components and intracellular signaling molecules that sense the extracellular chemotactic gradient (Gomez-Mouton and Manes, 2007; Kolsch et al., 2008). Actin dynamics is mainly controlled by signals from Rho GTPases, in which Rac activation leads to branched actin polymerization that is necessary for protrusion of the leading lamella, whereas myosin-dependent contraction is induced by Rho-A-derived signals (Vicente-Manzanares and Sanchez-Madrid, 2004). During antigen recognition, there is a dynamic balance between F-actin polymerization at the peripheral supramolecular activation cluster (p-SMAC) (Dustin, 2009) and myosin-driven waves of contraction that are necessary for the treadmilling of the F-actin filaments towards the center of the synapse, which powers the coalescence of T cell receptor (TCR) microclusters (Ilani et al., 2009). This fine-tuned actin dynamics is essential for the sustained signaling necessary for complete T cell activation.

Although many signals involved in T cell migration and activation have been described, the molecular machinery that organizes signaling molecules and their architecture in response to activation is not completely understood (Shaw and Filbert, 2009). Adaptor proteins containing PDZ domains (named for their occurrence in postsynaptic density protein, Drosophila Discs large and zonula occludens-1) are spatially segregated during T cell migration and formation of the IS (Ludford-Menting et al., 2005). PDZ domains have emerged as central organizers of protein complexes at the plasma membrane, linking extracellular signals to the cytoskeleton and intracellular signaling pathways.

Syntenin-1 (also known as MDA-9) is a cytosolic PDZ-containing protein that was originally identified as a molecule linking syndecan-mediated signaling to the cytoskeleton, and which has since been found to bind many other membrane receptors (Beekman and Coffer, 2008; Sarkar et al., 2008). Syntenin-1 contains two tandemly arranged PDZ domains preceded by a 122-residue N-terminal fragment of unknown structure (Grootjans et al., 1997). Phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P₂ or PIP2] contributes to the targeting of syntenin-1 to the plasma membrane through its PDZ domains, thereby regulating its availability to associate with membrane receptors (Zimmermann et al., 2005). Syntenin-1 also associates with the tyrosine kinase Src and the cytosolic actin-binding protein Merlin (Kang et al., 2003; Boukerche et al., 2008). Syntenin-1 plays an important role in cellular processes such as intracellular trafficking and cell surface targeting, transcription factor activation, cancer metastasis, neuronal synaptic transmission and axonal outgrowth (Sarkar et al., 2008).

We have studied the role of syntenin-1 in T cell chemotaxis and in the formation of immune conjugates. Our data reveal an unexpected role for this adaptor protein in controlling the targeting of the small GTPase Rac-1 to the plasma membrane. Although the subcellular localization of syntenin-1 is regulated by its PDZ domains, we have defined an ITIM-like motif in syntenin-1 that can be phosphorylated by Src and regulates syntenin-1 binding to myosin phosphatase Rho interacting protein (M-RIP), thereby regulating actin polymerization induced by extracellular cues. Syntenin-1 thus acts as a limiting factor in the signaling cascade linking Src to Rac-1 activation, which controls the asymmetric activation of this pathway.

To address the role of syntenin-1 in regulating T cell polarization, we first studied its subcellular localization in lymphocytes. Endogenous syntenin-1 is distributed evenly in non-stimulated primary peripheral blood lymphocytes (PBLs). Upon stimulation with the chemokine CXCL12/SDF-1α, it polarizes to dense clusters at the uropod, showing partial colocalization with uropod markers such as ICAM-3, and to the leading edge with more diffuse staining (Fig. 1A). In polarized T cell lines, syntenin-1 was also found at the uropod, and upon SDF-1α stimulation, partially relocalized to the leading edge within 30 seconds (supplementary material Fig. S1A). Time-lapse confocal microscopy of GFP–syntenin-1 in constitutively polarized T-lymphoblastic cell lines showed accumulation at the leading edge and uropod, sustained throughout cell migration (Fig. 1B and supplementary material Movie 1). Redistribution of syntenin-1 was highly active at the leading edge, which correlated with protrusive activity. This bipolar distribution was similar to that found with a GFP-tagged pleckstrin homology domain of PLCγ that binds PIP2, but different from that observed with other control GFP proteins such as GFP, farnesylated GFP or a different pleckstrin homology domain, BTK–PH–GFP (supplementary material Fig. S1B).

We next examined the role of this adaptor protein in T cell chemotaxis by suppressing its expression in T cells with two independent siRNA oligonucleotides (supplementary material Fig. S1C). Efficiency of endogenous protein silencing was assessed in every experiment and was greater than 60%. Chemotaxis of primary T lymphoblasts (Fig. 2A) or leukemic CEM T cells (Fig. 2B,C), towards CXCL12/SDF-1α was significantly impaired by both siRNA oligonucleotides. This inhibition could be also observed using a different chemokine such as CCL19/MIP-3β and in basal random migration in the absence of chemotactic stimulus (Fig. 2C). The chemokine SDF-1α induced a rapid increase in polymerized actin in primary T lymphoblasts (Fig. 2D). Knockdown of syntenin-1 almost completely abrogated the SDF-1α-induced actin polymerization response, as determined by flow cytometry in both T cells and primary T lymphoblasts (Fig. 2D,E). The magnitude and progression of actin polymerization induced by the chemokine differs in leukemic cell lines compared with primary PBLs, because primary PBLs display very low basal levels of F-actin.

Moreover, overexpression of GFP- or HA-tagged syntenin-1 significantly increased chemotaxis (Fig. 2F) and enhanced chemokine-induced actin polymerization (Fig. 2G), suggesting that syntenin-1 acts as a limiting factor in chemokine-induced signaling and migration.

The small Rho GTPase Rac is essential for chemokine-induced actin polymerization at the leading edge (Vicente-Manzanares et al., 2005), and is therefore a candidate syntenin-1 effector. SDF-1α stimulation of control oligonucleotide-transfected cells triggered an increase in GTP-loaded Rac that was abrogated in syntenin-1-depleted cells (Fig. 3A). Confocal analysis in primary T lymphoblasts detected a diffuse staining for Rac-1, which, upon SDF-1α stimulation, became polarized and colocalized with endogenous syntenin-1 at the leading edge (Fig. 3B,C). By contrast, SDF-1α stimulation of syntenin-1-depleted cells did not relocate Rac-1 to the leading edge (Fig. 3B,C). In cell fractionation experiments, translocation of Rac-1 to the membrane fraction was impaired by stimulation with SDF-1α after knockdown of syntenin-1 (Fig. 3D).

Rac activity lies downstream of Src activation in several cellular contexts, including cell migration and chemotaxis (Ouyang et al., 2008; Sai et al., 2008), and direct association of syntenin-1 with Src kinase has been recently reported in melanoma cells (Boukerche et al., 2008). Basal tyrosine phosphorylation of syntenin-1 was detected in syntenin-1-immunoprecipitated lysates (Fig. 4A), which was rapidly and transiently increased by stimulation with SDF-1α (peaking between 30 seconds and 1 minute; Fig. 4A). Treatment of T cells with the Src family kinase (SFK) inhibitor PP2, selectively inhibited syntenin-1 phosphorylation (Fig. 4B), indicating that SFKs are required for this process.

To directly determine whether syntenin-1 regulates Rac-1 activation downstream of Src, we examined the four potentially phosphorylatable Tyr residues contained in the syntenin-1 N-terminal domain (Beekman and Coffer, 2008). These residues occur in sequences consistent with Tyr-based domains described for functional immunoreceptors and are prototypical target sequences for Src family tyrosine kinases: one ITAM motif [YxxI/Lx(6–12)YxxI/L], which includes an additional hemi-ITAM, and one ITIM motif (S/I/V/LxYxxI/V/L) (Fig. 4C) (Beekman and Coffer, 2008). We generated mutant versions of GFP–syntenin-1, in which these tyrosine residues were substituted by phenylalanine: Y3, in which the three tyrosine residues included in the ITAM-like domain were mutated, and Y1, which had a mutation in the single Tyr in the ITIM-like domain (Fig. 4C). When expressed on polarized CEM T-cells, these mutants showed a distribution that was similar to that of GFP-tagged wild-type syntenin-1 during SDF-1α-induced migration (Fig. 4D and supplementary material Fig. S1D and Movie 2). When we analyzed the phosphorylation of the mutants upon treatment with SDF-1α, we observed that Y1 mutation almost completely abrogated syntenin-1 phosphorylation, suggesting that Tyr4 in the ITIM-like motif is the main substrate residue for Src kinases in syntenin-1 (Fig. 4E).

Expression of the Y3 mutant increased the chemotactic response to SDF-1α, although to a lesser extent than overexpression of wild-type syntenin-1 (Fig. 5A). By contrast, the Y1 mutation completely abolished the enhancing effect on cell migration, with Y1-transfected cells migrating similarly to control GFP-transfected cells (Fig. 5A). The effects of the syntenin-1 Tyr to Phe substitution on the ITIM-like motif on T cell chemotaxis were related to an effect on SDF-1α-induced actin polymerization. Overexpression of the Y1 mutant did not enhance F-actin polymerization to levels above that in control cells transfected with GFP upon treatment with SDF-1α (Fig. 5B). Moreover, Y1 overexpression in syntenin-1-depleted cells did not rescue the inhibition of chemokine-induced actin polymerization produced by silencing of syntenin-1 (Fig. 5C).

To directly determine whether syntenin-1 phosphorylation affects Rac-1 activation, we analyzed antibody-stained GTP-Rac by flow cytometry after SDF-1α stimulation. GTP-Rac staining increased with time in SDF-1α-stimulated GFP-transfected cells (Fig. 5D). This effect was greatly enhanced in cells expressing GFP–syntenin-1 and the Y3 mutant. Rac-1 activation was inhibited to below control levels in Y1-transfected cells (Fig. 5D). In classical pull-down experiments, the Y1 mutant was ineffective in the rescue of Rac activation after syntenin-1 depletion (Fig. 5E), whereas transfection of wild-type syntenin-1 increased Rac activity in both resting and stimulated conditions. Together, our results suggest that syntenin-1, through a mechanism requiring phosphorylation of its ITIM motif, regulates Src-induced Rac-dependent actin polymerization during T cell migration.

To investigate how Src-phosphorylated syntenin-1 mediates Rac-1 regulation, we conducted a proteomic screen for proteins that bind wild-type GFP–syntenin-1 but not the Y1 mutant. SDF-1α-stimulated J77 T cells transiently expressing GFP–syntenin-1 or GFP–Y1-mutant were immunoprecipitated with anti-GFP antibody. Proteomic comparisons identified specific binding of GFP–syntenin-1 to regulators of Rho-GTPases, including Rho-GAP, obscurin (a Rho GEF) and myosin phosphatase Rho interacting protein (M-RIP) (supplementary material Table S1). M-RIP, which is also called p116^RIP is a multidomain scaffold protein that binds to Rho-A and the myosin-binding subunit (MBS) of myosin light chain phosphatase (Surks et al., 2003), inhibiting Rho-A stimulated contractility (Gebbink et al., 1997; Koga and Ikebe, 2005). The interaction of wild-type syntenin-1 with M-RIP was confirmed by immunoblotting of GFP immunoprecipitates of GFP–syntenin-1 (Fig. 6A). Binding of M-RIP to the Y1 mutant could not be detected. Inducible association of endogenous syntenin-1 was observed in immunoprecipitates of endogenous M-RIP (Fig. 6B), which showed that this interaction peaked at 1–5 minutes of SDF-1α stimulation. We found that endogenous M-RIP had a partially submembranous distribution (supplementary material Fig. S1E) that relocalized to the leading edge upon SDF-1α stimulation, where it colocalized with Rac-1 and F-actin (Fig. 6C). Pull-down assays with GST-fusion proteins of different M-RIP domains showed that syntenin-1 binds the second coiled-coil domain (residues 728–878), overlapping with the reported binding site for MBS (Surks et al., 2003) (Fig. 6D and supplementary material Fig. S1F). Although M-RIP binding to RhoA through an adjacent RBD site has been reported (Surks et al., 2003), the possible relationship with Rac-1 GTPase has not been explored. We detected binding of Rac-1 to the first N-terminal half of M-RIP (residues 1–528), comprising the two PH domains (Fig. 6D).

To demonstrate that M-RIP provides a molecular link between Src-mediated syntenin-1 phosphorylation and the downstream regulation of Rac, we suppressed the expression of this adaptor protein in T cells by RNAi (Fig. 6E) and quantified the Rac-GTP content (Fig. 6F) after stimulation with SDF-1α. Silencing of M-RIP inhibited Rac activation (Fig. 6F) and actin polymerization after SDF-1α stimulation (Fig. 6G). The same pattern of inhibition was observed in chemotaxis experiments in both cell lines and primary T lymphoblasts (Fig. 6H,I). No additive effect was observed in any of these functional readouts when both syntenin-1 and M-RIP molecules were simultaneously silenced, further demonstrating that they act on the same signaling pathway.

Rac-1 is activated during antigen presentation (Arrieumerlou et al., 2000; Yu et al., 2001), a context in which dynamic asymmetric actin polymerization is also crucial for proper functioning of the immune synapse (Billadeau et al., 2007; Gordón-Alonso et al., 2010). However, Src activation occurs early after TCR stimulation (Rudd, 1990). We therefore assessed whether the syntenin-1–M-RIP axis could be also involved in TCR-induced activation of Rac. We first observed that both endogenous syntenin-1 and M-RIP polarized to the contact area in cognate conjugates of T cells and APCs (Fig. 7A,B and supplementary material Movie 3). Knockdown of either syntenin-1 or M-RIP, impaired actin polymerization and Rac activation in response to TCR crosslinking (Fig. 7C,D and supplementary material Fig. S2A). Moreover, syntenin-1 knockdown reduced Rac translocation to the plasma membrane (supplementary material Fig. S2B). Overexpression of GFP-tagged syntenin-1 was able to increase TCR-stimulated actin polymerization, and this effect was abolished by mutation of the tyrosine residue in the ITIM-like motif (supplementary material Fig. S2C).

When we analyzed the conjugates formed by syntenin-1-depleted cells with superantigen-loaded APCs, we observed no effect on the number of conjugates formed or on the translocation of the MTOC to the intercellular contact (supplementary material Fig. S2D). However, antigen-driven actin polymerization at the IS was locally reduced (supplementary material Fig. S2E), and as a consequence, CD3 congregation at the central SMAC was significantly impaired (Fig. 7E,F). When signals derived from antigen stimulation were analyzed, lower levels of phosphorylation of Vav-1, PLC-γ and ERK1/2 were detected in syntenin-1-silenced cells that were exposed to superantigen-loaded B cells at late time points of activation (10 minutes, Fig. 7G and supplementary material Fig. S2F).

This study reports the involvement of syntenin-1 in Rac-1 activation and actin polymerization downstream of SFK during both chemokine-induced T cell migration and cognate interactions with APCs. In both phenomena, cell polarity plays a crucial role, determining the front-to-rear axis during chemotaxis, and reorienting signaling and adhesion receptors towards the APC contact site in cognate interactions. Our data indicate that syntenin-1 combines a polarized subcellular localization, dictated by its binding to PIP2 (Zimmermann et al., 2002), with a Src-dependent scaffold activity. Thus, phosphorylated syntenin-1, through binding to the adaptor protein M-RIP, leads to polar translocation of Rac-1 to the plasma membrane and subsequent asymmetric actin polymerization. Syntenin-1-depleted cells have impaired Rac-1 activation and membrane targeting, which resulted in almost complete abrogation of stimuli-induced actin polymerization.

Syntenin-1 was recently shown to bind directly to Src (Boukerche et al., 2008), and Src acts in parallel with PI3K during CXCR2-directed neutrophil chemotaxis (Sai et al., 2008). The functional relationship between Src and Rac activities is complex. Rac activation in response to PDGF has been shown to depend on Src kinase activity (Ouyang et al., 2008). However, PDGF-induced Src activity is not polarized, whereas the consequent Rac activity is clearly localized to the leading edge. It is therefore conceivable that chemokine- or CD3-induced Src activation triggers a polarized signal through its binding to polarized syntenin-1. Syntenin-1 has also been reported to act upstream of the related GTPase Cdc42 in a non-canonical Wnt signaling pathway (Luyten et al., 2008); however, the implications of syntenin-1 phosphorylation were not analyzed.

The functional analysis of Y to F point mutations in the syntenin-1 ITAM-like and ITIM-like domains revealed that mutation of Tyr4 (Y1), in the ITIM-like domain, abolished syntenin-1 phosphorylation and the enhancing effect of syntenin-1 overexpression on chemotaxis, actin polymerization and Rac-1 activation. Overexpression of the Y1 mutant might even have a dominant-negative effect on Rac activation as suggested by the flow cytometry results in which the GFP-positive population can be separately analyzed. Moreover, this mutant, in contrast to wild-type GFP-tagged syntenin-1 was not able to rescue silencing of syntenin-1 in chemokine-induced Rac activation and actin polymerization, confirming syntenin-1 as a link in Src-induced Rac-1 activation in T cells. This tyrosine residue is crucial for the binding to M-RIP in T cells. Mutation of the tyrosine residues in the ITAM-like sequence did not statistically affect syntenin-1 function, suggesting that the effects might be due to small conformational changes. Although ITAM (Ivashkiv, 2009) and ITIM motifs (Daëron et al., 2008) have been extensively studied in immune cell recognition receptors, where they integrate signals emanating from different extracellular stimuli, other cytosolic adaptor proteins, such as ERMs (ezrin–radixin–moesin proteins), also contain functional Tyr-based motifs (Urzainqui et al., 2002). Thus, by harboring equivalent signaling motifs, membrane actin adaptors might share some of the signaling cascades elicited by the receptors themselves. It is intriguing that an inhibitory motif, such as an ITIM, which is potentially bound to phosphatases, can actually positively regulate Rac activity. However, positive regulation of Ras or Src has been previously reported to occur downstream of SHP-2 activation (Barrow and Trowsdale, 2006).

The mutation of phosphorylatable tyrosine residues did not affect the subcellular polarization of syntenin-1, suggesting that this is governed by the two PDZ domains, which is consistent with the finding that deletion of these domains abolishes the enhancing effect of syntenin-1 upon tumor cell invasion (Meerschaert et al., 2007). Syntenin-1 PDZ domains have been demonstrated to interact with the ERM protein merlin (Kang et al., 2003) and with PIP2 (Zimmermann et al., 2002). These interactions probably contribute to plasma membrane anchoring of syntenin-1 at lamellipodia and the T-cell–APC contact area, which are sites of active PIP2 production. Other transmembrane receptors, such as CD6 or tetraspanins (Gimferrer et al., 2005; Latysheva et al., 2006), might also contribute to subcellular localization of syntenin-1 in T cells. The fact that syntenin-1 overexpression enhances chemotaxis and actin polymerization suggests that it acts as a limiting factor in the pathway that links Src activity with Rac-1 activation. This effect is crucial for the polarization of the signal, ensuring that Rac-1 activation and subsequent actin polymerization will be optimal only at specific subcellular localizations containing syntenin-1.

Syntenin-1 has also been reported to be associated with membranes in intracellular compartments and to regulate intracellular trafficking of associated receptors (Sarkar et al., 2008). In T cells, however, most endogenous staining was submembranous. We observed no alterations to the basal expression of chemokine receptors or CD3 at the plasma membrane of syntenin-1-depleted cells (data not shown). However, it is also conceivable that vesicular syntenin-1 contributes to Rac activation during recycling (Palamidessi et al., 2008).

Our proteomic analyses aimed to search for proteins that could interact with wild-type syntenin-1 but not with mutant syntenin-1. In this screening, some other regulators of the Rho family of GTPases were also identified; however, only the association with M-RIP was assessed further and confirmed by immunoprecipitation and pull-down analyses. Because of the nature of M-RIP as a scaffolding protein, some of these other molecules might derive from the interaction between syntenin-1 and M-RIP. Further studies will be needed to address the hierarchies and kinetics of these interactions.

Our data identify M-RIP as a crucial local regulator of the Rho–Rac balance during chemotaxis and antigen recognition. Syntenin-1 binds to the same M-RIP domain reported to be essential for the binding to the MBS of myosin light chain phosphatase (Surks et al., 2003). Therefore, syntenin-1 binding might compete with MBS, shifting the balance towards Rac-induced actin polymerization, rather than Rho-mediated contraction. Myosin function in lymphocytes is not restricted to regulation of cell contractility during migration, but also regulates immune synapse maturation (Ilani et al., 2009), cytotoxicity (Andzelm et al., 2007) and chemokine receptor internalization (Rey et al., 2007). MLC phosphorylation is detected at the IS (Ilani et al., 2009) and at the leading edge of migrating polarized lymphocytes (Vicente-Manzanares et al., 2002), both sites of active Rac-induced actin polymerization.

Upon antigen presentation, actin treadmilling driven by myosin-IIA has been shown to be responsible for aggregation of CD3 at the c-SMAC in the IS (Ilani et al., 2009). Thus, syntenin-1 silencing only mildly attenuates the initial signaling events in antigen-driven cognate interactions, and MTOC translocation occurs normally. However, actin dynamics is perturbed, CD3 remains dispersed on the plasma membrane and the activation magnitude and duration are not maintained.

The involvement of syntenin-1 in T cell migration is in agreement with other observations indicating that syntenin-1 might be an important determinant of the malignant phenotype of epithelial cancers (Helmke et al., 2004; Sarkar et al., 2004; Boukerche et al., 2005). Moreover, M-RIP silencing has been reported to inhibit cell invasivity (Ono et al., 2008) and to lie downstream of TGF-β during breast cancer dissemination (Giampieri et al., 2009), so the syntenin-1–M-RIP axis might also be relevant for tumor cell migration. The homing of tumor cells to specific metastatic sites under the influence of chemokines and their receptors is a hallmark of tumorigenesis (Ben-Baruch, 2006), although the SDF–CXCR4 axis acts in conjunction with other pro-malignancy mechanisms. Overexpression of syntenin-1 might be one of the amplifiers of SDF–CXCR4 signals, enhancing the metastatic potential of tumor cells, and in malignancies of the haematopoietic system.

Lymphoblastoid cell lines of the T (CEM, J77, CH7C17, HSB2) and B (Raji, Hom2) lineages were cultured in RPMI 1640 (Flow Laboratories, Irvine, UK) supplemented with 10% fetal calf serum (FCS, Flow Laboratories). Human PBLs were obtained from buffy coats from healthy volunteer blood donors and derived to T-lymphoblasts by treatment with PHA/IL-2 (Sala-Valdes et al., 2006).

Anti-CD3 (T3b) and anti-ICAM-3 (HP2/19) monoclonal antibodies were obtained as previously described (Mittelbrunn et al., 2002). For detection of Tyr phosphorylation, we used a combination of 4G10 and PY72 antibodies (kindly provided by Balbino Alarcón, Centro de Biologia Molecular Severo Ochoa, Madrid, Spain). Antibodies againt phosphorylated Vav-1 and total Vav-1 were kindly provided by Xose Bustelo (Centro de investigacion del Cancer, Salamanca, Spain). Rabbit anti-syntenin-1 was purchased from Synaptic Systems (Göttingen, Germany), anti-α-tubulin and anti-M-RIP from Sigma, anti-Rac-1 from Millipore (Billerica, MA), anti-Rac-GTP from NewEastBio (Malvern, PA). p-ERK, total ERK, p-ZAP70, ZAP70, p-PLCγ1 and PLCγ1 antibodies were from Cell Signaling (Danvers, MA) and p-Src and Src antibodies from Invitrogen (Carlsbad, CA). Anti-GFP was purchased from Clontech and GFP-trap monobodies from Chromotek (Planegg-Martinsried).

SDF-1α and MIP-3β were purchased from Peprotech (London, UK). PP2 was purchased from Sigma. Jurkat cells were activated with HIT3a antibody (10 mg/ml) (Biolegend, San Diego, CA).

Psrα-HAGFP-syntenin-1 was generated by cloning cDNA encoding syntenin-1, obtained from EcoRI-digested PMT2–HA–syntenin-1 (Gimferrer et al., 2005), into psrα-HAGFP. Single point mutations of GFPHA–syntenin-1 were made by directed mutagenesis using the Quikchange Multi-Site Directed Mutagenesis Kit from Stratagene (La Jolla, CA) with the following sequences: Y3 mutation, 5′-GATGGAAATCTGTTTCCTAAACTGTTTCCAGAGCTCTCTCATTCATGGGACTGTC-3′; Y1 mutation, 5′-GGAATTCATGTCTCTTTTTCCATCTCTTGAAGATTTGAAGGTAGAC-3′.

GST constructs of M-RIP (Surks et al., 2003) were provided by Michael Mendelsohn (Tufts Medical Center, Boston, MA). GST–PAK–CRIB was donated by John Collard (The Netherlands Cancer Institute, Amsterdam, The Netherlands). GFP-tagged pleckstrin domains were provided by Tamas Balla (National Institutes of Health, Bethesda, MA).

CEM/J77 cells (2×10⁷) were transiently transfected with 20 mg of plasmids in OPTIMEM medium (Invitrogen) by electroporation at 250 V and 1200 mF (Gene Pulser II, Bio-Rad). To selectively knockdown the expression of endogenous syntenin-1, two target sequences were used: oligo 1 and oligo 2 (reference numbers J-008270-08 and J-008270-06, Dharmacon Research, Lafayette, CO). For rescue experiments, oligo 1 (which pairs at the non-coding region) was used. For silencing of M-RIP, Dharmacon reference J-014102 siRNA oligonucleotide was used. RNA duplexes (2 μM per sample) or a negative control oligonucleotide (Eurogentec, Seraing, Belgium) were transfected by electroporation or by nucleofection using the Nucleofector kit T (program T-23) (Amaxa, Cologne, Germany). Efficiency of knockdown was routinely assessed by western blot to be greater than 50% for each transfection.

Migration assays were performed 24 hours after transfection with GFP-tagged constructs and 48 hours after transfection with siRNA oligonucleotides, either on primary T lymphoblasts or CEM T cells. Cells (5×10⁵/100 ml) were placed in the upper chamber of 5 mm pore-size Transwell inserts (Costar, Cambridge, MA). The lower chambers contained 600 ml complete RPMI medium, with or without SDF-1α (100 nM) or MIP3β (1 μM) (Peprotech, London, UK). After 2–3 hours at 37°C, migrated cells were recovered from the lower chamber and counted by flow cytometry.

Cells (primary T lymphoblasts or J77 T cells) were stimulated, fixed with 4% formaldehyde, permeabilized with 0.5% Triton X-100 (5 minutes) and stained with Phalloidin Alexa Fluor 647 or Phalloidin Alexa Fluor 488 (Invitrogen). Mean fluorescence intensity of F-actin staining was analyzed in a FACS CANTO cytofluorometer (BD Biosciences).

Total cell lysates were obtained in Tris-buffered saline containing 1% Nonidet P-40, 2 mM CaCl₂, 2 mM MgCl₂, phosphatase inhibitors (1 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 10 mM sodium fluoride), and protease inhibitors (Complete; Roche Applied Science). For immunoprecipitation, cell lysates were incubated with the indicated antibodies coupled to Protein-G–Sepharose. After washing six times with lysis buffer, proteins bound to Sepharose beads were eluted by boiling in sample buffer. Pull-down experiments with GST proteins were performed as described (Sala-Valdes et al., 2006).

Target proteins were detected by chemiluminescence with a LAS3000 luminescent imager (Fuji Photo Film, Elmsford, NY). Quantitative densitometric analysis of immunoblots was performed with Image Gauge v.346 software (Fujifilm).

GTP-Rac loading was measured by classical pull-down experiments with GST–PAK–CRIB (Vicente-Manzanares et al., 2005). Alternatively, cells were fixed, permeabilized and stained with anti-Rac-GTP antibody and analyzed by flow cytometry.

Membrane and cytosolic protein extracts were prepared with the Proteo Extract Subcellular Proteome Extraction Kit (Calbiochem, San Diego, CA), which is based on TX114 solubilization, according to the manufacturer’s instructions.

For antigen-specific stimulation, Hom-2 APCs (5×10⁶) were incubated at 37°C for 3 hours with HA307-319 peptide (200 μg/ml) and mixed with a HA-specific T cell clone (CH7C17) at a ratio of 1:3. For superantigen stimulation, Raji B cells were labelled with the fluorescence cell trackers CMAC or CM-TMR (Invitrogen), and then incubated for 20 minutes at 37°C with 1 μg/ml SEE (Superantigen E). Antigen-driven conjugates were processed for immunofluorescence and superantigen stimulation used for western blot analysis of signaling molecule activation.

T cell and B cell cognate conjugates, PBLs or T lymphoblasts (seeded onto 50mg/mL fibronectin) were stained as described (Mittelbrunn et al., 2002; Sala-Valdes et al., 2006). Samples were fixed with 4% formaldehyde and permeabilized for 5 minutes with 0.5% Triton X-100, stained and mounted with Prolong (Invitrogen). Confocal images were obtained with a Leica TCS-SP5 confocal scanning laser microscope with an HCX PL APO lambda blue 63×/1.4 oil immersion objective and analyzed with Leica LAS confocal image processing software. For time-lapse confocal fluorescence microscopy, cells were maintained at 37°C in a 5% CO₂ atmosphere.

The synapsemeasure plug-in for ImageJ software was used to analyze the redistribution of fluorescent signal to the T-APC contact area (Calabia-Linares et al., 2011). Quantification of Rac translocation was calculated as the ratio of the signal at the leading edge membrane region versus the cytosolic signal in four confocal planes for each cell.

Statistical significance was calculated using the Student’s t-test or ANOVA, and significant differences are indicated by ^*P<0.05 and ^**P<0.01.

The authors wish to thank P. Zimmermann for help and advice with silencing of syntenin-1, Soraya López-Martín for technical assistance, Maria-Rosa Sarrias for helpful discussions and S. Bartlett for English editing.