Stromal cell–derived factor-1α/CXCL12–induced chemotaxis of T cells involves activation of the RasGAP-associated docking protein p62Dok-1

Chemokines play a central role in lymphocyte trafficking and homing. The chemokine stromal cell–derived factor-1α (SDF-1α/CXCL12) binds to CXC chemokine receptor 4 (CXCR4).^1,2  CXCR4 is a 7-transmembrane surface structure linked to G proteins.³  CXCR4 is expressed on a number of cell types, including T cells, hematopoietic stem cells, and progenitor cells. SDF-1α/CXCL12 is a highly efficient chemoattractant for T lymphocytes and other cells. Targeted disruption of SDF-1α/CXCL12 or CXCR4 is lethal in mice and is associated with the absence of lymphoid and myeloid hematopoiesis in the fetal bone marrow.^4,5 

The p120RasGAP-associated p62 protein Dok-1 (downstream of tyrosine kinase) was originally defined as a tyrosine-phosphorylated 62-kDa protein that coimmunoprecipitated with p21Ras GTPase-activating protein (RasGAP).⁶  Dok-1 is a docking protein purified from v-Abl or BCR-Abl–transformed hematopoietic cells. It consists of pleckstrin homology (PH) and phosphotyrosine binding (PTB) domains in the amino-terminal and the tyrosine phosphorylation site.⁷  Dok-1 is tyrosine phosphorylated by several cytokines, such as kit-ligand, platelet-derived growth factor (PDGF), and vascular endothelial growth factor (VEGF), and couples with the cytoplasmic protein tyrosine kinase (PTK).^8,9  Dok-1 is considered to be a downstream target of PTKs and to play a negative role in various signaling pathways.¹⁰ 

RasGAP is an essential component of Ras-activated signaling pathways. RasGAP has GTPase stimulating activity in the carboxyterminus and has 2 src homology domains (SH2) and 1 SH3 domain in the amino-terminal region. RasGAP down-regulates Ras activity, converting the activate form of RasGTP to the inactive form, RasGDP, and it plays a role in cell growth and differentiation.^11,12 

Proline-rich tyrosine kinase 2 (Pyk2), also known as cellular adhesion kinase β (CAKβ), related adhesion focal tyrosine kinase (RAFTK), and calcium-dependent tyrosine kinase (CADTK) are predominantly expressed in cells derived from hematopoietic lineages and in the central nervous system.^13-16  Pyk2 is one of the signaling mediators critical for signaling through G protein–coupled receptors, is activated by signals that elevate intracellular calcium concentrations, and is required for activation of mitogen-activated protein kinase (MAPK) signaling. Pyk2 is tyrosine phosphorylated after T-cell receptor (TCR) stimulation.^17,18 

The aims of our study were to evaluate intracellular effects mediating the chemotaxis induced by SDF-1α/CXCL12 and to determine whether this intracellular signaling pathway involved Dok-1, RasGAP, or Pyk2. Our results suggest that Dok-1, RasGAP, and Pyk2 are involved in SDF-1α/CXCL12 signaling.

Recombinant human SDF-1α/CXCL12, antiphosphotyrosine monoclonal antibody (mAb; 4G10) and anti–Crk-L mAb were purchased from Upstate Biotechnology (Lake Placid, NY). Rabbit anti–Dok-1 antibody (Ab), agarose-conjugated goat anti–Dok-1 Ab, anti-RasGAP mAb, anti-Vav Ab, phopho-ERK1 Ab, and protein A/G agarose were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti–Dok-1 Ab was purchased from Prosci Incorporated (Poway, CA). Antiphosphotyrosine mAb (PY20), anti-Pyk2 mAb, and anti-Nck mAb, and anti–Zap-70 mAb were obtained from Transduction Laboratories (Lexington, KY). Src kinase inhibitor PP2 was obtained from Calbiochem-Novabiochem (San Diego, CA). Anti-CXCR4 mAb was from R&D Systems (Minneapolis, MN). Other reagents were from Sigma (St Louis, MO).

The plasmid encoding mouse dok-1 was kindly provided by Dr Y. Yamanashi (Tokyo University, Japan). The complementary DNA (cDNA) encoding full-length dok-1 was amplified by polymerase chain reaction and was cloned into a mammalian GFP-expressing vector, GFP Fusion TOPO TA Expression Kits (Invitrogen, Carlsbad, CA), or a retroviral vector (pQCXIR; Clontech, Palo Alto, CA) at the NotI/PswI site.

Human leukemic T-cell line Jurkat, Lck-deficient T-cell line J.CaM1.6, and J.CaM1.6 cells engineered to express Lck were maintained in RPMI 1640 medium supplemented with 10% heat-inactivated fetal bovine serum (FBS) with 1% penicillin/streptomycin in a humidified incubator at 37°C. Murine fibroblast NIH3T3 cells expressing human CXCR4 and CD4 were maintained in Dulbecco modified Eagle minimal medium (DMEM) supplemented with 2 mM l-glutamine and 1% penicillin/streptomycin. Transfections were performed using LipofectAMINE 2000 (Invitrogen) according to the manufacturer's protocol. After transfection, cells were sorted using flow cytometry and were used for chemotaxis assay. For retroviral production, Phoenix-echo cells were transfected with empty vector alone or with full-length dok-1 using LipofectAMINE 2000 (Invitrogen). Supernatant was collected 24 hours and 48 hours after transfection. NIH3T3 cells were infected using supernatant and were analyzed by Western blot. Murine IL-3–dependent cell line Baf-3 and Dok-1–transfected Baf-3 Dok cells were cultured with 10% fetal calf serum (FCS) RPMI 1640 with 0.1 ng/mL murine interleukin-3 (IL-3). Lck cDNA expressing J.CaM/Lck cells were kind gifts from A. Weiss (University of California, San Francisco) and were cultured with 10% FCS RPMI 1640.

Jurkat, J.CaM1.6, and NIH3T3 cells were factor starved overnight and treated with 100 ng/mL SDF-1α/CXCL12, a predetermined optimal concentration, at the indicated times and were washed once with ice-cold phosphate-buffered saline (PBS). Cells were lysed in lysis buffer containing 20 mM Tris-HCl, pH 8.0, 137 mM NaCl, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride (PMSF), 10 μM ethylenediaminetetraacetic acid (EDTA), 10 μg/mL leupeptin, 100 mM sodium fluoride, 2 mM sodium orthovanadate, and 1% NP-40 for 20 minutes on ice. Lysates were centrifuged at 12 000 rpm (10 000 × g) for 20 minutes at 4°C. The protein content of lysates was determined with a protein assay kit (Bio-Rad Laboratories, Hercules, CA). Equivalent amounts of protein in cell lysates were boiled with 2 × SDS (sodium dodecyl sulfate) sample buffer for 5 minutes. For immunoprecipitation, cell lysates were incubated at 4°C overnight with the indicated precipitating antibody. Immunoprecipitates were collected using 40 μL protein A/G–agarose for 2 hours at 4°C. After 4 washings in lysis buffer, immunocomplexes were eluted and boiled for 5 minutes in 2 × sample buffer. Proteins or immunocomplexes were loaded onto polyacrylamide gels (BioWhittaker, Walkersville, MD) and then were transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, Bedford, MA). The membranes were blocked by 3% skim milk, PBS-Tween 20 (PBST) or 1% bovine serum albumin (BSA) PBST and probed with the indicated primary antibody at appropriate dilutions for 2 hours at room temperature (RT) or overnight at 4°C. Blots were probed with secondary antibody–conjugated horseradish peroxidase, and were developed using the enhanced chemiluminescence (ECL; Amersham Phamacia Biotech, Bucks, United Kingdom) system with ECL film according to the manufacturer's specification.

Chemotaxis assays for Jurkat, J.CaM1.6, and NIH3T3 cells were performed with 48-well microchemotaxis chambers with 8-μm pore size polycarbonate filters (PVP free; NeuroProbe, Gaithersburg, MD). Filters were soaked overnight in 0.1% wt/vol gelatin solution (Sigma). The chemotaxis medium (0.2% BSA in DMEM with and without 100 ng/mL SDF-1α/CXCL12) was placed in the lower chamber, and 50 μL cell suspension (1.6 × 10⁵ cells/mL) in chemotaxis medium (without SDF-1α/CXCL12) was placed in the upper wells. After incubation of the apparatus at 37°C for 3 to 4 hours in humidified air with 5% CO₂, filters were removed, fixed, and stained with the use of DiffQuik staining kit (Dade Behring, Newark, DE). Cells were counted in 3 high-power fields (HPFs; 400 ×). Results were expressed as the mean number of migrating cells in 1 HPF.

Splenic T cells from Dok^–/– and control mice¹⁹  were isolated by negative selection using the pan T-cell isolation kit (Miltenyi Biotech, Auburn, CA), which depletes cells expressing B220, Gr-1, and Ter119. The purity of T cells was evaluated by staining CD4 and CD8 and was greater than 95%. The chemotaxis assay to SDF1α/CXCL12 in the face of positive, negative, and zero gradient was performed as described.²⁰  After a 4-hour migration assay, input cells and cells that migrated to the bottom chamber were enumerated and stained with fluorescein isothiocyanate (FITC)–conjugated anti-CD4 and phycoerythrin (PE)–conjugated anti-CD8 antibodies (BD Biosciences, San Diego, CA) to determine the migration of T-cell subsets by flow cytometry. Migration index was calculated by the number of cells migrated divided by the background migration without SDF1α/CXCL12. Expression of CXCR4 was determined by staining the isolated T cells with biotinylated antimouse CXCR4 mAb (clone 2B11/CXCR4; BD Biosciences) with or without anti-CD4 and anti-CD8 Abs, followed by cytochrome-conjugated streptavidin staining.

NIH3T3 cells were detached by incubation with 10 mM EDTA in DMEM for 10 minutes, transferred to tubes, and sedimented (10 000 rpm, 1 minute [7000 × g]). Cells were stimulated with 100 ng/mL SDF-1α/CXCL12 at 37°C for the indicated times, washed once in ice-cold PBS, and fixed using the Cytofix/Cytoperm kit (BD PharMingen, San Diego, CA) or 1% paraformaldehyde PBS. Cells were incubated with phospho-Erk1 Ab for 1 hour and then incubated with phycoerythrin (PE)–conjugated secondary antibody. MAPK activities were monitored by flow cytometric analysis, and data obtained from 5 independent measurements were evaluated with the Student t test. To evaluate cell surface CXCR4, cells were fixed with 1% paraformaldehyde PBS and incubated with anti-CXCR4 mAb. Cell were stained with FITC-conjugated secondary antibody and analyzed by flow cytometry.

Dok-1 is activated by different proteins, such as CD2.²¹  To characterize signaling pathways activated by SDF-1α/CXCL12, we used the Jurkat T-cell line, which expresses the SDF-1α/CXCL12 receptor CXCR4. Jurkat cells were stimulated with 100 ng/mL SDF-1α/CXCL12 or control medium for 5 minutes. Cell lysates were subjected to anti–Dok-1 immunoprecipitation and to immunoblotting with antiphosphotyrosine antibody (Figure 1A). SDF-1α/CXCL12 stimulated the tyrosine phosphorylation of Dok-1. Nck is a ubiquitously expressed protein composed entirely of a single SH2 and 3 SH3 domains that fit into the adaptor class of signaling molecules.²²  Crk-L is in the family of Crk adaptor proteins and was discovered in the form of an oncogene carried by 2 sarcoma-inducing retroviruses.^23,24  Crk-L and Crk play a role in the signaling pathways of the T-cell receptor.^25,26  Dok-1 associates with RasGAP and the adaptor proteins Nck and Crk-L after stimulation.^27,28  To determine whether SDF-1α/CXCL12 induces the association of such molecules, coimmunoprecipitation experiments were performed. SDF-1α/CXCL12 triggering induced significant association of Dok-1 with RasGAP, Nck, and Crk-L (Figure 1B). Tyrosine phosphorylation of SDF-1/CXCL12–stimulated cells was not seen when control immunoglobulin G (IgG) was used for immunoprecipitation (Figure 1A, lower panel), and neither CrK-L, NcK, nor RasGAP was detected in control IgG immunoprecipitates (Figure 1C).

Dok-1 protein is phosphorylated by a wide range of PTKs, and Src family kinases are involved in SDF-1α/CXCL12 signaling.²⁹  We tested the potential relationships of Src to Dok-1 for SDF-1/CXCL12 signaling by using the specific Src kinase inhibitor, PP2. Jurkat cells were pretreated with 10 μM PP2 for 30 minutes and then were stimulated with SDF-1α/CXCL12. Pretreatment of Jurkat cells with PP2 completely blocked Dok-1 phosphorylation (Figure 2A). PP2 was not toxic to Jurkat cells, as assessed by trypan blue exclusion and cell numbers (data not shown). Using GST fusion proteins, we found that Dok-1 directly binds to the src kinase family members, Fyn and Lck, after SDF-1α/CXCL12 stimulation but does not bind without SDF-1/CXCL12 treatment (Figure 2B). Lck regulates T-cell surface receptors, such as CD2 and CD4.^30,31  We used J.CaM1.6 cells, a derivative of Jurkat cells that are defective in Lck, to determine whether Lck played a role in the activation of Dok-1. Dok-1 phosphorylation was completely blocked in J.CaM1.6 cells (Figure 2C), but not in J.CaM1.6 cells engineered to express Lck (Figure 2D), in response to SDF-1α/CXCL12 (Figure 2C), suggesting that Lck is involved in the regulation of Dok-1 phosphorylation in J.CaM1.6 cells in response to SDF-1α/CXCL12. Cell surface expression of CXCR4 was not different between J.CaM1.6 cells and the parental Jurkat cells (Figure 2E), demonstrating that differences between the 2 cell lines were not attributed to surface expression of CXCR4.

Pyk2 is expressed mainly in the central nervous system and in cells derived from hematopoietic lineages. Pyk2 is activated by a variety of extracellular signals that elevate intracellular calcium concentration and by stress signals.^13,16  Phosphorylation of Pyk2 leads to the recruitment of src family kinases and activation of ERKs.^13,15  We evaluated a role for Pyk2 in SDF-1α/CXCL12 signaling. Pyk2 was tyrosine phosphorylated after SDF-1α/CXCL12 stimulation, without an effect on the total amount of Pyk2 (Figure 3A). Vav is a hematopoietic cell-specific guanine nucleotide exchange factor for small guanosine triphosphate (GT)–binding proteins.³²  Zap-70 is a PTK that associates with the ζ subunit of the TCR. Zap-70 is tyrosine phosphorylated after TCR stimulation.^33,34  As seen in Figure 3B, tyrosine-phosphorylated Pyk2 associates with Vav and Zap-70 after SDF-1α/CXCL12 stimulation.

RasGAP is an essential component of Ras-activated signaling pathways and down-regulates Ras activity. The association of RasGAP with Pyk2 was enhanced after SDF-1α stimulation (Figure 4). This suggests that RasGAP association with Pyk2 is involved in SDF-1α/CXCL12 signaling.

We demonstrated that Dok-1 associates with the adaptor proteins Nck and Crk-L (Figure 1B). Because Nck and Crk-L are involved in cell migration to SDF-1α/CXCL12, ^35,36  we studied whether overexpression of Dok-1 influenced chemotaxis of CXCR4-expressing NIH3T3 and Baf3 cells in response to SDF-1α/CXCL12. Dok-1 cDNA was transfected to NIH3T3 cells expressing human CXCR4. Surface CXCR4 expression levels were not different between the parental CXCR4-expressing NIH3T3 cells transfected with empty vector and their dok-1 cDNA–transfected counterparts, as determined by flow cytometry and Western blot analysis (Figure 5A-B). Dok-1 overexpressing NIH3T3 cells were significantly decreased in response to SDF-1α/CXCL12–induced chemotaxis compared with empty vector–transfected NIH3T3 cells (Figure 5Ci). Baf3 cells and Baf3 cells overexpressing Dok-1 express similar levels of CXCR4 (data not shown). As seen in Figure 5Cii, Baf3 cells overexpressing Dok-1 were significantly suppressed in their chemotactic response to SDF-1/CXCL12. Dok-1 is considered a negative regulator of cell proliferation and Ras/MAPK signaling pathways.¹⁰  To evaluate Dok-1 regulation of MAPK activity after SDF-1α/CXCL12 signaling, we used Dok-1–overexpressing NIH3T3 cells (Figure 5D-E). Dok-overexpressing NIH3T3 cells manifested reduced MAPK activity (Figure 5D) and decreased phosphorylation of Erk1 (Figure 5E) in response to SDF-1α/CXCL12 compared with mock-transfected cells. This suggests that Dok-1 plays a role in SDF-1α/CXCL12–induced chemotaxis and regulation of MAPK activity.

To determine the relevance and importance of Dok-1 in SDF-1/CXCL12–induced chemotaxis, we evaluated the chemotactic responses of T-lymphocyte subsets from the spleens of Dok-1^–/– and Dok-1^+/+ mice¹⁹  to graded amounts of SDF-1/CXCL12. As shown in Figure 6, total T cells, CD4⁺ T cells, and CD8⁺ T cells from the spleens of Dok-1^–/– mice were enhanced in their chemotactic response to SDF-1/CXCL12; maximum enhancement occurred for Dok-1^–/– T cells at a concentration of SDF-1/CXCL12 that maximally stimulated chemotaxis for Dok-1^+/+ T cells. Chemotaxis of Dok-1^–/– and Dok-1^+/+ T cells to SDF-1CXCL12 was blocked by placing SDF-1/CXCL12 in the upper chamber or in both the upper and the lower chambers, demonstrating that the effects seen (migration to the lower chamber) were caused by chemotaxis and not by chemokinesis. Differences in chemotactic responses of Dok-1^–/– and Dok-1^+/+ T cells to SDF-1/CXCL12 were not caused by differences in levels of expression of CXCR4 (Figure 6B). These results demonstrate that Dok-1 negatively regulates primary T-cell chemotaxis to SDF-1/CXCL12. Along with the data shown (Figure 5A, Ci-ii), Dok-1 may also negatively regulate the migration of other cell types to SDF-1/CXCL12.

Chemokines belong to a large family of chemoattractant molecules and have been implicated in a number of different functions mediated through chemokine receptors. SDF-1α/CXCL12 plays a role in regulating the migration and homing of hematopoietic cells and is a highly efficient chemotactic factor for T cells. Several studies have evaluated the intracellular molecules involved in SDF-1α/CXCL12–induced chemotaxis of cells. In T cells, SDF-1α/CXCL12 stimulation results in activation of the Janus kinase/signal transduction and activation of transcription (Jak/STAT) pathways.³⁴  Phosphatidylinositol 3-kinase (PI3-K), Crk-associated substrate (p130Cas), focal adhesion kinase (FAK), and protein kinase C (PKC) are activated by SDF-1α/CXCL12, ³⁵  but the signaling mechanisms involved are not yet fully determined. Our present report implicates the docking protein, Dok-1, as a mediator of SDF-1α/CXCL12–induced migration of T cells and associated intracellular signals; Dok-1 links with downstream effectors of SDF-1α/CXCL12/CXCR4 signaling. Dok-1 has typical features of multiadaptor proteins, such as a membrane localization sequence (PH domain), receptor interaction domain (PTB domain), and several putative binding sites for downstream substrates (phosphotyrosine and PXXP elements). Dok-1 was identified as a tyrosine-phosphorylated protein of 62-kDa associated with p120RasGAP in fibroblasts transfected with v-src.⁶  Dok-1 is phosphorylated by several receptor tyrosine kinases and is regulated by tyrosine kinase. Src family kinases, such as Src, Fyn, and Lck, regulate Dok-1 phosphorylation.²¹  Lck is required for CD2-mediated phosphorylation of Dok-1. We have shown that Dok-1 tyrosine phosphorylation is completely blocked by use of the specific src kinase inhibitor PP2, implicating a reliance of Dok-1 on src family kinases (Figure 2A). We also found that Dok-1 phosphorylation was completely blocked after SDF-1α/CXCL12 stimulation in the Lck-deficient T-cell line J.CaM1.6 (Figure 2C), but not in J.CaM1.6 cells engineered to express LcK (Figure 2D), demonstrating that Lck regulates Dok-1 phosphorylation in this T-cell line.

Coimmunoprecipitation studies demonstrated that tyrosine-phosphorylated Dok-1 binds directly to RasGAP, Nck, and Crk-L. We had previously reported Nck involvement in SDF-1α/CXCL12 signaling and chemotaxis in Jurkat cells, ³⁷  thereby implicating RasGAP, Nck, and CrK-L in SDF-1α/CXCL12–induced cell chemotaxis. To study the role of Dok-1 regulation of cell migration and MAPK activity in response to SDF-1α/CXCL12, Dok-1 cDNA was transfected into NIH3T3 cells expressing human CXCR4 and Baf3, which express CXCR4. Dok-1–transfected NIH3T3 and Baf3 cells were significantly less responsive to SDF-1α/CXCL12-induced chemotaxis than empty vector–transfected cells, directly demonstrating Dok-1 as one of the intracellular molecules involved in SDF-1α/CXCL12–induced migration. Analysis of Dok-1 knockout mice reveals that Dok-1 is a negative regulator of cell proliferation.¹⁰  Cells derived from Dok-1 knockout mice hyperproliferate in response to a number of cytokines and growth factors. Moreover, we found that total T cells, as well as CD4⁺ T cells and CD8⁺ T cells, from the spleens of Dok-1^–/– mice were enhanced in response to SDF-1/CXCL12–induced chemotaxis compared with Dok-1^+/+ T cells (Figure 6). Dok-1 negatively regulates MAPK activity in T cells. We also found that Dok-1 regulates MAPK activity in response to SDF-1α/CXCL12 signaling (Figure 5D-E).

Pyk2 is a non–receptor tyrosine kinase belonging to the focal adhesion kinase family. Pyk2 interacts with several signaling intermediates. Pyk2 has no SH2 or SH3 domains, but it is proposed that proline-rich stretches in the C-terminus act as ligands for SH3 domain–containing signaling proteins. Pyk2 is constitutively expressed in human T cells and is rapidly phosphorylated on the activation of TCR. This is associated with its increased association with Src and Grb2.^17,18  We found that Pyk2 is tyrosine phosphorylated after SDF-1α/CXCL12 stimulation in Jurkat cells. Moreover, the present data also show that Pyk2 associates with Zap-70 and Vav. Pyk2 has been shown to participate in the activation of MAPKs. RasGAP is implicated as a negative regulator of ras because it is capable of stimulating the intrinsic GTPase-inactive guanosine diphosphate (GDP)–bound form of the molecule. Interestingly, Pyk2 can bind to RasGAP after SDF-1α/CXCL12 stimulation. Thus, Pyk2 also appears to play a role in SDF-1α/CXCL12 signaling and perhaps in the regulation of MAPK activity.

In summary, we have demonstrated that SDF-1α/CXCL12 action leading to Dok-1 activation is dependent on src kinases and on Pyk2. Several signaling pathways are induced by SDF-1α/CXCL12. These signaling pathways are regulated by positive and negative regulators. Dok-1 appears to play a negative role in SDF-1α/CXCL12 signaling and chemotaxis, and Pyk2 may play a positive role.