WAVE2 Regulates High-Affinity Integrin Binding by Recruiting Vinculin and Talin to the Immunological Synapse

Abstract

T-cell-receptor (TCR)-mediated integrin activation is required for T-cell-antigen-presenting cell conjugation and adhesion to extracellular matrix components. While it has been demonstrated that the actin cytoskeleton and its regulators play an essential role in this process, no mechanism has been established which directly links TCR-induced actin polymerization to the activation of integrins. Here, we demonstrate that TCR stimulation results in WAVE2-ARP2/3-dependent F-actin nucleation and the formation of a complex containing WAVE2, ARP2/3, vinculin, and talin. The verprolin-connecting-acidic (VCA) domain of WAVE2 mediates the formation of the ARP2/3-vinculin-talin signaling complex and talin recruitment to the immunological synapse (IS). Interestingly, although vinculin is not required for F-actin or integrin accumulation at the IS, it is required for the recruitment of talin. In addition, RNA interference of either WAVE2 or vinculin inhibits activation-dependent induction of high-affinity integrin binding to VCAM-1. Overall, these findings demonstrate a mechanism in which signals from the TCR produce WAVE2-ARP2/3-mediated de novo actin polymerization, leading to integrin clustering and high-affinity binding through the recruitment of vinculin and talin.

Reorganization of the actin cytoskeleton is an essential process required for many aspects of T-cell biology (45). Besides providing the mechanical force for various basic biological functions, including cell migration and polarization, the actin cytoskeleton serves a specialized role in T cells during the formation of the immunological synapse (IS) between the T cell and an antigen-presenting cell (APC) with an appropriate peptide-major histocompatibility complex. Signaling cascades originating from the engaged T-cell receptor (TCR) initiate IS formation, causing dynamic cytoskeletal reorganization, changes in gene transcription, and activation of cell surface integrins (25).

Integrins are heterodimeric cell surface receptors that are responsible for cell adhesion, including T-cell-APC conjugation and attachment to extracellular matrix components such as fibronectin. TCR-mediated activation of integrins occurs as a result of physical conformational changes within the receptor (affinity) as well as clustering of individual subunits on the cell surface (avidity) in response to signals generated from TCR ligation, a process known as “inside-out” signaling (25). Among the various integrin heterodimers expressed by T cells, LFA-1 (α_Lβ₂) plays a critical role during T-cell-APC conjugation and eventually localizes to the pSMAC of the IS, where it binds to its ligand ICAM found on the surface of the APC (9). The α₄β₁ integrin (VLA-4) also localizes to the pSMAC (33). In addition, VLA-4 binds to extracellular matrix proteins, such as fibronectin, and cell surface ligands, such as VCAM-1, in response to TCR stimulation (34). An array of signaling proteins are required for TCR-mediated integrin activation, including the guanine nucleotide exchange factor VAV1 (26), the small GTPase RAP1 (8, 40), the adhesion- and degranulation-promoting adaptor protein (ADAP; also known as SLAP-130/Fyb) (18, 38), and the nonconventional protein kinase C (PKC) isoform PKD (also known as PKCμ) (30). However, the exact molecular mechanism by which these proteins regulate TCR-mediated integrin activation remains largely unknown.

Actin cytoskeletal reorganization is also required for TCR-mediated integrin activation (39). Interestingly, T cells from WASP^−/− mice are unimpaired in their ability to form T-cell-APC conjugates, adhere to fibronectin, or cluster integrins in response to TCR stimulation, whereas T cells lacking VAV1 are defective in all three processes (26). Recently, we have demonstrated that the actin regulatory protein WAVE2 is an essential component of “inside-out” signaling required for TCR-mediated integrin activation (36). Additionally, WAVE2 participates in TCR-stimulated actin cytoskeletal dynamics needed for lamellipodial formation and accumulation of F-actin at the IS. Structurally, WAVE2 contains an N-terminal WAVE-homology domain, which links WAVE2 to a complex of associated protein components that includes ABI-1/ABI-2, NAP/HEM-1, and SRA-1/PIR121 (13). A basic region (BR) and a proline-rich region (PR) in WAVE2 play roles in both the localization and the activation of WAVE2 (32, 37). Finally, in similarity to WASP, WAVE2 has a C-terminal verprolin (also known as WH2)-connecting-acidic (VCA) domain, which binds the ARP2/3 complex and G-actin and promotes de novo actin polymerization. However, the exact mechanism or structural feature of WAVE2 that is required to link TCR-initiated signaling to integrin activation is not known.

In this report, we use biochemical, cellular, and genetic approaches to define the molecular mechanism by which WAVE2 regulates integrin activation downstream of the TCR. We show that the interaction of the WAVE2 VCA domain with the ARP2/3 complex links WAVE2 to the integrin scaffolding proteins vinculin and talin. The formation of a WAVE2-ARP2/3-vinculin complex leads to talin recruitment to the IS and high-affinity integrin binding. Overall, these findings establish a molecular mechanism by which WAVE2 and de novo actin polymerization control integrin activation in response to TCR stimulation.

MATERIALS AND METHODS

Reagents and antibodies.

Unless otherwise stated, all chemicals were obtained from Sigma-Aldrich. The antisera against WAVE2 (36) and VAV1 (15) have been previously described. Antibodies for RAC1 (clone 23A8), p34-Arc/ARPC2, vinculin (clone V284), and phosphotyrosine (clone 4G10) were purchased from Upstate Biotechnology. The antibody against β₁ integrin (clone P5D2) was purchased from Chemicon, the antibodies for talin (clone 8d4) and FLAG epitope were purchased from Sigma, the antibody for CDC42 was purchased from BD Biosciences, and the antibody for ARP2 was purchased from Santa Cruz Biotechnology. The antibody against β₂ integrin was clone MHM23. The OKT3 monoclonal antibody (MAb) was obtained from the Mayo Pharmacy, and the anti-CD28 MAb was purchased from BD Biosciences.

Plasmids and cloning.

The pCMS4 “suppression/reexpression” vectors used for short hairpin RNA (shRNA) silencing and reexpression of resistant cDNAs have been described previously (15). The following 19-nucleotide sequence was generated to target human WAVE2: GAGAAGAGAAAGCACAGGAA. The DNA sequence encoding full-length WAVE2 was amplified from Jurkat cDNA by use of the following set of oligonucleotides: 5′-GCACACGCCGACGCGTATGCCGTTAGTAACGAGGAACATCGAGCCA-3′ and 5′-GACGATGCGAGCGGCCGCTTAATCGGACCAGTCGTCCTCATCAAATTC-3′ (underlined sequences indicate unique restriction sites used for subcloning). An shRNA-resistant form of WAVE2 was created using a QuikChange site-directed mutagenesis kit from Stratagene in which the targeting sequence within the DNA was changed to GAaAAaAGgAAaCACAGGAA (lowercase letters indicates a changed nucleotide that does not affect amino acid sequence). The ΔBR mutant of WAVE2 was created by deleting amino acids 172 to 209 (37) and the ΔPR mutant by deleting amino acids 247 to 419 by site-directed mutagenesis. The ΔVCA version of WAVE2 was created by deleting the C-terminal end of the protein beginning with amino acid 436. The following 19- and 21-nucleotide sequences were generated to target human vinculin: A (GGTAGCCATCCCATGAACA) and B (GCCTTCCTCCACATCCTTTCT). Human vinculin cDNA was a gift from Tina Izard (St. Jude's Children's Hospital, Memphis, TN) and was subcloned into the pCMS4 vector. Since both of the shRNAs used to target human vinculin laid in the 3′ untranslated region of the primary transcript, generation of resistant cDNA for vinculin was not necessary. The P878A mutant of human vinculin was generated using site-specific mutagenesis. The following sequence was generated to target human ARP2: GTGGGTAAATCTGAGTTTA.

Cell culture and transfection.

Jurkat T cells, NALM6 B cells, and Raji B cells were grown in RPMI 1640 supplemented with 5% fetal bovine serum, 5% fetal calf serum, and 4 mM l-glutamine. Human CD4⁺ T cells were purified using buffy coats obtained from the Mayo Clinic Blood Bank and RosetteSep purification (StemCell Technologies). Following purification, human CD4⁺ cells were cultured in serum-containing media along with 5 μg/ml phytohemagglutinin and interleukin-2 for 24 h. Cells were then washed and cultured in serum-containing media with interleukin-2 for 72 h. Transient transfections in Jurkat were performed using 1 × 10⁷ cells per sample along with 30 to 40 μg of plasmid DNA as previously described (5). Transfected Jurkat cells were used 48 to 72 h following transient transfection.

Cell stimulation and immunoblot analysis.

For the stimulation time course studies, 10 × 10⁶ Jurkat T cells or 25 × 10⁶ human CD4⁺ T cells were stained on ice with 5 μg/ml anti-CD3 (OKT3; MAb) and 5 μg/ml anti-CD28 and then cross-linked using goat anti-mouse antibody over the indicated time course at 37°C. After each time point the cells were immediately washed in ice-cold phosphate-buffed saline (PBS) and lysed in NP-40 lysis buffer (20 mM HEPES [pH 7.9], 100 mM NaCl, 5 mM EDTA, 0.5 mM CaCl, 1% NP-40, 1 mM phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 5 μg/ml aprotinin, 1 mM Na₃VO₄, 5 μM MG-132) for 10 min on ice. Lysates were clarified by centrifugation at 18,000 × g for 5 min at 4°C and then transferred to antibody-coated beads. The protein complexes were then washed twice with NP-40 lysis buffer, eluted in 60 μl of sodium dodecyl sulfate (SDS)-sample buffer, resolved by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to Immobilon-P membranes (Millipore). For activation of GTPases, cells were lysed in GTPase activation buffer (50 mM Tris [pH 7.5], 500 mM NaCl, 5 mM MgCl₂, 0.5% NP-40, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 10 μg/ml leupeptin, 5 μg/ml aprotinin, 1 mM Na₃VO₄), vortexed, and immediately clarified and transferred to glutathione agarose previously bound to glutathione S-transferase-Pak GTPase binding domain. Lysates were allowed to rotate for 10 min and were washed once before elution and analysis as described above. In cases in which whole-cell lysates were prepared, 50 to 100 μg of protein was resolved by SDS-PAGE. For immunoblot analysis, MAbs were detected using goat anti-mouse immunoglobulin G (IgG) coupled to horseradish peroxidase (Santa Cruz), and polyclonal rabbit antisera were detected using goat anti-rabbit antibody coupled to horseradish peroxidase (Santa Cruz) and SuperSignal enhanced chemiluminescence (Pierce, Rockford, IL).

Immunofluorescence microscopy.

B-cell/T-cell conjugates were formed essentially as described previously (4). Briefly, NALM6 or Raji cells were stained with CellTracker Blue CMAC (7-amino-4-chloromethylcoumarin; Molecular Probes) and pulsed with or without 2 μg/ml staphylococcal enterotoxin E (SEE) or 2 μg/ml of a cocktail of staphylococcal superantigens (SEA, SEB, SEC3, SEE; Toxin Technology, Inc.). B cells were centrifuged together with the same number of T cells, incubated at 37°C for 15 min, plated onto poly-l-lysine-coated coverslips, and fixed with 4% paraformaldehyde-PBS for 10 min at room temperature. Fixed cells were quenched with 50 mM NH₄Cl and permeabilized in 0.3% Triton X-100. Blocking and antibody incubations were performed in PBS-0.05% saponin-0.25% fish skin gelatin. Slides were labeled with primary antibodies followed by goat anti-rabbit fluorescein isothiocyanate antibody or goat anti-mouse tetramethyl rhodamine isocyanate antibody (Molecular Probes). F-actin was visualized with fluorescein or rhodamine phalloidin (Molecular Probes). Coverslips were mounted in Mowiol 4-88 (Hoeschst Celanese) containing 10% 1,4-diazobicyclo[2.2.2]octane. Quantification of F-actin polarization and protein localization were performed by an individual blinded to the experimental conditions. To minimize bias, 50 conjugates were chosen at random based upon differential interference contrast and CMAC images (disregarding cell morphology or protein distribution) and only conjugates consisting of one green (green fluorescent protein [GFP]-transfected) T cell and one blue (CMAC-stained) B cell were scored. Those conjugates showing a distinct, bright band of labeling at the cell-cell contact site were scored as representing a positive result. Typically, this band was much brighter than any other portions of either cell; however, where necessary, the pixel intensity was determined, and only those interfaces with pixel intensity greater than the sum of pixel intensities of the cell surfaces two cell surfaces away from the interface were scored as polarized.

Conjugate analysis.

Conjugate assays were performed essentially as described previously (35). Briefly, Raji B cells were stained with PKH26 according to the manufacturer's directions. After being quenched with media containing serum, cells were incubated in the presence or absence of 2 μg/ml SEE for 1 h, washed, and resuspended at 0.5 × 10⁶ cells/ml in RPMI medium. Jurkat T cells transfected with GFP-expressing plasmids were also resuspended at 0.5 × 10⁶ cells/ml in RPMI medium. For conjugation, equal volumes of B and T cells were pelleted together at a speed of 500 rpm for 5 min and then incubated at 37°C for 10 to 15 min. Cells were vortexed for 5 to 10 s and then fixed by addition of an equal volume of 4% paraformaldehyde. The relative proportions of red, green, and red-green events in each tube were determined by two-color flow cytometric analysis using a FACSCalibur flow cytometer (BD Biosciences). The number of gated events counted per sample was at least 15,000.

Adhesion assays.

Adhesion assays were performed as previously described (10, 46). Briefly, Jurkat T cells were transfected with the indicated GFP-suppression vector and adhesion analysis was performed 48 to 72 h later using a 96-well plate precoated with 0.3 μg fibronectin/well. For TCR stimulation, the cells were preincubated with OKT3 and then added to wells containing secondary anti-IgG as a cross-linker. Phorbol myristate acetate (PMA) stimulation was used at a concentration of 10 ng/ml. Adhesion was quantified by flow cytometry as previously described (10, 46).

Single-cell calcium analysis.

Jurkat T cells were loaded with the cell permeant calcium indicator fura-2 AM (Molecular Probes) (3.0 μM) in RPMI medium for 15 min at room temperature (25°C). For the final 10 min of Fura-2 loading, OKT3 (5 μg/ml) was added to the incubation mixture. Cell suspensions containing Fura-2 and OKT3 were placed into the recording chamber on an inverted fluorescence microscope (Nikon) and allowed to adhere to poly-l-lysine-treated coverslips for 5 min in a solution which contained 155 mM NaCl, 4.5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 10 mM glucose, and 10 mM HEPES (pH 7.4). Excess fura-2 AM (Molecular Probes) and OKT3 were removed by perfusing the chamber with additional extracellular solution. Before stimulation, the chamber was perfused with a Ca²⁺-free bath solution containing 155 mM NaCl, 4.5 mM KCl, 1 mM MgCl₂, 0.5 mM EGTA, 10 mM glucose, and 10 mM HEPES (pH 7.4). Intracellular Ca²⁺ mobilization was initiated by the addition of goat anti-mouse IgG in Ca²⁺-free bath solution. Fura-2 fluorescence of individual cells was measured by digital imaging microscopy as previously described (28) and plotted as the ratio of fluorescence emission at 510 nm following sequential fura-2 excitation at 340 nM and 380 nM.

Soluble VCAM-1 binding.

This procedure was modified as previously described (6, 47). Briefly, 1 × 10⁶ Jurkat T cells were washed with modified Tyrode's buffer (12 mM NaHCO₃, 20 mM HEPES [pH 7.4], 1 mg/ml glucose, 150 mM NaCl, 2.5 mM KCl, 1 mg/ml bovine serum albumin, 1 mM Ca²⁺, 1 mM Mg²⁺) and incubated in 50 μl of Tyrode's buffer with 1.0 μg of recombinant chimeric human seven-domain VCAM-1 Fc (R&D Systems, Minneapolis, MN). The cells were either left untreated or stimulated with PMA (50 ng/ml) or Mn²⁺ (1 mM) for 10 min at 37°C and then diluted in 3 ml of Tyrode's buffer and immediately fixed with 0.5 ml of 4% paraformaldehyde for 20 min. The cells were then washed with Tyrode's buffer and stained with biotinylated goat anti-human Fc and streptavidin-allophycocyanin (eBioscience, San Diego, CA). Flow cytometry with a FACSCalibur system (BD Biosciences, San Jose, CA) was used to determine the degree of VCAM-1 binding.

RESULTS

WAVE2 regulates integrin activation downstream of Vav1, Rac1, and Cdc42.

We have previously demonstrated that WAVE2 regulates TCR-mediated integrin activation required for both T-cell-APC conjugation (primarily β₂ integrin mediated) and adhesion to fibronectin (primarily β₁ integrin mediated) (36). Even though WAVE2 and de novo actin polymerization are required for TCR-induced integrin activation, the exact mechanism by which WAVE2 regulates this process is unknown. We have previously demonstrated that suppression of WAVE2 does not alter the activation of the TCR-stimulated proximal tyrosine kinase LCK or ZAP-70, formation of the LAT signalosome, or activation of PLCγ1 (36), which are all essential components of signaling pathways leading to TCR-mediated integrin activation (see Fig. 1). VAV1, which is a guanine nucleotide exchange factor for the Rho GTPases RAC1 and CDC42, is tyrosine phosphorylated in response to TCR stimulation (44). In addition, T cells from VAV1^−/− mice are unable to form conjugates, adhere to fibronectin, or cluster integrins in response to TCR stimulation (26). Thus, we determined whether loss of WAVE2 affects the activation of TCR-stimulated molecular pathways leading to the activation of VAV1, RAC1, and CDC42. In order to determine whether WAVE2 regulated VAV1 and RHO GTPase activation, gross tyrosine phosphorylation of VAV1 was analyzed in WAVE2-suppressed T cells. TCR stimulation of both control and WAVE2-suppressed T cells induced tyrosine phosphorylation of VAV1 that peaked at 1 min and was reduced to background levels by 15 min (Fig. 2A). In addition, the activation of the cytoskeletal regulators RAC1 and CDC42, two RHO GTPases activated by VAV1, was not impaired in the absence of WAVE2 (Fig. 2B and C). In fact, it appeared that the activation of both RHO GTPases in the absence of WAVE2 was augmented. Together, these data indicate that WAVE2 affects TCR-stimulated integrin activation at a point distal to the regulation of signaling cascades that regulate the activity of these small GTP-binding proteins.

FIG. 1.

Signaling cascades required for TCR-mediated integrin activation. Ligation of the TCR activates the proximal tyrosine kinases Lck (not shown) and ZAP-70. These kinases can then activate several pathways known to regulate integrin activation, including the LAT-ITK-PLCγ1 pathway, required for both TCR-mediated calcium flux and activation of the small GTPase of the Ras family, RAP1. Targets of Rap1 (PKD, RAPL, MST1) regulate integrin activation primarily through the appropriate clustering of integrins on the cell surface. ADAP and SKAP55 are both required for TCR-mediated integrin activation and have been shown to affect both cytoskeletal dynamics as well as the localization of RAP1. ADAP and SKAP55 have also been shown to be indirectly linked to active RAP1 through another RAP1 effector, RIAM, which constitutively interacts with SKAP55. Finally, the RHO GTPases RAC1 and CDC42 (while never formally shown to be required for TCR-mediated integrin activation) can be activated by VAV1, leading to cytoskeletal reorganization, possibly through the activation of WAVE2. The integrin scaffolding protein Talin binds to F-actin filaments and also directly to the tail of the integrin β chain and has been shown to control both integrin affinity and clustering. Importantly, several of these proteins (*) are not required for PMA-stimulated integrin activation, suggesting that integrins can be activated independently of proximal TCR-stimulated signaling pathways when stimulated with PMA. Whether WAVE2 regulates cytoskeletal rearrangement leading to integrin activation is the focus of this study.

FIG. 2.

WAVE2 is not required for TCR-mediated activation of VAV1, RAC1, or CDC42. (A) Jurkat T cells were transfected with either control or WAVE2-suppression vector. At 72 h posttransfection, cells were stimulated with anti-CD3/goat anti-mouse antibody for 0, 1, 2, 5, or 15 min. Cells were immediately lysed, and after clarification, VAV1 was immunoprecipitated (IP). Bound proteins were eluted, separated by SDS-PAGE, transferred, and blotted with 4G10 MAb (pTyr) and VAV1. Whole-cell extracts (WCE) were also analyzed for expression of WAVE2 and VAV1. (B and C) Cells were transfected as described for panel A. Cells were stimulated for 0, 1, or 5 min and immediately lysed. Following clarification, GTP-bound RAC1 (B) or GTP-bound CDC42 (C) was detected using glutathione S-transferase-PAK GTPase binding domain and immunoblotting.

The VCA domain of WAVE2 regulates integrin activation.

WAVE2 is a multidomain protein that interacts with several other proteins and can directly modulate actin dynamics through an association of its VCA domain with ARP2/3. Since depletion of WAVE2 did not seem to impact known TCR-stimulated signaling pathways leading to integrin activation, we decided to directly examine the role of WAVE2 in this actin-regulated process. In order to determine which domain(s) of WAVE2 regulated integrin activation, “suppression/reexpression” plasmids were generated that would suppress endogenous levels of WAVE2 by driving shRNA behind the H1 RNA polymerase III-dependent promoter and reexpress resistant forms of FLAG-tagged WAVE2 cDNA from a separate cytomegalovirus promoter. Transfection of these plasmids into Jurkat T cells consistently yielded a >80% GFP-positive population. As shown in Fig. 3A, endogenous WAVE2 protein levels were suppressed 72 h following transient transfection compared to vector control levels. In addition, reexpression of wild-type (WT) WAVE2, along with that of deletion mutants lacking the BR (ΔBR), PR (ΔPR), or VCA region (ΔVCA), was comparable to endogenous levels of WAVE2 in the control transfection (Fig. 3A).

FIG. 3.

The VCA domain of WAVE2 regulates TCR-mediated integrin activation. (A) WAVE2 contains an N-terminal WAVE-homology domain, followed by a small BR, a large PR, and a C-terminal VCA domain. Jurkat T cells were transfected with “suppression/reexpression” constructs which suppress endogenous levels of WAVE2 and reexpress FLAG-tagged shRNA-resistant forms of WAVE2 and various deletion mutants. At 72 h posttransfection, whole-cell extracts were harvested, separated by SDS-PAGE, transferred, and blotted for WAVE2, FLAG, and ARPC2 as a loading control. The immunoreactive portion of WAVE2 is not present in the ΔPR mutant and can only be detected with FLAG immunoblotting. Multiple bands in the ΔVCA and ΔBR versions of WAVE2 are most likely due to different phosphorylation states of the proteins (36). (B) Cells were transfected as described for panel A and analyzed for their ability to adhere to fibronectin in response to stimulation of the TCR (+) or in the absence of stimulation (−). Adherent cells (GFP negative, GFP low, and GFP high) were detected using flow cytometry. GFP-negative cells (untransfected) served as an internal control to demonstrate adequate stimulation required for adhesion. (C) Cells were transfected as described for panel A and analyzed for their ability to form conjugates with PKH26-stained Raji B cells that were either pulsed with SEE or left unpulsed. Conjugate formation was determined using two-color flow cytometry as described in Materials and Methods. (D) Cells were transfected as described for panel A with either control or WAVE2-suppression vector and analyzed for surface expression of both β₁ and β₂ integrins. Cells were gated as GFP positive or GFP negative and stained with control IgG (white histogram), β₁ (black histogram), or β₂ (gray histogram) integrin in the different cell populations.

Using the “suppression/reexpression” system, we next determined which domains of WAVE2 were required for TCR-induced adhesion to fibronectin. As previously demonstrated, Jurkat T cells lacking WAVE2 are inhibited in their ability to adhere to fibronectin in response to TCR stimulation (Fig. 3B). However, reexpression of the WT-resistant form of WAVE2 in shWAVE2-transfected cells rescued the ability of the cells to adhere to fibronectin, demonstrating that the shRNA activity against WAVE2 is specific and that this result is not an artifact or an off-target effect. Two mutants of WAVE2 (ΔBR and ΔPR) partially rescued the adhesion phenotype. However, reexpression of the ΔVCA version of WAVE2 did not restore integrin activation and, in fact, was found to be similar to that seen in cells lacking WAVE2 altogether (Fig. 3B). In addition, reexpression of the ΔVCA version of WAVE2 was not able to rescue the ability of WAVE2-suppressed T cells to form LFA-1 integrin-dependent conjugates with superantigen-pulsed B cells (Fig. 3C). The defect in integrin activation was not a result of altered β₁ or β₂ integrin expression, as levels of these integrin subunits were comparable between control and WAVE2-suppressed T cells (Fig. 3D).

The results described above suggest that WAVE2 may be the critical mediator linking the TCR to actin polymerization and integrin activation. Indeed, cells suppressed for WAVE2 demonstrate defective actin accumulation at the IS, and this is rescued by expression of WT WAVE2 but not by the VCA domain deletion mutant (Fig. 4A). We have previously demonstrated that WAVE2 is required for calcium mobilization at a point distal to inositol trisphosphate-mediated storage release (36). Since calcium is an important regulator of intracellular pathways required for many biological functions (12), including actin dynamics and integrin activation, we obtained single-cell calcium measurements to determine whether the VCA domain of WAVE2 is also required for TCR-mediated calcium mobilization. As shown in Fig. 4B, shWAVE2-transfected cells exhibit a defect in extracellular calcium entry following TCR engagement compared to control vector-transfected cells. However, single-cell calcium imaging revealed that both the WT and ΔVCA versions of WAVE2 rescued the deficiency in TCR-mediated calcium entry seen with cells lacking WAVE2 (Fig. 4B). Thus, WAVE2 regulates TCR-mediated integrin activation independently of its role in regulating calcium entry.

FIG. 4.

The VCA domain of WAVE2 regulates actin polymerization and integrin activation downstream of proximal signals, but this occurs independently of calcium mobilization. (A) Jurkat T cells were transfected with control, WAVE2-suppression, WT WAVE2-reexpression, or ΔVCA WAVE2-reexpression vectors and allowed to recover for 72 h. T cells (green) were then allowed to form conjugates with CMAC-stained Raji B cells (blue) that were pulsed with SEE, allowed to adhere to poly-l-lysine-coated coverslips, fixed, and stained with rhodamine phalloidin (red) to visualize the accumulation of F-actin at the IS. Accumulation of F-actin was quantified in 50 random conjugates as described in Materials and Methods; data represent the average results of three independent experiments. (B) Jurkat T cells were transfected as described for panel A. After 72 h, calcium measurements were performed using single-cell fluorescence ratio (Fura-2 imaging) of GFP-positive control and WAVE2-suppressed Jurkat cells. In a calcium-free bath solution, increases in the Fura-2 ratio reflect Ca²⁺ release from intracellular stores. Extracellular Ca²⁺ entry via activated calcium release-activated calcium channels was subsequently assessed by reintroduction of extracellular calcium (2 mM Ca²⁺). Data represent the results of three independent experiments for each condition, and each trace represents the average response of at least 100 cells in the recording chamber. (C) Cells were transfected as described for panel A and analyzed for their ability to adhere to fibronectin in response to stimulation with 10 ng PMA/ml. Adhesion was determined as described for Fig. 2B.

Stimulation with the phorbol ester PMA also induces integrin clustering and activation in T cells and bypasses the requirement for intact TCR-proximal signals (9, 41). In fact, cells that lack ZAP-70 (10), LAT (14), VAV1 (26), ITK (11), PLCγ1 (24), or ADAP (18, 38) show no impairment of integrin activation in response to PMA stimulation (see Fig. 1). Since WAVE2 does not appear to regulate any of these upstream signaling pathways required for integrin activation, we hypothesized that WAVE2 regulates adhesion at a point downstream to these processes. In agreement with this, PMA stimulation was unable to rescue the ability of WAVE2-suppressed cells or cells reexpressing a ΔVCA version of WAVE2 to adhere to fibronectin (Fig. 4B). Overall, these data suggest that the VCA domain of WAVE2 is required for F-actin accumulation at the IS and, although not required for calcium mobilization, does regulate integrins at a point distal to the TCR signaling pathways outlined in Fig. 1.

WAVE2 and ARP2/3 are required for localization of integrins to the IS.

LFA-1 (α_Lβ₂ integrin) is required for conjugation between a T cell and an APC and localizes to the pSMAC area of the IS, where it associates with its ligand ICAM-1 found on the surface of the APC. Since the VCA domain of WAVE2 was required for F-actin accumulation at the IS, we next determined whether WAVE2 and the ARP2/3 complex were required for localization of LFA-1 to the IS. As shown in Fig. 5A, localization of CD18 (β₂ chain of LFA-1) to the IS occurred readily in control transfected cells interacting with SEE-pulsed Raji B cells. However, localization of LFA-1 to the IS in both WAVE2- and ARP2-suppressed T cells was severely impaired (Fig. 5A). In fact, the number of WAVE2- and ARP2-suppressed T cells that localized LFA-1 to the IS was similar to that seen with control cell conjugates formed in the absence of SEE.

FIG. 5.

Localization of both LFA-1 and β₁ integrins to the IS requires WAVE2 and ARP2/3. (A) Jurkat T cells were transfected with control, WAVE2 suppression, or ARP2 suppression vectors and allowed to recover for 72 h. T cells (green) were then allowed to form conjugates with CMAC-stained Raji B cells (blue) that were pulsed with SEE, allowed to adhere to poly-l-lysine-coated coverslips, fixed, and stained with antibody against the β₂ chain of LFA-1. Localization of LFA-1 to the synapse was quantified using 50 random conjugates as described in Materials and Methods; data represent the average results of three independent experiments. (B) Cells were transfected and conjugates were formed as described for panel A. Conjugates were then analyzed for their ability to localize β₁ integrins to the IS, and quantification was determined as described for panel A.

VLA-4 (α₄β₁ integrin) also plays a critical role in T-cell biology and can bind to both VCAM-1 and the extracellular matrix component fibronectin. In addition, VLA-4 can serve as a costimulatory molecule and also localizes to the IS of a T-cell-APC conjugate (33), where it may interact with CD14 on the APC (20). Since LFA-1 localization and adhesion to fibronectin require both WAVE2 and the ARP2/3 complex, we next determined whether the localization of this integrin also required these proteins. Similar to the result obtained for LFA-1, localization of β₁ integrins was impaired in both WAVE2- and ARP2/3-suppressed T cells in response to TCR stimulation (Fig. 5B). Overall, these data indicate that both WAVE2 and the ARP2/3 complex are required for localization of integrins to the IS during T-cell-APC conjugation.

WAVE2 forms a complex with ARP2/3 and vinculin in response to TCR stimulation.

Vinculin is an integrin scaffolding protein known to regulate integrin activation in other cell types. Vinculin becomes activated downstream of cell surface receptors and binds both talin and F-actin, providing a physical link between the actin cytoskeleton and integrins (48). Since the biological function of vinculin has not been extensively studied in T cells, we examined the localization of this scaffolding protein in T cell-APC conjugates. Vinculin localizes to the IS in human CD4⁺ T cells conjugated to superantigen-pulsed NALM6 B cells, where it colocalized with both F-actin (Fig. 6A) and WAVE2 (Fig. 6B). Localization to the IS was dependent upon TCR stimulation, since conjugates formed in the absence of superantigen did not localize vinculin or WAVE2 or polymerize actin at the T-cell-APC interface (−SAg; Fig. 6A and B).

FIG. 6.

The ARP2/3 complex is required to link WAVE2 to vinculin. (A) Human CD4⁺ T cells were allowed to form conjugates with CMAC-labeled NALM6 B cells (blue) that were pulsed with a superantigen cocktail containing SEE, SEA, SEB, and SEC (+ SAg) or left unpulsed (− SAg). Conjugates were then allowed to adhere to poly-l-lysine-coated coverslips, fixed, and subsequently stained for vinculin (red) and F-actin (green) or (B) vinculin (red) and WAVE2 (green). (C) Jurkat T cells were transfected with either control vector or ARP2-suppression vector and allowed to recover for 72 h. Cells were then left unstimulated or were stimulated with anti-CD3/CD28 for 30 min and immediately lysed. Lysates were clarified, and vinculin was immunoprecipitated (IP). Bound proteins were eluted, separated by SDS-PAGE, transferred, and subsequently blotted for expression of WAVE2, ARP2, and vinculin. Whole-cell extracts (WCE) were also analyzed for protein levels of WAVE2, ARP2, and vinculin. (D) Cells were transfected with control, WAVE2 suppression, WT WAVE2 reexpression, or ΔVCA WAVE2 reexpression vector. Cells were stimulated, and lysates were immunoprecipitated as described for panel C and subsequently blotted for WAVE2, ARPC2, and vinculin.

In addition to binding F-actin and talin, activated vinculin has recently been shown to bind the ARP2/3 complex, and this association is enhanced in the presence of a VCA domain (7). We therefore investigated whether a complex might form between WAVE2, ARP2/3, and vinculin in response to TCR/CD28 ligation. Indeed, following receptor engagement, an increased association of vinculin, ARP2/3, and WAVE2 was observed (Fig. 6C). Since vinculin is known to interact with ARP2/3, we next utilized shRNA against ARP2 to test whether WAVE2 requires ARP2/3 in order to form a complex with vinculin. Indeed, compared to control cells, T cells suppressed with respect to ARP2 were unable to form the WAVE2-vinculin complex (Fig. 6C), indicating that the ARP2/3 complex is a necessary component linking WAVE2 to vinculin. In addition, WAVE2 is required for vinculin to associate with the ARP2/3 complex in response to TCR stimulation (Fig. 6D). Reexpression of the WT form of WAVE2, but not the ΔVCA mutant, reconstituted the association of ARP2/3 and WAVE2 with vinculin (Fig. 6D). Overall, these data demonstrate that the VCA domain of WAVE2, through its association with ARP2/3, is necessary to link WAVE2 to vinculin and that vinculin is unable to bind the ARP2/3 complex without the VCA domain of WAVE2.

Vinculin regulates TCR-stimulated integrin activation but not F-actin accumulation or integrin localization to the IS.

The data presented above suggest that vinculin might be recruited to areas of de novo actin polymerization through an association with WAVE2 and ARP2/3. Since vinculin is known to be involved in integrin activation in other cell types, we next investigated whether vinculin is involved in integrin activation in T cells. In order to functionally characterize the possible role for vinculin in regulating TCR-mediated integrin adhesion, two shRNAs against vinculin that suppress vinculin protein levels in Jurkat T cells were generated (Fig. 7A). Depletion of vinculin in T cells resulted in a reduction in the formation of stable T-cell-APC conjugates as determined by flow cytometry (Fig. 7B). Since vinculin is required for TCR-mediated conjugation to occur, we next analyzed vinculin-suppressed T cells for F-actin accumulation and integrin localization to the IS. In contrast to WAVE2- and ARP2/3-suppressed T cells (16, 36), which show dramatic defects in both F-actin polymerization at the IS and localization of integrins to the T-cell-APC contact site, vinculin-suppressed Jurkat T cells conjugated to SEE-pulsed Raji B cells efficiently polymerized actin at the IS similarly to what is seen with control cells (Fig. 7C). The localization of LFA-1 to the IS also occurs in the absence of vinculin (Fig. 7D), suggesting that vinculin affects TCR-stimulated integrin activation independently of de novo actin polymerization and does not regulate integrin localization to the IS.

FIG. 7.

Vinculin is required for TCR-mediated integrin activation occurring independently of actin and integrin accumulation at the IS. (A) Jurkat T cells were transfected with either vector control or with suppression vectors against vinculin. At 72 h posttransfection, cells were analyzed by immunoblotting for expression of vinculin and WAVE2. (B) Cells were transfected as described for panel A and analyzed for their ability to form conjugates with PKH26-stained Raji B cells that were either pulsed with SEE or left unpulsed. Conjugate formation was determined using two-color flow cytometry as described in Materials and Methods. (C) Jurkat T cells were transfected as described for panel A. T cells (green) were then allowed to form conjugates with SEE-pulsed Raji B cells (blue), allowed to adhere to poly-l-lysine-coated coverslips, fixed, and stained with rhodamine phalloidin to visualize accumulation of F-actin at the IS. Actin polymerization was quantified using 50 random conjugates as described in Materials and Methods; data represent the average results of two independent experiments. (D) Same as panel C except that the cells were stained for β₂ integrin.

The ARP2/3 binding region of vinculin regulates integrin activation.

Since WAVE2 and ARP2/3 both associate with vinculin in response to TCR stimulation, we next sought to determine whether vinculin binding to ARP2/3 is required for TCR-mediated integrin activation. To test this, “suppression/reexpression” constructs were generated for vinculin, which reexpressed WT vinculin or a previously defined mutant of vinculin that does not bind the ARP2/3 complex (7). Importantly, transfection of these plasmids into Jurkat T cells results in the suppression of endogenous vinculin and reexpression of either the WT or the ARP2/3-binding mutant protein (Fig. 8A).

FIG. 8.

The association between ARP2/3 and vinculin is required for TCR-mediated integrin activation. (A) Jurkat T cells were transfected with control, vinculin suppression, WT vinculin reexpression, or P878A vinculin reexpression vectors. At 72 h posttransfection, whole-cell extracts were harvested, separated by SDS-PAGE, and transferred, and membranes were subsequently blotted for vinculin, FLAG, and ARPC2 as a loading control. (B) Cells were transfected as described for panel A and were subjected to TCR stimulation (+) or were left unstimulated (−) and were analyzed for their ability to adhere to fibronectin. (C) Cells were transfected as described for panel A and were subjected to stimulation with 10 ng/ml PMA or were left unstimulated and were analyzed for their ability to adhere to fibronectin. Adherent cells were detected using flow cytometry as described for Fig. 3B.

To determine whether vinculin binding to ARP2/3 was required for TCR-mediated integrin activation, Jurkat T cells expressing the different “suppression/reexpression” vinculin plasmids were analyzed for their ability to adhere to fibronectin in response to TCR stimulation. As previously shown, Jurkat T cells transfected with control vector adhered readily to fibronectin in response to TCR stimulation. In contrast, cells that were transfected with the vinculin suppression vector were impaired in their ability to adhere to fibronectin compared to the control cells (Fig. 8B). In addition, the ability of vinculin to bind the ARP2/3 complex is required for TCR-mediated integrin activation, because reexpression of the P878A mutant of vinculin failed to rescue the adhesion defect, whereas reexpression of WT protein resulted in adhesion similar to that of control transfected cells (Fig. 8B). Similar to the phenotype seen in WAVE2-suppressed cells, PMA did not rescue the ability of vinculin-suppressed cells, or of cells expressing the P878A mutant of vinculin, to adhere to fibronectin (Fig. 8C). Overall, these data demonstrate that the TCR-mediated association of WAVE2, ARP2/3, and vinculin controls integrin activation required for T-cell adhesion.

Talin is part of the TCR-stimulated WAVE2-ARP2/3-vinculin ternary complex.

Many scaffolding proteins that link to integrins also bind to the actin cytoskeleton, such as α-actinin, filamin, and talin. Talin, a large 250 kDa protein that binds directly to the β chain of the integrin heterodimer and F-actin branches (2), localizes to the interface of the T-cell-APC conjugate in response to TCR stimulation (27) and binds directly to vinculin. These qualities of talin prompted us to investigate whether WAVE2 and vinculin played a role in the regulation of this important integrin activator. Importantly, when Jurkat T cells were stimulated with anti-CD3/CD28, a complex consisting of WAVE2, ARP2/3, vinculin, and talin could be detected in cell lysates (Fig. 9A). This complex was mostly undetectable in unstimulated cells, but the association continued to form with increasing times of stimulation and still persisted after 30 min of TCR/CD28 stimulation. Formation of this complex could also be induced in primary CD4⁺ human T cells (Fig. 9B). We next investigated whether talin and vinculin binding to the β₁ integrin tail following TCR ligation would be affected by the loss of WAVE2. Indeed, both talin and vinculin could be detected with β₁ integrin following receptor ligation, which was abrogated in WAVE2-suppressed cells (Fig. 9C). These observations suggest that in T cells, WAVE2 might regulate TCR-mediated integrin activation through the formation of an ARP2/3-vinculin-talin signaling complex.

FIG. 9.

Vinculin associates with WAVE2, ARP2/3, and talin in response to TCR stimulation. (A) Jurkat T cells were stimulated with anti-CD3/CD28 antibody for 0, 5, 15, or 30 min and lysed, and immunoprecipitations (IP) were performed using either an isotype control IgG antibody or an antivinculin antibody. Bound proteins were eluted, separated by SDS-PAGE, and transferred, and membranes were subsequently blotted for talin, WAVE2, ARPC2, and vinculin. (B) Same as panel A except that human CD4⁺ T cells were used. (C) Jurkat T cells were transfected with either control or WAVE2 suppression vectors and allowed to recover for 72 h. Cells were then stimulated with anti-CD3/CD28 for 0, 15, or 30 min and lysed, and β₁ integrin was immunoprecipitated. Bound proteins were eluted, separated by SDS-PAGE, and transferred, and membranes were subsequently blotted for talin, vinculin, and β₁ integrin. Whole-cell extracts (WCE) were also analyzed for expression of talin, vinculin, β₁ integrin, and WAVE2.

WAVE2 and vinculin are required for talin localization to the IS.

Our finding that WAVE2 is required for both integrin localization to the IS and the interaction of talin with the β₁ integrin prompted us to investigate the requirement of WAVE2 in the localization of talin to the IS. Since the VCA domain of WAVE2 regulated actin accumulation at the IS, it seemed plausible that these mechanisms could be linked and that talin localization may be dependent on WAVE2-mediated actin polymerization. To test this hypothesis, both control cells and WAVE2-suppressed T cells were allowed to form conjugates with SEE-pulsed Raji B cells and were scored for their ability to localize talin to the IS. In control cells, talin localization was apparent, with nearly all of the protein localizing to the synapse in response to TCR stimulation (Fig. 10A). In contrast, in T cells lacking WAVE2, talin remained in both the cytoplasm and around the entire periphery of the cell. In addition, the VCA domain of WAVE2 regulated this process, since a WT version of WAVE2 rescued the ability to localize talin to the IS, whereas the ΔVCA version did not (Fig. 10A). Overall, these data suggest that the VCA domain of WAVE2, which regulates F-actin polymerization and integrin activation, is also required for talin localization to the IS.

FIG. 10.

The VCA domain of WAVE2 and vinculin are required for talin localization and activation-dependent high-affinity integrin binding. (A) Jurkat T cells were transfected with control, WAVE2 suppression, WT WAVE2 reexpression, or ΔVCA WAVE2 reexpression vectors and allowed to recover for 72 h. T cells (green) were then allowed to form conjugates with CMAC-stained Raji B cells (blue) that were pulsed with SEE, allowed to adhere to poly-l-lysine coated coverslips, fixed, and stained with antibody against talin (red) to visualize accumulation at the IS. Localization of talin was quantified using 50 random conjugates as described in Materials and Methods; data represent the average results of three independent experiments. (B) Same as panel A except that cells were transfected with control or vinculin suppression vectors. Localization was quantified as described for panel A. (C) Jurkat T cells were transfected with control, WAVE2 suppression, or vinculin suppression vectors and allowed to recover for 72 h. Cells were then either left unstimulated or stimulated with 50 ng/ml PMA or 1 mM Mn²⁺ for 10 min in the presence of soluble VCAM-1-Fc. Cells were then diluted and immediately fixed with paraformaldehyde and subsequently stained to detect binding of soluble VCAM-1 as described in Materials and Methods. Cells were analyzed by flow cytometry to detect binding of VCAM-1 and separated based on GFP expression levels. Gated populations indicate cells that bind with either high (right) or low (left) affinity to VCAM-1. High, Low, and Neg (negative) (listed to the left of each stimulation condition panel) refer to the GFP-expressing populations being analyzed in the experiment.

It is of interest that vinculin suppression did not affect the accumulation of F-actin or integrins at the IS but did affect TCR-stimulated integrin-mediated adhesion. Since vinculin forms a complex with both WAVE2-ARP2/3 and talin, we reasoned that vinculin might affect integrin activation in T cells by localizing talin to the IS. Indeed, in contrast to vector-transfected cell, localization of talin to the contact site between T cells and APCs is significantly diminished in vinculin-suppressed T cells (Fig. 10B). These data suggest that vinculin regulates integrin-dependent adhesion in T cells in part through its ability to stably recruit talin to the IS.

WAVE2 and vinculin are required for integrin affinity modulation.

Integrin activation is regulated by both the clustering of individual subunits on the cell surface (avidity) and conformational changes within the integrin heterodimer itself (affinity) (25). To test whether WAVE2 and vinculin regulated changes in integrin affinity, we analyzed these cells for their ability to bind soluble VCAM-1 in response to PMA stimulation. As shown in Fig. 10C, stimulation with PMA results in increased binding of VCAM-1 to vector control cells compared to unstimulated cells in all (high-GFP, low-GFP, and GFP-negative) cell populations analyzed. However, both WAVE2- and vinculin-suppressed high-GFP-expressing T cells are less efficient at binding soluble VCAM-1 in response to PMA stimulation than control cells (Fig. 10C). In contrast, all three transfected-cell populations were able to bind soluble VCAM-1 at similar levels when treated with Mn²⁺ (regardless of GFP expression level), which directly induces the high-affinity conformation of β₁ integrins. These data suggest that in T cells, both WAVE2 and vinculin are required to regulate intracellular signaling pathways leading to changes in integrin affinity.

DISCUSSION

T-cell activation is dependent upon stable conjugation with an APC that may persist for several hours (21). During this time, TCR engagement with a peptide-major histocompatibility complex found on the APC initiates signaling cascades that lead to the transcription of various target genes required for activation of the responding T cell. Besides the effects on gene transcription, TCR stimulation also initiates the activation of cell surface integrins, which in turn mediate tight conjugation between the T cell and APC required for sustained signaling. The data presented here indicate that the adaptor protein WAVE2 is a critical regulator of integrin function in T cells through its activity in directly linking ARP2/3-mediated de novo actin polymerization initiated by the TCR to the integrin scaffold proteins vinculin and talin (Fig. 11).

FIG. 11.

Proposed model for the role of WAVE2, ARP2/3, vinculin, talin, and actin polymerization in TCR-mediated integrin activation. (A) In resting cells, WAVE2, ARP2/3, vinculin, and talin are free of association and cell surface integrins are distributed evenly on the cell surface. (B) Stimulation of the TCR causes the activation of several signaling cascades leading to the activation of WAVE2, which then binds ARP2/3 and promotes de novo actin polymerization for the forming IS. Vinculin localizes to the IS through a direct interaction with WAVE2 and the ARP2/3 complex, and integrins begin to localize to the pSMAC. (C) Finally, talin (probably through its association with vinculin and F-actin) localizes and binds both β₁ and β₂ integrins, resulting in the high-affinity conformation at the IS. The newly formed F-actin generated by WAVE2 probably also serves as a binding partner for both vinculin and talin, providing the structural integrity required for stabilizing the integrin at the pSMAC.

Our data demonstrate that several key signaling pathways required for T-cell activation of integrins remain intact in WAVE2-suppressed cells and thus suggest that WAVE2 controls integrin activation downstream of these TCR signals. This finding is supported by the fact that PMA stimulation is unable to induce adhesion in T cells lacking WAVE2 or expressing WAVE2 that lacks the VCA domain. In fact, PMA induces adhesion in cells missing TCR signaling components, including ZAP-70 (10), LAT (14), VAV1 (26), ITK (11), PLCγ1 (24), and ADAP (18, 38), and is thought to initiate adhesion through the direct activation of PKCs (9, 41) and possibly other signaling molecules. Our results further suggest that WAVE2 may be one of the target molecules stimulated by PMA, leading to integrin activation. Interestingly, WAVE2 phosphorylation can directly result from PMA stimulation, independently of TCR ligation, through a mechanism downstream of extracellular signal-regulated kinase and PKC activation (31, 36). Although it remains to be formally demonstrated that phosphorylation of WAVE2 by PMA leads to its activation, the TCR-stimulated phosphorylation kinetics of WAVE2 mirror that of its association with vinculin and talin.

Our data demonstrate that the ability of vinculin to associate with WAVE2 and ARP2/3 is essential for TCR-mediated integrin activation. This is in contrast to a previous report of investigations of fibroblasts suggesting that the association of vinculin with ARP2/3 was important for lamellipodial formation but not adhesion (7). However, that study examined cell adhesion in the absence of any receptor stimulation, whereas TCR stimulation results in a substantial increase in adhesion to integrin ligands. Clearly, vinculin is providing a link between the actin cytoskeleton and components of integrin activation such as talin. Both vinculin and talin bind F-actin filaments, but the origin of these filaments has not been established. The data presented herein suggest that WAVE2-mediated F-actin nucleation through ARP2/3 is required for integrin clustering and the recruitment of vinculin and talin. Although it is clear that the actin cytoskeleton is necessary for integrin function, it is also apparent that a simple absence of gross F-actin at the IS does not necessarily result in decreased adhesion. For example, T cells in which the actin regulators dynamin 2 (15) and HS1 (17) have been suppressed display decreased F-actin accumulation at the IS while having no effect on conjugation. Therefore, WAVE2, through its association with vinculin, may be providing the specific de novo F-actin polymerization required for stabilization of talin and the integrin in the cell membrane.

The interaction between WAVE2, ARP2/3, vinculin, and talin provides a physical link between the actin cytoskeleton and the integrin scaffold network downstream of TCR engagement. Moreover, our data suggest that WAVE2-mediated ARP2/3 F-actin nucleation is critical for integrin clustering at the IS as well as providing the interaction with both vinculin and talin. Moreover, vinculin, although not required for F-actin and integrin accumulation at the IS, is required for the recruitment of talin to the IS. This most likely causes the defect in integrin-mediated adhesion, as talin is a critical regulator of integrin-mediated adhesion in both T cells and other cell types and is believed to induce the active conformation of the integrin through direct binding to integrin β subunit cytoplasmic tails (1, 42, 43). In contrast to our results, RNA interference-mediated suppression of talin has recently been shown to be critical for integrin localization at the IS (42). This suggests additional complexity in the functions of vinculin and talin in integrin activation. However, talin, like vinculin, is not required for F-actin accumulation at the IS. This is consistent with our model suggesting that vinculin is functioning to regulate TCR-stimulated integrin activation by recruiting talin to areas of de novo F-actin polymerization which are being generated by newly formed WAVE2-ARP2/3 complexes.

The present report also provides additional evidence for different biological functions of WAVE2 and another VCA domain-containing protein, WASP. Previously it has been demonstrated that WASP^−/− mouse T cells display no defects in cell adhesion, including the formation of TCR-APC conjugates, TCR-mediated adhesion to fibronectin, polymerization of F-actin at the IS, or the ability to cluster integrins in response to TCR stimulation (3, 26). However, these cells are still impaired in certain aspects of T-cell activation and cannot efficiently internalize TCRs after stimulation (29), another F-actin-dependent process. Interestingly, our data also suggest that the VCA domain of WAVE2 is specifically required to link the Arp2/3 complex to vinculin and does not occur in the absence of WAVE2 even though the VCA domain of WASP is still present. Therefore, WAVE2 and WASP appear to control distinct actin-dependent cellular functions that are required for T-cell activation. Our data further suggest that integrin avidity is a WAVE2-ARP2/3-dependent process. It is of interest that integrin avidity relies heavily on the activation of the Ras family GTPase, RAP1, which induces integrin clustering through the activation of recently characterized downstream effectors (22, 23). It will be of interest to determine whether the activation of RAP1 and WAVE2-mediated de novo actin polymerization are mutually exclusive or whether the two processes are linked in order to induce integrin clustering.

Surprisingly, it is now apparent that the ability of WAVE2 to control calcium release-activated calcium channel-mediated calcium entry appears to be VCA domain independent. This is consistent with our recent observation that suppression of ARP2/3 in T cells does not affect TCR-stimulated calcium mobilization (16). While this finding clearly establishes the importance of the VCA domain in controlling integrin activation, it now raises questions as to how WAVE2 regulates calcium flux independent of its established effector functions. It has previously been demonstrated that treatment of T cells with the actin-destabilizing agent cytochalasin D impairs NFAT-mediated gene transcription (19) but has been shown to augment calcium signaling and nuclear localization of NFAT in other studies (39). Although these studies demonstrate opposing roles for the actin cytoskeleton in regulating TCR-mediated calcium flux, they still suggest that the actin cytoskeleton plays a role. It will be of interest to uncover the VCA-independent mechanism by which WAVE2 or its associated complex members regulate calcium signaling and also to identify other actin nucleators that may play a role in this process.

In summary, this report establishes a molecular mechanism by which WAVE2 regulates integrin activation in response to TCR stimulation. We propose that T cells spatially and temporally regulate integrin avidity and affinity by linking WAVE2-mediated, ARP2/3-dependent de novo actin polymerization to the recruitment of vinculin and talin. This permits both integrin recruitment to the IS and affinity modulation of integrins at the IS. Continued research examining the spatial and temporal regulation of actin-regulatory proteins involved in inside-out signaling to integrin activation, as well as the identification of new binding partners, will no doubt continue to enhance our understanding of the contribution of the actin cytoskeleton to this important cellular process.