Calnexin Association Is Not Sufficient to Protect T Cell Receptor Proteins from Rapid Degradation in CD4 CD8 Thymocytes*

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 273, No. 37, Issue of September 11, pp. 23674 23680, 1998 Printed in U.S.A. Calnexin Association Is Not Sufficient to Protect T Cell Receptor Proteins from Rapid Degradation in CD4 CD8 Thymocytes* (Received for publication, March 18, 1998, and in revised form, May 8, 1998) Michael J. Bennett, Jeroen E. M. Van Leeuwen, and Kelly P. Kearse From the Department of Microbiology & Immunology, East Carolina University, School of Medicine, Greenville, North Carolina 27858-4354 and Experimental Immunology Branch, NCI, National Institutes of Health, Bethesda, Maryland 20892-1360 During T cell development, assembly of the mutisubunit T cell receptor (TCR) complex is regulated by the differential stability of newly synthesized TCR molecules, having a half-life of approximately 20 min in immature CD4 CD8 thymocytes compared with >75 min in mature T cells. The molecular basis for TCR instability in CD4 CD8 thymocytes is unknown but has been postulated to involve abnormalities in N-glycan processing and calnexin assembly as perturbation of these pathways markedly destabilizes TCR proteins in all other T cell types examined. Here, we compared the processing of TCR glycoproteins and their assembly with calnexin and calreticulin chaperones in CD4 CD8 thymocytes and splenic T cells. These studies show that TCR glycoproteins synthesized in CD4 CD8 thymocytes were processed in a similar manner as those made in splenic T cells and that TCR proteins stably associated with calnexin in both cell types. Interestingly, however, TCR association with the calnexin-related molecule calreticulin was decreased in CD4 CD8 thymocytes compared with splenic T cells. Finally, TCR degradation in CD4 CD8 thymocytes was impaired by inhibitors of proteasome activity, which was correlated with stabilization of calnexin TCR complexes. These data demonstrate that calnexin association is not sufficient to protect TCR proteins from rapid degradation in CD4 CD8 thymocytes, suggesting that additional components of the quality control system of the endoplasmic reticulum operate to ensure the proper folding of nascent TCR glycoproteins. Coincident with their synthesis and translocation into the lumen of the endoplasmic reticulum (ER), 1 many newly synthesized polypeptides are modified by the addition of oligosaccharide chains, having the structure Glc 3 Man 9 GlcNAc 2 (Glc glucose; Man mannose; GlcNAc N-acetyl glucosamine) (1, 2). Immature N-glycan chains are initially processed by the sequential action of glucosidase I and glucosidase II enzymes (the Glc-trimming pathway), creating monoglucosylated Glc 1 - * This work was supported by a faculty development research grant from the East Carolina University School of Medicine and National Institutes of Health Grant R29 AI42104 (to K. P. K.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Current address: Cell Biology and Metabolism Branch, NICHD, National Institutes of Health, Bethesda, MD 20892-5430. To whom correspondence should be addressed. Tel.: 252-816-2703; Fax: 252-816-3104; E-mail: kearse@brody.med.ecu.edu. 1 The abbreviations used are: ER, endoplasmic reticulum; TCR, T cell receptor; Ab, antibody; mab, monoclonal Ab; TPCK, N-tosyl-L-phenylalanine chloromethyl ketone; Endo H, endoglycosidase H; PAGE, polyacrylamide gel electrophoresis; Glc, glucose; cas, castanospermine. 23674 Man 9 GlcNAc 2 glycans important for interaction with the molecular chaperones calnexin and calreticulin (3, 4). Calnexin and calreticulin are transmembrane and luminal proteins, respectively, that are proposed to function in the ER quality control system of protein folding and assembly (5 11). Assembly of most newly synthesized glycoproteins with calnexin and calreticulin occurs by a two-step process involving initial recognition of monoglucosylated glycans, followed by protein-protein interactions, which stabilize these associations (3, 5, 11). The mechanisms important for release of proteins from calnexin and calreticulin are not well understood, although for certain molecules cleavage of the final (innermost) Glc residue by glucosidase II appears to be important (11 14). Studies using glucosidase inhibitors and mutant cell lines deficient in expression of glucosidase enzymes have established a strong correlation between Glc trimming, calnexin association, and the survival of newly synthesized T cell receptor (TCR) molecules in the ER (13, 15 17). Indeed, persistence of Glc residues on N-glycans precludes formation of calnexin TCR complexes and results in the rapid degradation of TCR proteins (15, 16, 18), occurring much faster than is typically observed for TCR protein turnover under conditions where glycan processing is unperturbed. These results are specific in that the stability of related TCR glycoproteins is unaffected by impaired glucosidase activity and the failure to assemble with calnexin (15, 17 19). Thus, calnexin association appears to be uniquely important for the survival of nascent TCR glycoproteins in the ER. Detailed investigation into the production of TCR complexes in CD4 CD8 thymocytes showed that inefficient assembly of complete TCR results from limited assembly of intermediates, which, in turn, results from rapid degradation of TCR proteins (16, 20, 21). Instability of TCR proteins in CD4 CD8 thymocytes is independent of clonotypic specificities as even a matched pair of TCR, transgenes fails to associate efficiently in CD4 CD8 thymocytes (16) and is developmentally regulated as TCR proteins synthesized in immature CD4 CD8 thymocytes have a significantly shorter half-life than those made in progeny CD4 and CD8 T cells (16, 21). The molecular basis for TCR instability in CD4 CD8 thymocytes is unknown but has been postulated to involve abnormalities in N-glycan processing and calnexin assembly as perturbation of these pathways markedly destabilizes TCR proteins in other T cell types (15, 16, 18). To enhance our understanding of the molecular basis for the limited survival of newly synthesized TCR proteins in immature CD4 CD8 thymocytes, we undertook the present study to examine and compare the processing of TCR glycoproteins and their assembly with lectin-like chaperones in CD4 CD8 thymocytes and splenic T cells. This paper is available on line at http://www.jbc.org

Calnexin Association with TCR Proteins in CD4 CD8 Thymocytes 23675 EXPERIMENTAL PROCEDURES Animals, Cell Preparation, and Reagents C57BL/6 (B6) mice were obtained from the Jackson Laboratory (Bar Harbor, ME). T cell receptor V 11 AND transgenic mice (22) were generously provided by Dr. Stephen Hedrick. CD4 CD8 thymocytes were isolated by their adherence to plastic plates coated with anti-cd8 mab (83-12-5) and were typically 95% CD4 CD8 as described (16). Splenic T cells were purified by incubation of single cell suspensions of spleen cells on tissue culture plates coated with rabbit anti-mouse immunoglobulin (Ig) (Organo-Technika-Cappel, Malvern, PA) for 60 min at 37 C, followed by isolation of nonadherent cells. The resultant cell populations were typically 80 85% CD3 as determined by flow cytometry analysis. Castanospermine (Calbiochem) was used at a final concentration of 100 g/ml. Diamide (Sigma) was used at a final concentration of 1 mm; N-acetyl-L-leucyl-L-leucyl-L-norleucinal (ALLN) (Calbiochem) was used at a final concentration of 100 g/ml; N-tosyl-L-phenylalanine chloromethyl ketone (TPCK) (Sigma) was used at a final concentration of 10 g/ml; lactacystin (Calbiochem) was used at a final concentration of 20 M. Antibodies The following monoclonal antibodies (mab) were used in this study: 145-2C11, specific for CD3 proteins associated with CD3 or CD3 chains (23); H28-710, specific for TCR (24); H57-597, specific for TCR (25); and anti-v 11 Ab (Pharmingen, San Diego, CA). The following antisera were used: SPA-860 anti-calnexin (Stressgen Biotechnologies, Victoria, BC); and PA3 900 anti-calreticulin (Affinity BioReagents, Neshanic Station, NJ). Metabolic Labeling and Immunoprecipitation Metabolic pulse labeling with [ 35 S]methionine was performed as described previously (16). Briefly, cells were pulse-labeled in methionine-deficient medium containing [ 35 S]-methionine (Trans 35 S-label; ICN, Irvine, CA) for 30 min at 37 C and chased in medium containing excess nonradioactive methionine for various time periods at 37 C. Cells were lysed by solubilization in 1% digitonin (Calbiochem) lysis buffer (20 mm Tris, 150 mm NaCl, plus protease inhibitors) at 1 10 8 cells/ml for 20 min at 4 C; lysates were clarified by centrifugation to remove insoluble material and immunoprecipitated with the appropriate antibodies preabsorbed to protein A-Sepharose beads (Pharmacia, Uppsala, Sweden). Sequential immunoprecipitation and immunoprecipitation/release/recapture procedures were performed as previously detailed (16, 20). Glycosidase Digestion and Gel Electrophoresis Digestion with endoglycosidase H (Endo H) was performed by resuspending precipitates in glycosidase digestion buffer (75 mm sodium phosphate, ph 6.1, 75 mm EDTA, 0.1% Nonidet P-40) containing 10 milliunits of Endo H (Genzyme, Cambridge, MA) for 16 h at 37 C; digestion with jack bean mannosidase (Oxford Glycosystems, Rosedal, NY) was performed according to the manufacturer s instructions. For experiments examining the effects of oligosaccharide release on calnexin association with TCR proteins, lysates were precipitated with anti-calnexin Abs and precipitates either mock-treated (buffer only) or digested with Endo H as described above. Precipitates were then washed 3 times with phosphate-buffered saline, boiled in 1% SDS, and TCR proteins specifically recaptured with anti-tcr mabs as described (20). SDS-PAGE gel electrophoresis was performed as previously detailed (16). RESULTS Calnexin Assembly with Newly Synthesized Proteins in CD4 CD8 Thymocytes Initially, we examined the assembly of newly synthesized glycoproteins with calnexin in CD4 CD8 thymocytes. For these studies, CD4 CD8 thymocytes were radiolabeled with [ 35 S]methionine for 30 min, solubilized in 1% digitonin, and lysates immunoprecipitated with anti-calnexin Ab; alternatively, lysates were sequentially immunoprecipitated with anti-cd3 mab followed by anti-tcr mab. Precipitates were digested with Endo H glycosidase to identify molecules containing (immature) N-linked glycan chains (26). As shown in Fig. 1A, numerous Endo H-sensitive proteins were associated with calnexin in CD4 CD8 thymocytes (Fig. 1A), including molecules that comigrated with TCR and CD3 glycoproteins (Fig. 1A). Analysis of anti-calnexin precipitates on two-dimensional nonequilibrium ph gradient electrophoresis SDS-PAGE gels showed that numerous radiolabeled proteins assembled with calnexin in CD4 CD8 thymocytes (Fig. 1B), most of which were not associated at the conclusion of a 60-min chase period (Fig. 1B). Taken together, FIG. 1. Calnexin assembly with newly synthesized glycoproteins in CD4 CD8 thymocytes. A, CD4 CD8 thymocytes were radiolabeled with [ 35 S]methionine for 30 min, solubilized in 1% digitonin, and lyates immunoprecipitated with anti-calnexin Ab or sequentially immunoprecipitated with anti-cd3 mab, followed by anti-tcr mab. Precipitates were digested with Endo H glycosidase as indicated and analyzed on 13% SDS-PAGE gels under reducing conditions. The positions of Endo-H-sensitive, deglycosylated TCR proteins (TCR Degly) and deglycosylated CD3, CD3 proteins (CD3, Degly) are marked. B, CD4 CD8 thymocytes were radiolabeled with [ 35 S]methionine for 15 min, chased in medium containing excess nonradioactive methionine for 60 min, and digitonin lysates immunoprecipitated with anti-calnexin Ab. Precipitates were analyzed on two-dimensional nonequilibrium ph gradient electrophoresis SDS-PAGE gels under reducing conditions. The positions of TCR and CD3 proteins are marked, and calnexin is indicated by an arrow. These data are representative of four independent experiments. these data indicate that newly synthesized glycoproteins interact transiently with calnexin in CD4 CD8 thymocytes, similar to what has been reported regarding calnexin-protein associations in other cell types (5, 10 12). Glc Trimming and CD3 Assembly of TCR Glycoproteins in CD4 CD8 Thymocytes Previous studies in splenic T cells showed that Glc residues are removed from newly synthesized TCR glycoproteins before their assembly with CD3 subunits

23676 Calnexin Association with TCR Proteins in CD4 CD8 Thymocytes FIG. 2.Glc trimming and assembly of TCR glycoproteins in CD4 CD8 thymocytes. Digitonin lysates of [ 35 S]methionine-radiolabeled CD4 CD8 thymocytes from AND transgenic mice were immunoprecipitated with anti-calnexin Ab or were sequentially immunoprecipitated with anti-cd3 mab, followed by anti-v 11 mab. Precipitated material was released by boiling in SDS, TCR proteins recaptured by precipitation with anti-tcr mab, and samples digested with jack bean (JB)and Endo H (EH) glycosidases as indicated. Precipitates were analyzed on 13% SDS-PAGE gels under reducing conditions. The positions of TCR glycoforms (A and B) and Endo H-sensitive, deglycosylated TCR proteins (TCR EH S ) are marked. These data are representative of three independent experiments. and that calnexin associates exclusively with unassembled, free TCR proteins containing incompletely trimmed glycan chains (17). To determine whether TCR proteins were equivalently processed in CD4 CD8 thymocytes, similar experiments were performed; because TCR proteins are clonotypic and quite heterogeneous, as in previous studies using splenic T cells (17), CD4 CD8 thymocytes were obtained from AND transgenic mice expressing the rearranged TCR gene product, V 11 protein (22). CD4 CD8 thymocytes were pulse-labeled with [ 35 S]methionine for 30 min, solubilized in 1% digitonin, and lysates sequentially immunoprecipitated with anti-cd3 mab to isolate CD3 -associated TCR proteins, followed by anti-v 11 mab to capture unassembled TCR chains. Alternatively, lysates were precipitated with anti-calnexin Ab. Precipitates were boiled in SDS to release bound material, TCR proteins specifically recaptured by precipitation with anti- TCR Ab, and recapture precipitates digested with jack bean mannosidase and Endo H glycosidases. Jack bean mannosidase digestion is useful for evaluating the Glc-trimming status of newly synthesized glycoproteins as it removes eight mannoses from fully trimmed N-glycan chains devoid of Glc residues (Man 8 9 GlcNAc 2 ), but only five mannoses from incompletely trimmed N-glycans containing 1 3 Glc saccharides (Glc 1 3 Man 8 9 GlcNAc 2 ) (5, 17). In contrast, Endo H removes all but a single GlcNAc from N-glycan chains irrespective of their Glc content (26). Similar to what was previously observed in splenic T cells (17), two major glycoforms of V 11 proteins were present in CD4 CD8 thymocyte lysates, denoted A and B, that are indicative of TCR proteins containing incompletely trimmed and fully trimmed glycan chains, respectively (Fig. 2). As demonstrated, V 11 proteins that were assembled with CD3 subunits in CD4 CD8 thymocytes possessed fully trimmed oligosaccharide chains (Fig. 2, middle, B glycoform) unlike free, unassembled V 11 proteins, which existed as both A and B glycoforms (Fig. 2, right-hand side). Newly synthesized V 11 proteins bearing partially trimmed oligosaccharides (A glycoform) coprecipitated with calnexin in CD4 CD8 thymocyte lysates, whereas those containing fully trimmed oligosaccharides (B glycoform) did not (Fig. 2, left-hand side). Thus, in agreement with previous studies on the processing of V 11 proteins in splenic T lymphocytes (17), these data show that Glc residues were removed from TCR glycoproteins prior to their assembly with CD3 chains in CD4 CD8 thymocytes. Moreover, these data demonstrate that newly synthesized TCR molecules containing partially trimmed oligosaccharides FIG. 3.N-Glycan chains are not necessary to maintain stable association of TCR glycoproteins with calnexin in CD4 CD8 thymocytes. Digitonin lysates of [ 35 S]methionine-radiolabeled CD4 CD8 thymocytes and splenic T cells were immunoprecipitated with anti-calnexin Ab; precipitates were resuspended in Endo H digestion buffer in the presence or absence of Endo H (EH) enzyme and incubated at 37 C for 14 h. Precipitates were then washed 3 times in phosphate-buffered saline, and supernatants containing material released by Endo H digestion were removed; the remaining pellets (precipitates) were boiled in SDS and TCR proteins recaptured with anti- TCR specific mab. The positions of fully glycosylated and Endo H-sensitive, deglycosylated TCR proteins (TCR Degly) are indicated. These data are representative of three independent experiments. Immpt., immunoprecipitated. were associated with the molecular chaperone calnexin in CD4 CD8 thymocytes. Oligosaccharide Chains Are Not Required to Maintain Calnexin Association with TCR Glycoproteins in CD4 CD8 Thymocytes Interaction of calnexin with most newly synthesized glycoproteins is proposed to occur by a two-step process involving initial recognition of monoglucosylated glycans, followed by protein-protein interactions, which stabilize these associations (3, 17). Recent studies show, however, that for certain glycoproteins, polypeptide contacts do not appear necessary for stable calnexin assembly (27, 28). We reasoned therefore that TCR proteins might precariously assemble with calnexin in CD4 CD8 thymocytes via predominantly carbohydrate binding, such that TCR polypeptide domains were particularly vulnerable to proteolytic degradation mechanisms. To investigate this issue, digitonin lysates of radiolabeled CD4 CD8 thymocytes and splenic T cells were immunoprecipitated with anti-calnexin Ab; precipitates were either mock-treated or digested with Endo H, and association of TCR proteins with calnexin before and after deglycosylation was compared. Calnexin interaction with TCR proteins was unaffected by removal of oligosaccharides in CD4 CD8 thymocytes and splenic T cells as similar amounts of TCR molecules were associated with calnexin in mock and Endo H digested groups of both cell types (Fig. 3). These data demonstrate that TCR proteins are stably associated with calnexin in CD4 CD8 thymocytes and show that similar to splenic T cells, once formed, calnexin TCR protein complexes do not require oligosaccharide chains to maintain their association in CD4 CD8 thymocytes. Assembly of Newly Synthesized TCR Proteins with Calnexin and Calreticulin in CD4 CD8 Thymocytes and Splenic T Cells To investigate the quantitative assembly of newly synthesized TCR proteins with lectin-like chaperones in CD4 CD8 thymocytes and splenic T lymphocytes, cells were metabolically labeled with [ 35 S]methionine for 30 min and digitonin lysates exhaustively immunoprecipitated with anti-calnexin Ab, anti-calreticulin Ab, and anti-tcr Abs; sufficient Ab was used in these experiments to capture 95% of the respective antigens (data not shown). CD4 CD8 thymocyte and splenic T lysates were immunoprecipitated with anti- TCR Ab, anti-calnexin Ab, or anti-calreticulin Ab; precipitates were boiled in SDS to release bound material and TCR pro-

Calnexin Association with TCR Proteins in CD4 CD8 Thymocytes 23677 FIG. 4. Assembly of newly synthesized TCR proteins with lectin-like chaperones in CD4 CD8 thymocytes and splenic T cells. A, digitonin lysates of [ 35 S]methionine-radiolabeled CD4 CD8 thymocytes and splenic T cells were immunoprecipitated with anti- TCR mab, anti-calnexin (Cx) Ab, or anti-calreticulin (Crt) Ab. Precipitates were boiled in SDS and TCR proteins recaptured by precipitation with anti-tcr mab. The position of TCR proteins is indicated. B, digitonin lysates of [ 35 S]methionine-radiolabeled CD4 CD8 thymocytes and splenic T cells were sequentially immunoprecipitated (Immpt.) with two rounds of anti-calnexin Ab, followed by anti-tcr mab to capture remaining TCR proteins; alternatively, lysates were precipitated with anti-tcr mab to isolate total TCR molecules. Precipitates were boiled in SDS, and TCR proteins were specifically recaptured as described in panel A). The position of TCR proteins is indicated. These data are representative of four independent experiments. teins specifically recaptured using anti-tcr mab. As demonstrated, TCR assembly with calnexin was quantiatively similar in CD4 CD8 thymocytes and splenic T cells (Fig. 4A). These results were confirmed in sequential immunoprecipitation experiments in which lysates were precipitated with anticalnexin Ab, followed by anti-tcr mab to capture remaining TCR molecules (Fig. 4B). In agreement with previous studies on calreticulin-tcr association in other T cell types (13), much fewer TCR proteins coprecipitated with calreticulin relative to calnexin in splenic T cells (Fig. 4A), presumably because of the transient nature of calreticulin-tcr interactions compared with calnexin-tcr interactions (13). Suprisingly, few, if any TCR proteins were present in anti-calreticulin precipitates of CD4 CD8 thymocyte lysates (Figs. 4A), even upon prolonged exposure of autoradiographs; calreticulin expression was comparable in both cell types (data not shown). Taken together, these data show that calnexin association with newly synthesized TCR proteins is quantitatively similar in CD4 CD8 thymocytes and splenic T cells. In addition, these data contained the surprising finding that calreticulin association with TCR proteins was specifically decreased in CD4 CD8 thymocytes compared with splenic T cells. Involvement of Proteasome Activity in TCR Degradation in CD4 CD8 Thymocytes Recent studies in various nonlymphoid cell types expressing transfected TCR gene products and murine BW thymoma cells have implicated proteasomes in the degradation of TCR proteins (29 32). Because TCR proteins synthesized in CD4 CD8 thymocytes are significantly less stable than TCR proteins expressed in other cell types, including T lymphocytes (15, 16, 18), we wished to determine the involvement of proteasome activity in TCR degradation in CD4 CD8 thymocytes. For these studies we utilized ALLN, an inhibitor of various proteases including the proteasome and lactacystin, a highly specific proteasome inhibitor (32, 33). In addition, we examined the effects of TPCK, a cysteine protease inhibitor (29) and diamide, a thiol-oxidizing agent previously reported to retard TCR degradation in nonlymphoid cells (29) and to both induce and inhibit dissociation of calnexin with newly synthesized proteins (11, 33). In agreement with previous studies on TCR degradation in other cell types (29 32), the stability of unassembled TCR proteins in CD4 CD8 thymocytes was significantly enhanced by these reagents (Fig. 5, A and B). As shown in Fig. 6, inhibition of TCR degradation in CD4 CD8 thymocytes was correlated with increased stabilization of calnexin-tcr protein complexes (Fig. 6). These data document that proteasome activity is important for the degradation of unassembled TCR proteins in CD4 CD8 thymocytes and show that impairment of proteasome activity results in protracted association of TCR proteins with calnexin. Failure to Remove Glc Residues from N-glycan Chains Does Not Preclude Assembly or Intracellular Transport of TCR Proteins in CD4 CD8 Thymocytes Finally, we wished to examine the effects of the persistence of Glc residues on N- glycans on the assembly and degradation of TCR proteins in CD4 CD8 thymocytes. For these studies, CD4 CD8 thymocytes were radiolabeled in the presence or absence of the glucosidase inhibitor castanospermine (cas), which blocks removal of Glc residues from N-glycan chains (34). As expected, TCR glycoproteins synthesized in cas-treated CD4 CD8 thymocytes migrated with decreased mobility compared with those made in untreated CD4 CD8 thymocytes (Fig. 7A), because of persistence of Glc residues on N-glycan chains. Most notably, these results demonstrate that TCR proteins effectively assembled with CD3 components in cas-treated CD4 CD8 thymocytes and that the stability of unassembled TCR proteins was similar in media- and cas-treated CD4 CD8 thymocytes (Fig. 7A); these findings were confirmed by immunoprecipitation/release/recapture experiments (Fig. 7B). These studies show that TCR assembly with CD3 proteins is not precluded by the failure to remove Glc residues from N-glycans on TCR proteins and that degradation of unassembled TCR proteins is not influenced by glucosidase activity in CD4 CD8 thymocytes. chain addition is the terminal stage of TCR assembly and depends on prior formation of TCR intermediates (20). Because TCR proteins are inherently unstable in CD4 CD8 thymocytes, few, if any, TCR intermediates or unassembled TCR proteins exist within the ER that are available for assembly with newly translated proteins (16, 20). Accordingly, the assembly of radiolabeled chains with CD3 proteins approximates the rate of formation of complete TCR complexes in CD4 CD8 thymocytes (16, 20). As shown in Fig. 7A, similar amounts of CD3-assembled proteins were present in mediaand cas-treated CD4 CD8 thymocytes. Thus, we further conclude that persistence of Glc residues on TCR proteins does not significantly affect the formation of complete TCR complexes in CD4 CD8 thymocytes. Glc-containing high mannose oligosaccharides that escape the ER can be processed by endomannosidase enzymes localized in the intermediate ER/cis-Golgi compartment that cleave

23678 Calnexin Association with TCR Proteins in CD4 CD8 Thymocytes FIG. 6.Inhibition of TCR degradation is associated with increased stabilization of calnexin TCR protein complexes in CD4 CD8 thymocytes. CD4 CD8 thymocytes were radiolabeled with [ 35 S]methionine for 30 min, and chased in medium containing excess nonradioactive methionine and diamide (DIAM) (1 mm), TPCK (10 g/ml), or ALLN (100 g/ml) as indicated. Digitonin lysates were immunoprecipitated with anti-calnexin Ab; precipitates were boiled in SDS, and TCR proteins were recaptured by precipitation with anti- TCR mab. The position of TCR proteins is indicated. These data are representative of three independent experiments. thymocytes, radiolabeled cells were chased for 150 min. As shown in Fig. 7A, CD3-associated TCR proteins in cas-treated CD4 CD8 thymocytes acquired Endo H resistance during longer chase times (Fig. 7A), indicative of their transport from the ER to at least the media Golgi compartment (2, 26). Consistent with these findings, Endo H-resistant CD3, glycoproteins were also visible in anti-cd3 precipitates of chase groups of cas-treated CD4 CD8 thymocytes (Fig. 7A, see * in figure legend). Note that glucosidase activity was effectively inhibited throughout the course of these experiments as unassembled TCR proteins in cas-treated CD4 CD8 thymocytes showed increased mobility relative to media-treated CD4 CD8 thymocytes during the entire chase period (data not shown). These data demonstrate that TCR complexes formed under conditions of cas-induced glucosidase blockade in CD4 CD8 thymocytes effectively transit to at least the medial Golgi compartment. FIG. 5. Degradation of TCR proteins in CD4 CD8 thymocytes is impaired by inhibitors of proteasome activity. A, CD4 CD8 thymocytes were radiolabeled with [ 35 S]methionine for 30 min, and chased in medium (MED) containing excess nonradioactive methionine and diamide (DIAM)(1mM), TPCK (10 g/ml) or lactacystin (20 M) as indicated. Digitonin lysates were sequentially immunoprecipitated with CD3 mab, followed by anti-tcr mab; precipitates were boiled in SDS, and TCR proteins were recaptured by precipitation with anti-tcr mab. The position of TCR proteins is indicated. C, CD4 CD8 thymocytes were cultured for 30 min at 37 C in medium, ALLN (100 g/ml), or lactacystin (20 M), and radiolabeled with [ 35 S]methionine for 30 min and chased in medium containing excess nonradioactive methionine for 30 min. Digitonin lysates were sequentially immunoprecipitated with CD3 mab, followed by anti-tcr mab; precipitates were boiled in SDS, and TCR proteins were recaptured by precipitation with anti-tcr mab. These data are representative of three independent experiments. Immpt., immunoprecipitated; c, chased; p, pulsed;, minutes. 1 3 Glc saccharides in association with a single mannose residue (2, 34), creating high mannose intermediates that are identical to those generated via conventional processing routes (2); unlike ER glucosidase activity, endomannosidase action is insensitive to castanospermine treatment (35). To determine whether CD3-assembled TCR proteins were ultimately transported from the ER to the Golgi in cas-treated CD4 CD8 DISCUSSION The current study has examined and compared the processing of TCR proteins and their association with the lectin-like chaperones calnexin and calreticulin in immature CD4 CD8 thymocytes and mature splenic T cells. Our data demonstrate that TCR glycoproteins are processed in a similar manner in both cell types, with Glc residues being removed from TCR proteins before their assembly with CD3 components in CD4 CD8 thymocytes, as previously reported for splenic T cells. TCR molecules were stably associated with calnexin in both cell types, but interestingly, calreticulin TCR associations were decreased in CD4 CD8 thymocytes relative to splenic T cells. Finally, these studies show that TCR proteins effectively assemble with CD3 subunits and egress from the ER to the Golgi complex under conditions where Glc residues persist on N-glycans and demonstrate that proteasome activity is important for TCR degradation in CD4 CD8 thymocytes. Because perturbation of Glc-trimming and calnexin assembly pathways markedly destabilizes TCR proteins in other T cell types (15 17), it was hypothesized that inefficient Glc removal from N-glycan chains and impaired calnexin assembly might represent the molecular basis for rapid degradation of TCR proteins in CD4 CD8 thymocytes (15, 16, 18, 19). The data in the current report effectively rule out this postulate, however, as these studies show that TCR proteins synthesized in CD4 CD8 thymocytes are effective substrates for Glc-trimming enzymes and associate with calnexin in a manner that is qualitatively and quantitatively similar to that which is observed in splenic T cells. Importantly, these studies

Calnexin Association with TCR Proteins in CD4 CD8 Thymocytes 23679 FIG. 7.Persistence of Glc residues on N-glycan chains does not preclude assembly and intracellular transport of TCR proteins in CD4 CD8 thymocytes. A, CD4 CD8 thymocytes were cultured in medium (MED) or castanospermine (cas) (100 g/ml final concentration) for 30 min, radiolabeled with [ 35 S]methionine for 30 min and chased for 150 min in medium containing excess nonradioactive methionine. The presence of inhibitor was maintained throughout the entire experiment. Digitonin lysates were immunoprecipitated with CD3 mab and digested with Endo H glycosidase as indicated; precipitates were analyzed on 13% SDS-PAGE gels under reducing conditions. The positions of TCR molecules before and after deglycosylation are marked (, Degly,, Degly). * indicates the position of Endo H-resistant forms of CD3 and CD3 glycoproteins; CD3 contains three N-glycan chains, only one of which is processed to complex, Endo H-resistant type oligosaccharides by Golgi maturation enzymes; therefore, mature CD3 chains migrate coincident with Endo H-resistant CD3 proteins. B, CD4 CD8 thymocytes were cultured in medium or castanospermine (CAS) (100 g/ml) for 30 min, radiolabeled with [ 35 S]methionine for 30 min, and chased for 40 min in medium containing excess nonradioactive methionine. The presence of inhibitor was maintained throughout the entire experiment. Digitonin lysates were sequentially immunoprecipitated with CD3 mab, followed by anti-tcr mab; precipitates were boiled in SDS, and TCR proteins were recaptured by precipitation with anti-tcr mab. The position of TCR proteins is indicated. These data are representative of three independent experiments. establish that rapid degradation of TCR proteins is not invariably correlated with the failure to associate with calnexin, suggesting that additional components in the ER quality control system (calreticulin?) operate to ensure the proper folding of nascent TCR glycoproteins. Consistent with this concept, recent studies by Hebert et al. (12) demonstrate that calnexin and calreticulin interact with distinct regions on newly synthesized influenza hemagglutinin molecules and associate with distinct hemagglutinin-folding intermediates. Calnexin and calreticulin have also been shown to associate with major histocompatibility complex class I proteins at different stages of their maturation and assembly (14, 36). Thus, it is reasonable to speculate that calnexin and calreticulin may perform specialized functions in the folding and assembly of TCR proteins and that interaction with both chaperones is requisite for TCR stability. Alternatively, it is possible that calreticulin association may intrinsically confer stability on TCR proteins, an interaction that was shown to be specifically decreased in CD4 CD8 thymocytes in this report. Data from previous studies on TCR instability are concordant with all of these ideas as TCR interaction with both calnexin and calreticulin chaperones is precluded under conditions of impaired glucosidase activity (15, 16, 19). It remains to be determined, however, why calreticulin association with newly synthesized TCR proteins is specifically decreased in immature CD4 CD8 thymocytes. Because TCR proteins are inherently unstable in CD4 CD8 thymocytes and their survival is unaffected by the persistence of Glc residues on N-glycan chains, we were able to significantly extend previous results on N-glycan processing and TCR assembly in other T cell types (15, 16, 18) that were confounded by the differential stability of TCR proteins under these conditions. Indeed, the data in the current report document that TCR molecules effectively assemble with CD3 proteins and are transported to the Golgi complex when glucosidase activity is inhibited in CD4 CD8 thymocytes. These results were hitherto unrealized because blockade of Glc removal from N- glycan chains dramatically destabilizes TCR proteins in other T cell types (15, 16, 18), making it difficult to determine the efficacy of TCR assembly under these conditions. The current data suggests, therefore, that although persistence of Glc residues on N-glycans severely limits the survival of newly synthesized TCR proteins, TCR molecules that survive under these conditions can effectively assemble into TCR complexes that are capable of ER to Golgi transport. Recent data from numerous laboratories indicate that misfolded proteins localized in the ER are degraded by cytoplasmic proteasomes, previously implicated only in the disposal of cytosolic proteins (37). Consistent with this idea, studies in nonlymphoid cells expressing a transfected TCR gene product and murine BW thymoma cells suggest a model in which unassembled TCR proteins are dislocated from the ER and degraded by cytosolic proteasomes (30 32). In the current study we extend these findings to primary lymphocytes by showing that TCR degradation in CD4 CD8 thymocytes was markedly impaired by inhibitors of proteasome activity. Interestingly, inclusion of proteasome inhibitors stabilized calnexin TCR complexes in CD4 CD8 thymocytes, suggesting that TCR proteins may be degraded when associated with calnexin, or alternatively, that feedback mechanisms exist that regulate the dislocation of TCR proteins in relation to proteasome activity (30). Calnexin association is not requisite for disposal of unasssembled TCR proteins in CD4 CD8 thymocytes, however, as TCR degradation was similar in the presence or absence of the glucosidase inhibitor, castanospermine, which effectively blocks the initial assembly of TCR proteins with calnexin (Refs. 13 and 15; and data not shown). Finally, it is noteworthy to mention that instability of TCR proteins and their interaction with lectin-like chaperones in CD4 CD8 thymocytes shows several interesting parallels to recent observations on the biosynthesis of tyrosinase molecules in normal versus malignant melanocytes (38). Analogous to the differential survival of nascent TCR proteins in immature and mature T cells, the stability of tyrosinase proteins is significantly decreased in melanoma cells compared with normal melanocytes. Moreover, tyrosinase stably associates with both calnexin and calreticulin in normal melanocytes but solely with calnexin in malignant melanocytes (38). Thus, the differential stability of TCR proteins in CD4 CD8 thymocytes and splenic T cells represents the second example where rapid degradation of a newly synthesized protein is correlated with impaired calreticulin association but not calnexin association. Importantly, we would also note that similar to what has been reported regarding wild type versus mutant proteins that con-

23680 Calnexin Association with TCR Proteins in CD4 CD8 Thymocytes tain defined singular amino acid changes, such as cystic fibrosis transmembrane conductor regulator molecules and 1-antitrypsin (39 42), calnexin assembly with newly synthesized TCR proteins in immature and mature T cells does not distinguish between the dramatically different fates of TCR proteins in these two cell types. Thus, undefined mechanisms must exist that regulate the assembly of TCR proteins with CD3 components and TCR survival in CD4 CD8 thymocytes and splenic T cells. Further identification of quality control components necessary for the proper folding and stability of nascent TCR proteins should provide important information regarding the regulation of TCR stability in these cell types. Acknowledgments We thank Drs. Mark Mannie, Tom McConnell, and Allan Weissman for critical reading of the manuscript and Dr. Stephen Hedrick for the gift of V 11 Tg mice. REFERENCES 1. Kornfeld, R., and Kornfeld, S. (1985) Annu. Rev. Biochem. 54, 631 664 2. Moremen, K. W., Trimble, R. B., and Herscovics, A. (1994) Glycobiology 4, 113 125 3. Ware, F. E., Vassilakos, A., Peterson, P. A., Jackson, M. R., Lehrman, M. A., and Williams, D. B. (1995) J. Biol. Chem. 270, 4697 4704 4. Spiro, R. G., Zhu, Q., Bhoyroo, V., and Soling, H.-D. (1996) J. Biol. Chem. 271, 11588 11594 5. Hammond, C., Braakman, I., and Helenius, A. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 913 917 6. Peterson, J. R., Ori, A., Van, P. N., and Helenius, A. (1995) Mol. Biol. Cell 6, 1173 1184 7. Otteken, A., and Moss, B. (1996) J. Biol. Chem. 271, 97 103 8. Nauseef, W. M., McCormick, S. J., and Clark, R. A. (1995) J. Biol. Chem. 270, 4741 4747 9. Vassilakos, A., Cohen-Doyle, M. F., Peterson, P. A., Jackson, M. R., and Williams, D. B. (1996) EMBO J. 15, 1495 1506 10. Bergeron, J. J. M., Brenner, M. B., Thomas, D. Y., and Williams, D. B. (1994) Trends Biochem. Sci. 19, 124 128 11. Hebert, D. N., Foellmer, B., and Helenius, A. (1995) Cell 81, 425 433 12. Hebert, D. N., Zhang, J. X., Chen, W., Foellmer, B., and Helenius, A. (1997) J. Cell Biol. 139, 613 623 13. van Leeuwen, J. E. M., and Kearse, K. P. (1996) J. Biol. Chem. 271, 25345 25349 14. van Leeuwen, J. E. M., and Kearse, K. P. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 13997 14001 15. Kearse, K. P., Williams, D. B., and Singer, A. (1994) EMBO J. 13, 3678 3686 16. Kearse, K. P., Roberts, J. L., Munitz, T. I., Wiest, D. L., Nakayama, T., and Singer, A. (1994) EMBO J. 13, 4504 4514 17. van Leeuwen, J. E. M., and Kearse, K. P. (1996) J. Biol. Chem. 271, 9660 9665 18. Kearse, K. P., Roberts, J. L, Wiest, D. L., and Singer, A. (1995) BioEssays 17, 1049 1054 19. van Leeuwen, J. E. M., and Kearse, K. P. (1997) J. Biol. Chem. 272, 4179 4186 20. Kearse, K. P., Roberts, J. L., and Singer, A. (1995) Immunity 2, 391 399 21. Kearse, K. P., Takahama, Y., Punt, J. A., Sharrow, S. O., and Singer, A. (1995) J. Exp. Med. 181, 193 202 22. Kaye J., Hsu, M. L., Sauron M. E., Jameson S. C., Gascoigne, N. R., and Hedrick, S. M. (1989) Nature 341, 746 749 23. Leo, O., Foo, M., Sachs, D. H., Samelson, L. E., and Bluestone, J. A. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 1374 1378 24. Becker, J. L. B., Near, R., Mudgett-Hunter, M., Margolies, M. M., Kubo, R. T., Kaye, J., and Hedrick, S. M. (1989) Cell 58, 911 921 25. Kubo, R. T., Born, J. W., Kappler, J., Marrack, P., and Pigeon, M. (1989) J. Immunol. 142, 2736 2742 26. Tarentino, A. L., and Maley, F. (1974) J. Biol. Chem. 249, 811 816 27. Rodan, A. R., Simmons, J. E., Trombetta, E. S., and Helenius, A. (1996) EMBO J. 15, 6921 6930 28. Zapun, A., Petrescu. S. M., Rudd, P. M., Dwek, R. A., Thomas, D. Y., and Bergeron, J. J. M. (1997) Cell 88, 29 38 29. Inoue, S., and Simoni, R. D. (1992) J. Biol. Chem. 267, 9080 9086 30. Yu, H., Kaung, G., Kobayashi, S., and Kopito, R. R. (1997) J. Biol. Chem. 272, 20800 20804 31. Huppa, J. B., and Ploegh, H. L. (1997) Immunity 7, 113 122 32. Yang, M., Omura, S., Bonifacino, J. S., and Weissman, A. (1998) J. Exp. Med. 187, 835 836 33. Wada, I., Ou, W. J., Liu, M. C., and Scheele, G. (1994) J. Biol. Chem. 269, 7464 7472 34. Elbein, A. D. (1991) FASEB J. 5, 3055 3063 35. Moore, S. E. H., and Spiro, R. G. (1992) J. Biol. Chem. 267, 8443 8451 36. Sadasivan, B., Lehner, P. J., Ortmann, B., Spies, T., and Cresswell, P. (1996) Immunity 5, 103 114 37. Kopito, R. (1997) Cell 88, 427 430 38. Halaban, R., Cheng, E., Zhang, Y., Moellmann, G., Hanlon, D., Michalak, M., Setaluri, V., and Hebert, D. N. (1987) Proc. Natl. Acad. Sci. U. S. A. 94, 6210 6215 39. Pind, S., Riordan, J. R., and Williams, D. B. (1994) J. Biol. Chem. 269, 12784 12788 40. Thomas, P. J., Qu, B. H., and Pedersen, P. L. (1995)Trends Biol. Sci. 20, 456 459 41. Le, A., Steiner, J. L., Ferrell, G. A., Shaker, J. C., and Sifers, R. N. (1994) J. Biol. Chem. 269, 7514 7519 42. Liu, Y., Choudhury, P., Cabral, C. M., and Sifers, R. N. (1997) J. Biol. Chem. 272, 7946 7951