Linker for Activation of T Cells, ζ-Associated Protein-70, and Src Homology 2 Domain-Containing Leukocyte Protein-76 are Required for TCR-Induced Microtubule-Organizing Center Polarization¹ Free

When a T cell encounters Ag on an APC, the TCR is stimulated, initiating a cascade of signaling events that eventually culminates in the expression of cytokines and growth factors (1, 2). Previous studies have shown that, like the polarized secretion of granules by a cytolytic T cell, the secretion of lymphokines by a helper cell is also polarized toward the APC. The secretion of these TCR-induced cytokines toward the specific APC is required for full activation of the APC (3). In order for the T cell to generate directed secretion of cytokines, signaling events must direct the reorganization of the T cell’s cytoskeleton. One of the cytoskeletal events which is reorganized is the polarization of the microtubule-organizing center (MTOC)³ from which microtubules emanate (4). Upon interaction with an APC, a signal is generated within the T cell that triggers the movement of the MTOC to the proximity of the interface between the T cell and the APC. This phenomenon is specific to the T cell, because the MTOC of the APC does not exhibit any directed polarization toward the site of contact (3). The movement of the MTOC in the T cell is required to polarize the secretory machinery toward the relevant APC. Our previous studies have established that stimulation of a functional TCR is necessary and sufficient to initiate the reorientation of the MTOC toward the site of TCR stimulation (5). However, little is known about the signaling events that emanate from the TCR that are required to polarize the MTOC.

Following stimulation of the TCR, the Src family protein tyrosine kinase (PTK) Lck is activated and phosphorylates tyrosines within the immunoreceptor tyrosine-based activation motifs (ITAMs) on the CD3 (γ, δ, ε) and ζ-chains of the TCR (1). Phosphorylation of these tyrosines creates docking sites for the PTK ζ-associated protein-70 (ZAP-70), which in turn binds to the ITAMs via its Src homology 2 (SH2) domains, placing ZAP-70 in proximity to Lck. Lck phosphorylates and activates ZAP-70, which in turn phosphorylates T cell-signaling components including the adaptor proteins linker for activation of T cells (LAT) and SH2 domain-containing leukocyte protein-76 (SLP-76) (6, 7, 8).

LAT is a transmembrane protein found within lipid rafts that contains multiple tyrosine-based motifs that, when phosphorylated by ZAP-70, can function as docking sites for SH2 domain-containing proteins (6). As a consequence of phosphorylation, LAT recruits other adaptor proteins including Grb2, Grap, and Gads as well as phospholipase Cγ1 (PLCγ1) (9). Recruitment of PLCγ1 to LAT facilitates the phosphorylation and activation of the lipase by Tec and Syk PTKs to generate the second messengers inositol trisphosphate and diacylglycerol, which are responsible for the mobilization of Ca²⁺ and the activation of protein kinase C, respectively (10). The recruitment of the adaptor protein Gads, which is a Grb2/Grap-related protein, to LAT acts to link the adaptor protein SLP-76 with LAT (11, 12). SLP-76 is a cytoplasmic protein with three YxxP motifs, an extensive central proline-rich region and a C-terminal SH2 domain. The significant role that these adaptor proteins play in TCR signaling has been demonstrated by studies of Jurkat T cell lines deficient in these adaptors as well as targeted gene disruption of either LAT or SLP-76 in mice (10, 13, 14, 15). The LAT- and SLP-76-deficient Jurkat cell lines do not efficiently mobilize Ca²⁺ or activate the Ras pathway in response to TCR stimulation and, consequently, are unable to produce IL-2. In mice deficient in LAT or SLP-76, early thymocyte development is arrested.

In this study, we demonstrate that ZAP-70 as well as LAT and SLP-76 are necessary for the TCR stimulation-induced polarization of the MTOC. As these proteins play a key role in regulating calcium flux, which has previously been shown to be required for MTOC polarization, we also used inhibitors to explore the function of the calcium-regulated enzymes calcineurin and calcium/calmodulin-dependent kinase (CaMK). However, although we confirmed that a calcium-dependent event is required for MTOC polarization, we failed to identify a role for either calcineurin or CaMK. Therefore, our studies implicate a role for proximal components of the TCR-signaling apparatus and a downstream unidentified calcium-regulated process in controlling MTOC polarization.

The Raji B cell line, the Jurkat T cell subclone E6-1, as well as the Jurkat-derived mutants and transfectants P116, P116.c39 (referred to in this article as P116.ZAP70), JCaM2, JCaM2.LAT, JCaM2.LAT.ALLF (10), J14.V29, and J14.76-11 (14) were maintained in RPMI 1640 medium supplemented with 5% FBS, 2 mM glutamine, 50 U/ml penicillin, and 50 μg/ml streptomycin. The Daudi B cell line was maintained in the same medium except with 10% FBS. Cells were maintained in a humidified incubator at 37°C with 5% CO₂. Poly-l-lysine, FK506, and KN-93 were purchased from Sigma-Aldrich (St. Louis, MO). Fura 2-AM was purchased from Molecular Probes (Eugene, OR). Staphylococcal enterotoxin D (SED) was purchased from Toxin Technology (Sarasota, FL). The transfection protocols and NFAT transcriptional reporter assays have been described previously (16).

The mAb Leu 4 (IgG1) is directed against the human CD3ε chain. The mAb 6.7 (IgG1) directed against human β₂ integrin and the mAb HIB19 (IgG1) directed against human CD19 were both purchased from BD PharMingen (San Diego, CA). Goat anti-mouse IgG-Cy5 (Jackson ImmunoResearch Laboratories, West Grove, PA) was used to detect anti-CD19. A rat mAb to α-tubulin (YOL1/34 or YL 1/2) was obtained from Harlan Sera-Laboratories (Loughborough, U.K.) and was detected with a donkey anti-rat IgG-FITC or -Cy3 (Jackson ImmunoResearch Laboratories). Biotinylated Abs were visualized with streptavidin conjugated to Alexa-488 (Molecular Probes).

Jurkat T cells or derived clones and mutants were resuspended in PBS and either left untreated or treated for 30 min at 37°C with either 100 ng/ml FK506 or 1 μM KN-93. Cells (1 × 10⁵) were plated onto Leu 4-coated slides, or anti-CD18 coated slides in the case of Jurkat, and incubated for 15 min at 37°C. Cells were fixed for 20 min in 3.4% paraformaldehyde at room temperature. Fixed cells were permeabilized for 4 min in PBS/0.1% Triton X-100 and blocked for 10 min in PBS/0.2% BSA. Cells were labeled for 20 min with Ab diluted in PBS/0.2% BSA, followed by one wash of 10 min in PBS/0.2% BSA. Coverslips were mounted onto the slides with Mowiol 4-88 mounting solution (Calbiochem, La Jolla, CA) supplemented with 0.2% DABCO anti-fade (Sigma-Aldrich). Optical 1-μm image sections of the cells were collected under a laser-scanning microscope (MRC 600; Bio-Rad, Hercules, CA; Axioplan; Zeiss, Thornwood, NY).

These studies were performed on a Zeiss Axiovert-100TV microscope with a Plan Neo-Glaur ×40/1.3 numerical aperture objective. The microscope is fitted with a high-speed piezoelectric z-motor (Polytech PI, Tustin, CA), dual excitation and emission filter wheels (Sutter Instruments, Novato, CA), and a Princeton Instruments Interline camera (Roper Scientific, Trenton, NJ). We used a high intensity xenon light source (Sutter Instruments). Time lapse acquisitions consisted of a differential interference contrast (DIC) image, followed by 340- then 380-nm fura 2 excitation images, followed by a green fluorescent protein (GFP) z-stack taken over 20 μm at 1-μm intervals. All acquisition and data analyses were done with Metamorph software (Universal Imaging, Downingtown, PA). Background calcium levels were obtained from three frames before activation. GFP intensity data were corrected for background and for photobleaching. Individual cells were analyzed for maximal GFP pixel intensities along the z-planes. The z-planes containing the most intense signal were collected for incorporation into the movies.

Conjugates were made by preloading Daudi B cells with 1 μg/ml SED for 30 min before mixing 1:1 with T cells in complete medium. Cells were then gently pelleted for 30 s and incubated at 37°C for 30 min. Conjugates were placed onto poly-l-lysine-coated slides and allowed to settle for 5 min. Paraformaldehyde was added to yield a 4% final concentration for 30 min. Cells were then blocked in 1% BSA and 10% FBS in PBS. Cells were stained with anti-CD3ε, anti-α-tubulin, and anti-CD19 Abs with the appropriate secondary Abs. Slides were visualized on a Marianas Turn-Key system from Intelligent Imaging (Denver, CO), and images were analyzed using SlideBook software (Intelligent Imaging, Denver, CO). Images were deconvolved by nearest neighbor and exported as TIFF files.

To study MTOC reorientation in real time, a Jurkat cell line stably expressing GFP-tubulin (JTUB) was created. The GFP-tubulin was incorporated into microtubules in these cells and exhibited an especially intense GFP signal at the site of the MTOC. The JTUB cells allowed us to monitor the MTOC both spatially and temporally, while T cell stimulation was occurring. This provided the advantage of viewing a more physiologically relevant interaction between a T cell with a superantigen-loaded B cell (17). JTUB cells were plated on poly-l-lysine-coated slides and then loaded with a fluorescent calcium indicator, fura 2. Relative changes in cytoplasmic free calcium were monitored by observing the ratios of absorbed light of unbound vs calcium-bound fura 2 over time. The ability to monitor the well-characterized calcium response provided a means by which to monitor in real time at least one of the biochemical signal transduction events associated with TCR stimulation. The human B cell line Raji was used as the APC and was incubated with the superantigen SED for 30 min before mixing with JTUB cells. The interactions between the JTUB and the Raji-SED cells were monitored using high-speed time-lapse epifluorescence microscopy.

As Jurkat and Raji cells interacted, time-lapse images were acquired for 15 min at 15-s intervals. Each acquisition consisted of a DIC image and a mid-cell GFP z-section and as a ratio of fura 2 wavelengths (Fig. 1). The ratio of fura 2 wavelengths was pseudocolored to reflect the relative concentration of free calcium in the JTUB cells. More than 100 conjugate pairs were monitored over time in separate experiments. Twenty of these conjugates were monitored continuously for at least 15 min each. In some cases, photobleaching limited the length of time each conjugate could be monitored. Following contact of the JTUB cells with the SED-loaded Raji cells, the JTUB cells rapidly (within 1 min) increased cytoplasmic free calcium and subsequently exhibited an oscillating or plateau pattern of calcium release throughout the course of the experiment. Approximately 5 min after initial contact with the SED-bound Raji cells, the MTOC of the JTUB cells began to move toward the site of cell-cell contact. The MTOC appeared adjacent to the site of B cell/T cell contact by 10 min and did not appear to move again during the subsequent 5 min of observation. No repolarization of the MTOC occurred in JTUB cells stimulated with Raji cells that were not loaded with SED. These real-time observational studies clearly indicated that calcium flux and therefore upstream signaling events precede MTOC polarization.

The kinetics of the calcium response of the LAT-deficient and SLP-76 deficient JCaM2 and J14 cells in response to interactions with Raji-SED was also investigated with this imaging system (Fig. 2). In the case of these two cell lines, only the DIC and calcium ratios were observed simultaneously. The parental Jurkat cell line is shown for comparison. The JCaM2 and J14 cells are known to have calcium defects based on fluorometry studies of populations of cells (10, 14). However, they have not been studied at the single-cell level or with superantigen stimuli. We were also interested in whether the defects in these cells impaired their abilities to form stable conjugates with SED-loaded Raji cells. Somewhat surprisingly, conjugation between the JCaM2 cell and the Raji-SED cell could be observed and was maintained throughout the 15 min of observation during the experiment. In contrast to the Jurkat cell line and consistent with previous results, the JCaM2 cell line did not exhibit any calcium increase in response to contact with the SED-loaded Raji cells.

The J14 cells reacted somewhat differently from either the Jurkat or the JCaM2 cells. Following contact with the SED-loaded Raji cells, the J14 cells exhibited an initial rapid rise in cytoplasmic free calcium, which returned to near-baseline levels and then failed to exhibit the previously observed oscillations of calcium seen with Jurkat cells. This pattern of the initial calcium increase was more moderate and less sustained and is consistent with the changes we previously reported at the population level by fluorometry when J14 cells are stimulated with anti-TCR mAb (14). This leaky phenotype could be secondary to low levels of residual SLP-76, redundancy by other adaptors such as Clnk (18), or lack of an absolute requirement for SLP-76 for low levels of PLCγ1 activation. As with Jurkat and JCaM2, the conjugation of J14 cells with Raji-SED cells was maintained throughout the 15 min of the experiment.

To further characterize the role of SLP-76 in other proximal T cell activation pathways at the single-cell level, we examined MTOC reorientation in J14.V29-SE-coated Daudi cell (as APCs) conjugates. Conjugates were visualized by immunofluorescence microscopy. Stable T cell-APC interactions were identified by tight clustering of CD3 at the cell-cell interface. As previously described, the MTOC of T cells reorient to the site of APC contact (4) (Fig. 3,A). Interestingly, the MTOC of J14.V29 cells did not appear to polarize toward the APC more frequently than by random chance, despite the ability of their CD3 chains within the TCR to cluster adjacent to the SED-coated Daudi cells. To quantitate the percentage of cells that displayed MTOC reorientation, the T cell of a stable conjugate was divided into thirds. If the MTOC was observed in the third of the cell closest to the APC, the conjugate was scored as a positive. Fig. 3 B shows the mean percentage of correctly polarized MTOCs of the indicated cell type and supports the notion that J14.V29 cells do not polarize their MTOC despite the tight CD3 clustering induced by engagement with SE-coated APCs.

MTOC reorientation in T cell-APC conjugates was difficult to study, because conjugate formation occurred somewhat asynchronously and not all Jurkat mutants formed conjugates with the same efficiency. This may be due, in part, to variation in adhesion molecule expression. Also, scoring individual conjugates was quite labor intensive. Therefore, we modified the MTOC reorientation assay to examine the polarization of a large number of cells simultaneously and in a way that allowed us to focus on only the role of TCR signals in inducing MTOC polarization. For this purpose, we used microscopy slides that were coated with anti-TCR Abs. This also allowed us to focus on the isolated contribution of the TCR signaling pathway for MTOC polarization. We have found that the behavior of the MTOC in this assay closely mimics what we have observed in more limited numbers of conjugate pairs of Jurkat/Raji-SED that are individually and asynchronously interacting. This assay allows for a more quantitative approach toward studying MTOC reorientation populations of cells. Jurkat cells or derivative mutants were dropped onto coated slides, and after a 15-min incubation at 37°C, the cells were fixed, permeabilized, and stained with anti-α-tubulin Ab, which was detected by a fluorescent secondary Ab. This method of MTOC identification has been used widely by other laboratories (19, 20, 21) as well as by our own (5). The cells were then scored for polarization by analyzing 1-μm optical sections with a confocal microscope as previously described (20). Cells were scored positive for MTOC reorientation if the MTOC was seen within 2 μm of the slide-bound surface of the cell (Fig. 4,A). Fig. 4 shows an example of Jurkat cells plated on anti-TCR-coated slides with optical slices at 2 and 8 μm (B). About 85% of Jurkat cells were found to polarize toward the Ab-coated surface (Fig. 4 C). For comparison, Jurkat were plated onto anti-β₂ integrin (CD18) Ab-coated slides. Although β₂ integrin is a cell surface protein involved in adhesion and is highly expressed on Jurkat, cells plated onto anti-β₂ integrin Ab-coated slides exhibit randomly oriented MTOC, with only ∼25% polarized toward the Ab-coated surface. These results, using a modified assay, confirm our previous findings that MTOC reorientation is a specific consequence of TCR-induced signaling and is not an effect of accessory receptor cross-linking (5). Using this modified MTOC orientation assay, we assessed three cell lines in which known signaling components are missing.

Studies in a LAT-deficient Jurkat cell line, JCaM2, have revealed a LAT requirement for several downstream effects of TCR stimulation, including the tyrosine phosphorylation of other TCR targets, Ras activation, calcium and phosphatidylinositol production, and actin-ring-associated cell spreading (10, 22, 23), but not for Pak1 activation (24). Using the method described in Fig. 4, we studied whether LAT is also required for MTOC orientation. We found that JCaM2 cells exhibited a randomly oriented MTOC toward anti-TCR-coated slides (Fig. 5 A). When JCaM2 cells were reconstituted with LAT, the orientation was rescued to levels similar to those seen in the parental Jurkat cells. To investigate the necessity of tyrosine phosphorylation of LAT, we analyzed JCaM2 cells that had been stably transfected with a mutant form of LAT, in which all 10 tyrosines were mutated to phenylalanine, LAT.ALLF. The expression of LAT.ALLF failed to reconstitute MTOC orientation in JCaM2 cells. These results clearly demonstrate not only a requirement for the adapter protein LAT in the signaling pathway linking TCR stimulation to MTOC polarization but also establish a dependence on the tyrosine phosphorylation of LAT for correct polarization of Jurkat T cells.

The requirement for LAT tyrosine phosphorylation led us to suspect that the tyrosine kinase responsible for phosphorylating LAT, ZAP-70, would also be required for correct MTOC orientation. To confirm this, we took advantage of a Jurkat cell line in which ZAP-70 is not expressed, P116 cells, to ask whether ZAP-70 is necessary for MTOC orientation. Cells lacking ZAP-70 exhibited a random orientation of MTOC when plated on anti-TCR Ab-coated slides (Fig. 5 B). P116 cells reconstituted with ZAP-70, P116.ZAP70 cells, rescued MTOC orientation. These results clearly demonstrate the requirement for ZAP-70 in TCR-induced polarization of the MTOC.

The initial observations of the J14.V29 cells at a single-cell level raised an interesting question. At the single-cell level, the SLP-76-deficient cells can make an initial weak flux of calcium in response to engagement with an Ag-loaded APC. This is consistent with the weaker and more transient calcium response observed by calcium fluorometry that we reported previously (14). We were curious to determine whether this initial, albeit weaker, calcium burst would be sufficient to drive MTOC polarization. J14.V29 cells exhibited a random orientation of MTOC when plated onto anti-TCR-coated slides compared with the parental Jurkat cells (Fig. 5,C). Analysis of a SLP-76 stably reconstituted J14 cell line, J14.76-11 cells, restored the MTOC polarization of J14 cells to a level resembling that of the parental Jurkat cell line. These data demonstrate the necessity of the adapter protein SLP-76 for polarization of Jurkat cells. Studies to determine which domains of SLP-76 are required for MTOC orientation are ongoing. Preliminary data reveal that mutation of any single domain within SLP-76 is not sufficient to completely disrupt MTOC polarization, consistent with our observations on NFAT responses (25). Together with the data presented in Figs. 3–5, we have demonstrated that superantigen-induced polarization of Jurkat T cells requires components common to the TCR signaling complex including the adapter molecules LAT and SLP-76, as well as the PTK ZAP-70.

Our data reveal a strong correlation between robust calcium fluxes and MTOC orientation. To confirm the necessity of a calcium flux for MTOC polarization, Jurkat cells were washed and resuspended in calcium-free PBS before plating onto anti-TCR mAb-coated slides. The MTOC of these cells failed to polarize correctly (Fig. 6,A). To address which calcium effectors might be involved in driving the polarization of MTOC, we used chemical inhibitors to calcineurin and CaMK and to determine whether these would inhibit MTOC orientation. We pretreated Jurkat cells with FK506, an inhibitor of the calcium-sensitive phosphatase calcineurin. Although FK506 treatment inhibited expression of an NFAT-driven reporter (Fig. 6,B), we saw no difference in the polarization of the MTOC between FK506-treated cells and untreated Jurkat (C). Similarly, inhibition of CaMK with the inhibitor KN-93, at 1 μM, a commonly used working concentration and one that has been shown to inactivate CaMK in Jurkat cells (26, 27, 28), did not alter the polarization of MTOC in Jurkat cells (Fig. 6 C). In contrast to the effects of FK506 on NFAT induction, KN-93 had no effect on NFAT induction (data not shown). Thus, the calcium effectors that are required for MTOC polarization do not appear to depend on calcineurin or CaMK function.

The engagement of a T cell with an APC triggers a number of signaling events, which culminate in the activation of the T cell. These signaling events result in the expression and targeted secretion of cytokines and growth factors toward the interacting APC that promote complete activation of the APC. In order for the T cell to accomplish polarized secretion, the cell-to-cell contact must affect an intricate restructuring of the cytoskeleton. Previous work from our laboratory established that TCR engagement, specifically an ITAM-containing receptor, was sufficient to induce T cell polarization, whereas engagement of accessory cell surface proteins is unnecessary (5). In this study, we extended those studies to address what other proximal intracellular signaling components are required to accomplish this polarization. Our real-time imaging of T cell-B cell interactions allowed us to visualize the kinetics of MTOC polarization and correlate the movement of the MTOC with an early and sustained calcium response. It is very clear that calcium fluxes and therefore early signaling events precede MTOC polarization. These studies are consistent with recently published studies that suggest that Lck-inducible phosphorylation precedes the formation of the immunological synapse (29), although the requirements for MTOC polarization and formation of the immunological synapse is not clear. The observation that signaling mutants of Jurkat, defective in components of the proximal signaling machinery, retained the ability to form stable conjugates with the APC prompted us to ask whether these mutants could polarize the MTOC.

To analyze the polarization of large numbers of cells, we used anti-TCR-coated slides to mimic the interacting surface of an APC. This is analogous to our previous studies, as well as those of others, using anti-TCR-coated beads to study the polarization of T cells (5, 30). With this method, we analyzed the requirement for the tyrosine kinase ZAP-70 and its adaptor substrates, LAT and SLP-76, for polarization. By using the cell lines P116, JCaM2, and J14, which are lacking ZAP-70, LAT, and SLP-76, respectively, we have shown in this study that these molecules are necessary to orient the MTOC toward the site of TCR engagement. Furthermore, the tyrosine phosphorylation of LAT is essential for LAT to rescue the MTOC polarization of JCaM2 cells.

The dependence on the tyrosine phosphorylation of LAT to rescue the polarization defect in JCaM2 cells is consistent with the requirement of LAT phosphorylation for TCR-stimulated transcriptional events and at least some other signaling events, including PLCγ1 activation, Ras activation, Rac activation, and actin-ring-associated spreading (10, 22, 24). It is important to note that Pak1 activation is intact in JCaM2 cells (24). Therefore, we can deduce that activation of Pak1, an important regulator of the actin cytoskeleton, is insufficient for MTOC polarization. Unlike transcriptional responses, the MTOC orientation defect in JCaM2 and J14.V29 cells could not be overcome by treatment with the phorbol ester PMA and the calcium ionophore ionomycin (data not shown). Thus, TCR-induced cytoskeletal reorganization cannot be rescued by mechanisms that bypass the TCR signaling complex to generate calcium fluxes and protein kinase C and Ras activation. Our results would suggest that a calcium-dependent event is necessary but insufficient for MTOC polarization. Thus, there may be a concomitant requirement for the detection of the generation of a localized signal such as the physical protein-protein interactions emanating from the TCR ITAMs to the signaling complex of LAT and SLP-76 and then to the microtubules themselves.

Calcium-dependent events have been reported previously to be required for MTOC polarization (4). However, the events required have not been defined. Because ZAP-70, LAT, and SLP-76 are required for a robust and sustained calcium signal, we asked whether calcium-sensitive enzymes are involved in TCR-driven cytoskeletal restructuring. We addressed this question by using the inhibitors FK506 and KN-93, which inhibit calcineurin and CaMK, respectively. Although treatment of Jurkat cells with the calcineurin inhibitor FK506 inhibited NFAT-driven transcription, treatment of Jurkat cells with FK506 did not inhibit the polarization of MTOC. Similarly, treatment of Jurkat with the CaMK inhibitor KN-93 had no inhibitory effect on MTOC polarization. Therefore, these two calcium-sensitive enzymes are not required for MTOC orientation. Thus, our data reveal that there are differences between the signaling requirements of TCR-induced transcriptional events and TCR-induced MTOC polarization. Likely candidates for the calcium-dependent component might involve myosin motors, which are dependent on calcium for their function. Indeed, studies have suggested that myosin motors may play a role in restructuring T cell membrane components during Ag recognition (31). These motors are likely candidates in connecting the actin cytoskeleton and microtubule cytoskeleton. Future studies will be required to examine the involvement of these motors and define how TCR-regulated events might direct their activities toward sites of Ag recognition.

We thank Dr. Matthew Krummel for assistance with imaging. We thank the members of the Weiss and Davis labs for helpful discussions.