Recruitment of Transferrin Receptor to Immunological Synapse in Response to TCR Engagement¹

T cell receptor engagement by an APC induces the formation of a highly organized network of receptors and intracellular signaling molecules (1), accompanied by a profound reorganization of the actin cytoskeleton (2) and extensive remodeling of surface proteins and lipids (3, 4). The formation of this network, known as the immunological synapse, is a multistep process involving lateral movement of proteins along the plane of the plasma membrane that eventually coalesce at the contact area between the T cell and the APC (5, 6, 7). The synapse is organized into two differentiated concentric rings or supramolecular activation clusters (SMAC)⁴ (5), termed the central SMAC, which contains the TCR, CD28, CD4, and associated signaling molecules (1, 8), and the peripheral SMAC (p-SMAC), which contains the adhesion LFA-1 and ICAM molecules and the cytoskeletal molecule talin (1, 9). Membrane microdomains with a high content of both glycolipid and cholesterol, called lipid rafts (10), appear to participate in the formation of the synapse during the spatial-temporal remodeling of the membrane at the contact site (11). In addition, lipid rafts are involved in membrane trafficking to the rear pole of polarized T lymphocytes (12), and in the assembly of the signaling machinery for T cell activation (4, 13).

The membrane glycoprotein transferrin receptor (TfR, CD71) is a housekeeping receptor that constitutively cycles between the plasma membrane and the endosomal compartment. The TfR binds iron-loaded transferrin at the cell surface and is internalized and transported to early sorting endosomes, where the complexed iron is rapidly separated from transferrin. Subsequently, the TfR-transferrin complex accumulates within recycling endosomes and then recycles to the plasma membrane (14). In addition to this general function, the TfR appears to play a costimulatory role in T cell activation (15, 16). In this study, we show that TCR/CD3 triggering up-regulates surface expression of the TfR during the first minutes of stimulation. This increase took place by translocation of internal TfR from the endosomal compartment, and was concomitant with the partial incorporation of the receptor into lipid rafts. The endosomal and surface TfR pools polarized to the contact site when T cells were activated with anti-CD3 and anti-CD28 Abs immobilized onto beads, or with Raji B cells presenting staphylococcal enterotoxin (SEE) superantigen acting as APCs. In the latter case, surface TfR redistributed to the p-SMAC ring of the synapse formed at the interface between the T cell and the APC. Blockade of the TfR with specific Abs interfered with the formation of T cell-APC conjugates and with the recruitment of Lck and CD3ε to the contact site. Our results suggest an important role for the TfR in early events in the formation of the immunological synapse.

Mouse mAb MEM-57 (anti-CD3ε, IgG2a) and MEM-92 (anti-CD3ε, IgM) were kindly provided by V. Horejsi (Institute of Molecular Genetics, Prague, Czech Republic). Rabbit polyclonal Ab to Lck (used for immunofluorescence analysis) was kindly provided by A. Veillette (McGill University, Montreal, Canada); mouse mAb to Lck (used for Western blotting) and anti-phosphotyrosine Abs were from Transduction Laboratories (Lexington, KY). The mouse IgG anti-TfR FG 1/5, FG 1/6, FG 2/12 mAbs (17) and the polyclonal rabbit Abs to CD3ζ were generous gifts from B. Alarcón (Centro de Biología Molecular “Severo Ochoa,” Madrid, Spain); the anti-TfR mAb H68.4 (used for Western blotting) was from Zymed Laboratories (San Francisco, CA); the mouse anti-TfR mAb L5.1 (used for immunofluorescence analysis) was purchased from the American Type Culture Collection (Manassas, VA); the biotinylated BV8 anti-human Vβ8 mAb was from BD Biosciences (San Diego, CA); the fluorescent anti-TCRαβ Ab αβ was from Immunotech (Marseille, France). Other Abs used were the anti-LFA-1 mAb TP40.1.2 (18) and the anti-CD28 mAb 9.3 (19). The PE-coupled mouse mAb anti-CD3 and anti-CD4 were kindly provided by M. Toribio (Centro de Biología Molecular “Severo Ochoa”). The Alexa fluor 594 Ab-labeling kit and the fluorescent cell tracker chloromethyl derivative of aminocoumarin (CMAC) were purchased from Molecular Probes (Eugene, OR). We obtained SEE from Toxin Technology (Sarasota, FL). HRP-conjugated Abs were obtained from Pierce (Rockford, IL); fluorochrome-labeled secondary goat Abs were from Southern Biotechnology Associates (Birmingham, AL).

Human T lymphoblastoid Jurkat cells were grown in RPMI 1640 supplemented with 5% FBS (Sigma-Aldrich, St. Louis, MO), 50 U/ml penicillin, and 50 μg/ml streptomycin at 37°C in an atmosphere of 5% CO₂/95% air. Cells (1.0 × 10⁸) were washed, incubated for 10 min at 37°C in RPMI 1640 medium plus 1% FBS, and, when indicated, stimulated for 3 min at 37°C with a mixture of 0.5 μg/ml each anti-CD3 mAb MEM-92 and MEM-57. Cells were then lysed for 15 min in 25 mM Tris-HCl, pH 7.2, 150 mM NaCl, and 0.2% Triton X-100 at 4°C in the presence of phosphatase inhibitors (10 mM NaF, 1 mM sodium orthovanadate, and 10 mM β-glycerol phosphate) and protease inhibitors. The lysate was homogenized by passing the sample through a 22-gauge needle, and the extract was centrifuged to equilibrium in a continuous sucrose density gradient (20). The lysate was then brought to 40% sucrose (w/w) in a final volume of 4 ml and placed at the bottom of an 8-ml 5–30% linear sucrose density gradient. Gradients were centrifuged for 18 h at 39,000 rpm at 4°C in a Beckman SW41 rotor (Beckman Coulter, Fullerton, CA). Fractions of 1 ml were harvested from the bottom of the tube. Equivalent aliquots from each fraction were subjected to SDS-PAGE and transferred to Immobilon-P membranes (Millipore, Bedford, MA). After blocking with 5% nonfat dry milk, 0.05% Tween 20 in PBS, blots were incubated with the appropriate Abs for 1 h. After several washings, blots were incubated for 30 min with anti-mouse or anti-rabbit IgG Abs coupled to HRP, washed extensively, and developed using an ECL kit (Amersham, Arlington Heights, IL).

Jurkat cells (1.0 × 10⁶ cells) stimulated with an anti-CD3 mAb mixture consisting of MEM-57 and MEM-92 (0.5 μg/ml) were analyzed at different times by flow cytometry for the expression of TfR, CD3ε, and CD4. The mean fluorescence intensity obtained in each case was compared with that obtained for cells that had not been stimulated that was taken as 100%. CD3ε and CD4 were directly detected with PE-coupled mAbs; the TfR was detected in parallel with anti-TfR mAb L5.1 (IgG1) mAb and fluorescent secondary anti-IgG1 Abs, thus preventing recognition of the anti-CD3 Abs (IgG2a and IgM) used for activation. The activating anti-CD3 Abs did not affect the binding of the fluorescent anti-CD3 Ab used to measure the CD3 surface levels, as assayed in cells incubated at 4°C. Results similar to those observed for CD3 were obtained with fluorescent anti-TCRαβ Ab αβ, which was used to measure the expression of the TCR/CD3 complex (our unpublished observations). Controls to assess the specificity of the labeling included incubations with control primary Abs or omission of the primary Abs.

Cells were fixed in 4% (w/v) paraformaldehyde in PBS for 20 min, rinsed, and treated with 10 mM glycine for 5 min to quench the aldehyde groups. Cells were then washed, permeabilized or not with 0.2% Triton X-100 in PBS at 4°C for 5 min, rinsed, incubated with 3% (w/v) BSA for 15 min, and incubated with the primary Ab. After 1 h at room temperature, cells were washed and incubated with the appropriate fluorescent secondary Ab. For double-labeling experiments, the same procedure was repeated for the second primary Ab. In the cases in which two different mouse mAbs were used, the analysis was done either by using fluorochrome-labeled, isotype-specific, secondary Abs (double-label analysis of TfR and CD3ε) or with the second primary Ab biotinylated and subsequent detection with fluorochrome-labeled streptavidin (double-label analysis of TfR and LFA-1 and TfR and Vβ8). In the experiments indicated, the TfR was directly visualized using the anti-TfR mAb L5.1 covalently linked to Alexa-594. Controls to assess the specificity of the labeling included incubations with control primary Abs or omission of the primary Abs. Images were obtained using a Bio-Rad (Hercules, CA) Radiance 2000 confocal laser microscope. To quantify the recruitment of the TfR to the contact area of the T cell with the bead, 10 different random fields per experiment in three different triplicates were analyzed using a conventional fluorescence microscope (Zeiss, Oberkochen, Germany). Cells were divided into four quadrants. Cells with >75% of the TfR staining in the quadrant contacting the bead were scored as positive. Fluorescence intensity was quantified using the Metamorph program (Universal Imaging, Downingtown, PA).

Polystyrene latex microspheres were coated with 10 μg/ml anti-CD3 Ab and/or anti-CD28 Ab. Jurkat cells (1 × 10⁴ cells) were cultured with 1 × 10⁵ beads in 200 μl of RPMI 1640 containing 10% FBS in round-bottom microplates. Cells were fixed and processed for immunofluorescence analysis.

To distinguish Raji cells from Jurkat cells in the cell-cell conjugates, we loaded Raji cells with blue fluorescent tracker CMAC. Briefly, Raji cells were preincubated in RPMI 1640/5% FBS containing 10 μM CMAC for 20 min at 37°C, washed, and resuspended (5 × 10⁶ cells/ml) in RPMI 1640/5% FBS. The Raji cells were then incubated for 20 min in the presence or absence of 5 μg/ml SEE, mixed with an equal number of Jurkat J77 cells (2 × 10⁶ cells/well) in a final volume of 600 μl, and immediately plated onto poly(l)lysine-coated slides. After incubation for 10 min at 37°C, cells were fixed, permeabilized or not, stained with the appropriate Abs, and analyzed under a confocal fluorescence microscope. Conjugates were first identified by directly observing both cell morphology under differential interference contrast and blue fluorescent CMAC-labeled cells.

Engagement of TCR/CD3 with an anti-CD3 Ab mixture induces the capping of the complex and its redistribution to lipid rafts. Simultaneously, a specific subset of surface proteins redistributes to rafts and colocalizes with the TCR/CD3 complex. This membrane reorganization can be monitored biochemically by using a flotation assay, based on the resistance of lipid rafts to solubilization by detergents (18). As the TfR appears to play a costimulatory role in T cell activation (15, 16), we addressed whether engagement of the TCR produces compartmentalization of the TfR into rafts. Fig. 1,A shows that in Jurkat T cells that were stimulated with anti-CD3-specific Abs for 3 min ∼50% of the TfR redistributed into raft fractions, whereas it was basically undetectable in these fractions in unstimulated cells. In addition, active Lck, recognized by its marked mobility shift, was also recruited to lipid rafts. The ε and ζ CD3 subunits were also incorporated into the same fractions. As an internal control, we observed that most CD3-induced tyrosine phosphorylation was detected in the buoyant fractions containing the rafts (Fig. 1,B). Fig. 1,C shows that in parallel with the compartmentalization of the TfR, engagement of TCR/CD3 led to increased surface expression of the TfR, as measured by flow cytometry. After 3 min of stimulation, the surface expression of the TfR showed a 2-fold increase relative to that in resting cells and then progressively declined to recover the initial levels of surface expression. As controls, we observed that the surface expression of CD4 was not significantly altered and CD3 was rapidly down-modulated. An identical result was obtained when the cells were pretreated with 100 μg/ml cycloheximide 15 min before CD3 engagement and maintained thereafter (our unpublished observations), implying that the increase in surface expression of the TfR takes place by mobilization of internal pre-existing molecules. A large amount of TfR was distributed intracellularly in endosomes, in keeping with its constitutive role in transferrin recycling (Fig. 1 D).

Lipid raft coalescence defines membrane subdomains with polarized distribution (13). The changes observed in the incorporation of the TfR into rafts prompted us to examine the polarization of the TfR under simultaneous engagement of the TCR/CD3 complex and CD28 using beads coated with anti-CD3 and anti-CD28 Abs, which act as surrogate APCs. Cells were incubated for 3 min in the presence of beads coated with anti-CD3 and/or anti-CD28 Abs, fixed, and permeabilized, and the distribution of the TfR was analyzed by conventional immunofluorescence microscopy. The number of bead-cell conjugates greatly increased in the presence of anti-CD3 or anti-CD28 Abs (Fig. 2,A, top panel). The conjugated cells contacted were then scored for the orientation of the intracellular TfR relative to the site of contact. Fig. 2,A (bottom panel) shows that the intracellular TfR was polarized to the contact site in 65% of cells in the presence of beads coated with both anti-CD3 and anti-CD28 Abs (Fig. 2,B). In contrast, in the presence of anti-CD3 or anti-CD28 Abs alone, only 30–35% of the cells displayed the TfR oriented to the contact site, a percentage similar to that found when using a negative control Ab and close to that obtained by random attachment of the cell to the bead in the absence of coupled Ab (25%) (Fig. 2,A). Results similar to those for the intracellular TfR pool were found for surface TfR (Fig. 2,C). This indicates that, in addition to the general reorientation of the exocytic and endocytic compartments, surface TfR also polarized to the contact site. Fig. 3 shows that the redistributed intracellular pool of TfR partially colocalized with Lck and CD3ζ at the Jurkat cell-bead interface in cells stimulated simultaneously with anti-CD3 and anti-CD28 Abs.

Jurkat cells expressing the Vβ8 gene segment of the TCRβ chain (Jurkat J77) are able to recognize the SEE superantigen bound to the human B cell line Raji, which acts as an APC, and forms conjugates similar to genuine T cell-APC pairs (9, 21). The TCR and many other membrane proteins and signaling machinery redistribute to the contact interface formed between the Jurkat and the Raji cell (9, 21). Fig. 4,A (top and middle panels) shows that, while surface TfR was evenly distributed on Jurkat cells in the absence of SEE, the TfR polarized to the immunological synapse in the conjugates formed in the presence of SEE. Moreover, consistent with the results obtained with coated beads, the endosomal TfR pool reoriented in the conjugates to face the T cell-APC interface (Fig. 4,A, bottom panels). The observed redistribution of surface TfR to the interface formed by the Jurkat and Raji cells prompted us to investigate the cellular origin of TfR. We did this by labeling the TfR either on the Jurkat or the Raji cell surface for 20 min at 4°C with anti-TfR mAb L5.1 coupled to Alexa-594, and mixing them with unlabeled Raji or Jurkat cells, respectively, in the presence of SEE. This Ab affects neither the efficiency of conjugate formation nor the translocation of CD3ε to the immunological synapse (our unpublished observation). After 10 min at 37°C, the conjugates formed under each set of conditions were monitored for the redistribution of fluorescent label at the site of contact. Fig. 4,B shows that the labeled TfR accumulated at the T cell-B cell interface only when the T cell was labeled, implying that most of the TfR at the immunological synapse is provided by the Jurkat cell. To establish whether the translocation of the TfR takes place in normal T cells, human peripheral T cells were incubated with Raji cells in the presence or absence of SEE, and the distribution of the TfR in the SEE-responding Vβ8⁺ T cell population was determined by immunofluorescence microscopy. The TfR accumulated at the contact site in Vβ8⁺ T cells in the presence of SEE (Fig. 4,C), but not in its absence (our unpublished observations). To confirm that the recruitment of the TfR to lipid rafts observed in Jurkat cells upon treatment with mitogenic anti-CD3 Abs was not a particular aspect of this cell line or of the use of cross-linking Abs, we compared the TfR content in the insoluble raft membrane fraction obtained from Jurkat J77 cells conjugated with Raji cells in the presence or absence of SEE, and in primary T cells treated or not with mitogenic anti-CD3 Abs. Fig. 4 D shows that compartmentalization of the TfR in lipid rafts took place in Jurkat cell only in the presence of SEE (top panel) and when primary T cells were activated with mitogenic anti-CD3 Abs (bottom panel), but not in the absence of SEE or anti-CD3 Ab treatment.

Fig. 5 shows that the accumulation of surface TfR occurred preferentially at the p-SMAC, as revealed by double-label immunofluorescence analysis of TfR and LFA-1 or CD3ε, used as markers of the p-SMAC and central SMAC, respectively. Quantitative analysis showed a significant enrichment in the levels of the TfR (∼4.3) and CD3ε (∼9.8), but not of CD45, in the contact zone in the conjugates formed in the presence of SEE, whereas all of them were evenly distributed in its absence (Fig. 6).

To investigate the functional role of the TfR in the formation of the immunological synapse, we analyzed the effect of perturbing the TfR by treatment with either transferrin or with mAbs FG 1/5, FG 1/6, or FG 2/12 directed to different epitopes of the TfR molecule (17). The number of T cell-APC contacts showed a 100% increase in the presence of SEE relative to those formed in its absence (Fig. 7,A). This increase was slightly reduced by treatment with transferrin or mAb FG 1/6 or FG 1/5, but treatment with FG 2/12 mAb caused a 60% drop in the number of T cells contacted by APCs. Under the experimental conditions used, preincubation with transferrin did not significantly affect the steady state surface levels of the TfR or the redistribution of the TfR to the immunological synapse, as assayed by flow cytometry or immunofluorescence analysis, respectively (our unpublished observations). To investigate further the effect of the different treatments used in Fig. 7,A on the assembly of the immunological synapse, the presence of CD3ε and Lck in the contact zone of the T cell-APC conjugates was determined 10 or 30 min after mixing the cells, as a measure of the efficiency of the formation of immunological synapses. Fig. 7,B shows that, whereas the rest of the treatments only caused a modest decrease in the number of conjugates with CD3ε polarized to the synapse, only 60% of the T cell-APC contacts contained CD3ε after 10 min in the cultures treated with mAb FG 2/12. Similar results were obtained after 30 min of conjugation and for the polarization of Lck (our unpublished observations). Consistent with our previous observation that the TfR accumulated at the contact site is supplied by the T cell (Fig. 4 B), the preincubation of SEE-loaded Raji cells with the blocking anti-TfR mAb FG 2/12 affected neither the number of conjugates formed with Jurkat cells nor the presence of the TfR, CD3, and Lck at the contact site (our unpublished observations).

The formation of an immunological synapse between a T cell and an APC is a complex process involving the redistribution of specific surface proteins (TCR/CD3, LFA-1, ICAM-1, ICAM-3) to the site of contact, the translocation of intracellular signaling machinery (Lck, protein kinase C-θ, ZAP70) to that site, and the reorientation of the secretory apparatus to face the T cell-APC interface (1, 2, 3, 4, 5, 6, 7, 22). These processes are triggered upon engagement of the TCR/CD3 complex by the Ag presented by the APC, and are mediated by the reorganization of the cytoskeleton and extensive remodeling of specialized raft membranes. In this study, we have used a combination of experimental approaches to show that the T cell surface levels of the TfR are up-regulated from the internal TfR pool after TCR/CD3 engagement. The TfR on the T cell surface relocates to the p-SMAC, and the TfR endosomal pool becomes placed in front of the immunological synapse.

Our observation that treatment with anti-TfR FG 2/12 mAb impaired the efficient formation and the quality of T cell-APC contacts is consistent with an important role for the TfR in this process. Alternatively, because the TfR does not have a counterreceptor in the APC, it is possible that the binding of the FG 2/12 mAb perturbs conjugate formation and, as a consequence, CD3ε and Lck accumulation by steric hindrance. However, although it cannot be ruled out, this explanation seems unlikely because all the anti-TfR mAb assayed were of the same Ig class (IgG). Our findings are consistent with other results showing that FG 2/12 mAb strongly blocks target killing by NK cells, whereas FG 1/5 has only a moderate effect, and FG 1/6 mAb and transferrin are ineffective (17). The observation that FG 1/6 mAb did not interfere with the formation of conjugates or the recruitment of CD3ε and Lck to the contact site is consistent with previous reports that the perturbation of the TfR with this Ab transduces costimulatory signals (15, 16). It is of note that mAb FG 2/12, but not mAb FG 1/5 or FG 1/6, blocks the binding of transferrin to its receptor (17). This indicates that a region in the TfR close to, but different from, the transferrin binding site is important for the efficient formation of T cell-APC contacts. Functional cross talk of the TfR with the TCR complex was evidenced by the finding that TfR stimulation results in tyrosine phosphorylation of CD3ζ and, conversely, stimulation of the TCR increases tyrosine phosphorylation of the TfR (16). The recruitment of the TfR to the immunological synapse in the T cell-APC conjugates brings the TfR into the proximity of the TCR complex, and this may facilitate cross talk between the two receptors. It is of particular note that, even in the reduced number of T cell-APC pairs formed in the presence of mAb FG 2/12, the percentage of cells with CD3ε and Lck recruited to the cell-cell contact decreased by 40%. This indicates that, in addition to its role in the formation of the T cell-APC contacts, surface TfR probably participates in events occurring before the assembly of the immunological synapse, such as in the TCR-mediated signaling process preceding the formation of the synapse (23).

The spatial and temporal recruitment of Lck to the immunological synapse has been recently investigated (24). Lck are initially recruited to the synapse periphery. Later on, CD3ζ becomes enriched in the center, while Lck is enriched at the synapse periphery. This pattern is similar regardless of activation with strong or weak agonists, although the efficiency of the conjugation is reduced in the latter case (24). The translocation of internal TfR to the cell surface and its incorporation into rafts observed soon after TCR engagement might reflect an early supply of endosomal components to the immunological synapse, including the raft-associated Lck molecule (24) and raft lipids (25). The inefficient recruitment of Lck to the T cell-APC contact area in the presence of the anti-TfR mAb FG 2/12 is consistent with transport of Lck being coupled to TfR trafficking. Furthermore, the fact that recycling endosomes, as visualized by staining of internal TfR, reoriented to face the T cell-APC contact area suggests that these endosomes may intervene in additional processes occurring at the synapse, such as a late supply of Lck (24) or transport of endocytosed TCR/CD3 complexes for degradation (26).

We thank S. Gómez for technical help. We express our gratitude to B. Alarcón and K. Bernouin for critically reading the manuscript.