Receptor Engagement Transiently Diverts the T Cell Receptor Heterodimer from a Constitutive Degradation Pathway*

In the absence of ligand, the T cell receptor (TCR)/CD3 complex is continuously internalized and recycled to the cell surface, whereas receptor engagement results in its down-regulation. The present study shows that the TCR and CD3 components follow different fates accompanying their constitutive internalization. Although the CD3 moiety is recycled to the cell surface, the TCR heterodimer is degraded and replaced by newly synthesized chains. Since the TCR heterodimer cannot reach the cell membrane on its own, we propose a model in which recycling CD3 is transported along a retrograde pathway to the endoplasmic reticulum, where it associates with newly made TCR. Interestingly, engagement of the TCR·CD3 complex by superantigen resulted not only in the down-regulation of the TCR and CD3 components but also caused a transient stabilization of the TCR heterodimer. This suggests that TCR engagement diverts the TCR heterodimer from a degradation to a recycling pathway. Contrary to CD3, the intracellular fate of the TCR heterodimer is thus regulated, providing a mechanism for rapidly replacing nonfunctional TCR during intrathymic development of T cells.

T cells respond to antigen via a polypeptide complex composed of the ligand binding T cell receptor (TCR) 1 ␣ and ␤ chains (or ␥ and ␦ in ␥␦ T cells) and the CD3 subunits CD3␥, CD3␦, CD3⑀, and CD3 (1,2). Unlike the TCR chains, the CD3 components have long cytoplasmic tails that are responsible for the association with cytoplasmic-signaling proteins. These association properties are mediated, at least in part, by a double tyrosine and leucine motif present in a single copy in the CD3␥, -␦, and -⑀ chains and in three copies in CD3 (3). This immune receptor tyrosine-based activation motif (ITAM) becomes tyrosine-phosphorylated during T cell activation (reviewed in Refs. 4 -6). Cross-linking of the TCR⅐CD3 complex with anti-TCR or anti-CD3 antibodies or its interaction with peptide/ MHC and superantigens results in the rapid down-regulation of the complex.
Receptor down-regulation is a common phenomenon shared with other membrane receptors with intrinsic or associated tyrosine kinase activity (7). It is thought that the internalization of signal transducing receptors could have a dual effect. First, it may contribute to signal transduction by favoring the encounter with intracellular signaling molecules (8 -10). Second, it may contribute to terminating cellular responses by reducing the number of receptors at the cell surface or by uncoupling receptors from membrane-anchored signaling molecules (11). This latter effect is supported by the observation that TCR⅐CD3 down-regulation results in a loss of cellular sensitivity to subsequent stimulation (12)(13)(14)(15), and vice versa, the inhibition of receptor down-regulation by expression of a dominant negative Rab5 mutant leads to enhanced signaling (16). On the other hand, it is believed that TCR⅐CD3 internalization plays an important role during T cell activation by allowing serial triggering of multiple TCR⅐CD3 complexes by a few antigen-MHC complexes (17).
Using a neuraminidase-protection assay, it was shown that the TCR⅐CD3 complex is constitutively internalized and recycled back to the cell surface (18). Both antibody-mediated TCR⅐CD3 cross-linking and phorbol ester treatment increase internalization (18,19). Whereas phorbol ester treatment only results in increased internalization, ligand-induced stimulation results in increased internalization and degradation (17, 18, 20 -22). Not all TCR⅐CD3 chains seem to have the same fate; it has thus been shown that the CD3 chain has a more rapid turnover than the other TCR⅐CD3 subunits (23). In addition, it has been shown that the TCR heterodimer dissociates from the CD3 complex after stimulation of splenic T cells with anti-CD3 antibodies and is thereafter degraded (24).
In this study, we analyzed the fate of the TCR heterodimer and the CD3 complex after superantigen-mediated stimulation. We found that in nonstimulated cells, the TCR heterodimer is constitutively degraded in lysosomes, whereas the CD3 complex remains stable. In superantigen-stimulated cells, the degradation of the TCR heterodimer is transiently halted first, resulting in increased recycling to the cell surface, to become accelerated later. The degraded TCR is continuously replaced by newly synthesized receptors that then associate to recycling CD3. Since the TCR heterodimer cannot reach the cell membrane on its own, we propose a model in which recycling CD3 is retrotransported to the endoplasmic reticulum to associate with newly made TCR. JOVI.3 (IgG 2a ) and anti-C␤1 TCR monoclonal antibody JOVI.1 (IgG2a) were generously provided by Dr. M. Owen (Imperial Cancer Research Fund, London, UK). The hamster monoclonal anti-murine TCR-␤ antibody H57-597 was a gift from Dr. Ralph Kubo (National Jewish Center, Denver, CO). Fluorescein-labeled anti-mouse IgG 1 was from Southern Biotechnology, Inc. (Birmingham, AL).
Flow Cytometry Analysis of TCR⅐CD3 Down-regulation-CH7C17 cells (10 6 cells/ml) were incubated for different time periods at 37°C in the presence of the various stimuli. Incubations were performed in RPMI medium supplemented with 10 mM HEPES, pH 7.4. The cells were then washed twice with PBS, and cell surface expression of TCR⅐CD3 was assessed by flow cytometry. Cells were stained for 1 h at 4°C using saturating concentrations of the appropriate murine monoclonal antibodies, washed twice in cold PBS, and incubated for one additional hour with fluorescein-conjugated goat anti-mouse Ig. Cells were washed twice with PBS and analyzed in a EPICS-XL flow cytometer (Coulter Immunology, Hialeah, FL). The mean fluorescence intensity was obtained from the recorded data, and the results are expressed as the percentage of the fluorescence intensity of unstimulated cells.
Surface Labeling, Immunoprecipitation, and SDS-PAGE-Briefly, 150 ϫ 10 6 cells were washed twice with PBS and resuspended in 300 l of PBS. The cells were then 125 I-labeled using the lactoperoxidase method as described previously (26). After terminating the reaction, the cells were resuspended in RPMI medium supplemented with 10 mM HEPES, pH 7.4, and divided into 1-ml aliquots. Stimulation was performed in a 37°C water bath for the indicated periods of time by adding enterotoxin B (10 g/ml), leaving one sample at 4°C as control. In some experiments the cells were lysed at this time point in 1% Brij 96, and immunoprecipitation was performed as described previously (26). The samples were then resuspended in nonreducing sample buffer and subjected to SDS-PAGE in 13% polyacrylamide gels. In some experiments, to isolate only the molecules present on the surface at each time point, the cells were preincubated with anti-TCR and anti-CD3 antibodies for 1 h at 4°C, lysed, immunoprecipitated by the addition of protein A/G-Sepharose, and the experiment was finished as above.
Metabolic Labeling, Surface Biotinylation, and Western Blotting-A total of 150 ϫ 10 6 cells were washed twice with PBS and incubated for 1 h in Dulbecco's modified Eagle's medium without cysteine and methionine. [ 35 S]Methionine (0.5 mCi/ml/15 ϫ 10 6 cells) was then added for 45 min at 37°C. After labeling, the cells were washed twice with PBS and incubated at 4°C with sulfo-NHS-biotin (Pierce) at 0.5 mg/ml/10 ϫ 10 6 cells in PBS containing 0.1 mM CaCl 2 and 1 mM MgCl 2 . The cells were then washed, resuspended in RPMI medium supplemented with 10 mM HEPES, pH 7.4, divided into 1-ml aliquots, and stimulated with SEB at 37°C for the indicated times. The cells were then lysed, and immunoprecipitation was performed as above. After the last wash in immunoprecipitation buffer, the immunoprecipitated proteins were eluted by incubation with 20 mM triethylamine, pH 11.7, for 15 min, then neutralized by the addition of Tris-HCl, pH 8.0, to 100 mM. The neutralized samples were diluted 2-fold in 1% Brij 96 immunoprecipitation buffer and were finally precipitated with streptavidin-agarose beads (Pierce).The precipitated proteins were subjected to 13% SDS-PAGE gels and transferred to a nitrocellulose membrane (Bio-Rad) by standard procedures. The membrane was blocked in a solution of 10% nonfat dry milk in PBS for 1 h at room temperature. The membrane was rinsed three times with 100 ml of 0.1% Tween 20 in PBS and incubated with streptavidin-horseradish peroxidase conjugate (Amersham Pharmacia Biotech) in PBS-Tween for 1 h. After 5 washes with 100 ml of PBS-Tween, the membrane was processed by the enhanced chemiluminescence (ECL) method according to the manufacturer's specifications (Amersham Pharmacia Biotech). Protein bands were quantitated directly from the film in a Computing Densitometer model 300 A Molecular Dynamics.

Receptor Engagement Results in a Transient Stabilization of the TCR Moiety of the TCR⅐CD3 Complex-
To determine the fate of the TCR and CD3 components of the TCR⅐CD3 complex after ligand engagement, the V␤3 ϩ Jurkat cell transfectant CH7C17 was surface-labeled with 125 I and incubated at 37°C in the presence or absence of 10 g/ml SEB. Cells were lysed in 1% Brij96, and immunoprecipitation was performed with the anti-CD3 antibody OKT3. In the absence of stimulus, the amount of 125 Ilabeled TCR coprecipitated with the anti-CD3 antibody decreased with incubation time, whereas labeled CD3 was stable (Fig. 1A). Densitometric quantitation of the TCR and CD3␦ϩ⑀ bands at each time point showed a gradual decline in the TCR⅐CD3 ratio in the absence of ligation (Fig. 1B). In contrast, superantigen stimulation resulted in a stabilization of the amount of 125 I-labeled TCR associated to CD3 at 15 and 30 min after stimulation (Fig. 1, panels A and B). This stabilization was transient, as stimulation for 60 min resulted in a decrease of the TCR⅐CD3 ratio to unstimulated levels. Although the experiment shown in Fig. 1 was performed with soluble SEB, similar results were obtained when the superantigen was presented by antigenpresenting cells (data not shown).
These results suggest that at 37°C the TCR dissociates from the CD3 complex and that superantigen stimulation causes a transient stabilization of the association. To determine the validity of these observations using a different cell system as well as different labeling and stimulation procedures, resting murine spleen cells were surface-labeled with biotin at 4°C and stimulated at 37°C with 10 g/ml soluble anti-TCR-␤ antibody H57-597. After biotinylation, an aliquot of the cells was maintained at 4°C for 1 h, whereas all other samples were incubated at 37°C for the same period. The anti-TCR-␤ stimulus was provided during the last 15 min, during the last 30 min, or for the whole hour of incubation. The cells were lysed in 1% Brij96, and immunoprecipitation was performed with H57-597. As in Fig. 1, the incubation for 1 h at 37°C in the absence of stimulus resulted in a decrease in the amount of labeled TCR heterodimer recovered as compared with the 4°C control ( Fig.  2A). TCR engagement with anti-TCR-␤ antibody resulted in stabilization of the biotinylated TCR heterodimer. This was especially clear when the stimulating antibody was added during the last 30 min of incubation. The TCR heterodimer is constitutively degraded in resting spleen T cells, whereas TCR⅐CD3 engagement results in its transient stabilization ( Fig. 2A). Compared with the TCR heterodimer, the coprecipitated CD3␥, CD3␦, and CD3⑀ chains were more stable and did not increase upon stimulation with the anti-TCR-␤ antibody. The relative increase of labeled TCR versus CD3 upon TCR engagement is best seen when the TCR⅐CD3 ratios are calculated from densitometric scans (Fig. 2B). Unlike the experiment shown in Fig. 1 in which the labeling method used (surface radioiodination) does not allow visualization of CD3, this chain was clearly detected under the conditions used in the experiment shown in Fig. 2. Interestingly, the amount of biotinylated CD3 also decreased with incubation at 37°C.
TCR Engagement Transiently Increases TCR Recycling and Down-regulation-These results nevertheless do not indicate whether the stabilization of the TCR obtained upon ligand-induced stimulation takes place at the cell surface or inside the cell. To follow the fate of the surface-expressed TCR heterodimer, CH7C17 cells were 125 I-labeled and incubated for 1 h at 37°C in the absence or presence of SEB. SEB was added during the last 15 min, the last 30 min, or for the whole hour of incubation. After the 37°C incubation, cells were cooled and incubated for 1 h at 4°C with OKT3 or with the anti-TCR antibody JOVI.1 at saturating concentrations. Cells were lysed in 1% Brij96, and the surface receptors were immunoprecipitated with protein A-or protein G-Sepharose beads. These immunoprecipitates corresponded to proteins on the cell surface at the time of lysis. The supernatants from these immunoprecipitations were subsequently incubated with protein A-or protein G-Sepharose beads previously coated with JOVI.1 or OKT3 to immunoprecipitate internal proteins. A scheme of the procedure is shown in Fig. 3A. The direct immunoprecipitation of surface-expressed 125 I-labeled TCR with JOVI.1 showed an increase of the amounts of labeled TCR in SEB-stimulated samples (Fig. 3B, 15, 30, and 60 min) compared with the unstimulated sample (time 0). In contrast, the amount of coprecipitated CD3␦ϩ⑀ was comparable in all cases. A similar result was obtained when the iodinated cells were bound to OKT3. Superantigen stimulation resulted in a transient increase in the amount of 125 I-labeled TCR associated to CD3, whereas the amount of radiolabeled CD3 was constant (Fig. 3B).
Once surface TCR and CD3 components were immunoprecipitated, the second immunoprecipitation with the anti-TCR and anti-CD3 antibodies should have isolated radiolabeled complexes that had been internalized and were not exposed on the cell surface. The second immunoprecipitation of the unstimulated sample with JOVI.1 showed three protein bands with a relative molecular mass of 42-50 kDa (Fig. 3C) that were not present in the first immunoprecipitate (Fig. 3B). These proteins may correspond to partial degradation forms of the TCR heterodimer that are found inside the cell. Whereas the intact TCR heterodimer (90 kDa) was not found in the unstimulated sample, it was clearly detected in the three stimulated samples (15,30, and 60 min) with a maximum in the sample stimulated for 30 min. The second immunoprecipitation with the anti-CD3 antibody showed equivalent levels of CD3 chains in all cases, whereas the amount of coprecipitated TCR heterodimer was higher in the sample stimulated for 30 min. The intermediate TCR degradation products were not coprecipitated with the anti-CD3 antibody, indicating that they are not associated to the CD3 complex. The experiments shown in Fig. 3 suggest that the TCR heterodimer in CH7C17 cells is constitutively internalized and degraded at 37°C, whereas the CD3 components are spared. Stimulation with superantigen transiently increases the amount of labeled TCR heterodimer on the cell surface. To test whether this effect could be due to inhibition of TCR⅐CD3 internalization and downregulation, CH7C17 cells were stimulated with anti-TCR anti- All stimulated samples were incubated for 60 min at 37°C, of which the last 15 min, the last 30 min, or the total 60 min were in the presence of the stimulus. An aliquot of the biotinylated cells was maintained at 4°C for 1 h. After incubation, cells were lysed in 1% Brij96, and immunoprecipitation was performed with H57-597. SDS-PAGE was performed under nonreducing conditions, and biotinylated proteins were visualized after immunoblotting with streptavidin peroxidase. The positions of the TCR heterodimer and CD3 chains are indicated. As a loading control, the membrane was stripped and immunoblotted with anti-CD3to detect the total amount of this protein (surface and internal) coprecipitated with the anti-TCR antibody. B, densitometric quantitation of protein bands shown in A. The TCR:CD3 ratios were obtained as in Fig. 1B. The TCR is constitutively degraded in resting splenic T cells, whereas the CD3 components, except CD3-, are spared. TCR engagement by an anti-TCR-␤ antibody transiently rescues the TCR heterodimer from degradation. body or with superantigen, and the level of TCR and CD3 expression was analyzed by flow cytometry. Both stimuli caused an equivalent decline in surface TCR and CD3 levels, which reached 45% for the superantigen stimulus and 80% for the anti-TCR stimulus (Fig. 4). According to these data, the stimulation of the TCR⅐CD3 complex potentiated TCR down-regulation. These results are in agreement with those of Alcover and co-workers (21), who showed that SEB-induced down-regulation of the TCR⅐CD3 complex in Jurkat cells resulted from increased internalization. Thus, whereas the net effect of antibody-or superantigen-mediated stimulation was down-regulation of the TCR heterodimer (Fig. 4), the amount of preactivation-labeled TCR increased on the cell surface (Fig. 3B). Both sets of data can be reconciled by postulating that stimulation of the TCR⅐CD3 complex is accompanied by increased internalization of the TCR and CD3 components but also by increased recycling of the TCR heterodimer to the cell surface, thus diverting the TCR from its constitutive degradation pathway. Degradation of the TCR heterodimer in unstimulated cells was prevented in the presence of the lysosomotropic drug chloroquine (data not shown), suggesting that the TCR is degraded in lysosomes.
Recycling CD3 Subunits Assemble with Newly Synthesized TCR Heterodimer-It is puzzling how the stoichiometry of the TCR⅐CD3 on the cell surface is maintained constant if the TCR heterodimer is constitutively degraded, whereas the CD3 complex is stable. One possibility is that once dissociated from the internalized TCR heterodimer, recycling CD3 complex could associate with newly synthesized TCR before the complex is reexported to the cell surface. To demonstrate this point, a double labeling experiment was performed. To trace the TCR heterodimer in the exocytic route, CH7C17 cells were first metabolically labeled with 1-h incubation. After incubation, cells were prebound with the indicated antibody at 4°C and then lysed. Immunoprecipitation was performed by adding protein A/G-Sepharose to the lysates. The immunoprecipitates were analyzed in 12% SDS-PAGE under nonreducing conditions. The positions of the TCR⅐CD3 chains are indicated by arrows. C, the supernatants of the immunoprecipitates (ip) in Fig. 3B were reprecipitated with specific anti-TCR and anti-CD3 antibodies to visualize internalized proteins. The 42-50-kDa protein bands observed in the immunoprecipitations with JOVI.1 probably correspond to degradation products of the TCR heterodimer. Superantigen stimulation transiently increases the amount of prelabeled TCR heterodimer expressed at the cell surface and prevents its constitutive degradation. then cooled and surface-biotinylated at 4°C. After double labeling, the cells were incubated at 37°C for 1 h, and as in Fig.  3, the stimulation with superantigen was performed during the last 15 min, the last 30 min, or for the whole hour of incubation. After stimulation, the cells were lysed in 1% Brij96, and immunoprecipitation was performed with the anti-TCR antibody. These immunoprecipitates should consist of a mixture of internal 35 S-labeled and surface-biotinylated TCR⅐CD3 complexes. To isolate the surface complexes, the immunoprecipitates were treated at high pH to elute the bound material, and the supernatant was neutralized and precipitated with streptavidin-Sepharose beads. The precipitates were subjected to SDS-PAGE and blotted onto a nitrocellulose membrane. A double image of the membrane was obtained after exposing an x-ray film to show the 35 S-labeled proteins and after incubation with streptavidin-peroxidase to show the biotinylated proteins. A scheme of the procedure is shown in Fig. 5A. Compared with the cells maintained at 4°C, the unstimulated sample incubated for 1 h at 37°C had a 2-fold lower amount of biotinylated TCR; the stimulation with superantigen (Fig. 5B, left panel, 15 and 30 min) transiently stabilized the heterodimer to become completely degraded in 1-h-stimulated cells. In contrast, the amount of TCR-associated-biotinylated CD3 chains remained constant. Unlike the biotin-labeled TCR, the amount of 35 Slabeled TCR increased 2.9-fold during the incubation at 37°C without stimulus compared with the cells maintained at 4°C. The export of the TCR⅐CD3 complex from the ER calculated for Jurkat cells is slow, with a t1 ⁄2 of 45 min (27). During the 45 min of metabolic labeling, therefore, only a minor fraction of the TCR⅐CD3 complex had time to reach the surface. It was thus possible that the increase of 35 S-labeled TCR in the sample incubated for 1 h at 37°C without stimulus could be due to increased time, which allowed the labeled complexes to exit the ER and reach the surface. It should nevertheless be borne in mind that all the 35 S-labeled TCR shown in Fig. 5B was precipitated with streptavidin-agarose. Therefore, either the 35 Slabeled TCR molecules were also directly labeled with biotin or they were associated to biotinylated proteins. Since the increase in 35 S-labeled TCR during the incubation at 37°C without stimulus was concomitant with a decrease in the amount of biotinylated TCR, it can be concluded that newly synthesized TCR becomes associated with CD3 components that are or have been at the cell surface. Of note, although biotinylated CD3␦ϩ⑀ chains were clearly coprecipitated with the anti-TCR antibody, 35 S-labeled CD3 was not detected. Indeed, very little 35 S-labeled CD3␦ϩ⑀ was detected after direct immunoprecipitation with anti-CD3 antibody (not shown). These data reinforce the idea that the turnover of the TCR heterodimer at the cell surface is much higher than that of the CD3 chains and that internalized and degraded TCR has to be replaced by newly made TCR. DISCUSSION In this study, we used surface radioiodination and immunoprecipitation to trace the fate of TCR and CD3 components of the TCR⅐CD3 complex in unstimulated and superantigen-or anti-TCR-␤ antibody-stimulated cells. We found that both components have different fates in Jurkat and in murine spleen T cells; whereas the CD3 complex remains stable, the TCR heterodimer is constitutively degraded. On the other hand, TCR⅐CD3 engagement has a dual effect. It promotes the downregulation of the TCR and CD3 components and simultaneously increases the amount of prelabeled TCR on the cell surface. TCR stabilization was transient, and longer incubations with superantigen or anti-TCR-␤ antibody resulted in degradation of the complex. Both phenomena, TCR heterodimer down-regulation and stabilization of prelabeled TCR levels at the cell surface, can be explained if superantigen or antibody stimulation simultaneously caused an increased internalization rate of the TCR⅐CD3 complex, as has previously FIG. 5. Recycling CD3 proteins associate with newly synthesized TCR heterodimer. A, scheme of the experimental procedure. B, the cell line CH7C17 was first metabolically labeled with [ 35 S]methionine, then surface-biotinylated at 4°C. The cells were subsequently stimulated for the indicated time periods with SEB, lysed, and immunoprecipitated with JOVI.1. After elution at high pH, the biotinylated proteins were recovered by precipitation with streptavidin-agarose beads and separated on a 12% polyacrylamide gel under nonreducing conditions. The proteins were transferred to a nitrocellulose membrane that was first exposed to x-ray film to show 35 S-labeled proteins and later incubated with streptavidin peroxidase to show biotinylated proteins. Although the amount of biotinylated TCR heterodimer decreases in unstimulated cells with the incubation at 37°C, the amount of [ 35 S]methionine-labeled TCR associated to surface proteins increases. This suggests that newly synthesized TCR assembles with biotinylated CD3 complex that is or has been at the cell surface. been shown (21), and an increase in the recycling versus degradative pathways for the TCR. Although the mechanisms responsible for the latter effect are not yet understood, it is quite likely that the signaling pathways initiated upon TCR⅐CD3 cross-linking influence protein sorting at the endosomal/lysosomal interface. In this regard, phosphatidylinositol 3-kinase, which is activated by the TCR⅐CD3 complex, has been implicated in transport from endosomes to lysosomes (reviewed in Refs. 7 and 28). Furthermore, it has been shown that the expression of a constitutively active form of the Src-family kinase Lck causes the down-regulation of the TCR⅐CD3 complex and the degradation of the TCR heterodimer and the CD3 chain in lysosomes (29). Thus, it is quite likely that this kinase is responsible for the down-regulation and enhanced degradation of the TCR heterodimer that we have observed after long (60 min) stimulations. On the other hand, it has been shown that phorbol esters stimulate the internalization and recycling, rather than degradation, of the TCR⅐CD3 complex (18,22), suggesting that PKC elicits a recycling pathway. Thus, the effects of TCR⅐CD3 engagement that we have observed, i.e. first a transient increase of TCR recycling to the cell surface followed by increased degradation, could be the result of the sequential intervention of PKC and Lck on the endocytic machinery. Nevertheless, experiments that specifically address this issue need still to be performed.
A model for the constitutive-and stimulation-mediated fates of the TCR⅐CD3 complex is shown in Fig. 6. In unstimulated T cells, the TCR⅐CD3 complex is constitutively internalized and recycled back to the cell surface, as initially described by Krangel (18). The CD3 components are efficiently returned to the surface; however, the TCR heterodimer is mostly diverted to lysosomes, where it is degraded. The different fate of the TCR and CD3 moieties of the complex implies that they become dissociated at some point in the endocytic route. Interestingly, this effect was initially described by Kishimoto et al. (24) in cells stimulated with anti-CD3 antibodies but not in unstimulated cells.
In addition to the TCR⅐CD3 complex, the B cell receptor is also constitutively internalized, probably as a mechanism to internalize and present antigens. Interestingly, although as with all other ITAM-bearing polypeptides, Ig␣ and Ig␤ contain two tyrosines that conform to classical tyrosine-based endocytosis signals, only the N-terminal tyrosine of the ITAM of Ig␣ mediates the constitutive internalization of the B cell receptor (30). Although the signals responsible for the constitutive internalization of the TCR⅐CD3 complex have not yet been characterized, it is likely that some tyrosines within the ITAMs of CD3 are involved. Indeed, CD3␥, CD3␦ (31), and CD3⑀ (32) have been shown to contain endocytosis signals within their ITAMs. Nevertheless, other signals such as the double leucine motifs of CD3␥ and CD3␦ that mediate the PKC-mediated internalization of the TCR⅐CD3 complex (33) are also potential candidates.
To maintain constant levels of the complex, the constitutive dissociation of the intracellular TCR and CD3 endocytic routes makes necessary the existence of a compensatory mechanism that replenishes the degraded TCR with newly synthesized receptor. In this study, we have shown that newly synthesized TCR associates with surface-labeled CD3 (Fig. 5). Although newly made TCR could in principle contact the CD3 complex at the cell surface or in endosomes, it has been shown that free (non-CD3-associated) TCR heterodimer is not exported from FIG. 6. Model for the constitutive and stimulated endocytic pathways for the TCR⅐CD3 complex. A, constitutive internalization in unstimulated cells. The TCR⅐CD3 complex, which is continuously internalized in unstimulated T cells, reaches the early endosomal compartment. At some point in the endocytic route, the TCR heterodimer (black arrows) dissociates from the CD3 complex (gray arrows). Thus, although the CD3 complex is mostly recycled to the cell surface through recycling endosomes, the TCR heterodimer is mostly diverted to lysosomes, where it is degraded. The internalized and degraded TCR must be replaced by newly synthesized receptor. This newly made TCR reaches the cell surface either directly or associated to recycling CD3 complex, which will have been transported to the ER along a retrograde pathway. B, receptor engagement increases the internalization of the TCR⅐CD3 complex, causing its down-regulation. The stimulation also promotes a transient enhancement of TCR recycling at the expense of the degradative pathway. the ER in CD3 Ϫ Jurkat mutants (34 -36) nor in a human T-cell lymphotrophic virus-I-transformed T cell clone (37). One may then postulate that newly synthesized TCR⅐CD3 complex is transported to an endosomal compartment, where it associates with internalized CD3. Alternatively, internalized surface CD3 may be transported to the ER, where it enters in contact with new TCR before it is returned to the cell surface (Fig. 6). If the first possibility were correct, additional assumptions would have to be made, namely that once it reached the endosomal compartment, the recently synthesized TCR⅐CD3 complex would have to dissociate and the CD3 component be degraded before the newly made TCR becomes associated with recycling CD3. The second possibility, of recycling CD3 reaching the ER, would be simpler. Nevertheless, although this possibility is appealing, the retrograde transport from the cell surface to the ER has not yet been shown for proteins other than certain bacterial toxins (Ref. 38; reviewed in Ref. 39). Notwithstanding, the CD3 subunits contain sequences that potentially can mediate their retrograde transport to the ER. It has been shown that a peptide can be targeted to the trans-Golgi network (TGN) when it is covalently bound to TGN38, a membrane protein that cycles between the TGN and the plasma membrane. The peptide can subsequently be retrotransported to the ER depending on the presence of an ER retrieval signal (40). Several proteins, including TGN38, depend on tyrosinebased signals for their localization in the TGN (41,42). In this regard, the tyrosine-based sequences contained within the ITAMs of the CD3 subunits could mediate the transport of the CD3 complex to the TGN. In addition, CD3⑀ contains an ERretrieval sequence located in its C-terminal tail (43,44) that could promote the retrograde transport of the CD3 complex to the ER.
Previous reports describe a high turnover rate for the CD3 chain independent of the TCR⅐CD3 complex (23). In this study, we have detected the constitutive degradation of surface-expressed CD3 in spleen T cells incubated at 37°C. In most experiments we were not able to trace the fate of individual CD3 chains due to the nature of the antibodies and the labeling methods employed. Although it seems clear that the TCR, CD3 (␥, ␦, ⑀), and CD3 components are loosely bound at the cell surface and can be isolated independently from the others (23,45), the down-regulation of the TCR heterodimer paralleled that of the CD3 complex (Fig. 4), suggesting that the fates of TCR and CD3 divert after they have been internalized.
The disassembly of multisubunit receptors and the selective degradation of some subunits that follow endocytosis have been previously described for the high affinity interleukin-2 receptor (46). Whereas the interleukin-2 receptor ␣ chain is internalized and recycled to the cell surface, the ␤ and ␥ chains are degraded. The different fates of the interleukin-2 receptor subunits were explained by the need to maintain extended expression of the high affinity receptor, which contains the recycled interleukin-2 receptor ␣ chain, despite the brief induction of its gene transcription. More difficult to explain is the requirement for the T cell to maintain a constitutive degradation pathway for the TCR moiety of the TCR⅐CD3 complex while preserving CD3. It may be reminiscent of thymic development, as the TCR heterodimer also has a high turnover rate in immature thymocytes (data not shown). The high turnover of the surface-expressed TCR heterodimer would permit a positive-selecting thymocyte to rapidly replace a nonselected TCR for the products of the rearrangement of the second allele. The stabilization and recycling of the engaged TCR heterodimer to the cell membrane may also aid in replacing the nonfunctional receptor. It would be of interest to determine whether the pre-TCR is also subjected to constitutive degradation.