The N-terminal 20-Amino Acid Region of Guanine Nucleotide Exchange Factor Vav1 Plays a Distinguished Role in T Cell Receptor-mediated Calcium Signaling*

Background: Vav1 is indispensable for T cell calcium mobilization despite of its co-existing isoforms. Results: The N-terminal 20-amino acid region of Vav1 was identified to be essential for calmodulin-binding and TCR-induced calcium signaling. Conclusion: The interaction with calmodulin distinguishes Vav1 from other Vav family members in TCR-mediated calcium release. Significance: This study highlights the irreplaceable role of Vav1 in T cell calcium signaling. Vav1 is a guanine nucleotide exchange factor (GEF) specifically expressed in hematopoietic cells. It consists of multiple structural domains and plays important roles in T cell activation. The other highly conserved isoforms of Vav family, Vav2 and Vav3, are ubiquitously expressed in human tissues including lymphocytes. All three Vav proteins activate Rho family small GTPases, which are involved in a variety of biological processes during T cell activation. Intensive studies have demonstrated that Vav1 is indispensable for T cell receptor (TCR)-mediated signal transduction, whereas Vav2 and Vav3 function as GEFs that overlap with Vav1 on TCR-induced cytoskeleton reorganization. T cells lacking Vav1 exhibited severe defect in TCR-mediated calcium elevation, indicating that the co-existing Vav2 and Vav3 did not compensate Vav1 in calcium signaling. What is the functional particularity of Vav1 in lymphocytes? In this study, we identified the N-terminal 20 amino acids of Vav1 in the calponin homology (CH) domain to be essential for its interaction with calmodulin (CaM) that leads to TCR-induced calcium mobilization. Substitution of the 1–20 amino acids of Vav1 with those of Vav2 or Vav3 abolished the association with CaM, and the N-terminal mutations of Vav1 failed to potentiate normal TCR-induced calcium mobilization, that in turn, suspended nuclear factor of activated T cells (NFAT) activation and IL-2 production. This study highlights the importance of the N-terminal 20 aa of Vav1 for CaM binding, and provides new insights into the distinguished and irreplaceable role of Vav1 in T cell activation and signal transduction.


Vav1 is a guanine nucleotide exchange factor (GEF) specifically expressed in hematopoietic cells. It consists of multiple structural domains and plays important roles in T cell activation. The other highly conserved isoforms of Vav family, Vav2 and Vav3, are ubiquitously expressed in human tissues including lymphocytes. All three Vav proteins activate Rho family small GTPases, which are involved in a variety of biological processes during T cell activation. Intensive studies have demonstrated that Vav1 is indispensable for T cell receptor (TCR)mediated signal transduction, whereas Vav2 and Vav3 function as GEFs that overlap with Vav1 on TCR-induced cytoskeleton reorganization. T cells lacking Vav1 exhibited severe defect in TCR-mediated calcium elevation, indicating that the co-existing Vav2 and Vav3 did not compensate Vav1 in calcium signaling. What is the functional particularity of Vav1 in lymphocytes?
In this study, we identified the N-terminal 20 amino acids of Vav1 in the calponin homology (CH) domain to be essential for its interaction with calmodulin (CaM) that leads to TCR-induced calcium mobilization. Substitution of the 1-20 amino acids of Vav1 with those of Vav2 or Vav3 abolished the association with CaM, and the N-terminal mutations of Vav1 failed to potentiate normal TCR-induced calcium mobilization, that in turn, suspended nuclear factor of activated T

cells (NFAT) activation and IL-2 production. This study highlights the importance of the N-terminal 20 aa of Vav1 for CaM binding, and provides new insights into the distinguished and irreplaceable role of Vav1 in T cell activation and signal transduction.
Vav1 is a guanine nucleotide exchange factor (GEF) 2 specifically expressed in hematopoietic cells. It contains multifunc-tional domains, a calponin homology (CH) domain, an acidic domain, a Dbl homology (DH) domain, a pleckstrin homology domain, a cysteine-rich motif, and Src homology domains (SH3-SH2-SH3). Upon T cell activation, Vav1 is activated and functions as a GEF for Rho family small GTPases, which play essential roles in actin cytoskeleton rearrangement and the formation of immune synapse (1)(2)(3)(4). Besides Vav1, two other isoforms of Vav family, Vav2 and Vav3, exist abundantly in human T cells and exhibit functional redundancy with Vav1 in regard to the activation of small GTPases (5-7). As reported previously, T cells derived from knock-in mice with GEF-negative Vav1 mutant displayed intact TCR-stimulated calcium mobilization, ERK1/2 activation, T cell proliferation, and cytokine secretion (8,9), suggesting that the above events were GEF independent. On the other hand, Vav1 lacking its CH domain failed to potentiate calcium signaling and NFAT activation in T cell and NK cells (10 -12), even though its GEF activity was enhanced by removing the autoinhibitory CH domain (13). Thereby, the CH domain of Vav1 may act in TCR-mediated calcium signaling independently of the GEF activity.
Calcium mobilization is one of the earliest and pivotal events in TCR-induced cell activation. The intracellular calcium concentration is an essential factor to facilitate cellular signaling and gene transcription in immune responses (14,15). Accumulated evidences demonstrated that Vav1 played a critical role in regulating TCR-induced calcium mobilization (5,6,12), whereas the roles of Vav2 and Vav3 in calcium flux have rarely been reported. Data from vav1/2/3 triple knock-out mice suggested limited involvement of Vav2 and Vav3 in calcium mobilization (6). And those two isoforms appeared little relevancy to the TCR-induced NFAT activation (16 -19).
To date, studies have been focused on the CH domain of Vav1 and several CH-associated proteins were revealed, such as lymphoid-specific guanine dissociation inhibitor (20,21) and phospholipase C␥1 (22). We previously reported that the CH domain was indispensable for Vav1-CaM interaction, and that the complex potentiated calcium release in a GEF-independent manner (11). However, the precise role of Vav1 that is different from other Vav proteins still remains to be investigated.
In this study, we identified the N-terminal 1-20 amino acids (aa) of Vav1 to be an essential region for its interaction with CaM and function in calcium flux upon TCR engagement. Mutations of Vav1 with a substitution of the 1-20 aa fragment by that of Vav2 or Vav3 failed in binding to CaM and potentiating TCR-induced calcium mobilization. These data highlight a key region within the CH domain of Vav1 that endows Vav1 with a distinguished function from the other coexisting Vav family members. This study provides new insights into the irreplaceable role of Vav1 in T cell activation and signal transduction.

EXPERIMENTAL PROCEDURES
Reagents-Antibodies against Vav1 and CaM were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-Vav2 antibody and anti-Vav3 antibody were described previously (5). The anti-␣-tubulin antibody, CaM-agarose beads, and thapsigargin (TG) were purchased from Sigma. The anti-CD3 monoclonal antibody OKT3 and goat anti-mouse IgG were purchased from Biolegend (San Diego, CA). Fura-2-AM was purchased from Invitrogen.
Plasmid DNA Constructs-The plasmids encoding Vav1 and the CH-truncated mutation, Vav1⌬CH, were prepared in our laboratory as described (11). Other constructs for the expression of N-terminal-truncated mutants of Vav1 were derived from plasmid encoding wild type Vav1 by polymerase chain reactions (PCR) with the corresponding oligonucleotides, and the mutant fragments were subcloned into pcDNA4/HisMax.C vector at BamHI and XhoI restriction sites. The cDNA of N-terminal substituted Vav1 mutants were amplified by PCR and subcloned into pcDNA4/HisMax at BamHI and XhoI. FLAGtagged Vav2 and Vav3 or their N-terminal-substituted mutations were constructed by subcloning the PCR-amplified cDNA into the pCMV-Tag2B vector at EcoRI and HindIII restriction sites. For the lentiviral transduction of Jurkat or J.Vav1 cells, Vav1 or Vav1 mutations were subcloned into pCDH-MCS1-EF1-Puro lentivector (provided by Dr. R. Xiang, School of Medicine, Nankai University), respectively. The shRNA sequences targeting Vav family members or random nontargeting control RNA were designed in reference to the published sequences (23)(24)(25). Oligonucleotides were synthesized accordingly and cloned into pLKO.1 vector, respectively. All the constructs were confirmed by DNA sequencing.
Cell Culture, Transfection, and Stimulation-293T and HeLa cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS). Jurkat (clone E6) T cells were originally from ATCC, and J.Vav1 cells were obtained by somatic gene targeting of Jurkat as described previously (5), and cultured in RPMI 1640 medium containing 10% (v/v) FBS. For transfection of Jurkat or J.Vav1 cells, 2 ϫ 10 7 cells were collected and resuspended in RPMI 1640 medium containing 30 g of total DNA, and subjected to electroporation using a ECM830 Electro-square Porator (BTX Inc., San Diego). The stimulation of Jurkat and J.Vav1 T cells were performed by adding goat anti-mouse IgG with or without (as control) OKT3 to the final concentration of 1 g/ml, respectively. And intracellular calcium release was induced by TG at 1 M.
Lentivirus-based Transduction-For reconstitution of J.Vav1 cells with Vav1 and its variants, or knockdown of the Vav proteins in Jurkat cells, the lentiviral-based particles were made with 293T cells. Basically, 293T cells were transfected with the coding DNA of Vav1 mutants constructed in pCDH-MCS1-EF1-Puro or shRNA in pLKO.1, respectively, together with 4 g of pCMV-VSV-G, 4 g of pMDLg/pRRE, and 4 g of pRSV-REV to produce pseudoviral particles. At 48 h post-transfection, the supernatants were collected and viruses were harvested by centrifugation at 20,000 ϫ g for 1.5 h. The virus-containing pellets were resuspended in RPMI 1640 supplemented with 10% FBS and incubated with J.Vav1 cells at 37°C for 24 h. To obtain cells stably expressing variant mutations or shRNA, the transduced cells were selected with 0.25 g/ml of puromycin for 7 days.
Intracellular Calcium Measurement-Cells were loaded with 5 M Fura-2 AM in RPMI 1640 medium for 30 min at 37°C, then washed with 2 mM Ca 2ϩ Ringer's solution containing 145 mM NaCl, 1 mM MgCl 2 , 4.5 mM KCl, 10 mM glucose, 10 mM Hepes, pH 7.4, and 2 mM CaCl 2 , and kept at room temperature for 20 min. Cells were subjected to spectrophotometer analyses with excitation wavelengths of 340 and 380 nm, and an emission intensity at wavelength of 508 nm was simultaneously recorded by Synergy4 spectrometer (BioTek). For each measurement, the cells were pre-warmed for 5 min at 37°C and a baseline was recorded before stimulation by OKT3 as described above. The intracellular free Ca 2ϩ concentration was presented as the intensity ratio of 340 nm/380 nm.
Luciferase Reporter Assay-J.Vav1 T cells were transfected with reporter constructs, pNFAT-luc and TK-Renilla as described (26,27), together with plasmids encoding Vav1 or mutants, respectively. At 24 h post-transfection, cells were harvested and stimulated with or without 1 g/ml of OKT3 for 6 h. Cell lysates were prepared and Luciferase activity was measured with Dual Luciferase assay kit (Promega, Madison, WI) by TD20/20 luminometer (Turner Designs Inc, Sunnyvale, CA) as instructed by the manufacturer. Normalized luciferase activity was presented as the ratio of firefly luciferase activity to Renilla luciferase activity.
Bimolecular Fluorescence Complementation (BiFC) Assay-The BiFC method was adopted to visualize protein interactions in live cells (28 -30). CaM was fused to the carboxyl-terminal fragment (YC) of YFP. The wild type Vav1 or Vav1 mutations were fused to the N-terminal fragment (YN) of YFP. YC-CaM and the YN-Vav1 variants were co-transfected into individual HeLa cells. At 24 h post-transfection, the transfected cells were incubated for another 4 h at room temperature for the maturation of the fluorophore, and then subjected to fluorescence microscopy and FACS analysis. The mean fluorescence intensity was determined by FACS analysis.
Measurement of IL-2 Production-Cells were cultured in 24-well tissue culture plates at 1 ϫ 10 6 per ml and remained unstimulated or stimulated with anti-CD3 antibody for 24 h. The cultured media were then collected by centrifugation and the secreted IL-2 was determined by ELISA using human IL-2 ELISA kit (NeoBioscience, China). The data from three independent experiments were subjected to statistical analysis.

TCR-mediated Calcium Mobilization Requires Vav1, but Not
Vav2 or Vav3-There are three Vav family proteins, Vav1 is specific for hematopoietic cells, whereas Vav2 and Vav3 are also expressed. All three Vav proteins share high homology in structure and bare functional redundancy as guanine nucleotide exchange factors (6, 7). However, it has been reported that T cells and NK cells lacking Vav1, even though Vav2 and Vav3 were present, exhibited severe defect in receptor-induced cell signaling, especially the calcium related events, such as the activation of NFAT (5, 6, 10). Thus the defective calcium profiles in vav1-null cells reflect a GEF-independent function of Vav1, which is not possessed by Vav2 and Vav3. In this regard, we applied an lentiviral system carrying shRNA to knockdown the endogenous Vav1, Vav2, or Vav3, respectively, in Jurkat cells. The homogeneity of vav-knockdown cells was achieved by puromycin-resistant selection and the calcium profiles in response to TCR stimulation were recorded (Fig. 1). Upon OKT3 stimulation, vav1-knockdown cells (shVav1) displayed a severe defect in calcium mobilization (dash-dotted line) in comparison with the cells harboring scrambled shRNA (shCtrl, Fig. 1A, solid line). The cells with reduced Vav2 or Vav3 appeared normal in calcium elevation (Fig. 1A, dashed line and dotted line) as control cells, suggesting that Vav2 and Vav3 were not necessary in TCR-induced calcium elevation. The expressions of the indicated Vav proteins were dramatically disrupted by shRNAs (Fig. 1B, right lanes) in contrast to that in control cells (Fig. 1B, left lanes). The tubulin was monitored as loading controls. We also verified that the suppression effects of shRNAs on Vav proteins could be maintained for at least 20 days (data not shown). The above data demonstrate an irreplaceable role of Vav1 in TCR-mediated calcium mobilization.
It was reported previously that TCR-stimulated calcium flux was defective in cells expressing CH-deleted Vav1, and the CH domain featured Vav1, a GEF-independent function in T cell calcium signaling (5,10). We therefore compared the amino acid sequences of CH domains as aligned in Fig. 2A. The CH domain similarity of Vav1 versus Vav2 is 81%, and Vav1 versus Vav3 is 88%. The secondary structure prediction was performed using PRALINE multiple sequence alignment software, and no difference between Vav1 and Vav2/3 was revealed. Concurrently, our work reported that Vav1 binds to CaM via the CH domain to potentiate calcium signaling upon TCR stimulation (11). We thus examined the interaction between individual Vav proteins and CaM by CaM-agarose pulldown assay. As shown in Fig. 2B, three Vav members were detected in Jurkat cell lysates (Fig. 2B, left lanes). Only Vav1 was immobilized by CaM-agarose, whereas Vav2 and Vav3 appeared negative in CaM binding (Fig. 2B, right lanes), regardless of the high similarities in their CH domain. These data suggest that the CH domain provides a unique structural basis of Vav1 to interact with CaM, and thereby renders Vav1 a distinguished role in TCR-activated calcium mobilization.
The N-terminal 20 Amino Acids of Vav1 Are Identified for CaM Binding-As the CH domain of Vav1 contains a stretch of 116 amino acids, we aim to identify the region responsible for its interaction with CaM. A series of Vav1 mutations at its CH domain were constructed, and the representative mutations   FEBRUARY 8, 2013 • VOLUME 288 • NUMBER 6 JOURNAL OF BIOLOGICAL CHEMISTRY 3779 depicted in Fig. 3A, ⌬N20 (a deletion of amino acids 1-20), ⌬N70 (a deletion of amino acids 1-70), and CH domain-deleted mutant (Vav1⌬CH). CaM-agarose pulldown assays were performed and the immobilized contents were resolved by Western blot analysis. As shown in Fig. 3B (upper panel), wild type Vav1 was immobilized by CaM-agarose (left lane), whereas the truncated variants showed no association with CaM. The expressions of the Vav1 variants were monitored and comparable amounts of cell lysates were loaded in the pulldown assay (Fig. 3B, lower panel). As the minimum truncation of 20 aa abrogated Vav1-CaM association, we concluded that the stretch of 20 aa of Vav1 was essential for the interaction between Vav1 and CaM.

N Terminus of Vav1 Plays a Role in Calcium Signaling
To verify the association of Vav1 and CaM in living cells, we adopted a BiFC assay, which has been widely applied to examine intracellular protein interactions (28 -30). The wild type Vav1 and Vav1⌬N20 were expressed as fusion proteins with the N-fragment of YFP (YN), respectively, and CaM was fused with YC (YC-CaM). The cells transfected with the indicated constructs were visualized by fluorescence microscopy and complex formation was analyzed by flow cytometry. The mean fluorescence intensity reflected the relative binding efficiency of the two fusion proteins (29). As shown in Fig. 3C, fluorescence in the control samples was undetectable (Fig. 3C, first and third  top panels), indicating no association between YN and YC fragments. In contrast, intense fluorescence was observed in cells co-transfected with YN-Vav1 and YC-CaM (Fig. 3C, second top  panel), with a mean fluorescence intensity of 512, manifesting the complexation of Vav1 and CaM in the cytoplasma. Meanwhile, barely detectable fluorescence was seen in cells expressing YN-Vav1⌬N20 and YC-CaM (Fig. 3C, bottom panel), reflecting a defect of Vav1⌬N20 in binding to CaM. The com-parable expression levels of the fusion proteins were analyzed by Western blot (Fig. 3D). The above analyses by the CaMagarose pulldown assay and BiFC approaches suggest that the N-terminal 20 aa of Vav1 are indispensable for the interaction with CaM.
The Segment of 1-20 aa Is Required for TCR-stimulated Calcium Flux-We next aimed to see whether the N-terminal 20 aa of Vav1 functionally affected calcium signaling during T cell stimulation. To obtain homogeneous expressions of Vav1 and mutations for monitoring cellular calcium fluctuation, we constructed lentivirus-based expression vectors encoding wild type Vav1, Vav1⌬CH, Vav1⌬N20, or vector alone as a control, respectively. The vav1-null cell line, J.Vav1, was used for transduction, and expression of the Vav1 variants was verified by Western blot (Fig. 4A). The intracellular Ca 2ϩ profiles were recorded upon stimulation with OKT3 (Fig. 4B). Control cells lacking Vav1 or cells with CH-truncated Vav1 showed a limited increase (Fig. 4B, solid or dotted line), and cells containing the full-length Vav1 presented a rapid elevation of intracellular calcium upon TCR activation (Fig. 4B, dashed line). In comparison, cells with Vav1⌬N20 exhibited attenuated elevation (Fig.  4B, dash-dotted line), which resembled the control group or the cells with Vav1⌬CH, indicating that the N-terminal 20 aa were the determinant region for the TCR-mediated calcium response.
Calcium profile is shaped by a combination of calcium release from intracellular stores, such as endoplasmic reticulum, and the extracellular calcium influx through the calcium release-activated calcium channel (31). We further investigated the effect of 1-20 aa of Vav1 on calcium release. T cell stimulation was carried out in buffer containing EGTA to deplete the extracellular Ca 2ϩ . The Ca 2ϩ spike in this condition would present only the Ca 2ϩ release from intracellular stores. As seen in Fig. 4C, J.Vav1 cells reconstituted with Vav1⌬CH and Vav1⌬N20 presented severe defects in OKT3-stimulated calcium release (dash-dotted line), whereas a dramatic peak was observed in cells with wild type Vav1 (dashed line). To rule out the possibility that reconstituted cell lines bared different levels of calcium stores, we treated cells with TG, a chemical that leads to passive calcium depletion from the endoplasmic reticulum lumen into the cytoplasm (32). As shown in Fig. 4D, in the absence of extracellular Ca 2ϩ , the magnitude of calcium elevation in TG-treated Vav1⌬N20 or Vav1⌬CH-reconstituted cells resembled that of cells bearing wild type Vav1 (Fig. 4D), indicating that the deletion of 1-20 aa or CH domain of Vav1 had no effect on the intracellular Ca 2ϩ stores. The above data suggest that the N-terminal 20 amino acids of Vav1 are essential for TCR-induced calcium release.
The N-terminal 20 Amino Acids of Vav1 Are Indispensable for TCR-induced NFAT Activation-The activation of NFAT is a major event following Ca 2ϩ elevation during T cell stimulation, and has been used as a functional indicator for calcium signaling (33). We therefore tested the effect of Vav1 and its N-terminal-truncated variants on the NFAT activity by luciferase reporter assays. As shown in Fig. 5, the normalized NFAT activity was induced by OKT3 (closed bars) to 6-fold higher than the basal level (open bars) in cells transfected with vector (pcDNA4), whereas nearly 25-fold above the basal level in cells transfected with wild type Vav1. In comparison, TCR-stimu-   FEBRUARY 8, 2013 • VOLUME 288 • NUMBER 6 lated NFAT activities were severely impaired in the presence of truncated mutants Vav1⌬N20 and Vav1⌬CH (Fig. 5A). The equivalent expression of each Vav1 variant was verified by Western blot (Fig. 5B). Therefore, the N-terminal 20 amino acids of Vav1 are required for the TCR-stimulated NFAT transcriptional activity.

N Terminus of Vav1 Plays a Role in Calcium Signaling
The N-terminal 20 Amino Acids of Vav1, but Not Other Vav Members, Determine the Association with CaM-As mentioned, there are three Vav family members in humans, Vav1, -2, and -3. They shared high similarity in primary structures, and their GEF functions for small GTPases exhibited redundancy to a certain extent (5-7). However, TCR-mediated calcium elevation was independent of GEF, and the existing Vav2 and Vav3 were incompetent to rescue calcium signaling in vav1-null cells (5,11). As 1-20 aa of Vav1 were involved in CaM binding, we next asked whether the stretches of the N-terminal 20 aa from the three Vav proteins were interchangeable. A variety of substitutions were designed to swap the N-terminal 20 aa of Vav proteins as depicted in Fig. 6, A and C. In Fig. 6A, the N-terminal 20 aa of Vav1 were replaced by those from Vav2 and Vav3, and designated as Vav1(Vav2-N20) and Vav1(Vav3-N20), respectively. These mutants, together with wild type Vav1 and Vav1⌬N20, were subjected to a CaM-agarose pulldown assay. As seen in Fig. 6B (upper panel), Vav1 with the N-terminal deletion or substitution by Vav2/3 failed to interact with CaM, whereas wild type Vav1 was immobilized by CaMagarose as positive control. The Vav1 and its mutants in the lysates were equally expressed as examined by Western blot (Fig. 6B, WCL panel).
Meanwhile, we wondered if Vav2 or Vav3 with the N-terminal 20 aa from Vav1 would gain the capability of CaM binding. The swapped mutations were constructed as shown in Fig. 6C, in which, the N-terminal 20 aa of Vav2 and Vav3 were replaced by 1-20 aa of Vav1, and designated Vav2(Vav1-N20) and Vav3(Vav1-N20), respectively. The CaM pulldown assay was performed as described above. As seen in Fig. 6D The N Terminus of Vav1 Is Essential for the Calcium Signal Lineage upon TCR Stimulation-Previous study demonstrated that the association of Vav1 with CaM is a prerequisite for TCR-stimulated Ca 2ϩ elevation. It was of interest to examine the roles of the two Vav1 N-terminal-substituted mutants in calcium signaling. First, we reconstituted J.Vav1 cells with lentiviruses carrying wild type Vav1, Vav1(Vav2-N20), Vav1(Vav3-N20), Vav1⌬N20, or vector to obtain homogenous expressions of the variants in J.Vav1 cells. The expressions of the indicated proteins were verified by Western blot (Fig. 7A). The calcium profiles of the reconstituted cells were recorded during T cell activation. As seen in Fig. 7B, compared with cells lacking Vav1 (vector, solid line), the wild type Vav1-reconstituted cells exhibited a significant elevation of intracellular Ca 2ϩ under OKT3 treatment (dashed line), whereas those cells harboring Vav1(Vav2-N20), Vav1(Vav3-N20), or Vav1⌬N20 exhibited similar Ca 2ϩ patterns as the cells with vector.
We further analyzed the NFAT activation by luciferase reporter assay in cells transfected by plasmids encoding wild type Vav1 or its N-terminal-substituted mutations as indicated (Fig. 7C). Following OKT3 treatment or no stimulation (NS), the NFAT luciferase activities of the transfected cells were measured. As shown in Fig. 7C, upon TCR stimulation, cells with wild type Vav1 presented 4-fold higher NFAT activity than cells with vector (pcDNA4) or Vav1⌬N20. The substituted mutation by either Vav2-N20 or Vav3-N20 had no significant restoration of NFAT as the wild type Vav1, even though the expressions of Vav1 derivatives were comparable as that of full-length Vav1 as monitored by the Western blot analysis (Fig. 7D).

The CaM Binding Region of Vav1(N-20) Is Indispensable for the Production of IL-2 in T Cell Activation-
The IL-2 secretion is a critical landmark following T cell activation (34). We therefore measured the secreted IL-2 by J.Vav1 cells reconstituted with wild type Vav1 or the mutations, which lack the functional N termini. As shown in Fig. 7E, all samples exhibited equivalent basal levels of IL-2 (open bars). Upon T cell activation, J.Vav1 cells expressing wild type Vav1 displayed at least 5-fold higher secreted IL-2 than those with the N-terminal 20 aa mutations or the control vector (dark bars). These data reveal that the N-terminal 20 aa of Vav1, rather than those of Vav2 or Vav3, are essential for TCR-induced IL-2 secretion.
Together, we identified the N-terminal 20 amino acids of Vav1 to be essential for its interaction with calmodulin and potentiation of TCR-induced calcium mobilization. Deletion or substitution of this 20-aa stretch of Vav1 abolished its functions in TCR-mediated transcriptional activity of NFAT and IL-2 production. The deficiencies of TCR-induced calcium mobilization and NFAT activation by the N-terminal mutants of Vav1 strongly suggest that the N-terminal 20 aa of Vav1 reckon for not only the unique feature among Vav proteins, but also the irreplaceable role of Vav1 in T lymphocytes.

DISCUSSION
It has been illustrated that the hematopoietic specific Vav1 plays a pivotal role in T cell activation by acting as a GDP-GTP exchange factor and a scaffold to recruit signal molecules to the immune synapse via its multiple structure domains (3,12). The CH domain of Vav1 was originally found to be related to transformation of fibroblasts (35). As the confined expression of Vav1 in lymphocytes, the indispensable role of the CH domain of Vav1 in intracellular calcium signaling has been suggested in NK and T cells (5,10). Several binding partners of the CH domain have been revealed, such as lymphoid-specific guanine The intracellular calcium concentrations were monitored for 5 min for each sample and plotted with the fluorescence emission ratio of the Ca 2ϩ bound to the free forms of Fura-2 as the vertical axis, and time as the horizontal axis. C, J.Vav1 cells were transfected with reporter plasmids, pNFAT-Luc and TK-Renilla, together with plasmid backbone (pcDNA4), or plasmids encoding Vav1, Vav1(⌬N20), Vav1(Vav2-N20), or Vav1(Vav3-N20) as indicated. At 24 h post-transfection, the cells were left unstimulated (NS) or stimulated with OKT3 for 6 h. The luciferase reporter assay was performed as described under "Experimental Procedures." The NFAT activity was presented as the ratio of the firefly luciferase activity to Renilla luciferase activity for each transfectant. Histograms represent the mean values from three independent experiments with S.D. D, cell lysates were analyzed by Western blot and one set of experiments was shown to indicate expression of the various Vav1 proteins for reporter assays. E, J.Vav1 cells were transduced with lentiviruses bearing various Vav1 proteins or vector backbone, respectively, and seeded at the same concentrations. Then the cells remained non-stimulated (NS) or stimulated with OKT3 for 24 h. The cell culture media were collected and IL-2 content was determined by ELISA. Three independent experiments were performed and are represented by histograms with the mean Ϯ S.D. from triplicate samples. FEBRUARY 8, 2013 • VOLUME 288 • NUMBER 6 dissociation inhibitor (20,21), phospholipase C␥1 (22), and CaM (11), and they were involved in TCR-induced calcium signaling. However, a line of evidence proposed that the Vav1 that participated in calcium signaling is GEF-independent (8,9,12). In addition, the CH-deleted Vav1, which possesses hyperactive GEF activity, could not potentiate calcium flux in response to T cell activation (10,11,36). Here, we identify the N-terminal 20-amino acid stretch of the CH domain to be responsible for Vav1 in the modulation of calcium flux via binding to CaM.

N Terminus of Vav1 Plays a Role in Calcium Signaling
Previous work (11) demonstrated that the interaction of Vav1 with CaM did not require its GEF activity. Recently, the high resolution structure by xtallography have revealed that the CH domain is a trans-inhibitory domain for Vav1 GEF activity, and amino acid residues in the interface of the CH and DH domains are key factors in autoinhibition (13). Nevertheless, the N-terminal 20 aa is located beyond the autoinhibitory region for the GEF, further supporting that CH binding to CaM is independent of its GEF activity.
Besides Vav1, two other Vav family members, Vav2 and Vav3, are expressed in all human cell types including lymphocytes, and their functions as GEF overlap with Vav1 in activating Rho/Rac GTPases, like Rac1 or Rac2 (37). It remains unsolved why Vav1 is particularly required in T cell activation. We herein addressed that Jurkat T cells with reduced expression of Vav2 or Vav3 by shRNA underwent normal calcium flux upon T cell activation, whereas knocking down the expression of Vav1 resulted in a defective calcium response (Fig. 1). Consistently, Vav2 was reported to have no function in controlling calcium signals in T cells (18), and Vav3 was demonstrated to play an unnecessary role on TCR-induced NFAT activation (7). Therefore the calcium signals evoked by TCR engagement are mainly contributed by Vav1, but not Vav2 or Vav3.
Although more than 80% similarity was revealed among the CH domains of Vav proteins, we only found Vav1 CH association with CaM. Nonetheless, as Vav proteins contain multiple functional domains, we cannot rule out the involvement of other regions or domains to contribute the binding to CaM. It was suggested that multiple domains of Vav1 synergize the formation and stabilization of the microclusters to promote T cell activation (12). Indeed, other motifs, such as the IQ consensus sequence for calmodulin interaction (38), does exist in Vav1. However, mutations of the IQ motif did not affect the association of Vav1 with CaM, nor the calcium profile following TCR ligation (data not shown). We speculate that the CH domain, especially the first 20-aa stretch of peptide, is essential for CaM binding, whereas the other domains of Vav1 may be required to cooperate with the CH domain for normal calcium flux. Further studies by crystallography of the Vav1-CaM complex would help to define the interaction sites between Vav1 and CaM.
In summary, we highlight the N-terminal 20 aa as a key region of Vav1 for association with CaM in T lymphocytes. As a result, this region endows Vav1 with a distinct function among the Vav proteins in antigen-induced T cell activation, including calcium flux, NFAT activation, and IL-2 production. Recently, the aberrant expressions of Vav1 were observed in several cancerous cells and associated with malignancy (39). The exploration for the distinct mechanisms of Vav1 may also contribute to the understanding of its pathological relevance.