Engaged and Bystander T Cell Receptors Are Down-modulated by Different Endocytotic Pathways*

T cell antigen receptor (TCR) engagement by stimulatory antibodies or its major histocompatibility complexantigen ligand results in its down-modulation from the cell surface, a phenomenon that is thought to play a role in T cell desensitization. However, TCR engagement results in the down-modulation not only of the engaged receptors but also of non-engaged bystander TCRs. We have investigated the mechanisms that mediate the down-modulation of engaged and bystander receptors and show that co-modulation of the bystander TCRs requires protein-tyrosine kinase activity and is mediated by clathrin-coated pits. In contrast, the down-modulation of engaged TCRs is independent of protein-tyrosine kinases and clathrin pits, suggesting that this process is mediated by an alternate mechanism. Indeed, down-modulation of engaged TCRs appears to depend upon lipid rafts, because cholesterol depletion with methyl-β-cyclodextrin completely blocks this process. Thus, two independent pathways of internalization are involved in TCR down-modulation and act differentially on directly engaged and bystander receptors. Finally, we propose that although both mechanisms coexist, the predominance of one or the other mechanisms will depend on the dose of ligand.

T cell antigen receptor (TCR) engagement by stimulatory antibodies or its major histocompatibility complexantigen ligand results in its down-modulation from the cell surface, a phenomenon that is thought to play a role in T cell desensitization. However, TCR engagement results in the down-modulation not only of the engaged receptors but also of non-engaged bystander TCRs. We have investigated the mechanisms that mediate the down-modulation of engaged and bystander receptors and show that co-modulation of the bystander TCRs requires protein-tyrosine kinase activity and is mediated by clathrin-coated pits. In contrast, the down-modulation of engaged TCRs is independent of protein-tyrosine kinases and clathrin pits, suggesting that this process is mediated by an alternate mechanism. Indeed, down-modulation of engaged TCRs appears to depend upon lipid rafts, because cholesterol depletion with methyl-␤-cyclodextrin completely blocks this process. Thus, two independent pathways of internalization are involved in TCR down-modulation and act differentially on directly engaged and bystander receptors. Finally, we propose that although both mechanisms coexist, the predominance of one or the other mechanisms will depend on the dose of ligand.
Regulation of TCR 1 expression at the plasma membrane is an important means of calibrating the T cell response to its MHC-antigen peptide (MHCp) ligands. At rest, the TCR is continuously internalized and recycled at the T cell surface, a cycle that is dependent on a dileucine endocytosis motif present in the CD3␥ subunit (1)(2)(3)(4)(5). The significance of the constitutive internalization-recycling process is unknown, although it could certainly add plasticity to the topological distribution of the TCR. Indeed, this may facilitate the rapid transport of TCRs to the immune synapse via recycling endosomes when an appropriate antigen-presenting cell is encountered (6). Furthermore, the endocytosis of the TCR is increased, and TCR recycling is reduced upon T cell activation leading to the down-modulation of the TCR at the cell surface. This is in part because of the phosphorylation of a serine residue upstream of the dileucine motif in CD3␥, which improves the presentation of the dileucine signal to the endocytic machinery. As a result, the dileucine motif binds to AP-2 that in turn recruits the clathrindependent endocytic machinery leading to the internalization of the TCR (7). However, although the serine-dependent dileucine motif of CD3␥ is required for phorbol ester-induced TCR down-regulation, it is not necessary for TCR downregulation induced by anti-CD3 antibodies or by antigenpresenting cells bearing the appropriate MHCp. Hence, it appears that other endocytotic signals may be involved in these cases (8 -11).
It remains unclear what the mechanisms are that underlie TCR down-modulation. Some studies suggest that TCR downmodulation is mediated by an increase in internalization and degradation (9,12), whereas others indicate that TCR downregulation is the result of the increased degradation of the internalized receptors (13). Whichever the case, TCR degradation through a combination of lysosome-and proteasome-dependent mechanisms does appear to occur (13,14).
The contribution of TCR signaling to TCR internalization and down-modulation is also unclear. Some studies suggest that the TCR internalization and down-modulation are protein-tyrosine kinase (PTK)-dependent processes that ultimately result in the degradation of the TCR (5,(15)(16)(17). However, in other studies PTK inhibitors did not affect TCR downmodulation (18,19).
By studying the T cells expressing two distinct TCRs, it was concluded that only TCRs directly triggered by the appropriate MHCp are down-regulated (20 -22). However, other studies have demonstrated that TCR engagement by antigen, superantigen, or anti-TCR antibodies not only led to the down-regulation of directly triggered TCRs but also to the co-modulation of a proportion of non-engaged (bystander) TCRs (23)(24)(25)(26)(27)(28).
We found previously that bystander TCRs are modulated by a mechanism that requires the activation of PTKs and protein kinase C by the engaged TCR (28). Furthermore, it was recently shown that in contrast to the internalization of engaged TCR, the modulation of bystander TCRs is highly dependent on protein kinase C and the dileucine endocytosis motif of CD3␥ (23). In contrast the engaged TCR is downmodulated by a mechanism independent of PTKs. We have further investigated the mechanisms that mediate the internalization of engaged and co-modulated TCRs and found that although the bystander TCR is internalized by a clathrin-dependent process, internalization of the engaged TCR is lipid-raft-dependent.  1 The abbreviations used are: TCR, T cell antigen receptor; M␤CD, methyl-␤-cyclodextrin; MHC, major histocompatibility complex; MHCp, MHC complexed with peptide antigen; NP, nitrophenol; PTK, proteintyrosine kinase; GFP, green fluorescent protein.

EXPERIMENTAL PROCEDURES
Cells-The cell line CH7␤Rex TT (TT clone) is a Jurkat transfectant that has been described previously and that expresses the TCR␣ and ␤ chains (V␤3.1) of HA 1.7, a T cell clone specific for influenza hemagglutinin peptide 307-319 (27). This cell line was additionally transfected with a chimeric gene encoding the extracellular and transmembrane domains of the human IL-2 receptor ␣ chain (Tac) fused to the cytoplasmic tail of the CD3 chain (29). The cell line Jk.scTCR␤ (clone 2D10) was obtained by transfection of Jurkat with the expression vector pSR␣-scTCRV␤3.
Antibodies and Chemicals-The mouse monoclonal antibody directed against human CD3, OKT3, was obtained from Ortho Diagnostics. The anti-Tac antibodies, MAR 108 and TP1/6 were generously provided by Drs. M. López-Botet and F. Sá nchez-Madrid, respectively (Hospital de la Princesa, Madrid, Spain). The CD3-specific antiserum, Ab 448, was generated by immunization of a New Zealand rabbit with a synthetic peptide corresponding to the amino acid sequence 109 -132 of human CD3 coupled to keyhole limpet hemocyanin. The anti-clathrin heavy chain antibody was purchased from Transduction Laboratories. The fluorescein isothiocyanate-conjugated antibodies specific for mouse Igs and fluorescein isothiocyanate-conjugated streptavidin were purchased from Southern Biotechnology, and the fluorescein isothiocyanate-conjugated antibodies specific for mouse V␤8 (F23.1) were purchased from Pharmingen. The antibody specific for V␤3, Jovi3, was a kind gift of Dr. M. Owen (Cancer Research). PP2 was purchased from Calbiochem, methyl-␤-cyclodextrin was purchased from Sigma, and nitrophenol (NP)-bovine serum albumin was purchased from Biosearch Technologies.
Down-regulation Experiments-Antibody-coated plates were generated by incubating plastic 96-well plates (Costar) overnight at 37°C with 100 l/well of the appropriate antibody (25 g/ml) in phosphatebuffered saline, and they were washed with phosphate-buffered saline before use. Using radiochemical techniques, the total bound antibody under these conditions was estimated as 20 ng/well. Cells (10 5 /well) were incubated with the desired antibodies in culture medium at 37°C for the times indicated. The cells were then collected, transferred to ice-cold phosphate-buffered saline, and washed twice. Subsequently, the cells were stained with the stimulating antibody, followed by a phycoerythrin-conjugated secondary antibody. In co-modulation studies, the cells were stained with the specific antibody labeled with biotin followed by streptavidin-phycoerythrin. The cells were ultimately analyzed in a FACSCalibur flow cytometer (BD Biosciences), and the mean fluorescence intensity was measured at each point.
Inhibition of Clathrin-mediated Endocytosis-Potassium depletion was carried out as described previously (32). Briefly, the cells were washed and resuspended in hypotonic medium (Dulbecco's modified Eagle's medium:water, 1:1), and after 5 min at 37°C the cells were washed in medium without K ϩ (100 mM NaCl, 50 mM HEPES, pH 7.4) and then resuspended in the same medium with 1 mg/ml bovine serum albumin. Cells were incubated at 37°C in wells previously coated with the desired antibodies for the times indicated. For acid treatment, the cells were washed and resuspended in RPMI supplemented with 2.5g/ liter glucose, 10% fetal bovine serum, and 10 mM acetic acid, as described (33). The cells were then incubated at 37°C in wells previously coated with the desired antibodies for the times indicated.

RESULTS
The Nature and Dose of the Ligand Determines the Sensitivity of Antibody-induced TCR Down-modulation to PTK Inhibitors-Conflicting results have been obtained from different studies on the PTK dependence of antibody-induced TCR downmodulation (15)(16)(17)(18)(19). One reason for such variation could be the use of soluble or immobilized (plate-bound) antibodies. We analyzed the effect of the Src kinase inhibitor PP2 on TCR downmodulation induced by stimulation of Jurkat T cells with different doses of soluble or plate-bound anti-CD3 antibodies. The dose-response curves in the absence of PP2 were very different when soluble or immobilized anti-CD3 was used (Fig. 1A). In the presence of soluble anti-CD3, TCR down-modulation was observed at concentrations as low as 0.1 g/ml, and it steadily increased in response to increasing antibody concentrations. When plate-bound antibody was used, TCR down-modulation was not observed until the antibody concentration reached 1 g/ml and above. The response to increasing concentrations of immobilized anti-CD3 was abrupt, resulting in a shift from none to almost complete down-modulation with a 10-fold increase in antibody concentration (Fig. 1A).
The effect of PTK inhibition by PP2 was also dependent on the way the stimulus was presented. Although TCR downmodulation induced by soluble anti-CD3 was partly inhibited at all antibody concentrations, the effect on TCR down-modulation induced by a plate-bound antibody was strongly dependent on the dose of the antibody (Fig. 1A). At relatively low concentrations (1-3 g/ml) TCR down-modulation was completely inhibited by PP2, whereas at high concentrations (10 -30 g/ml) TCR down-modulation was refractory to PTK inhibition. At the dose used, PP2 completely blocked TCR signaling, even at high doses of stimulating antibody, as measured by the tyrosine phosphorylation of total proteins and the induction of the CD69 activation marker (data not shown), suggesting that PTKs were completely inhibited. Therefore these results suggest that PTK-dependent and -independent mechanisms mediate the antibody-induced down-regulation of the TCR. Moreover, the preponderance of one or the other depends on the way the antibody is presented (soluble or immobilized) as well as on the dose of antibody.
At certain concentrations of plate-bound antibody, TCR engagement led to its down-modulation by a PTK-independent mechanism. A feature that characterized this mechanism was that the stimulated cells shift abruptly from a non-modulated to a fully modulated state (clearly seen in Fig. 1B). In the absence of PP2, stimulation with increasing concentrations of soluble anti-CD3 induced a gradual decrease of TCR expression on the cell surface. However, when stimulated with a high concentration of plate-bound anti-CD3 (10 g/ml) two cell populations were seen to exist, one in which the TCR was not downmodulated (Fig. 1B, population a) together with another in which TCR down-modulation was virtually complete (population b). In contrast, stimulation with a small amount of immobilized anti-CD3 (1 g/ml) induced only a small decrease in TCR expression in the whole T cell population (Fig. 1B). Incubation with PP2 completely prevented the decrease in TCR intensity in cells stimulated with low doses of immobilized anti-CD3 (or with soluble anti-CD3) but did not alter the discontinuous down-modulation induced by high doses of immobilized anti-CD3.
Directly Engaged and Bystander TCRs Are Down-modulated by Different Mechanisms-Along with others, we have demonstrated that engagement of a TCR with a specific ligand results not only in its own down-modulation but also in that of nonengaged bystander TCRs. We proposed that the differential sensitivity of TCR down-modulation to PP2 inhibition observed upon stimulation with different doses of immobilized anti-CD3 (shown in Fig. 1) might indicate the existence of different mechanisms underlying the down-modulation of directly engaged and bystander TCRs. We determined the sensitivity to PP2 of the down-modulation of engaged and non-engaged TCRs. Jurkat T cells were transfected with an immunoglobulin single chain (scFv)-TCR␤ construct in a way that the trans-fected cell expresses two TCR complexes, one with the endogenous V␤8-containing TCR␤ and the other with the transfected scFv-(and V␤3)-containing TCR␤. The immunoglobulin single chain fused to the transfected V␤3-TCR␤ confers specificity to the NP hapten, and a stable cell transfectant of scFv(V␤3)-TCR␤ in Jurkat T cells was generated. Stimulation of these cells with NP conjugated to bovine serum albumin induced near maximal down-modulation of the scFv(V␤3)-containing TCR at a concentration of 30 g/ml (Fig. 2). The dose-response curve and the sensitivity to PP2 inhibition made the stimulation with NP-bovine serum albumin appear very similar to stimulation with soluble anti-CD3 (Fig. 1A). Furthermore, stimulation with NP-bovine serum albumin induced the comodulation of the non-engaged V␤8-containing TCR. Interestingly, the co-modulation of the bystander V␤8-TCR was completely inhibited by PP2 treatment, whereas down-modulation of the directly engaged TCR was only partially inhibited. Therefore, these results indicate that the bystander non-engaged TCR is down-modulated by a PTK-dependent mechanism, whereas the stimulated TCR is down-regulated by a mixture of PTK-dependent and -independent mechanisms.
Down-modulation of Directly Engaged and Bystander TCRs Are Differentially Dependent on Endocytosis Mediated by Clathrin-coated Pits-To further study the mechanisms that mediate the down-modulation of engaged and bystander TCR, we proceeded to study Jurkat T cells transfected with a CD25(Tac)-CD3 chimera (TT). This chimera results from appending the cytoplasmic tail of CD3 to the extracellular and transmembrane domains of CD25. We preferred to use these cell transfectants instead of the scFv(V␤3)-TCR␤ transfectant, because the TT chimera does not associate with the TCR complex (28). On the contrary, it is still possible that a single TCR complex includes two different TCR␤s (24,25,35). We stimulated clone 4 (a Jurkat clone expressing low levels of TT relative to CD3) with an anti-CD3 antibody to study the direct modulation of the TCR complex and the co-modulation of the TT chimera. In parallel, clone 65 (a Jurkat clone expressing high levels of TT) was stimulated with an anti-Tac antibody to study the direct modulation of the TT chimera and the comodulation of the TCR complex. The stimulation of clone 4 with a high dose of plate-bound anti-CD3 antibody induced both the discontinuous down-modulation of the TCR complex, the "allor-nothing" effect described in Fig. 1B, and co-modulation of the TT chimera (Fig. 3). Conversely, the stimulation of clone 65 with a high dose of anti-Tac resulted in the discontinuous all-or-nothing down-regulation of the TT chimera and the co-modulation of the TCR. The stimulation of each clone with the inverse antibody also caused the co-modulation of the bystander receptor, although, probably because of the suboptimal TT/TCR ratio, the effect was not so dramatic (data not shown).
We then examined the role of clathrin-coated pits in the down-modulation of the engaged and non-engaged receptors. We first examined whether we could co-precipitate the engaged and non-engaged receptors with clathrin. The TCR has been shown to undergo AP2-and clathrin-dependent constitutive internalization (7), and accordingly, the TCR co-precipitated with clathrin in non-stimulated cells (Fig. 4A). However, stimulation of clone 4 with a high dose of immobilized anti-CD3 for 15 and 60 min did not increase the TCR that co-precipitated with clathrin, suggesting that down-modulation of the engaged TCR takes place without an increased association with clathrin-coated pits. Interestingly, stimulation of clone 4 with anti-CD3 induced the co-precipitation of clathrin with TT, suggesting that stimulation of the TCR could promote recruitment of the bystander receptor to clathrin-coated pits.
To further investigate the dependence of clathrin for TCR down-regulation, we overexpressed a dominant negative mutant of dynamin. This GTPase is necessary to close the neck of the coated pit, and therefore expression of the dominant negative mutant inhibits receptor internalization by preventing the formation of endocytic vesicles (30). The effect of expressing dominant negative dynamin in the down-modulation of engaged and non-engaged receptors was studied by transiently transfecting clone 4 either with the K44A mutant of dynamin 1 using a bicistronic GFP-expressing vector or with wild type dynamin 1. The transfected cells were stimulated with anti-CD3, and the expression of the TCR and TT was examined by flow cytometry upon gating on the GFP-positive population. Expression of the dominant negative dynamin partly inhibited the down-modulation of both the TCR and the bystander TT compared with expression of wild type dynamin 1 (Fig. 4B) or to untransfected cells (data not shown).
In addition to its role in clathrin-mediated endocytosis, dynamin may also control the endocytosis of receptors that use a clathrin-independent pathway (36). The most specific and effective method to inhibit clathrin-mediated endocytosis involves the overexpression of a dominant negative Eps15 mutant (37), but this appeared to be toxic in Jurkat cells. We therefore used alternative methods to inhibit clathrin-dependent endocytosis such as the use of an acidic medium and depletion of intracellular potassium, which detach clathrin from the internal side of the plasma membrane (32,33). When clone 4 was stimulated with a high dose of immobilized anti-CD3, co-modulation of TT was completely inhibited by acidification of the medium (Fig. 5A), whereas down-modulation of the engaged TCR remained unaffected. Conversely, upon depletion of intracellular potassium, stimulation with anti-Tac of clone 65 resulted in the inhibition of TCR co-modulation (Fig. 5B). Again, down-modulation of the engaged receptor (TT in this experiment) was not affected, except for a slight smoothening of the all-or-nothing effect 30 min after stimulation.
Together, the co-precipitation with clathrin, the inhibition by dominant negative dynamin, and the inhibition by acidification or depletion of intracellular potassium indicate that co-modulation of the bystander receptor is clathrin-dependent. In contrast, down-modulation of the directly engaged TCR appeared to be primarily mediated by a dynamin-sensitive mechanism independent of clathrin.
Down-modulation of Directly Engaged TCRs Is Dependent on Lipid Rafts-TCR engagement results in its redistribution into cholesterol-and glycosphingolipid-rich membrane microdomains known as lipid rafts (38,39). Lipid rafts have been shown to play an important role in T cell activation, because they serve to concentrate important signaling molecules (40).
To study the role of lipid rafts on the down-modulation of directly engaged and bystander receptors, we examined the  (Fig. 6A, bottom panel). Most interestingly, M␤CD pretreatment resulted also in a dose-dependent inhibition of the directly engaged receptor (TCR). Also noteworthy, the inhibition of TCR down-modulation only occurred at concentrations of M␤CD 3-10-fold higher than those required to inhibit TT co-modulation.
The differential dose-dependent effect of M␤CD on downmodulation of the directly engaged and bystander receptors was also observed when clone 65 was stimulated with anti-Tac antibody, suggesting that the effect of cholesterol depletion on TCR and TT down-modulation was independent of which receptor is stimulated (Fig. 6B). Again, co-modulation of the bystander receptor (TCR in this case) was more sensitive to M␤CD treatment than down-modulation of the directly engaged receptor (TT). Thus, TCR co-modulation was strongly inhibited at 0.3 mM M␤CD, whereas TT down-modulation was only inhibited at a concentration 10-fold higher. Given that co-modulation of the bystander receptors is PTK-dependent ( Figs. 1 and 2 (28)) and that lipid rafts play an important role in TCR signaling (38 -40), it is likely that the high sensitivity of TCR co-modulation to M␤CD treatment is due to the inhibition of TCR signaling. On the other hand, the resistance of engaged TCR down-modulation to PTK inhibitors and lower sensitivity to M␤CD suggests that engaged TCRs are downmodulated by a lipid raft-dependent mechanism different to the one that promotes co-modulation of the bystander TCRs. DISCUSSION In this study, we have extended our previous observations showing that TCR engagement results in the down-modulation of bystander TCRs. We have found that stimulation of the TCR with high doses of immobilized anti-CD3 induces its downmodulation by a PTK-independent mechanism, whereas stim-ulation with low doses of anti-CD3 results in TCR down-modulation by a PTK-dependent mechanism. This contrasts with the effect of soluble anti-CD3, which was PTK-dependent, albeit partially, at all antibody doses analyzed. We propose that antibody engagement of the TCR results in the down-modulation of directly engaged receptors as well as bystander receptors. Furthermore, we showed here that bystander and directly engaged TCRs are down-modulated by different mechanisms. We propose that the down-modulation of the directly engaged TCR is by a PTK-independent mechanism, whereas down-modulation of the bystander TCRs requires PTK-initiated signal spreading.
The differential sensitivity to PP2 of TCR down-modulation by immobilized anti-CD3, and the partial inhibition by PP2 of TCR down-modulation induced with soluble anti-CD3, might reflect the ratios of directly engaged:bystander receptors in each situation. Thus, at low ligand doses most down-modulated TCRs will be bystander TCRs that rather than being directly engaged will have received a signal to down-modulate transmitted by PTKs. This hypothesis is however counterintuitive when the effect of saturating doses (10 -30 g/ml) of soluble anti-CD3 are studied (Fig. 1). At these concentrations of anti-CD3 one should expect that all TCRs are directly engaged. However, only part of the TCR is down-modulated by a PTKindependent mechanism. We propose the hypothesis that at high doses of soluble antibody, monomeric binding predominates (i.e. without cross-linking) thus transforming most antibody-bound TCRs into mere bystander receptors. This hypothesis goes with the idea that direct PTK-independent downmodulation requires strong cross-linking.
In addition to PTKs, we have previously shown that protein kinase C is involved in transmitting the co-modulation signal (28). This finding was recently confirmed and extended when the co-modulation of the bystander TCR was shown to be protein kinase C-dependent and to involve the phosphorylation of the S 126 DxxxLL endocytosis motif in CD3␥ (23). Additionally, we have now shown that the bystander TCR is down-modu- FIG. 3. Engagement of the TCR or a TT chimera induces the co-modulation of the reciprocal receptor. Jurkat T cells were stably transfected with the TT chimera. Two clones expressing low (clone 4) or high (clone 65) levels of the chimera were used to study the co-modulation of bystander receptors. Clone 4 was stimulated for 1 h with 10 g/ml immobilized OKT3 to analyze the direct modulation of the engaged TCR and the comodulation of the bystander TT chimera. Clone 65 was stimulated for 1 h with 10 g/ml immobilized anti-Tac (CD25) antibody MAR108 to analyze the direct modulation of the TT chimera and the comodulation of the TCR. The populations of modulated (b) and non-modulated (a) cells are indicated as in the legend to Fig. 1. lated by a clathrin-mediated mechanism. Furthermore, we found that TCR triggering induced the association of clathrin with the bystander TT chimera, suggesting that TCR signaling somehow promotes the association of the bystander receptor with the clathrin machinery.
The clathrin-associated adapter AP-2 has been shown to be recruited to the CD3␥ double leucine motif upon phosphorylation of Ser-126 (7). Therefore, the sequence of events that precedes the down-modulation of bystander TCRs seems to likely involve the activation of PTKs by the engaged TCR, resulting in protein kinase C activation and the subsequent phosphorylation of CD3␥ in the bystander TCR. Finally, AP-2 and clathrin are recruited to the phosphorylated CD3␥, and the TCR is endocytosed. Although this sequence of events might occur, it seems likely to be only part of the picture. We detected the recruitment of clathrin to the TT chimera that is down-modulated in a protein kinase C-dependent fashion. However, the TT chimera does not bear protein kinase C phosphorylation motifs; therefore, additional mechanisms must mediate the recruitment of the clathrin endocytic machinery to motifs in the TCR other than the double leucine motif in CD3␥. For example, these could include the tyrosine-based motifs present in various copies in the immunoreceptor tyrosine-based activation motifs of the cytosolic region. In addition, it has been described recently that T cell stimulation with soluble anti-CD3 antibody induces the Lck-dependent tyrosine phosphorylation of the clathrin heavy chain (41), suggesting a possible mechanism for the transmission of the internalization signal to bystander TCRs. In this regard, the high sensitivity of TCR co-modulation to inhibition by M␤CD is in agreement with the proposed role for lipid rafts, Lck is associated with lipid rafts, in TCR signaling (42,43). FIG. 4. TCR engagement induces the association of clathrin to bystander TT chimera. A, clone 4 was stimulated on Petri dishes coated with 10 g/ml OKT3 for the times indicated. Cells were collected, lysed, and immunoprecipitated (IP) with anti-CD3 (OKT3) or anti-Tac (MAR108) antibodies. After SDS-PAGE, immunoblotting was performed with an anti-clathrin heavy chain antibody and then reprobed with anti-Tac antibody TP1/6. Arrows indicate the positions of clathrin heavy chain and of TT as a loading control. B, down-modulation of the engaged and bystander receptors is inhibited by a dominant negative dynamin-1 construct. Clone 4 cells were transiently transfected either with the K44A mutant of dynamin 1 in a GFP-expressing vector (Dyn DN) or with wild type dynamin 1 in the same vector (Dyn WT). After stimulation in wells coated with 10 g/ml OKT3 for the times indicated, TCR (direct modulation) and TT (comodulation) down-modulation was assessed by flow cytometry after gating the GFP ϩ cells.
Although we have made significant advances in understanding the mechanisms that mediate the co-modulation of the bystander TCR, our knowledge of the mechanism that mediates down-modulation of the directly engaged TCR is more limited. Down-modulation of the engaged TCR is not affected by inhibitors of PTK or protein kinase C or by treatments that inhibit clathrin-dependent endocytosis. Down-modulation of the engaged TCR is ATP-and temperature-dependent, it requires dynamin to close the endocytic vesicle and results in a characteristic all-or-nothing effect. The all-or-nothing effect may suggest the need for a critical number of engaged TCRs to accumulate before producing endocytosis. This accumulation FIG. 5. Co-modulation of the bystander receptor is mediated by a clathrin-dependent mechanism. A, comodulation of the bystander receptor but not of the engaged receptor is inhibited by treatment with acidified medium. Clone 4 cells were incubated in wells coated with 10 g/ml OKT3 in acidic or in neutral medium (control) for the times indicated. Down-modulation of the directly engaged (TCR) and bystander receptor (TT) was assessed by flow cytometry. B, co-modulation of the bystander receptor but not of the engaged receptor is inhibited by potassium depletion. Clone 65 cells were subjected to a protocol for depletion of intracellular potassium or left untreated (control) and incubated for the indicated times on wells coated with 10 g/ml anti-Tac antibody. Down-modulation of the engaged (TT) and co-modulated (TCR) receptors was assessed by flow cytometry.
could be the result of their translocation to lipid rafts. Indeed, we showed in this study that treating cells with a cholesteroldepleting agent, M␤CD, can prevent TCR down-modulation. This result strongly suggests that TCR down-modulation involves both clathrin-dependent and lipid raft-dependent mechanisms, as has been proposed for the IL-2 receptor (36). However, our data also indicate that the mechanism for down-modulation of engaged TCRs is PTK-independent. We therefore propose that TCR engagement results in its internalization and down-modulation by a process that involves TCR accumulation in lipid rafts without the participation of signaling components.
Is the PTK-dependent and -independent mechanism model obtained for antibody engagement relevant to MHCp-induced FIG. 6. Cholesterol depletion inhibits down-modulation of the engaged and bystander receptors. A, clone 4 cells were incubated in wells coated with 10 g/ml OKT3 for the times indicated (in minutes) in the presence or absence of 3 mM M␤CD. Cells were collected, and down-modulation of TCR and TT was examined by flow cytometry. A dose response to M␤CD is shown in the bottom of the panel. B, clone 65 cells were incubated in wells coated with 10 g/ml MAR108 (anti-Tac) for the times indicated in the presence or absence of 3 mM M␤CD. Cells were collected, and down-modulation of TT and TCR was examined by flow cytometry. A dose response to M␤CD is shown in the bottom.
TCR down-regulation? We previously demonstrated that at high concentrations of MHCp, the TCR is down-regulated by a PTK-independent mechanism, whereas TCR down-modulation was completely abrogated by a PTK inhibitor at low MHCp concentrations (28). Previous studies have generated conflicting results relating to the dependence of TCR down-modulation on PTK (15)(16)(17)(18)(19). Although this issues remains unclear, it does seem that PTKs play an important role in routing the endocytosed TCR for degradation in the lysosomes (13,14,17). Indeed, it has been proposed that TCR engagement induces TCR downmodulation by preventing recycling rather than by increasing internalization above constitutive levels (13). However, high doses of MHCp have similarly been shown to increase TCR internalization (13). Together, the data presented here show that TCR down-regulation is a much more complex process than first thought, involving both directly engaged and bystander TCRs, which are internalized through distinct endocytotic pathways.