Research ArticleImmunology

The Chemokine CXCL12 Generates Costimulatory Signals in T Cells to Enhance Phosphorylation and Clustering of the Adaptor Protein SLP-76

See allHide authors and affiliations

Science Signaling  30 Jul 2013:
Vol. 6, Issue 286, pp. ra65
DOI: 10.1126/scisignal.2004018

Abstract

The CXC chemokine CXCL12 mediates the chemoattraction of T cells and enhances the stimulation of T cells through the T cell receptor (TCR). The adaptor SLP-76 [Src homology 2 (SH2) domain–containing leukocyte protein of 76 kD] has two key tyrosine residues, Tyr113 and Tyr128, that mediate signaling downstream of the TCR. We investigated the effect of CXCL12 on SLP-76 phosphorylation and the TCR-dependent formation of SLP-76 microclusters. Although CXCL12 alone failed to induce SLP-76 cluster formation, it enhanced the number, stability, and phosphorylation of SLP-76 microclusters formed in response to stimulation of the TCR by an activating antibody against CD3, a component of the TCR complex. Addition of CXCL12 to anti-CD3–stimulated cells resulted in F-actin polymerization that stabilized SLP-76 microclusters in the cells’ periphery at the interface with antibody-coated coverslips and increased the interaction between SLP-76 clusters and those containing ZAP-70, the TCR-associated kinase that phosphorylates SLP-76, as well as increased TCR-dependent gene expression. Costimulation with CXCL12 and anti-CD3 increased the extent of phosphorylation of SLP-76 at Tyr113 and Tyr128, but not that of other TCR-proximal components, and mutation of either one of these residues impaired the CXCL12-dependent effect on SLP-76 microcluster formation, F-actin polymerization, and TCR-dependent gene expression. The effects of CXCL12 on SLP-76 microcluster formation were dependent on the coupling of its receptor CXCR4 to Gi-family G proteins (heterotrimeric guanine nucleotide–binding proteins). Thus, we identified a costimulatory mechanism by which CXCL12 and antigen converge at SLP-76 microcluster formation to enhance T cell responses.

Introduction

Chemokines are chemoattractant cytokines that bind to G protein (heterotrimeric guanine nucleotide–binding protein)–coupled receptors (GPCRs) (1), and they direct the migration of cells in lymph nodes to sites of infection and facilitate conjugate formation between T cells and dendritic cells (27). In this manner, chemokines can exacerbate autoimmune diseases such as rheumatoid arthritis (3, 8, 9). Of the various chemokines, CXCL12 [also termed stromal cell–derived factor 1 (SDF-1)] binds to the CXC chemokine receptor CXCR4 (CD184), which also serves as a co-receptor for HIV-1 (10). Mice lacking CXCL12 or CXCR4 show lethal defects that involve vascular deficiencies and impaired lymphopoiesis and myelopoiesis (1113). Classically, CXCR4 signals through G proteins such as Gαs, Gαi, Gαq, and Gα12 (1, 10, 14). Gαi inhibits adenylate cyclase, leading to a decrease in production of the second messenger cyclic adenosine monophosphate (cAMP), whereas Gαq activates isoforms of phospholipase C (PLC), such as PLC-β, leading to intracellular Ca2+ mobilization.

CXCR4 acts as a costimulator of T cell proliferation and effector functions (5, 6). Further, it can associate with the T cell antigen receptor (TCR) complex (15), whereas the TCR or B cell antigen receptor (BCR) can arrest motility that impairs chemotaxis to CXCL12 at a distal site (10, 16, 17). CXCL12 engagement of CXCR4 and its associated G proteins increases the activities of phosphatidylinositol 3-kinase (1820) and the protein tyrosine kinases ZAP-70 (ζ chain–associated protein kinase of 70 kD) (21), Fyn, and Lyn (22), as well as increases the extent of phosphorylation of various adaptor proteins (6, 21, 2328). These effects may be direct or, alternatively, indirect as a result of the activation of integrin-mediated adhesion and increased cell-to-cell adhesion (29). Further downstream, CXCL12 signaling increases the activities of the serine and threonine kinase Akt [also known as protein kinase B (PKB)], extracellular signal–regulated kinases (30) (18, 31), and the Janus-activated kinase (JAK)–signal transduction and activation of transcription (STAT) pathway (20). The activation of ZAP-70 participates in chemokine-dependent T cell migratory responses (21).

Modern imaging techniques have shown that the TCR induces the rapid clustering of the receptors, kinases, and adaptor proteins that are needed for T cell activation (3236). Clusters 200 to 500 nm in diameter form within seconds of TCR engagement, and they serve as sites of early tyrosine phosphorylation (3236). Clusters initially form in the periphery of the region of cell-cell contact with the substratum or antigen-presenting cells, which is followed by their migration to the central contact region (32, 33, 37). The formation of these clusters precedes the development of supramolecular activation clusters (SMACs) at the mature immune synapse (38). Of the adaptor proteins, linker of activated T cells (LAT) and Src homology 2 (SH2) domain–containing leukocyte protein of 76 kD [SLP-76; also known as lymphocyte cytosolic protein 2 (lcp2)] also readily form clusters (34, 39, 40). Furthermore, clusters of LAT or SLP-76 interact with individual ZAP-70 clusters during the T cell activation process (32, 41) in a manner that may contribute to their phosphorylation (42, 43). A fraction of LAT also resides in mobile, intracellular vesicles that interact with surface SLP-76 microclusters (41). SLP-76–deficient T cells show reduced formation of ZAP-70 clusters, which is indicative of the presence of a feedback loop between SLP-76 and ZAP-70 (44).

Despite its capacity to costimulate T cells, whether CXCL12 can mediate this event through an effect on the induction of SLP-76 microcluster formation by an anti-CD3 antibody or whether this event contributes to the costimulation of T cells is unclear. Here, by imaging individual T cells on coverslips, we showed that CXCL12 markedly stabilized anti-CD3–induced formation of SLP-76 microclusters and F-actin (polymerized actin), effects that were mediated by the specific increase in SLP-76 Tyr113 (Y113) and Tyr128 (Y128) phosphorylation by CXCL12. Whereas mutation of these phosphorylation sites had more variable effects on the stimulation of T cells by anti-CD3 alone in our cells, it consistently impaired the potentiating effects of CXCL12 on T cell activation, which were dependent on the coupling of its receptor CXCR4 to Gi proteins. Together, these data suggest that CXCL12 modulates SLP-76 microcluster formation for the costimulation of T cells.

Results

CXCL12 enhances the number and persistence of anti-CD3–dependent SLP-76 microclusters

Although TCR engagement stimulates the formation of SLP-76 microclusters, the effect of costimulatory chemokines such as CXCL12 on the process has yet to be explored. We transfected J14 Jurkat T cells that lack SLP-76 with plasmid encoding enhanced yellow fluorescent protein (EYFP)–tagged SLP-76 and imaged the cells on coverslips coated with activating antibody against CD3 (anti-CD3) in the absence or presence of immobilized CXCL12. We used mouse immunoglobulin G2a (IgG2a) as a negative control. We performed imaging of individual cells on coverslips with time-lapse confocal microscopy, as previously described (34, 41, 44). Focusing on individual isolated cells on coverslips enabled us to determine signaling events stimulated by only anti-CD3 and CXCL12 without the involvement of adjacent cells and their ligands. In addition, imaging was performed in the absence of fetal calf serum (FCS) to determine the effects of CXCL12 signaling independently of integrin adhesion to serum factors such as fibronectin. Under these conditions, anti-CD3 stimulated the formation of EYFP-SLP-76 clusters in the central and peripheral regions of the contact area between the cell and the coverslip, and these coalesced over time in the inner contact region of the T cell (Fig. 1A). In contrast, although CXCL12 had a nominal effect alone, it enhanced the abundance and stability of anti-CD3–dependent clusters, which remained localized in the peripheral contact region of the cell (Fig. 1A). Volocity software measurements showed that CXCL12 increased the mean size (P = 0.001) and persistence of the clusters (Fig. 1A). Over the interval of 60 to 1200 s, anti-CD3–dependent clusters decreased in number from 53 to <22 (that is, by 58%), whereas the number of clusters induced by anti-CD3 and CXCL12 decreased from 105 to 88 (that is, by 16%) (Fig. 1A and fig. S1).

Fig. 1 CXCL12 enhances the anti-CD3–stimulated formation of SLP-76 microclusters.

(A) Jurkat J14 cells transfected with plasmid encoding EYFP-SLP-76 were incubated with IgG2a isotype control antibody (control), CXCL12, anti-CD3 antibody, or anti-CD3 and CXCL12 on coated coverslips and were monitored by time-lapse confocal microscopy for the formation and movement of SLP-76 microclusters over 1200 s. Images were obtained at 10-s intervals. Scale bar, 10 μm. Left: Images taken from the digital movie at 60, 300, 600, and 1200 s are representative of three independent experiments. Middle: CXCL12 increases the number and stability of anti-CD3–stimulated clusters. The numbers of microclusters per cell were calculated from images taken at the indicated times. Data are from a representative experiment and are means ± SD (n = 4 experiments). Right: CXCL12 increases the size of anti-CD3–stimulated clusters. The sizes of clusters in each cell at 5 min were calculated by Volocity software. ***P < 0.001. (B) Kymographs displaying the trajectories of SLP-76 microclusters. Left: The trajectories of EYFP-SLP-76 microclusters in cells in response to anti-CD3 or anti-CD3 and CXCL12 were followed over time. Images are representative of four experiments. Top right: The average speeds and displacement distances of SLP-76 microclusters and the average percentages of SLP-76 microclusters that persisted over 5 min were determined. ***P < 0.001. Bottom right: Graph showing the relative numbers of mobile, intermediate (Intem), or static SLP-76 microclusters in cells treated with anti-CD3 alone or with anti-CD3 and CXCL12.

Kymograph analysis further underscored the stabilizing effects of CXCL12 on SLP-76 microclusters (Fig. 1B). Anti-CD3 stimulated the formation of SLP-76 clusters in the central and peripheral regions that moved inward over a 5-min period (Fig. 1B), as previously reported (32, 35). In contrast, clusters stimulated by the combination of anti-CD3 and CXCL12 remained stably positioned in the peripheral regions over the same period (Fig. 1B). CXCL12 reduced the average speed of clusters from 9.5 μm/s in anti-CD3–treated cells to 5.2 μm/s in cells treated with anti-CD3 and CXCL12 (Fig. 1B). Consistent with this, CXCL12 reduced the displacement of clusters (that is, their distance from the site of origin) from 7.9 to 2.4 μm, whereas the chemokine increased the persistence of clusters from 14 to 40% (Fig. 1B). We also determined the relative numbers of static versus mobile clusters (Fig. 1B). These observations suggested that CXCL12 increased the number, size, and stability of anti-CD3–stimulated SLP-76 clusters. We found that the presence of soluble CXCL12 in the context of immobilized anti-CD3 also increased the number of SLP-76 clusters compared to that stimulated by anti-CD3 alone (fig. S2).

CXCL12 increases the extent of F-actin ring formation

F-actin polymerization in T cells rapidly follows contact with the substratum or other adherent cells (33, 45, 46). It was therefore next of interest whether CXCL12 also affected filamentous (F-actin) formation in our system. We incubated wild-type Jurkat cells (a human CD4+ T cell line) on coverslips with anti-CD3 in the absence or presence of CXCL12 for 5 min and then stained the cells with tetramethyl rhodamine isothiocyanate (TRITC)–conjugated phalloidin to detect F-actin (Fig. 2A). Phalloidin binds to polymerized F-actin without binding to monomeric actin. Unlike in the case of cluster formation, CXCL12 alone stimulated an increase in F-actin, in accordance with previous reports (21, 47). As determined by Volocity software, CXCL12 alone increased the mean fluorescence intensity (MFI) of F-actin staining from 14 for the isotype control to 36 (Fig. 2A). Similarly, anti-CD3 alone increased the MFI of F-actin from 14 for the isotype control to 34. The combination of anti-CD3 and CXCL12 further increased the MFI to 66 (Fig. 2A). Furthermore, the increased abundance of F-actin was primarily in the peripheral contact region and was accompanied by an increase in cell size from 76 μm2 in the presence of anti-CD3 alone to 89 μm2 in the presence of both anti-CD3 and CXCL12 (Fig. 2A). We observed a similar effect on F-actin in experiments with J14 T cells that were cotransfected with plasmids encoding SLP-76 and monomeric red fluorescent protein (mRFP)–actin (Fig. 2B). The MFI of mRFP-actin in the presence of anti-CD3 alone was 32, whereas that in the presence of CXCL12 alone was 20; however, in the presence of both anti-CD3 and CXCL12, the MFI of mRFP-actin was increased to 62. In this case, the increased mRFP-actin staining was also seen primarily in the peripheral contact region (Fig. 2B).

Fig. 2 CXCL12 enhances the extent of F-actin polymerization in T cells.

(A) Jurkat cells were incubated with IgG2a isotype control antibody (control) or anti-CD3 on coated coverslips in the absence or presence of CXCL12. The cells were then stained for endogenous actin with TRITC-conjugated phalloidin. Left: Images of cells stained with TRITC-phalloidin. Middle: MFI values of TRITC-phalloidin staining. Right: Sizes of TRITC-phalloidin–stained cells. Data in the bar graphs are means ± SD (n = 3 experiments). (B) CXCL12 enhances the anti-CD3–stimulated polymerization of mRFP-actin and its stability. Jurkat cells were transfected with plasmid encoding mRFP-actin and were left untreated (control) or were stimulated on coverslips coated with anti-CD3 in the absence or presence of CXCL12. Left: Images of cells expressing mRFP-actin. Images are representative of five experiments. Scale bar, 10 μm. Right: MFI of mRFP-actin staining. Data are means ± SD (n = 3 experiments). **P < 0.01, ***P < 0.001. (C) Kymographs displaying the trajectories of EYFP-SLP-76 and mRFP-actin. J14 cells transfected with plasmids encoding EYFP-SLP-76 and mRFP-actin were incubated on coverslips coated with CXCL12 alone or with anti-CD3 in the absence or presence of CXCL12 and were imaged over time for the movement of EYFP-SLP-76 and mRFP-actin. Left: Cells incubated with anti-CD3 alone. Middle: Cells incubated with anti-CD3 and CXCL12. Images are representative of four experiments. Right: PCC values for the colocalization of EYFP-SLP-76 and mRFP-actin in the peripheral regions of T cells in contact with slides. Data are means ± SEM from four independent experiments. ***P < 0.001.

Kymograph analysis confirmed that CXCL12 stabilized EYFP-SLP-76 with mRFP-actin in the peripheral contact region when coexpressed in J14 cells (Fig. 2C). In response to anti-CD3 alone, EYFP-SLP-76 and mRFP-actin migrated together in both the central and peripheral regions. Over time, both EYFP-SLP-76 and mRFP-actin moved inward toward the central region, resulting in a reduction in signal intensity. In contrast, CXCL12 stabilized EYFP-SLP-76 and mRFP-actin in the peripheral region over the entire time course of imaging (Fig. 2C). This led to the colocalization of EYFP-SLP-76 and mRFP-actin that remained stable in this region. The Volocity-derived Pearson’s correlation coefficient (PCC) showed that whereas anti-CD3 increased the PCC value from 0.02 to 0.22, CXCL12 further increased the value to 0.47 at 2.5 min (Fig. 2C). These data indicated that CXCL12 and anti-CD3 together increased the colocalization of SLP-76 and F-actin and the stability of the interaction in the peripheral region of the T cell.

Calculations of the speed of migration of actin and SLP-76 microclusters from multiple kymographs showed that in the absence of CXCL12, actin moved to the center of the contact region at similar rates, at an average sustained rate of 16.7 ± 4.4 nm/s, whereas SLP-76 microclusters moved to the center at a rate of 12.5 ± 1.3 nm/s (fig. S3). These observations were consistent with the coupling of SLP-76 microclusters to actin flow at a duty ratio of 0.75. In contrast, CXCL12 slowed the movement of actin to 4.9 ± 2.3 nm/s and that of SLP-76 microclusters to an even greater extent to 1.8 ± 0.67 nm/s, yielding a duty ratio of 0.37. These data are consistent with the ability of CXCL12 to reduce the rate of flow of actin, which led to more stable SLP-76 microcluster formation in the peripheral region of cells. The increase in F-actin polymerization with reduced actin inner flow may also contribute to the increased size of the cells.

CXCL12 preferentially increases the extent of the anti-CD3–dependent phosphorylation of SLP-76 at Tyr113 and Tyr128

Given the effects of CXCL12 on anti-CD3–dependent SLP-76 clustering and F-actin polymerization, it was next of interest to assess the molecular basis of this effect. Two key N-terminal tyrosine residues in SLP-76 are phosphorylated by ZAP-70 (43) and are required for thymic differentiation and T cell activation (48, 49). To assess the possible involvement of these residues in CXCL12-dependent effects, we imaged individual Jurkat cells incubated on anti-CD3–coated coverslips in the absence or presence of immobilized CXCL12 for 5 min. Cells were then permeabilized and stained for phosphorylated SLP-76 (pSLP-76) at Tyr113 or Tyr128 with phosphospecific antibodies. Although CXCL12 alone had little, if any, ability to generate clusters detected with anti-pTyr113 or anti-pTyr128, CXCL12 greatly increased the number of pSLP-76–containing clusters stimulated by anti-CD3 (Fig. 3A). As determined by Volocity software, anti-CD3 increased the MFI of pSLP-76 at Tyr113 from 2 to 28, which was further increased to 63 in the presence of both anti-CD3 and CXCL12 (Fig. 3A). Similarly, anti-CD3 increased the MFI of pSLP-76 at Tyr128 from 2 to 29, which was further increased to 48 in the presence of both anti-CD3 and CXCL12 (Fig. 3A). When we applied the same analysis to experiments with antibodies specific for ZAP-70 phosphorylated at Tyr493 or Tyr319, LAT phosphorylated at Tyr191, or the TCR ζ-chain phosphorylated at Tyr83, we found that CXCL12 did not increase the extent of phosphorylation beyond that stimulated by anti-CD3 alone (Fig. 3B). These data showed that CXCL12 cooperated with anti-CD3 to preferentially phosphorylate SLP-76 relative to other potential targets under conditions in which single cells were imaged in the absence of serum factor binding to integrins.

Fig. 3 CXCL12 preferentially enhances the anti-CD3–dependent phosphorylation of SLP-76 Tyr113 and Tyr128.

(A) Jurkat cells were incubated with IgG2a isotype control antibody (control) or anti-CD3 on coated coverslips in the absence or presence of immobilized CXCL12 for 5 min. Cells were then permeabilized and incubated with antibodies specific for SLP-76 phosphorylated at Tyr113 or Tyr128. Left: Representative images (from four experiments) of cells stained with pTyr113-specific (top) and pTyr128-specific (bottom) antibodies. Right: Graphs show the MFIs of SLP-76 pTyr113 (top) and SLP-76 pTyr128 (bottom). Data are means ± SEM from four independent experiments. **P < 0.01, ***P < 0.001. (B) CXCL12 failed to enhance anti-CD3–stimulated phosphorylation of ZAP-70, LAT, and TCRζ as imaged in individual T cells. Jurkat cells were incubated with IgG2a isotype control (control) or anti-CD3 on coated coverslips in the absence or presence of immobilized CXCL12 for 5 min. Cells were then permeabilized and incubated with antibodies specific for pZAP-70 (pTyr493), pZAP-70 (pTyr319), pLAT (pTyr191), and TCRζ (pTyr83). Graphs show the MFIs of the indicated antibodies. Data are means ± SEM from four independent experiments. ***P < 0.001, when comparing CXCL12-treated cells to anti-CD3–treated cells. (C) Flow cytometric analysis of the preferential enhancement by CXCL12 of the anti-CD3–stimulated phosphorylation of SLP-76 at Tyr113 and Tyr128. Primary human T cells were left untreated or were stimulated with anti-CD3 (5 μg/ml) in the presence or absence of CXCL12 (200 ng/ml) for 5 min. Cells were then fixed, permeabilized, and incubated with antibodies specific for SLP-76 pTyr113 or SLP-76 pTyr128. Left: Representative flow cytometry profiles of staining with antibodies against SLP-76 pTyr113 or pTyr128 in the indicated cells. Data are representative of three experiments. Right: Bar graphs showing the MFIs of SLP-76 pTyr113 or pTyr128 in the indicated cell types. Data are means ± SEM from four independent experiments. **P < 0.01; ***P < 0.001. (D) CXCL12 does not enhance the phosphorylation extent of other TCR signaling components. Primary human T cells were left untreated or were stimulated with anti-CD3 (5 μg/ml) in the presence or absence of CXCL12 (200 ng/ml) for 5 min. Cells were then fixed, permeabilized, and incubated with antibodies specific for ZAP-70 pTyr493, ZAP-70 pTyr319, LAT pTyr191, or TCRζ pTyr83. Bar graphs show the MFIs of the indicated antibodies in cells under the indicated conditions. Data are means ± SEM from four independent experiments. *P < 0.05. (E) The CXCL12-dependent enhancement of anti-CD3–stimulated phosphorylation of SLP-76 Tyr113 is unaffected by engagement of LFA-1. Left: Jurkat cells were left untreated (lane 1) or were stimulated with anti-CD3 (lanes 2 to 5) in the presence (lanes 3 and 5) or absence (lanes 2 and 4) of CXCL12 for 2 and 5 min. Cells were then lysed, and equal amounts of cell lysates were analyzed by Western blotting with antibodies specific for SLP-76 pTyr113, ZAP-70 pTyr319, or LAT pTyr191. Antibodies specific for total SLP-76, ZAP-70, and LAT proteins demonstrated equal loading. Data are representative of three experiments. Top right: Jurkat cells were stimulated for 5 min with isotype control antibody (isotype, lane 1), anti-CD3 (lane 2), CXCL12 (lane 3), anti-CD3 and CXCL12 (lane 4), anti–LFA-1 (lane 5), or anti-CD3 and anti–LFA-1 (lane 6). Cell lysates were subjected to Western blotting analysis with an antibody specific for SLP-76 pTyr113. Bottom right: The CXCL12-dependent increase in the phosphorylation of SLP-76 Tyr113 and Tyr128 as detected by immunofluorescence staining is not affected by anti–ICAM-1 blockade. Cells were treated as described for the Western blotting analysis and then were fixed, permeabilized, and incubated with antibodies specific for SLP-76 pTyr113 (upper graph) or SLP-76 pTyr128 (lower graph) and were analyzed by flow cytometry. Data are means ± SEM from three independent experiments. **P < 0.01; ***P < 0.001.

In a second approach, we incubated human T cells with plate-bound anti-CD3 or CXCL12 in RPMI 1640 in the absence of FCS and then incubated the cells with phosphospecific antibodies before analyzing them by flow cytometry (Fig. 3C). We found that CXCL12 enhanced the anti-CD3–dependent phosphorylation of SLP-76 at Tyr113 and Tyr128 (Fig. 3C). The MFI of pTyr113 in the presence of anti-CD3 alone was 27, which was increased to 37 in the presence of both anti-CD3 and CXCL12. In the case of pTyr128, the MFI of 15 in the presence of anti-CD3 alone was increased to 30 in the presence of both anti-CD3 and CXCL12. In contrast, we did not detect any substantial increase in the extent of phosphorylation of the TCR-proximal signaling intermediates ZAP-70 (at either Tyr493 or Tyr319), LAT (at Tyr191), or TCRζ (at Tyr83) in the presence of anti-CD3 and CXCL12 compared to that in the presence of anti-CD3 alone (Fig. 3D).

The specific phosphorylation of SLP-76 was confirmed through Western blotting analysis of cell lysates with phosphospecific antibodies (Fig. 3E). CXCL12 increased the extent of anti-CD3–dependent phosphorylation of the 75- to 76-kD band corresponding to SLP-76 at 2 and 5 min after stimulation, as detected with an anti-pTyr113 antibody (Fig. 3E). We did not observe a consistent increase in the phosphorylation of ZAP-70 at Tyr493 or Tyr319 or of LAT at Tyr191 beyond that seen in cells treated with anti-CD3 alone (Fig. 3E). As a further control, we also assessed whether the binding of the integrin lymphocyte function–associated antigen 1 (LFA-1) to ligand could alter the observed effects (29).

Although we attempted to avoid integrin involvement by imaging individual cells on coverslips and using medium lacking serum containing fibronectin and other integrin ligands, it was not possible to conclusively exclude a contribution by outside-in integrin signaling to the phosphorylation of SLP-76. To assess this directly, we costimulated T cells with anti-CD3 (5 μg/ml) in the presence of an antibody against LFA-1 (5 μg/ml) or a blocking antibody to the LFA-1 ligand intercellular adhesion molecule 1 (ICAM-1; 1 μg/ml) and then analyzed the cells by either Western blotting or flow cytometry for the phosphorylation of Tyr113 and Tyr128. Western blotting analysis showed that whereas CXCL12 increased the extent of phosphorylation of SLP-76 at Tyr113, the combination of anti-CD3 and anti–LFA-1 failed to increase the extent of Tyr113 phosphorylation beyond that seen with anti-CD3 alone (Fig. 3E). In terms of immunofluorescence staining, although CXCL12 increased the extent of anti-CD3–stimulated phosphorylation of SLP-76 at Tyr113 and Tyr128, the blocking of ICAM-1failed to alter this phosphorylation (Fig. 3E). These data suggested that CXCR4-dependent activation of LFA-1–mediated adhesion was not likely to have contributed to the effects on SLP-76 phosphorylation that we observed. Instead, the enhancing effects on SLP-76 phosphorylation were more likely a result of a direct effect of CXCL12 signaling on anti-CD3–dependent T cell activation. Therefore, the different approaches used in our study to minimize cell-cell contact and to control for the involvement of integrin activation showed the preferential phosphorylation of SLP-76 at Tyr113 and Tyr128 in response to costimulation with CXCL12.

To further confirm this result with another approach, we performed an enzyme-linked immunosorbent assay (ELISA) of permeabilized T cells involving detection of substrates with phosphospecific antibodies followed by a horseradish peroxidase (HRP)–conjugated secondary antibody (fig. S4). CXCL12 enhanced the anti-CD3–dependent phosphorylation of SLP-76 Tyr113 from a value of 0.34 to 0.44 (P = 0.02). Similarly, CXCL12 increased the anti-CD3–dependent phosphorylation of SLP-76 Tyr128 from 0.18 to 0.33 (P = 0.04). In contrast, there was no statistically significant increase in the extent of phosphorylation of other TCR-proximal signaling molecules by CXCL12, including ZAP-70 at LAT at Tyr191 and TCRζ at Tyr83.

CXCL12 enhances the colocalization of SLP-76 with ZAP-70

The protein tyrosine kinase ZAP-70 phosphorylates SLP-76 at Tyr113 and Tyr128 (43, 50), and SLP-76 clusters can interact with ZAP-70 clusters for phosphorylation (3234, 41, 44). Given the effect of CXCL12 on SLP-76 phosphorylation, we next assessed whether CXCL12 could increase the interaction between clusters of ZAP-70 and SLP-76 (Fig. 4A). We cotransfected preactivated primary T cells with plasmids encoding EYFP-SLP-76 and ZAP-70-mRFP, rested the cells for 24 hours, and then restimulated them with anti-CD3 in the absence or presence of CXCL12 on coverslips. We found that the PCC for the colocalization of ZAP-70 and SLP-76 clusters in the presence of CXCL12 alone was 0.14 and that for anti-CD3 alone was 0.28 (Fig. 4A), whereas in the presence of both anti-CD3 and CXCL12, the PCC was increased to 0.47 (Fig. 4A). Furthermore, we observed an increase in cluster colocalization stimulated by CXCL12 in the presence of anti-CD3 over a period of 1 to 5 min (Fig. 4A).

Fig. 4 CXCL12 enhances the anti-CD3–stimulated colocalization of SLP-76 clusters with ZAP-70 clusters.

A) PCC values for the CXCL12-dependent colocalization of mRFP-ZAP-70 and EYFP-SLP-76. Primary human T cells transfected with plasmids encoding mRFP-ZAP-70 and EYFP-SLP-76 were stimulated on anti-CD3–coated coverslips in the absence or presence of CXCL12. The movement of microclusters was monitored by time-lapse confocal microscopy. Top left: PCC values for the colocalization of ZAP-70 and SLP-76 in cells under the indicated conditions. Data are means ± SEM from three independent experiments. ***P < 0.001. Bottom left: Measurement of the PCC values for the colocalization of ZAP-70 and SLP-76 clusters over time. Right: Representative images from the video generated by time-lapse confocal microscopy. Data are representative of three independent experiments and were analyzed by Volocity software. (B) PCC values for the colocalization of EYFP-SLP-76 and SLP-76 pTyr113 in the peripheral versus central regions of cells. T cells transfected with plasmid encoding EYFP-SLP-76 were stimulated on anti-CD3–coated coverslips in the absence or presence of CXCL12. Samples were then permeabilized and incubated with an antibody against SLP-76 pTyr113. Left: PCC values for the colocalization of EYFP-SLP-76 and SLP-76 pTyr113 in the peripheral regions (top) and central regions (bottom) of human peripheral T cells. Middle: PCC values for the colocalization of EYFP-SLP-76 and SLP-76 pTyr113 in the peripheral regions (top) and central regions (bottom) of Jurkat cells. Data are means ± SEM from three independent experiments. **P < 0.01; ***P < 0.001. Right: Representative images from the video generated by time-lapse confocal microscopy of Jurkat cells.

To then confirm that SLP-76 within this peripheral region showed a preferential increase in CXCL12-dependent phosphorylation, we fixed T cells transfected with plasmid encoding EYFP-SLP-76 and incubated them with antibody against pTyr113 (Fig. 4B). We used Volocity analysis to define the inner (central) and outer (peripheral) regions within individual human peripheral T cells. In response to anti-CD3, PCC values showing correlation between anti–SLP-76-pTyr113 and EYFP-SLP-76 increased from 0.001 to 0.18 and 0.13 in the central and peripheral regions, respectively (Fig. 4B). In contrast, when anti-CD3 was combined with CXCL12, the PCC values increased in the peripheral region to 0.56 but were only 0.23 in the central region. We observed similar results in experiments with Jurkat cells (Fig. 4B). The combination of CXCL12 and anti-CD3 increased the signal in the peripheral region from 0.32 to 0.82 but had no statistically significant effect in the central region (Fig. 4B). Thus, CXCL12 enhanced the anti-CD3–dependent interaction between microclusters of ZAP-70 and SLP-76, which led to an increase in the phosphorylation of SLP-76 in the peripheral region.

Mutation of Tyr113 and Tyr128 reverses the CXCL12-dependent enhancement of the anti-CD3–stimulated formation of SLP-76 microclusters

Given the sensitivity of the SLP-76 Tyr113 and Tyr128 sites to phosphorylation in response to CXCL12, we next assessed whether these sites had a role in the effect of CXCL12 on microcluster stability, F-actin formation, and the activity of promoters driven by activating protein 1 (AP-1) and nuclear factor of activated T cells (NFAT) (Fig. 5A). AP-1 and NFAT are key transcription factors responsible for the expression of genes encoding cytokines such as interleukin-2 (IL-2) that are needed for T cell proliferation. Expression of the individual Y113F or Y128F mutants of SLP-76 in J14 cells resulted in reduced cluster formation in response to anti-CD3 and CXCL12 to the same extent as that observed in cells treated with anti-CD3 alone (Fig. 5A). Neither mutant had any consistent effect on anti-CD3–dependent microcluster formation in cells cultured under serum-free conditions (Fig. 5A). Occasionally, the SLP-76 Y113F clusters were less stable in response to anti-CD3 than were wild-type SLP-76 clusters, as previously reported (35). These data indicate that Tyr113 and Tyr128 on SLP-76 play a preferential role in regulating the CXCL12-dependent costimulatory arm of anti-CD3–dependent cluster formation. We observed a similar effect on F-actin polymerization (Fig. 5B). Cells were cotransfected with plasmids encoding EYFP-SLP-76, EYFP-Y113F, or EYFP-Y128F together with plasmid encoding mRFP-actin. The MFI values for F-actin in cells expressing either the Y113F or the Y128F mutant SLP-76 proteins in response to stimulation with anti-CD3 and CXCL12 were reduced to that observed for stimulation with anti-CD3 alone (Fig. 5B). As before, the mutants had little, if any, effect on F-actin polymerization stimulated by anti-CD3 alone.

Fig. 5 The Y113F and Y128F SLP-76 mutants block CXCL12-dependent costimulation of T cells.

(A) J14 Jurkat cells expressing EYFP-tagged SLP-76 wild type (WT) or its Y113F or Y128F mutants were stimulated on anti-CD3–coated coverslips in the absence or presence of CXCL12. Cells were analyzed to determine the number of SLP-76 microclusters under each condition. Data are means ± SEM from four independent experiments. ***P < 0.001. (B) The Y113F and Y128F SLP-76 mutants block the CXCL12-dependent increase in actin polymerization. J14 Jurkat cells cotransfected with plasmid encoding mRFP-actin together with plasmids encoding EYFP-SLP-76, EYFP-Y113F, or EYFP-Y128F were left untreated or were stimulated on anti-CD3–coated coverslips in the absence or presence of CXCL12 for 5 min. Data are means ± SEM from three independent experiments. ***P < 0.001. (C and D) Trajectories of EYFP-tagged SLP-76 proteins. Kymographs displaying the trajectories of EYFP-SLP-76 and mRFP-actin (C) as well as those of EYFP-SLP-76 Y113F and mRFP-actin (D) in J14 cells stimulated with anti-CD3 alone or anti-CD3 in the presence of CXCL12. Data are representative of three experiments. (E) J14 cells were cotransfected with plasmids encoding SRα control vector (mock), hemagglutinin (HA)–tagged SLP-76 WT, HA-SLP-76 Y113F, or HA-SLP-76 Y128F together with plasmid encoding 3× NFAT/AP-1 and pRL-TK vector (control reporter plasmid). Cells were stimulated with plate-bound IgG isotype control antibody, CXCL12, anti-CD3 antibody, or anti-CD3 with CXCL12. Luciferase activity was assessed 6 hours after stimulation. Bar graphs show means ± SD. Data are representative of three independent experiments. Inset: Western blotting analysis of the abundances of SLP-76 WT, SLP-76 Y113F, and SLP-Y128F in cell lysates was performed with anti-HA antibody. Data are representative of three experiments.

Kymograph analysis further illustrated this finding (Fig. 5, C and D). Anti-CD3 stimulated the formation of EYFP-SLP-76 clusters in the peripheral and central regions that moved inward over time to the central region (Fig. 5C). We found that mRFP-actin also moved toward the central region, and that it dissipated as it entered the central region and occasionally in the periphery. We observed a similar pattern in cells expressing the EYFP-SLP-76 Y113F mutant (Fig. 5C). In contrast, whereas CXCL12 markedly stabilized the actin cytoskeleton and SLP-76 microclusters in the periphery, it failed to stabilize these patterns in the case of cells expressing the Y113 mutant (Fig. 5D). Instead, the pattern in these cells resembled that seen for wild-type SLP-76 stimulated by anti-CD3 alone (Fig. 5D).

The Y113F and Y128 SLP-76 mutants also preferentially blocked enhancement of the anti-CD3–induced transcriptional activity of AP-1–NFAT by CXCL12 (Fig. 5E). We cotransfected J14 T cells with plasmids encoding wild-type or mutant SLP-76 proteins together with the 3× NFAT/AP-1 luciferase promoter construct from the IL-2 promoter and then stimulated the cells with anti-CD3 in the absence or presence of CXCL12. Anti-CD3 led to increased transcriptional activity in cells expressing wild-type SLP-76, and this was further augmented by CXCL12 (Fig. 5E). As a negative control, we measured reporter activity in J14 cells lacking SLP-76. Y113F or Y128F SLP-76 mutants expressed in our transfected cells exhibited weak or variable effects on responses to plate-bound anti-CD3 alone. However, the same mutants statistically significantly impaired the costimulatory effect of CXCL12, resulting in a reduction from the 260 × 103 U observed in response to anti-CD3 and CXCL12 in cells expressing wild-type SLP-76 to 150 × 103 to 155 × 103 luciferase units, a value similar to that observed in response to anti-CD3 alone (148 × 103 U). Overall, these observations demonstrated that mutating Tyr113, Tyr128, or both in SLP-76 preferentially impaired the costimulatory effect of CXCL12 on various aspects of T cell activation, including SLP-76 microcluster formation, F-actin polymerization, and the transcriptional activity of NFAT and AP-1.

CXCL12-dependent costimulation of T cells is mediated by Gαi proteins

Last, although CXCL12 signals through various mechanisms, the predominant signaling pathway involves the activation of Gαi proteins (1). To assess whether Gαi was responsible for the potentiating effects of CXCL12 on T cell activation, we incubated Jurkat cells and primary human T cells with the Gαi inhibitor pertussis toxin (PTX) or its inactive analog PTX-B for 30 min before performing imaging (Fig. 6). Whereas the extent of SLP-76 phosphorylation in Jurkat cells in response to anti-CD3 alone was not affected by PTX, as determined by Western blotting analysis (Fig. 6A), the drug inhibited the increased SLP-76 phosphorylation in response to costimulation with CXCL12. We also showed that PTX-B had no such effect on CXCL12-dependent costimulation (Fig. 6A). Furthermore, PTX had little effect on the anti-CD3–dependent generation of SLP-76 clusters; however, PTX reduced the increase in cluster number in response to costimulation with CXCL12 (Fig. 6B). Consistent with these observations, PTX also reduced the proliferation of murine T cells activated by anti-CD3 and CXCL12 to that seen with anti-CD3 alone (Fig. 6C). As before, these effects were not shared by PTX-B. These observations indicate that CXCL12 mediates its enhancement of SLP-76 phosphorylation, cluster formation, and T cell proliferation by signaling through Gαi proteins.

Fig. 6 CXCL12 costimulates SLP-76 phosphorylation, cluster formation, and proliferation through Gαi.

(A) Jurkat cells were pretreated with either PTX (100 ng/ml) or PTX-B (100 ng/ml) as a control for 2 hours. Pretreated cells were left untreated or were treated for 5 min with soluble CXCL12, anti–LFA-1, or anti-CD3 and rabbit anti-mouse secondary antibody in the absence or presence of CXCL12. Cell lysates were analyzed by Western blotting with antibodies specific for total SLP-76 and pSLP-76 proteins. Treatments were as follows: lane 1, untreated; lane 2, anti-CD3; lane 3, CXCL12; lane 4, anti-CD3 and CXCL12; lane 5, anti–LFA-1. Data are representative of three experiments. (B) Jurkat cells were pretreated with either PTX or PTX-B (both at 100 ng/ml) for 2 hours. Cells were then plated on anti-CD3–coated coverslips in the absence or presence of CXCL12 and were analyzed by confocal microscopy and Volocity software. Data are the mean numbers of microclusters per cell ± SEM from three experiments. ***P < 0.001. (C) Primary mouse T cells were pretreated with either PTX or PTX-B (both at 100 ng/ml) for 2 hours. Cells were then left untreated or were stimulated with anti-CD3 in the absence or presence of CXCL12. After 48 hours, cellular proliferation was measured by [3H]thymidine incorporation. Data are from a single experiment and are representative of three independent experiments. Bar graphs show means ± SD from triplicate measurements within the given experiment. **P < 0.01.

Discussion

Despite the importance of CXCL12 in costimulating T cells, the underlying signaling mechanism has been unclear. CXCL12 enhances the phosphorylation of several proteins; however, a clear distinction between its direct and indirect effects has not been made because of the involvement of CXCL12 in integrin activation and increasing cell-cell contact. The advanced ability to image signaling proteins and microclusters in isolated cells has now enabled us to distinguish between these events. Here, with this imaging approach, we showed that CXCL12 markedly stabilized the SLP-76 microclusters that were formed in response to anti-CD3, and that this effect was dependent on the selective phosphorylation of Tyr113 and Tyr128 of SLP-76. Alone, CXCL12 did not stimulate the formation of SLP-76 clusters; however, it stimulated the formation of a peripheral F-actin ring independently of TCR signaling that stabilized the anti-CD3–stimulated SLP-76 clusters for signaling. Mutation of the critical tyrosine residues preferentially impaired CXCL12-dependent costimulation of various aspects of T cell function. The ability of CXCL12 to costimulate these responses was dependent on the coupling of its receptor CXCR4 to Gαi. Overall, our findings show that CXCL12 potentiates the generation and stability of SLP-76 microclusters as part of its costimulation of T cells.

Microclusters serve as centers of T cell signaling, and mutations that impair cluster formation also inhibit T cell activation (32, 38, 51). The action of CXCL12 in enhancing SLP-76 microcluster formation provides a previously uncharacterized mechanism that could account for the ability of CXCL12 to costimulate T cells. It is possible that CXCL12 may also operate in a similar manner to stimulate the formation of microclusters of receptors and signaling molecules in other cell types. CXCL12 did not stimulate the formation of SLP-76 microclusters on its own. Occasionally, we observed the formation of a small number of clusters in response to CXCL12; however, this was not reproducible. Instead, we found that CXCL12 functioned only in cooperation with anti-CD3 to increase the number and stability of SLP-76 clusters.

The ability of CXCL12 to increase cluster stability likely contributed to the increased numbers of clusters that we observed. The mechanism by which this occurred involved the independent ability of CXCL12 to polymerize F-actin, especially in the peripheral contact area of cells. CXCL12 alone increased the MFI of F-actin by 2.5-fold, consistent with previous reports (46, 52). Furthermore, kymograph analysis showed a correlation between the presence of F-actin and the stability of peripheral clusters. CXCL12 reduced the movement of actin toward the central region of contact, which led to increased stability of SLP-76 microclusters in the peripheral region. The peripheral contact region was also the site at which the extent of phosphorylation of SLP-76 on Tyr113 and Tyr128 was increased by CXCL12, and where there was an increase in the colocalization of SLP-76 clusters with those of ZAP-70, the kinase that phosphorylates SLP-76 (43). The increased interaction between SLP-76 and ZAP-70 is likely to be stochastic because of the increased stability and abundance of the SLP-76 clusters. The peripheral contact area is also the major general site of tyrosine phosphorylation (53). CXCL12 shares an ability to stabilize peripheral clusters with the costimulatory integrin CD29/49 (VLA-4) (54).

CXCL12 stimulated preferential phosphorylation of SLP-76 at Tyr113 and Tyr128 under the activation conditions used here. We did not consistently observe the phosphorylation of other early T cell signaling molecules, including TCRζ and ZAP-70. Our findings do not exclude the possibility that more robust conditions of CXCL12 presentation and stimulation (for example, involving different concentrations and culture conditions) might stimulate the phosphorylation of other substrates, especially in the context of mixed cell cultures with extensive cell-to-cell adhesion. Rather, we observed a greater propensity for SLP-76 to be phosphorylated by CXCL12 in single cells and under conditions that minimized the involvement of integrins. Consistent with this, mutation of Tyr113 and Tyr128 also specifically blocked the CXCL12-dependent enhancement of microcluster formation, F-actin polymerization, and AP-1– and NFAT-induced transcriptional activity. These mutations had a less consistent or substantial effect on anti-CD3–dependent activation alone (48, 55, 56). It is therefore a possibility that part of the reported dependency of anti-CD3 responses on the Tyr113 and Tyr128 sites of SLP-76 might be related to the secretion by T cells of chemokines that function in an autocrine loop (57). This mechanism might also account for the variation in the inhibition by the SLP-76 Tyr113 and Tyr128 mutants of anti-CD3–induced proliferation in previous reports (43, 48, 56, 58). We also speculate that it might be possible that chemokines play a role in the dependency of thymic differentiation on the Tyr113 and Tyr128 sites of SLP-76 (48, 49, 58).

Overall, a model for CXCL12-mediated costimulation of T cells would involve its independent stimulation of F-actin polymerization and ring formation, leading to an increased stability of SLP-76 clusters, interaction with ZAP-70, and phosphorylation of SLP-76 on Tyr113 and Tyr128. In a possible feedback loop, the increased phosphorylation of both tyrosines in turn would be needed to sustain the cytoskeletal integrity required for the continued stability of the SLP-76 clusters and signaling events, which would lead to increased AP-1– and NFAT-dependent production of IL-2 and T cell costimulation. The regulatory effect of SLP-76 on the cytoskeleton can be mediated by the binding of Vav-1 and Nck to Tyr113 and Tyr128 (43, 5961). Whether other sites in SLP-76, such as Tyr145, are affected in response to CXCL12 remains to be determined. In either case, the upstream event linking CXCR4 to the modulation of SLP-76 microcluster formation was Gαi-dependent because blocking Gαi coupling with the inhibitor PTX prevented the CXCL12-dependent enhancement of anti-CD3–dependent SLP-76 phosphorylation, cluster formation, and AP-1–NFAT–dependent transcriptional activity. PTX blocks the inhibition of adenylate cyclase by Gαi, and so PTX prevents CXCR4-dependent reduction in the generation of the second messenger cAMP; PTX had little effect on anti-CD3–induced signals alone. cAMP activates protein kinase A (PKA), which phosphorylates and activates C-terminal Src kinase (CSK), which inhibits Src-family kinases, such as p56lck (62). The kinase p56lck in turn operates upstream of SLP-76 in the activation of T cells. Other effects could be related to the reported association of CXCR4 with the TCR (15). Further studies will be needed to establish the signaling event that connects CXCL12-dependent activation of Gαi with the phosphorylation of SLP-76 and the formation of SLP-76 clusters in T cells.

Materials and Methods

Cell culture, constructs, and transfection

Jurkat cells [American Type Culture Collection (ATCC)] were grown in RPMI 1640 with 10% FCS, 2 mM l-glutamine, penicillin, and streptomycin. J14 cells (SLP-76–deficient Jurkat cells) were provided by A. Weiss (University of California, San Francisco). To generate the SLP-76-EYFP fusion construct, complementary DNA encoding SLP-76 was amplified by polymerase chain reaction and subcloned into the Xho I and Bam HI sites of the pEYFP-N1 vector (Clontech). The plasmid PMXs-mZAP-70-mRFP was a gift from T. Saito (Japan). The plasmid encoding actin-mRFP was a gift from G. Griffiths (Cambridge Institute for Medical Research, UK). J14 cells expressing SLP-76-EYFP were generated by microporation (Digital Bio Technology) with a single pulse of 30 ms at 1410 V followed by a number of rounds of cell sorting and continuous selection in medium containing G418 (2 μg/ml). SLP-76-EYFP mutants were constructed with a site-directed mutagenesis kit (QuikChange, Stratagene). Primers used to generate the Y113F mutant were 5′-cctttgaagaagacgattttgaaagtcccaatga-3′ and 5′-tcattgggactttcaaaatcgtcttcttcaaagg-3′, and primers used to generate the Y128F mutant were 5′-ggaggatgatggagactttgagtcccccaatga-3′ and 5′-tcattgggggactcaaagtctccatcatcctcc-3′.

Antibodies

Anti-human CD3 (OKT3) and anti-mouse CD3 (145-2C11) antibodies were obtained from the ATCC. CXCL12 (also known as SDF-1α) was purchased from R&D Systems. Antibodies against pTyr113 and pTyr128 of SLP-76 were purchased from BD Pharmingen, anti-ZAP-70 antibody was from Santa Cruz Biotechnology, and antibodies against pTyr319 and pTyr493 of ZAP-70 were purchased from Cell Signaling Technology. Antibody against pTyr191 of LAT was bought from Millipore. Antibody against pTyr83 of TCRζ was purchased from Epitomics. Anti–LFA-1 and anti–ICAM-1 antibodies were obtained from R&D Systems. Alexa Fluor 488–conjugated goat anti-rabbit IgG and Alexa Fluor 568–conjugated goat anti-mouse IgG1 were purchased from Invitrogen. Poly-l-lysine, PTX, and PTX-B were bought from Sigma. Chambered cover glasses were obtained from Nalgene Nunc.

Isolation, culture, and transfection of human T cells

Human T lymphocytes were prepared from the peripheral blood of healthy donors. Mononuclear cells were isolated by Ficoll density gradient centrifugation. After being washed, cells were stimulated for 24 hours at 37°C with phytohemagglutinin (5 μg/ml). After two washes, cells were maintained for 5 to 6 days in an exponential growth phase in RPMI 1640 containing 10% FCS supplemented with recombinant IL-2 (20 ng/ml), followed by washing and culture in the absence of IL-2 for 24 hours. Cells were suspended at 108 cells/ml in 250 μl of complete medium with 5 to 10 μg of the SLP-76-EYFP–encoding plasmid. Transfections were performed in 4-mm gap cuvettes with the BTX ECM 830 electroporator with a single 6-ms pulse of 385 V. Cells were immediately transferred to prewarmed complete medium and allowed to recover for 24 hours before analysis.

Isolation and culture of mouse T cells

Isolated mouse splenocytes were cultured after the removal of red blood cells with hypotonic buffer [0.15 M NH4Cl, 1 mM NaHCO3, 0.1 mM EDTA (pH 7.25)] in RPMI 1640 containing 10% FCS, 5% glutamine, 5% penicillin, streptomycin, and 50 μM 2-mercaptoethanol. To generate T cell blasts, splenocytes were cultured with concanavalin A (2.5 μg/ml) for 2 days, washed once, and cultured in growth medium containing IL-2 (20 ng/ml) for 2 to 3 days. After removing residual IL-2, T cells were rested in growth medium for 2 days and then were used for experiments.

Live-cell imaging

All imaging assays were performed in poly-l-lysine–treated chambered glass culture slides (Lab-Tek). Proteins were sequentially adsorbed onto the plates in the following order: First, the plates were coated with anti-CD3 antibody (OKT3, 5 μg/ml) overnight at 4°C. Second, the plates were coated with CXCL12 (200 ng/ml) for 3 hours at 37°C. Finally, the plates were blocked with 1% bovine serum albumin (BSA) for 1 hour. Cells were then layered on the surface of the coverslips while imaging in real time was performed. In the case of the addition of soluble CXCL12, the coverslips were coated with anti-CD3, incubated with cells in the presence of soluble CXCL12 (200 ng/ml), and monitored by time-lapse confocal microscopy for 5 min for the formation and movement of microclusters. Cells were imaged at the interface with the coverslips with a Zeiss LSM 510 confocal microscope with a 63× oil immersion objective (numerical aperture, 1.2) and excitation wavelengths of 514 nm for EYFP and 594 nm for mRFP. Images were collected at 10-s intervals. Simultaneous imaging of different fluorophores was acquired by sequential scanning. Single Z sections were captured over time to improve the rate of image acquisition; proper focus was maintained with guide cells with distinguishable fluorescence properties to mark the plane of the coverslip. For some live-cell imaging, cells were imaged with a spinning disk confocal imaging system mounted on a Zeiss Axio Observer Z1 inverted microscope. In these cases, live images were collected as vertical Z stacks and then subsampled over the plane of the coverslip. Images were collected continuously for 5 min. The temperature of the sample was maintained at 37°C. All of the images were processed by Volocity software (Improvision). SDs and SEs were calculated with the Microsoft Excel and GraphPad Prism software. Differences between means were tested with one-way analysis of variance (ANOVA) with the Newman-Keuls multiple comparison test. In all cases, the threshold P value required to be considered statistically significant was 0.05.

Phalloidin staining

To stain F-actin, 1 × 106 Jurkat cells were stimulated with anti-CD3 antibody (5 μg/ml) in the absence or presence of CXCL12 (200 ng/ml) on coverslips for 5 min. Cells were then fixed and permeabilized in Cytofix/Cytoperm (BD Pharmingen) for 20 min, incubated in blocking solution (5% FCS and 3% BSA in Perm/Wash buffer, BD Pharmingen) for 1 hour, and then incubated with TRITC-conjugated phalloidin (Sigma-Aldrich) for 1 hour at 4°C.

Immunofluorescence staining and cell proliferation assays

Human T cells were stimulated with anti-CD3 (5 μg/ml) in the absence or presence of CXCL12 (200 ng/ml) for 5 min, washed, and then fixed in 4% paraformaldehyde (PFA). After permeabilization and blocking, cells were incubated with anti-SLP pTyr113, anti-SLP pTyr128, anti-CD3ζ pTyr83, LAT pTyr191, anti-ZAP pTyr319, or anti-ZAP pTyr493 antibodies and the appropriate fluorochrome-labeled secondary antibodies for analysis by flow cytometry (BD FACSCalibur) and fluorescence microscopy. For T cell proliferation assays, cells were stimulated with CXCL12, anti-CD3, or both for 48 hours. To measure proliferation, cells were pulsed with 1 μCi of [3H]thymidine (Amersham Biosciences) for the last 6 hours, and [3H]thymidine uptake was measured in a β-scintillation counter.

Luciferase assays

J14 cells were transfected with an irrelevant plasmid (the SRα vector), plasmids encoding HA-SLP-76 wild type, HA-SLP-76 Y113F, or HA-SLP-76 Y128F together with the reporter plasmid 3× NFAT/AP-1 from the Il2 promoter and pRL-TK vector (control reporter plasmid). Eighteen hours after transfection, cells were stimulated with plate-bound isotype control IgG, CXCL12, anti-CD3, or anti-CD3 and CXCL12 for 6 hours. Cells were then lysed, and luciferase activity was determined with the Dual-Luciferase Reporter Assay Kit (Promega) according to the manufacturer’s instructions.

Phosphospecific intracellular ELISA

For the intracellular ELISA assay, 96-well flat-bottomed tissue culture plates (Greiner) were coated with fibronectin (2 μg/ml; Sigma) in phosphate-buffered saline (PBS) for 1 hour and then were washed twice with PBS. From a suspension of human peripheral blood leukocytes in RPMI 1640 and 10% FCS (2 × 106 cells/ml), 100 μl was added to each well, and cells were left to adhere for 16 hours. For stimulation, the supernatant was decanted, and 50 μl of anti-CD3 antibody (UCHT1, 5 μg/ml) with or without CXCL12 (200 μg/ml) in serum-free medium, prewarmed to 37°C, was added to the wells and left for 5 min in an incubator. Then, 100 μl of ice-cold 4% PFA in PBS containing 1 mM sodium orthovanadate was added, and cells were fixed for 20 min on ice, followed by a permeabilization step with ice-cold methanol for 5 min. Samples were then blocked with blocking buffer (PBS, 4% BSA, 0.1% saponin) for 1 hour, followed by incubation with the phosphospecific antibodies in blocking buffer for 1 hour as follows: anti-SLP pTyr113 (1:750 dilution), anti-SLP pTyr128 (1:300), anti-TCRζ pTyr83 (1:300), anti-LAT pTyr191 (1:500), anti-ZAP pTyr319 (1:300), or anti-ZAP pTyr493 (1:300). Samples were washed three times and then incubated with HRP-conjugated secondary anti-mouse antibody (GE Healthcare; 1:1000 dilution) or anti-rabbit antibody (Cell Signaling Technology; 1:1000) in blocking buffer. After three washes, 100 μl of ABTS [2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid)] substrate (Roche) was added to the wells until no further color development was visible, which was followed by quantification of the absorbance at 405 nm with an ELISA reader (Optimax). All conditions within one assay were performed in triplicates. Measurements were corrected with the absorbance of a control sample without primary antibody. Statistical analysis was performed with a Student’s t test between pairs and by a one-way ANOVA combined with the Newman-Keuls multiple comparison test.

Western blotting analysis

Stimulated cells (1 × 106) were lysed in ice-cold lysis buffer containing 1% (v/v) Triton X-100, 20 nM tris-HCl (pH 8.3), 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na4VO3, 10 mM NaF, and 1 mM Na4P2O7. Proteins were resolved by SDS–polyacrylamide gel electrophoresis and transferred onto nitrocellulose membranes. The membranes were then blocked with 5% BSA in PBS containing 0.1% Tween and incubated for 1 hour with the indicated antibodies. Bound antibody was then visualized with HRP-conjugated anti-rabbit or anti-mouse secondary antibodies, which was followed by detection with enhanced chemiluminescence (Amersham Pharmacia Biotech). For experiments with PTX, T cells were incubated with PTX or the control PTX-B (both at 100 ng/ml) for 2 hours at 37°C in 5% CO2. After incubation, cells were washed twice and were prepared for imaging or proliferation assays as described earlier.

Statistical analysis

Statistical differences between means were tested with two-tailed unpaired Student’s t test (GraphPad Prism 3.02). P < 0.05 was considered statistically significant; *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA (Prism) combined with the Newman-Keuls multiple comparison test.

Supplementary Materials

www.sciencesignaling.org/cgi/content/full/6/286/ra65/DC1

Fig. S1. CXCL12 enhances the anti-CD3–dependent formation of SLP-76 microclusters and their stability.

Fig. S2. Soluble CXCL12 increases the extent of anti-CD3–dependent SLP-76 microcluster formation.

Fig. S3. CXCL12 slows the movement of SLP-76 microclusters and actin.

Fig. S4. CXCL12 preferentially enhances the anti-CD3–dependent phosphorylation of SLP-76 at Tyr113 and Tyr128.

References and Notes

Acknowledgments: We thank P. Bereton for computer assistance in the Department of Pathology in the University of Cambridge. Funding: This work was funded by a grant from the Wellcome Trust (060111-Z-99-B). Author contributions: C.E.R. contributed to the overall conception of the project, experimental design, and the writing of the paper; X.S. performed the imaging experiments, helped design the experimental approach for microscopy, and prepared the figures; H.S. performed the biochemical and functional experiments, helped with the writing of the paper, prepared the figures, and contributed to the design of experiments; H.L. helped with confocal microscopy and generated constructs used in the experiments; Y.L. performed functional experiments on G protein signaling; and K.K. performed and analyzed the ELISA assays on the phosphorylation of substrates. Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Signaling

Navigate This Article