|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
Blood, 15 April 2005, Vol. 105, No. 8, pp. 3026-3034. Prepublished online as a Blood First Edition Paper on December 23, 2004; DOI 10.1182/blood-2004-07-2925.
CHEMOKINES Control of lymphocyte shape and the chemotactic response by the GTP exchange factor VavFrom the Servicio de Inmunología, Hospital Universitario de la Princesa, Madrid, Spain; Centro de Investigación del Cáncer (CIC), Campus Miguel de Unamuno, Salamanca, Spain; and Facultad de Medicina, Universidad Autónoma de San Luis Potosí (UASLP), San Luis Potosí, Mexico.
Rho GTPases control many facets of cell polarity and migration; namely, the reorganization of the cellular cytoskeleton to extracellular stimuli. Rho GTPases are activated by GTP exchange factors (GEFs), which induce guanosine diphosphate (GDP) release and the stabilization of the nucleotide-free state. Thus, the role of GEFs in the regulation of the cellular response to extracellular cues during cell migration is a critical step of this process. In this report, we have analyzed the activation and subcellular localization of the hematopoietic GEF Vav in human peripheral blood lymphocytes stimulated with the chemokine stromal cell–derived factor-1 (SDF-1  ). We show a robust activation of Vav and its redistribution to motility-associated subcellular structures, and we provide biochemical evidence of the recruitment of Vav to the membrane of SDF-1–activated human lymphocytes, where it transiently interacts with the SDF-1 receptor CXCR4. Overexpression of a dominant negative form of Vav abolished lymphocyte polarization, actin polymerization, and migration. SDF-1–mediated cell polarization and migration also were impaired by overexpression of an active, oncogenic Vav, although the mechanism appears to be different. Together, our data postulate a pivotal role for Vav in the transmission of the migratory signal through the chemokine receptor CXCR4.
Leukocyte migration in and out of target tissues during homeostasis and inflammation is a finely regulated process mediated by many receptors, which regulate rolling, adhesion and/or detachment, and motility. Chemotactic receptors play an important role in the modulation of cell adhesion as well as in controlling the morphology of migrating leukocytes.1 In particular, chemokines are chemotactic cytokines that, acting through heterotrimeric G-protein–coupled receptors (GPCRs), regulate cell adhesion through cross-talk with integrin receptors and also modulate the morphology of migrating leukocytes.1,2
The chemokine stromal cell–derived factor-1 (SDF-1
Small GTPases are devices controlling the actin cytoskeleton (Etienne-Manneville and Hall19). The function of the 3 "classic" GTPases—Cdc42, Rac, and Rho—was determined in the mid-1990s, showing that different aspects of cell migration, such as filopodia extension (Cdc42), lamellipodia formation (Rac), and cell retraction and adhesion (Rho), are operated by these molecules.20-22 Small GTPases are thus at the core of actin regulation in migrating cells. However, the mechanisms by which SDF-1 The Vav subfamily is composed of 3 different genes: Vav1, which is restricted to hematopoietic cells; and Vav2 and Vav3, whose expression pattern is broader.27,28 The 3 of them are involved in lymphocyte ontogeny (Bustelo25 and Turner and Billadeau26), and there seems to be some degree of redundancy among them; for example, Vav2 can compensate Vav1 in some contexts, such as B-cell development and function.29 However, redundancy is not absolute; for instance, Vav1-deficient mice have lower numbers of T cells, due to defective pre–T-cell receptor (pre–TCR) signaling, that cannot be replaced by Vav2, which does not seem to participate in T-cell ontogeny.30 Regarding their ability to activate small GTPases, it seems that Vav1 is primarily a GEF for Rac,31 whereas Vav2 and Vav3 would act on Rho.28,32 Their ability to activate Cdc42 is controversial.33 To avoid aberrant signaling to the GTPases, these molecules are finely regulated. In the case of Vav, tyrosine phosphorylation of specific residues induces exposure of the GTPase binding site.34 Thus, Vav appears a likely candidate to mediate signal transduction from CXCR4 to the actin cytoskeleton and therefore control lymphocyte morphology.
In this report, we demonstrate Vav activation by SDF-1
Reagents
Rabbit polyclonal anti-Vav and anti–phospho-Tyr174 Vav (pY174) have been described elsewhere.34 Mouse monoclonal anti–intercellular adhesion molecule 3 (ICAM-3) antibody HP2/19 (IgG2a, PBL purification PBLs were purified as described elsewhere.15 Briefly, freshly prepared buffy coats from healthy donors were subjected to gradient centrifugation on Histopaque-1077 (Sigma), followed by 3 rounds of plastic adherence to remove monocytes. Populations were shown to contain less than 2% monocytes as shown by staining with anti-CD14. Approval was obtained from the Hospital Universitario de la Princesa Institutional Review Board for these studies. Informed consent was provided according to the Declaration of Helsinki. Immunofluorescence and confocal microscopy Indirect immunofluorescence assays were performed as described.15 Briefly, cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) containing 2 M sucrose and 2 mM MgCl2 for 10 minutes at room temperature and stained for 30 minutes with anti–ICAM-3 or anti-CXCR4 at 37°C followed by incubation with a 1:200 dilution of anti–mouse IgGs coupled to rhodamine red-X (Jackson Immuno Research Laboratories, West Grove, PA). Cells were then permeabilized in 0.5% Triton X-100 + 2% paraformaldehyde in PBS for 5 minutes, followed by incubation with either rabbit anti-Vav or anti–phospho-Tyr174 Vav and staining with highly cross-absorbed anti–rabbit IgGs coupled to Alexa 488. Cells were observed using a Leica DMR photomicroscope (Leica, Mannheim, Germany) with a 100 x/1.40-0.7 OIL CS objective, coupled to a COHU 4912-5010 CCD Camera (COHU, San Diego, CA). The acquisition software was Leica QFISH V2.1, and images were processed with Adobe Photoshop 7.0. For quantification of cell polarization, at least 300 transfected cells were counted in 4 independent experiments as assessed by ICAM-3 redistribution and formation of 2 well-defined morphological poles.15 Quantitative analysis of cell spreading was performed with ImageJ software (National Institutes of Health, Bethesda, MD) in at least 50 transfected cells from 3 independent experiments, as described.39 Immunoprecipitation experiments and Western blot
For analysis of Vav phosphorylation, 107 PBLs were stimulated with 10 nM SDF-1
For immunoprecipitation, 50 x 106 PBLs were resuspended in RPMI 1640 medium containing 0.1% bovine serum albumin (BSA) at 37°C. Cells were then stimulated with 10 nM SDF-1 Mice Vav1-deficient C57/BL10 mice40 were kindly provided by Dr V. Tybulewicz (National Institute for Medical Research, London, United Kingdom) and kept in the animal facility of the Centro de Investigación del Cáncer (Salamanca). Control littermates were from Charles River Laboratories (Wilmington, MA). Mice were processed according to institutional guidelines in compliance with international laws and policies. T-cell populations isolated from spleen and lymph nodes were decreased in both the spleens and lymph nodes of Vav1-deficient mice (50% ± 5% vs 22% ± 8%, n = 8 in each case), in agreement with previous reports.40 Migration assay Migration assays were performed in Boyden-modified chambers (Transwell, Costar, Cambridge, MA) as described elsewhere.18 Quantification of cell migration was performed by counting transfected and nontransfected cells in a FACScalibur flow cytometer (Becton Dickinson, Mountain View, CA) for 60 seconds. Fluorescence intensity was determined in a FACScalibur flow cytometer. Four cell subsets were defined according to green fluorescent protein (GFP) expression (NULL, LOW, MEDIUM, and HIGH), and the percentage of migrating cells was calculated for each expression interval as previously described.41 F-actin determination assay
Determination of the levels of polymerized actin was performed as described.17 Briefly, lymphocytes from wild-type or Vav1-deficient mice or PBLs from buffy coats were collected as described and resuspended in prewarmed RPMI 1640 medium at 2 x 106/mL, and 100 µL of cells were incubated or not with 10 nM SDF-1 PBL nucleofection Transfection of human PBLs was performed as follows: 12 x 106 cells were washed twice in ice-cold sterile PBS and resuspended in 100 µL nucleofection T-cell buffer (Amaxa Biosystems, Cologne, Germany). A total amount of 12 µg of the indicated cDNA was added and subjected to the U-14 nucleofection program in an Amaxa Nucleofector I. Cells were resuspended in 4 mL RPMI 1640 supplemented with 10% FBS without antibiotics. Expression efficiency was measured 24 hours after nucleofection, and cells were employed in the experiments indicated. Rac activation assay
GST-PAK-CRIB, which recognizes active Rac and Cdc42, was kindly donated by Dr John Collard (The Netherlands Cancer Institute, Amsterdam) and was prepared as described.42 Pull-down experiments were performed as follows: 5 x 106 of CXCR4-CFP stably expressing HEK-293 cells, kindly provided by Dr M. Mellado (Centro Nacional de Biotecnología, Madrid), were transfected with the calcium phosphate precipitation method with the different constructs of Vav and stimulated for 10 minutes with 10 nM SDF-1
Activation of Vav by SDF-1 in human PBL
Activation of Vav by SDF-1
It has been previously reported that adhesion of Jurkat T cells to fibronectin induces Vav phosphorylation, thus decreasing the pool of Vav that can be phosphorylated by other stimuli.43 To ascertain whether lymphocyte adhesion affected the extent to which SDF-1 can phosphorylate Vav, studies were performed similar to that shown in Figure 1A and comparing SDF-1–stimulated cells either in suspension or adhered to human fibronectin. Interestingly, we found that adhesion by itself induced a 3-fold phosphorylation of Vav, which was further enhanced by addition of SDF-1 up to 6-fold (Figure 1C). Dose-response experiments showed that the dose of chemoattractant required for maximal phosphorylation of Vav (Figure 1D) is within the range of 5 to 10 x 10–9 M (5-10 nM, 50-100 ng/mL) in adhered cells as well as in the case of suspended cells, which is in the same range of the maximal induction of chemotaxis and cell polarization by this chemokine (Figure 1E and Vicente-Manzanares et al15). Together, our data indicate that the Vav phosphorylation induced by integrin ligation is complementary to that induced by chemoattractants, pointing to a role for this molecule in both signaling pathways. Polarized activation of Vav and interaction with the chemokine receptor CXCR4
The subcellular localization of Vav upon activation with SDF-1
We and others have previously shown that the chemokine CXCR4 is partially clustered at the leading edge of migrating cells during the chemotactic response.9-12 Since Vav was clustered at the leading edge of SDF-1 –responding cells, we studied its possible interaction with CXCR4. Indirect immunofluorescence experiments with either anti-Vav or anti–phospho-Tyr174 Vav suggested their partial colocalization in SDF-1–stimulated cells (Figure 3A, arrowheads in insets). Remarkably, strong biochemical evidence was provided by CXCR4 immunoprecipitates of SDF-1–stimulated PBLs, in which we found a clear-cut SDF-1–dependent interaction of Vav with CXCR4 (Figure 3B). Such biochemical interaction was further confirmed by reciprocal immunoprecipitation, in which we were able to detect CXCR4 in Vav immunoprecipitates (Figure 3C). This fact suggests that CXCR4 probably recruits Vav to the membrane as part of a putative signalosome to transmit signals to the actin cytoskeleton and may regulate the morphology of the cell.
Lymphocytes from Vav1-deficient mice are not defective in responses to SDF-1
Vav1-deficient mice have proved a useful tool to dissect the role of GEFs in the development of the immune system and its responses.40,44,45 To assess the ability of Vav1-deficient lymphocytes to respond to SDF-1
A dominant negative form of Vav impairs lymphocyte polarization
To investigate the role of Vav in the regulation of lymphoid chemotaxis and cell shape, we obtained different forms of Vav1 fused to GFP (Figure 5A). These constructs exert different effects on Rac activation when overexpressed in HEK-293 cells bearing the SDF-1
Human PBLs were nucleofected with the constructions indicated (Figure 5A), allowed to adhere to fibronectin in the presence of 10 nM SDF-1 , and stained for the polarization marker ICAM-3. Remarkably, overexpression of GFP-Vav L213Q, which has been shown not to interact with Rac1, thus interrupting Rac1 signaling,37 resulted in abrogated cell polarization monitored by ICAM-3 redistribution to the uropod (Figure 5C,E). However, the distribution of the mutated molecule was very similar to that of wild-type Vav, suggesting its normal recruitment to the membrane (Figure 5C, green). As control, nucleofection of either GFP or GFP-Vav wild type exerted a negligible effect on SDF-1–induced cell polarization (Figure 5E). These data indicate that Vav L213Q interrupts the signaling from CXCR4 to the actin cytoskeleton that is required to initiate cell polarization, likely by inhibiting further signal transduction to small monomeric GTPases, namely Rac. Overexpression of oncogenic Vav impairs cell polarization by a mechanism different from Vav L213Q
A mutant of Vav lacking its N-terminus, Vav (
Regarding the effect of Vav on the morphology of the cell, although it has been repeatedly demonstrated that ectopic Rac1 activation by Rac-GEF or activated Rac overexpression induces cell spreading,38 no such spreading was observed in Vav (
Overexpression of dominant negative or oncogenic Vav impairs chemotaxis
To investigate whether lymphocyte shape changes resulted in impaired cell migration, chemotaxis experiments were performed in Boyden-modified chemotaxis chambers (Figure 7), defining expression gates as previously described.41 We found that none of the constructs exerted any effect on cell migration at low expression. However, Vav (
Small GTPases are crucial mediators of cytoskeletal-based shape changes in eukaryotic cells, and their regulation appears as a fundamental step to control complex processes such as cell migration. Directional migration, on the other hand, requires sustained extracellular cues, normally adopting the form of a gradient, soluble or immobilized, which implies 2 fundamental issues: (1) amplification of the signal provided by the gradient; and (2) continuous intracellular cycling of the signaling intermediates to ensure a maintained response.47 It is well known that small GTPases cycle between active and inactive state; such cycling depends on both translocation of the small GTPase to the subcellular location where it exerts its biologic effect and also on interactions with catalytic activators or inhibitors.19 Regulation of the first step is now being unveiled by advances in our understanding of the mechanisms that retain the molecule on the cytoplasm48 or its interaction with second messengers that provide positional information, such as phospholipids or site-restricted phosphorylation.49 Such processes are usually coordinated, that is, GEFs can be phosphorylated at specific cellular locations, thus rendering the GTPase active within that location. In agreement with this notion, this work shows that the chemokine SDF-1 induced tyrosine phosphorylation of the GEF Vav at Tyr174, a residue that has been consistently shown to regulate Vav activity; as such phosphorylation destroys the intramolecular interaction of the N-terminus of Vav with its C-terminus, which allows the Rac binding site to remain hidden and inactive. This is a more consistent indicator of Vav activation than those experiments in which Vav is immunoprecipitated and total phosphotyrosine revealed, as Vav can be phosphorylated in other tyrosine residues, whose implication in Vav activation is not so clear, even leading to its down-modulation.34 Phosphorylation of Vav at this residue is likely to be dependent of ZAP-70, since this kinase has been described to activate Vav in leukemic cells.24 Vav phosphorylation occurs in response to different extracellular stimuli, that is, growth factors, adhesion, etc. For cell migration to occur, 2 signals are required, which are cell adhesion and a promigratory stimulus. It has been previously shown that adhesion of Jurkat T cells to fibronectin induces Vav phosphorylation.43 Consistent with this, we found fibronectin-induced Tyr174 Vav phosphorylation in human PBLs. In addition, SDF-1 induced further phosphorylation of Vav at Tyr174, which suggests that adhesion per se is not able to induce maximal activation of Vav during cell migration, which is achieved by the combination of adhesive and chemoattractant stimuli.
Furthermore, we found that phosphorylated Vav is restricted to the leading edge of migrating cells and to the trailing edge or uropod. Activated Vav at the leading edge probably regulates its formation and perpetuation inducing the local activation of Rac and thus actin polymerization through a WAVE/Scar-Arp2/3–dependent mechanism50,51 and/or relaxation of myosin-dependent tension by Pak-dependent phosphorylation of myosin-II heavy chain,52 but the role of Vav activation at the uropod is currently unknown. However, Vav activation at the uropod may be a consequence of its interaction with the Syk kinase, which is an upstream regulator of Vav.53 In this regard, Syk kinase is recruited to the uropod, where it associates to an ITAM-like motif within ERM adaptor proteins.54 Thus, it is likely that Vav activation at the uropod may occur at adhesion molecule–dependent signalosomes including both ERMs and Syk. However, we have been unable to detect Vav in immunoprecipitates of adhesion molecules clustered at the uropod of the cell (data not shown). Therefore, it is likely that Vav presence in the uropod of the cell is associated with the formation of actin cables at the trailing edge of the cell, which are involved in adhesion turnover and retraction. On the other pole of the cell, SDF-1 Contrary to what occurs in the GRK-dependent complex, actin polymerization–inducing complexes are mainly dependent on heterotrimeric inhibitory G (Gi) proteins, since pertussis toxin–treated cells are still able to undergo CXCR4 internalization but not actin polymerization.55,56 It is interesting to note that Vav still remains biochemically bound to CXCR4 when Vav phosphorylation is declining (20 minutes), which suggests the existence of in situ signal limitation processes that switch off Vav prior to its dissociation from the receptor. In this regard, both CXCR4 and Vav have been shown to interact with different membrane and cytoplasmic phosphatases, which may play a role in signal turn-off.57,58
The role of Vav in the propagation of CXCR4 signaling to the cytoskeleton is highlighted by overexpression of a GEF-deficient form of Vav, in which the critical residue for interaction with Rac has been point-mutated. This form is still recruited to the membrane but blocks the polarizing response to SDF-1
In recent years, cells from Vav-deficient mice have proved useful tools to explore the immunobiology of these molecules. Having access to Vav1-deficient mice, we performed ex vivo migration and actin polymerization experiments with T and B cells from such mice. These mice showed reduced numbers of T cells, as previously described,40 but were devoid of other obvious defects. Cell polarization experiments proved extremely difficult to perform, probably due to the lack of an appropriate integrin-dependent substrate, although monocyte spreading, which occurs spontaneously and in the absence of integrin-dependent interactions, was unaffected in Vav1-deficient cells (data not shown). A range from 0.1 nM to 1 µM SDF-1
A previous report clearly showed that overexpression of an activated mutant of Rac1 blocked leukemic cell polarization due to increased and omnidirectional cell spreading.38 Consistent with this, overexpression of Vav ( Together, our data demonstrate the key role of Vav proteins in the chemotactic response and provide a functional link between chemokine triggering and lymphocyte polarization and migration in response to extracellular cues.
We acknowledge Drs Isabel Olazabal and Manuel Gómez for critical reading of the manuscript and Dr Mario Mellado (Centro Nacional de Biotecnología, Madrid) for his kind donation of CXCR-4-293-HEK cells and technical advice. I.O. and M.G. are scientists of the Ramón y Cajal Program (Ministerio de Educación y Ciencia, Spain).
Submitted July 28, 2004; accepted December 15, 2004.
Prepublished online as Blood First Edition Paper, December 23, 2004; DOI 10.1182/blood-2004-07-2925.
Supported by grant BMC02-00563 from the Spanish Ministerio de Ciencia y Tecnología; grant 2002 for Basic Research from the Juan March Foundation; grant 36289/02 from FIPSE (Fundación para la Investigación y Prevención del SIDA en España) (F.S.-M.); grant LSHG-CT-2003-502935-MAIN Network of Excellence from European Union; and grant SAF2003-00028 from the Spanish Ministerio de Ciencia y Tecnología.
A.C.-A. and N.B.M.-C. contributed equally to this work.
The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked "advertisement" in accordance with 18 U.S.C. section 1734.
Reprints: Francisco Sánchez-Madrid, Servicio de Inmunología, Hospital Universitario de la Princesa, Diego de León 62, 28006-Madrid, Spain; e-mail: fsanchez.hlpr{at}salud.madrid.org.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
1) and leukocyte function associated antigen-1 (LFA-1, 
) was previously described.
1-186) and untagged Vav (
) or RANTES (regulated upon activation normal T cell expressed and presumably secreted; 
), used as a control, and migration was quantified by flow cytometry as stated in "Materials and methods." Data represent the mean ± standard deviation of 4 independent experiments performed in triplicate.
) or Vav1–/–-deficient mice (KO; 
) and allowed to migrate for 3 hours in 3-µm pore diameter Boyden-modified migration chambers in the presence of 10 nM SDF-1
) or 10 nM (100 ng/mL; 
