A role for the cytoplasmic tail of the pre-T cell receptor (TCR) alpha chain in promoting constitutive internalization and degradation of the pre-TCR.

Engagement of the alpha beta T cell receptor (TCR) by its ligand results in the down-modulation of TCR cell surface expression, which is thought to be a central event in T cell activation. On the other hand, pre-TCR signaling is a key process in alpha beta T cell development, which appears to proceed in a constitutive and ligand-independent manner. Here, comparative analyses on the dynamics of pre-TCR and TCR cell surface expression show that unligated pre-TCR complexes expressed on human pre-T cells behave as engaged TCR complexes, i.e. they are rapidly internalized and degraded in lysosomes and proteasomes but do not recycle back to the cell surface. Thus, pre-TCR down-regulation takes place constitutively without the need for extracellular ligation. By using TCR alpha/p Tau alpha chain chimeras, we demonstrate that prevention of recycling and induction of degradation are unique pre-TCR properties conferred by the cytoplasmic domain of the pT alpha chain. Finally, we show that pre-TCR internalization is a protein kinase C-independent process that involves the combination of src kinase-dependent and -independent pathways. These data suggest that constitutive pre-TCR down-modulation regulates pre-TCR surface expression levels and hence the extent of ligand-independent signaling through the pre-TCR.

During intrathymic development of ␣␤ T cells, thymocytes that have a successful rearrangement at the T cell receptor ␤ (TCR␤) 1 locus express a pre-TCR complex composed of the TCR␤ chain paired with the invariant pre-TCR␣ (pT␣) chain and associated with CD3 components (1)(2)(3). Surface expression of this pre-TCR (4) triggers the selection, expansion, and further differentiation of developing pre-T cells in a ligand-inde-pendent manner (5,6), finally resulting in the induction of rearrangements at the TCR␣ locus. Upon productive TCR␣ gene rearrangements, the TCR␣ chain pairs with TCR␤ and associates with CD3⑀, -␥, -␦, and -chains, and thymocytes undergo a second step of selection upon binding of the TCR␣␤ to self-peptide-major histocompatibility complex molecules (1)(2)(3). Despite experimental evidence on the similar biochemical compositions of the pre-TCR and TCR in terms of their associated CD3 subunits (7)(8)(9), current studies support the theory that mechanisms regulating the assembly and intracellular transport of these complexes may differ markedly, because the pre-TCR is expressed only transiently and very inefficiently during thymocyte development, at levels about 50 -100-fold lower than those of the TCR on mature T cells (3,10). By using TCR␣-pT␣ chain chimeras, we have shown recently (9) that limited expression of the human pre-TCR is pT␣ chain-dependent. Particularly, the pT␣ cytoplasmic (Cy) domain was found to serve an endoplasmic reticulum retention function that could contribute in part to the regulation of pre-TCR assembly and surface expression (9).
However, the level of expression of a cell surface receptor is the result of an equilibrium between the synthesis and transport of new polypeptides and their internalization, recycling, and degradation (reviewed in Ref. 11). Extracellular stimuli induce changes in one or several of these processes and therefore modify the level of expression of a given receptor. In the particular case of the TCR, receptor engagement by its natural ligands results in down-regulation of TCR-CD3 cell surface expression (11)(12)(13)(14), which is a critical event intimately associated with TCR signaling and T-cell activation (14). Different molecular mechanisms have been proposed to account for the down-modulation of ligated TCR complexes. Most studies support the position that TCR ligation results in a significant increase of the TCR internalization rate followed by the degradation of the internalized complex (15,16). However, TCR-CD3 complexes are continuously internalized and recycled back to the cell surface in nonstimulated T cells (17,18), suggesting that constitutive TCR recycling on resting T cells has to be affected by TCR ligation. In this regard, Wiest and co-workers (19) have recently proposed that TCR engagement has little effect on the TCR internalization rate; rather, it prevents TCR recycling back to the cell surface by inducing the intracellular retention of ligated complexes and their degradation by lysosomes and proteasomes.
In sharp contrast to the TCR, pre-TCR signaling occurs apparently without any need for ligation (5,6). This concurs with the finding that the pre-TCR spontaneously clusters and localizes on the cell surface into lipid rafts (20,21), in a manner similar to that found for the ␣␤ TCR following ligand binding (22,23). These findings support the view that surface pre-TCR complexes are constitutively activated without any need for ligation (20) and raise the question of what is the intracellular fate of such "activated" pre-TCR complexes. In this study, we have analyzed the dynamics of pre-TCR-CD3 cell surface expression and down-modulation in unstimulated pre-T cells and show that, similar to ligated TCR complexes, surface pre-TCR complexes are continually and rapidly endocytosed and degraded in the absence of extracellular ligation but do not recycle back to the cell surface. Moreover, we show that cell-autonomous pre-TCR down-modulation depends on the pT␣ cytoplasmic tail. The possibility that constitutive endocytosis and degradation of the pre-TCR is a self-safe mechanism responsible for its limited expression on the cell surface is discussed.

EXPERIMENTAL PROCEDURES
Cell Lines and Transfectants-The ␣wt and ␣/CypT␣ stable transfectants were derived as described elsewhere (9) from the pre-T cell line SUP-T1, which expresses an endogenous TCR␤ (V␤1) chain and the pT␣ chain but lacks TCR␣. Briefly, SUP-T1 cells were transfected, respectively, either with full-length cDNAs encoding a conventional TCR␣ (V␣12.1) chain or with a TCR␣/pT␣ chimeric construct in which the Cy domain of TCR␣ was replaced by the equivalent domain of pT␣ (CypT␣). G418-selected transfectants were grown in RPMI 1640 (Bio-Whittaker) supplemented with 10% fetal calf serum (Invitrogen). Likewise, TCR-GFP stable transfectants were derived from SUP-T1 cells transfected with a plasmid encoding a COOH-terminal fusion protein of the TCR chain with the green fluorescence protein (GFP). The TCR-GFP fusion was performed by PCR amplification of a complete TCR chain cDNA (kindly provided by Dr. B. Alarcón, Centro de Biología Molecular Severo Ochoa, Madrid, Spain) with the sense 5Ј-CGC GCG CCC GGG ATG AAG TGG AAG GCG CTT-3Ј and antisense 5Ј-GCG CGC CCG GGC CCC GCG AGG GGG CAG GGC-3Ј primers followed by digestion and ligation into the SmaI site of the pEGFP-N1 plasmid vector (Clontech). Pre-TCR surface expression levels analyzed on 48 selected TCR-GFP clones were similar to those on parental SUP-T1 cells.
Surface Biotinylation, Immunoprecipitation, and Immunoblotting-Cells (10 7 /ml) either untreated or pretreated for 2 h with 50 M chloroquine, 20 M lactacystin, or 1 M epoxomicin (Calbiochem) were washed three times in complete 1ϫ PBS at 4°C and labeled with 0.5 mg/ml Sulfo-NHS-biotin (Pierce) in complete 1ϫ PBS for 30 min on ice. Excess biotin was eliminated by washing the cells three times with 10 mM L-lysine (Sigma) in 1ϫ PBS. Then, cells were cultured (5-10 ϫ 10 6 /ml) in the absence or presence of the indicated lysosome or proteasome inhibitors. At each time point, cells (20 ϫ 10 6 /immunoprecipitation) were washed with 1ϫ PBS, lysed with 1% Brij, and immunoprecipitated as described above with the UCHT-1 or SP-34 (kindly provided by Dr. B. Alarcón) anti-CD3⑀ mAb. The immunoprecipitates were resolved by 12% SDS-PAGE under nonreducing conditions, and blots were probed with streptavidin-horseradish peroxidase (Pierce) and revealed by the ECL method. When indicated, blots were stripped and reprobed with the 448 anti-TCR antiserum or an anti-␣-tubulin mAb (Sigma) as above.
Confocal Microscopy and Colocalization Analysis-Confocal microscopy was performed essentially as described previously (9). Cells were adhered to poly-L-lysine (Sigma)-precoated coverslips, fixed with 4% paraformaldehyde in PBS for 15 min at 4°C, and stained at the cell surface with an anti-CD4 mAb (BD Biosciences). For intracellular staining, fixed cells were permeabilized with a commercial lysing solution (BD Biosciences) and stained with the anti-CD63 mAb TEA3/18 (VI International Leukocyte Typing Workshop), kindly provided by Dr. F. Sá nchez-Madrid (Hospital de la Princesa, Madrid, Spain). Rhodamine Red X-coupled goat anti-mouse Igs (Molecular Probes) was used as the second-step reagent for both surface and cytoplasmic staining. Samples were examined under a confocal microscope (Radiance 2000, Bio-Rad Laboratories) coupled to an Axiovert S100TV inverted microscope (Zeiss). Serial optical sections were recorded at 0.3-0.5 m intervals with a 63ϫ lens under an optimal iris setup. Colocalization analyses were performed using Metamorph, version 5.03, software (Universal Imaging).

Surface Pre-TCR-CD3 Complexes Are Continually Internalized but Do Not Recycle Back to the Cell Surface-
To study the dynamics of pre-TCR cell surface expression, we used a human pre-T cell line, SUP-T1, which has been shown to display low pre-TCR surface levels as found on primary human pre-T cells (Ref. 9 and Fig. 1A). The relative contribution of synthesis and secretion of new chains as compared with internalization and recycling of expressed ones to actual pre-TCR-CD3 surface levels was explored by comparative cycloheximide or brefeldin-A (BFA) treatment, as described recently for the mature TCR-CD3 complex (19). Pre-TCR surface expression was measured with an anti-TCRV␤1 mAb (9). The dynamics of TCR-CD3 surface expression was analyzed for comparison in the same cellular environment, namely SUP-T1 cells stably transfected with a TCR␣ chain (␣wt transfectants), which co-expressed the conventional TCR␣␤-CD3 complex (Ͼ99% TCR␣␤ ϩ ) together with the endogenous pre-TCR (Ref. 9, and Fig. 1A). The results confirmed that, as described (19), the mature surface TCR-CD3 was relatively synthesis-independent, that is, surface TCR-CD3 expression levels remained essentially stable within the studied 12-h time period (Fig. 1C). Therefore, mature TCR complexes are long-lived on ␣wt transfectants. As reported (19), the effect of BFA on TCR expression was very different from that of cycloheximide and resulted in a partial reduction of TCR membrane levels ( Fig. 1A), which fell rapidly during the first 2 h of treatment (about 30% reduction) and remained low (50% of control levels) for 8 h thereafter (Fig. 1B). Because surface expression of the mature TCR was found independent of newly synthesized complexes, this reduction in TCR surface levels cannot be due to the reported BFA-induced blockade of the anterograde transport from the endoplasmic reticulum to the Golgi complex (24,25). Rather, it may be caused by the documented capacity of BFA to induce tubulation and fusion of the trans-Golgi network with early endosomes, which, although previously reported to leave cycling between plasma membrane and endosomes of certain molecules such as transferrin unperturbed (24), has recently been shown to affect the endocytic transport of the TCR (19,26). Therefore, our data indicate that TCR-CD3 complexes expressed on SUP-T1 ␣wt transfectants are continually internalized and recycled back to the cell surface and thus behave as conventional TCR-CD3 complexes on resting T cells (19).
In sharp contrast to the TCR, surface expression of the pre-TCR was dramatically affected by cycloheximide, indicating that it is dependent on newly synthesized complexes. Pre-TCR levels fell to less than 20% of control levels after 6 h of treatment and remained essentially undetectable after 12 h (Fig. 1C). BFA also induced a rapid and marked decrease on pre-TCR surface expression (80% after 2 h of treatment), which resulted in the down-modulation of the complex (Fig. 1, A and  B). These data suggest that pre-TCR-CD3 complexes are continually internalized from the surface of SUP-T1 pre-T cells, with internalization rates that are apparently higher than those for the TCR-CD3. Surprisingly, however, endocytosed pre-TCR-CD3 complexes do not recycle back to the cell surface.
The pT␣ Chain Cytoplasmic Domain Is Responsible for the Impaired Recycling of the Pre-TCR-CD3 Complex-Because the pre-TCR and TCR are co-expressed on ␣wt transfectants, the differential intracellular fate of the two complexes cannot be cell type-dependent but, rather, is structure-dependent (i.e. the substitution of pT␣ in the pre-TCR with TCR␣ in the TCR). To uncover the structural properties of pT␣ that are responsible for the impaired recycling of the endocytosed pre-TCR, TCR␣/ pT␣ chimeric constructs involving distinct pT␣ and TCR␣ domains were stably transfected into SUP-T1 cells (9), and the surface expression dynamics of the resulting ChTCR was analyzed by flow cytometry. Particularly, we focused on ␣/CypT␣ chimeric constructs in which the Cy domain of TCR␣ was replaced by the equivalent domain of pT␣. As shown Fig. 1A, the effects of BFA and cycloheximide on surface expression of the resulting ␣/CypT␣-TCR␤ ChTCR, measured with an anti-TCR␣␤ mAb, were equivalent to those observed on endogenous pre-TCR expression. BFA induced a significant reduction of ChTCR expression levels after 2 h, which was followed by the complete loss of the complex after 6 h of treatment (Fig. 1, A  and B), and cycloheximide treatment also resulted in ChTCR down-modulation (Fig. 1C). These data indicate that the pT␣ Cy domain actively mediates impaired recycling of the pre-TCR-CD3 complex to the cell surface without the need of ligand binding.
TCR Chains Associated to Internalized pre-TCR and ChTCR Complexes Are Degraded Intracellularly-To investigate the mechanisms that could account for the impaired return of endocytosed pre-TCR (and ChTCR) complexes to the cell surface, we followed the intracellular fate of pre-TCRassociated TCR chains in SUP-T1 cells stably transfected with a plasmid encoding a COOH-terminal fusion protein of the TCR chain with the green fluorescence protein (-GFP). As shown in Fig. 2A for one representative -GFP ϩ clone of four, flow cytometry analysis allowed us to simultaneously measure the surface levels of the pre-TCR complex and the total cellular content of the -GFP protein. These studies revealed that, as shown in Fig. 1 for the parental SUP-T1 pre-T cell line, pre-TCR complexes expressed on -GFP transfectants are continually internalized from the cell surface and become down-modulated after treatment with cycloheximide, BFA, or both. Kinetic analysis showed that pre-TCR down-modulation paralleled a gradual decrease of the total content of -GFP when protein synthesis was blocked, so that green fluorescence became barely detectable (Ͻ20% of control expression levels) after 12 h of treatment with cycloheximide (Fig. 2B), suggesting that the -GFP chimeric chains, including those associated with the internalized pre-TCR, were degraded intracellularly. In contrast, pre-TCR down-modulation was coupled with a dramatic increase of green fluorescence in BFA-treated cells, indicating that -GFP protein chimeras had accumulated in the cytoplasm. More importantly, blocking of endosome to lysosome trafficking induced by BFA (24) counteracted the effect of cycloheximide and prevented the disappearance of GFP expression, which remained stable although partly reduced (Ն60% of control levels) after 12 h (Fig. 2, A and B). Therefore, loss of green fluorescence induced by cycloheximide could be the result of selective degradation of -GFP fusion proteins in lysosomes.
Supporting this possibility, confocal microscopy analysis showed that untreated -GFP transfectants had low levels of surface -GFP that colocalized with the membrane marker CD4 (Fig. 2C), whereas a significant amount of green fluorescence was expressed intracellularly and accumulated in the cytoplasmic structures that expressed the lysosomal marker CD63 (27). As shown in Fig. 2D a high proportion of -GFP (55 Ϯ 17%) colocalized with CD63 on different 0.2 m sections. According to flow cytometry data, -GFP expression was sensitive to BFA treatment, so that a significant increase of green fluorescence was observed in BFA-treated cells in which -GFP accumulated intracellularly and acquired a characteristic distribution in tubular structures typical of BFA-treated cells (Fig. 2E). Expectedly, no -GFP expression was detected upon cycloheximide treatment (data not shown), again supporting the possibility that pre-TCR-associated intracellular TCR was degraded. However, intracellular accumulation of -GFP was observed when cells were treated simultaneously with cycloheximide and BFA (Fig. 2F), indicating that degradation of cytoplasmic -GFP was blocked because of the impaired trafficking to lysosomes induced by BFA.
It has been shown that the primary mechanism mediating down-modulation of the TCR upon ligand binding involves targeting of endocytosed complexes for intracellular degradation, predominantly by lysosomes, but also by proteasomes (16,19). To investigate whether proteasome was also involved in degradation of endocytosed pre-TCR or ChTCR complexes, biochemical studies were performed aimed at analyzing the intracellular fate of TCR associated to internalized complexes in cells treated with drugs that block lysosome (NH 4 Cl and chloroquine) or proteasome (lactacystin) function. None of these drugs affected surface receptor expression levels or cellular viability (data not shown). However, as shown by immunoblot analysis (Fig. 3A), they induced a significant accumulation of intracellular TCR 2 dimers that were immunoprecipitated associated to the pre-TCR or the ChTCR from SUP-T1 cells or ␣/CypT␣ transfectants, respectively. In contrast, these drugs did not significantly affect the levels of TCR chain associated to the mature TCR complex in ␣wt transfectants, which is consistent with an active and selective degradation of TCR 2 dimers associated with the pre-TCR and the ChTCR but not with the conventional TCR. Accordingly, degradation of TCR 2 dimers associated to pre-TCR and ChTCR complexes was observed when protein synthesis was blocked with cycloheximide ( Fig. 3A), but TCR chain degradation was inhibited by NH 4 Cl and chloroquine, and to a lesser extent, by lactacystin (see densitometric analysis in Fig. 3B). Taken together, our data provide evidence that, as shown previously for ligated TCR-CD3 complexes, constitutive internalization of unligated pre-TCR-CD3 complexes is followed immediately by TCR chain degradation, mainly by lysosomes, but also by proteasomes, which prevents recycling to the cell surface. Moreover, they suggest that the pT␣ Cy tail is involved selectively in that process.
The pT␣ Cy Domain Is Sufficient to Divert the TCR␣-TCR␤ Heterodimer from a Recycling Pathway to Intracellular Degradation-To determine whether, as shown for TCR 2 dimers, internalized pT␣-TCR␤ heterodimers are targeted for intracellular degradation, we next analyzed biochemically the intracellular fate of biotin-labeled surface pre-TCR-CD3 complexes. Immunoblot analysis with avidin-peroxidase of anti-CD3⑀ immunoprecipitates confirmed that the pT␣-TCR␤ heterodimer was rapidly degraded in SUP-T1 pre-T cells. As shown in Fig.  4A, 40% of the biotinylated heterodimers had disappeared after 90 min, and less than 30% of the input pT␣-TCR␤ complexes remained after 4 h, as assessed by densitometric analysis (Fig.   FIG. 2. The pre-TCR-associated TCR chain localize intracellularly to lysosomes. A, Sup-T1 stable transfectants homogeneous for the expression of a TCR-GFP chimeric protein were cultured in the absence (shaded histogram) or presence of cycloheximide (CHX), BFA, or both (CHXϩBFA) for 8 h. Two-color flow cytometry analysis was performed to measure simultaneously the surface levels of pre-TCR expression with a phycoerythrin-coupled anti-CD3⑀ mAb (red fluorescence) and the total cellular content of TCR-GFP chimeric chains (green fluorescence). B, green fluorescence analyses were performed at the indicated time points in -GFP transfectants cultured in the presence of either cycloheximide, BFA, or both. Results are given relative to untreated cells. The results of confocal microscopy and colocalization analyses of -GFP transfectants, either surface-stained with an anti-CD4 mAb (C) or permeabilized and stained with a mAb against the lysosomal marker CD63 (D), are shown. Rhodamine Red X-coupled goat anti-mouse Ig was used as second step reagent. 4B). In contrast, no degradation of the mature TCR␣␤ heterodimer was observed in ␣wt transfectants, in which biotinylated complexes remained stable for 4 h. Strikingly, heterodimers composed of the TCR␤ chain bound to the chimeric ␣/CypT␣ chain behaved essentially as pT␣-TCR␤ heterodimers. Moreover, the simultaneous analysis of the ChTCR and the endogenous pre-TCR coexpressed in ␣/CypT␣ transfectants revealed that the total content of these two heterodimers decreased with identical kinetics (Fig. 4B), demonstrating that the pT␣ Cy tail is sufficient to determine the degradation fate of the pre-TCR components.
To investigate the intracellular degradation pathway followed by the pT␣-TCR␤ heterodimer, we next performed immunoblot analysis of anti-CD3⑀ immunoprecipitates from biotin-labeled SUP-T1 pre-T cells treated with drugs that affect either the lysosome (cloroquine) or proteasome (lactacystin and epoxomicin) function. Blotting with anti-␣-tubulin was used as an internal control of protein loading (Fig. 5A). These studies revealed that about 70% of the biotinylated pT␣-TCR␤ heterodimers was lost after a 6-h chase as assessed by densitometric analysis (Fig. 5B). Surprisingly, a complete inhibition of the degradation of pT␣-TCR␤ heterodimers was observed in cells treated with lactacystin, an degradation was likewise sensitive to epoxomicin, two specific and irreversible proteasome inhibitors, whereas a weak inhibitory effect was observed after a 6-h chase in the presence of the lysosome inhibitor chloroquine. Therefore, although these data can not rule out the possibility that a proportion of the pT␣-TCR␤ heterodimers is degraded in lysosomes, they are consistent with a prominent role for the proteasome in the constitutive degradation of the internalized pT␣-TCR␤ heterodimers.
Molecular Mechanisms Involved in Constitutive Pre-TCR Down-regulation-Phosphorylation of the cytoplasmic tail of CD3␥ by PKC is the mechanism responsible for constitutive TCR internalization in unstimulated T cells (28). Therefore, we analyzed whether constitutive pre-TCR internalization in unstimulated pre-T cells is also PKC-dependent. However, no effects on surface pre-TCR levels were observed after treatment of SUP-T1 cells with doses of phorbol 12-myristate 13-acetate, which induced 60% down-regulation of the mature TCR in ␣wt transfectants. Neither was ChTCR down-modulation induced by phorbol 12-myristate 13-acetate in ␣/CypT␣ transfectants even after 2 h of treatment (Fig. 6A). These data, together with the fact that bisindolylmaleimide and Ro-31-7549, two specific PKC inhibitors, did not affect pre-TCR surface expression (data not shown), suggest that PKC is not involved in the constitutive internalization of the pre-TCR.
An intriguing possibility is that protein tyrosine kinases (PTK) such as Lck, which might play a key role in cell-autonomous signaling through the pre-TCR (20) could participate in its down-regulation, as shown for engaged TCR-CD3 complexes (29). To address this possibility, we analyzed the effects of the src family (Lck/Fyn) PTK inhibitor PP2 on surface pre-TCR expression. As shown in Fig. 6B, inhibition of src kinases resulted in increased expression levels of the pre-TCR and the ChTCR in SUP-T1 and ␣/CypT␣ transfectants, respectively. In contrast, PP2 treatment had no effect on the expression levels of the mature TCR in Jurkat T cells (data not shown), although some increase in TCR expression was observed in ␣wt transfectants. To assess whether the increase in surface pre-TCR levels upon Lck/Fyn inhibition could represent a blockade in both pre-TCR internalization and degradation, we next analyzed the kinetics of pre-TCR down-regulation in cycloheximide-treated SUP-T1 pre-T cells, with or without PP2. As shown in Fig. 6C, inhibition of src kinases consistently delayed, but could not block, the constitutive internalization and downregulation of the pre-TCR. Therefore, mechanisms involving phosphorylation by src kinases are partially, but not fully, responsible for constitutive pre-TCR down-regulation. DISCUSSION In this study, comparative analyses on the dynamics of human pre-TCR and TCR cell surface expression and downmodulation revealed striking differences in the behavior and intracellular fate of unligated TCR and pre-TCR complexes. We have shown that TCR-CD3 complexes expressed on SUP-T1 cells, upon transfection with TCR␣, are constitutively internalized and recycled back to the cell surface in the absence of ligand binding, and thus behave as conventional unligated TCR complexes on resting mature T cells (17)(18)(19). In contrast, pre-TCR complexes expressed on unstimulated SUP-T1 pre-T cells are continually and rapidly endocytosed but do not recycle back to the cell surface. As reported for TCR complexes internalized following antigenic stimulation (16,19), we show here that intracellular degradation is the mechanism responsible for the impaired recycling of pre-TCR in unstimulated pre-T cells. Strikingly, we found that chain dimers associated to the internalized pre-TCR are sorted for degradation in lysosomes and proteasomes and thus follow the intracellular fate of TCR-chain complexes internalized following antigenic stimulation (16,19), whereas evidence is provided that the proteasome plays a prominent role in the constitutive degradation of the internalized pT␣-TCR␤ heterodimers. In this regard, it is worth noting that, as observed upon TCR ligation, internalized chains associated with the unligated pre-TCR are found mostly in a phosphorylated state, a characteristic event associated with T cell activation (30). Therefore, the human pre-TCR complex behaves constitutively as an activated TCR without any need for ligand binding. After submission of this manuscript, Panigada et al. (31) provided evidence of constitutive pre-TCR internalization and degradation in the mouse. Our results extend the peculiar behavior of the pre-TCR to the human, and map it to the cytoplasmic tail of the pT␣ chain by comparison with TCR␣, which does not share this capacity.
Ligand-independent activation of the pre-TCR has recently been proposed to result from its constitutive co-localization in membrane rafts with signaling molecules, such as Lck (20), which may trigger cell-autonomous activation of proximal signaling including CD3⑀ and Zap70 phosphorylation in a manner similar to that observed for the ligated ␣␤ TCR (22,23). Such a particular membrane distribution might depend on the unique biochemical structure of the pre-TCR. Particularly, it was proposed that palmitoylation of the conserved juxtamembraneous cysteine residue of the pT␣ Cy domain might be required for the cell-autonomous raft localization of the pre-TCR (20). However, a very recent study has proved that this is not an essential component for endowing the murine pre-TCR with cellautonomous signaling capability (32). Although this finding might support the current view that the Cy domain of the murine pT␣ molecule is dispensable for pre-TCR function, the same study provides direct evidence that the COOH-terminal proline-rich domain of the murine pT␣ Cy tail plays a crucial role in pre-TCR signaling and T cell development (32). This finding supports our proposal that the Cy tail of the human pT␣ molecule is an essential component of pre-TCR function (9) and concurs with the present finding that the pT␣ Cy tail is sufficient to confer constitutive internalization and degradation properties to the conventional TCR. It is thus likely that the same mechanisms involved in ligand-induced TCR signaling and down-regulation could control the pre-TCR-CD3 intracellular fate. Accordingly, phosphorylation of CD3␥ by PKC, which is currently believed to control the internalization and recycling of unligated TCR complexes but not ligand-induced TCR down-modulation (11,28), does not seem to play a role in pre-TCR down-modulation. Regarding the potential role of PTKs involved in TCR signaling such as Lck and Fyn in TCR down-regulation, the available data support the view that down-regulation of engaged TCR complexes involve both PTKdependent and -independent mechanisms, which are most likely controlled by the concentration of ligand and final receptor occupancy (11,13,30). Because pre-TCR internalization was found to be only partly dependent on src kinase activity, the situation may be equivalent to that reported for the TCR at maximal receptor occupancy (13).
Alternatively, it can be proposed that constitutive internalization and degradation of the pre-TCR depends on unique endocytosis and/or degradation motifs. In this regard, the CD3 and TCR components shared by pre-TCR and TCR display internalization/sorting motifs of both the dileucine-and the tyrosine-based types, which could mediate clathrin-dependent internalization and intracellular sorting to degradation (reviewed in Ref. 11). Interestingly, a consensus tyrosine-based motif ( . . . 226 YPTC 229 . . . ) exists within the Cy domain of the human pT␣ molecule as well, which could become cell-autonomously exposed in the activated pre-TCR conformation to fulfill both the internalization and degradation functions. It is also possible that association of the pre-TCR into lipid rafts could regulate a constitutive clathrin-independent endocytic pathway similar to that recently described for the interleukin-2 receptor (33). Whatever the mechanism involved, we would suggest that constitutive internalization and degradation of the pre-TCR is a key process that controls surface receptor levels and provides the cell with a self-safe mechanism to avoid sustained ligand-independent signaling through a potent, potentially oncogenic, cell growth receptor (34). FIG. 6. Down-modulation of surface pre-TCR and ChTCR is independent of PKC but partly dependent on src kinase activity. A, SUP-T1 cells, ␣wt, and ␣/CypT␣ transfectants were cultured in the absence or presence of phorbol 12-myristate 13-acetate (PMA) and analyzed by flow cytometry for receptor surface expression at the indicated time points as described in the legend for Fig. 1. B, cells were cultured overnight in the absence (shaded histogram) or presence of PP2 (thick line) and then analyzed for receptor surface expression as described in A. Background values (thin line) were determined by staining with isotype-matched irrelevant Abs. C, SUP-T1 cells and ␣wt transfectants were cultured in the presence of cycloheximide (CHX) with or without PP2 and analyzed at the indicated time points for receptor surface expression as described in A. The percentage of surface expression was determined from the mean fluorescence values of treated cells using the untreated controls as reference. Results are representative of three independent experiments.