TY - JOUR
AU - Williams, David B.
AB - Abstract Class I histocompatibility molecules, consisting of a heavy chain, β2‐microglobulin and peptide, are assembled in the endoplasmic reticulum (ER) with the assistance of several molecular chaperones and accessory proteins. Peptide binding occurs when assembling class I molecules associate with a loading complex consisting of the transporter associated with antigen processing (TAP) peptide transporter, tapasin, ERp57 and calreticulin (CRT)/calnexin. To assess the physical organization of this complex, we generated a series of mutants in the murine H‐2Dd heavy chain and assessed their association with components of the complex. Seven mutations, clustered between amino acids 122 and 136 in the heavy chain α2 domain plus one mutation at position 222 in the α3 domain, resulted in loss of interaction with tapasin. Association with TAP was always lost simultaneously, supporting the view that tapasin acts as an obligatory bridge between class I molecules and TAP. Compared with previous studies on the HLA‐A2 molecule, some differences in points of tapasin interaction were observed. Failure of the H‐2Dd mutants to bind tapasin resulted in low cell‐surface expression and altered intracellular transport. Most mutants retained a substantial degree of peptide loading, consistent with the view that although tapasin may promote peptide binding to class I, it is not required. A surprising observation was that all mutants lacking tapasin interaction retained normal association with CRT. This contrasts with previous observations on other class I molecules and, combined with differences in tapasin interaction, suggests that the organization of the ER peptide‐loading complex can vary depending on the specific class I molecule examined. antigen presentation, calreticulin, endoplasmic reticulum, H‐2Dd, histocompatibility, tapasin, transporter associated with antigen processing Introduction The assembly of class I histocompatibility molecules within the endoplasmic reticulum (ER) is a highly coordinated process that relies on the interplay of several molecular chaperones, folding enzymes and specialized accessory factors. The process begins with the association of nascent class I heavy (H) chains with the molecular chaperone calnexin (CNX). This serves to promote H chain folding, retard degradation and prevent premature exit of H chains from the ER (1–3). CNX also promotes more efficient assembly of H chains with β2‐microglobulin (β2m) resulting in the formation of an unstable H chain–β2m heterodimer (1), which in turn becomes part of a large multisubunit complex known as the peptide‐loading complex. In human class I molecules and in some mouse class I allotypes CNX is replaced by calreticulin (CRT) as the main chaperone associated with this complex (4,5). Other components of the peptide‐loading complex include the peptide transporter associated with antigen processing (TAP), a type I transmembrane glycoprotein termed tapasin and the thiol oxidoreductase ERp57 [reviewed in (6)]. Interaction with this complex facilitates the loading of peptides into the binding groove of class I molecules prior to the export of the fully assembled protein from the ER to the cell surface. The importance of tapasin in the peptide‐loading complex was first elucidated in studies involving the human tapasin‐negative cell line LCL721.220 (4,7–10) and subsequently confirmed in tapasin‐deficient mice (11,12). In both systems, the surface expression of class I molecules is substantially reduced, but levels can be up‐regulated by the addition of class I‐binding peptides. Both intracellular and cell‐surface class I molecules are more thermolabile than in wild‐type cells, and they can be stabilized by peptide addition. Furthermore, presentation of viral antigens to cytotoxic T lymphocytes (CTL) is impaired and, in tapasin‐deficient mice, significant defects are apparent in both positive and negative selection of CD8+ T cells (11). Collectively, these findings suggest a role for tapasin in the loading of peptides onto class I molecules. In addition, the association between class I molecules and the TAP transporter is lost in the absence of tapasin (4,13,14). This latter observation, combined with the finding that tapasin binds to class I heterodimers in the absence of TAP and to TAP in the absence of class I, led to the suggestion that tapasin forms a bridge between the TAP transporter and class I heterodimers (4). However, this bridging role does not appear to be crucial to tapasin’s functions since a soluble version of tapasin that binds to class I but not to TAP, and hence cannot serve as a bridge, is fully capable of restoring normal peptide loading to tapasin‐deficient 721.220 cells (15). Although tapasin facilitates peptide loading, it is not absolutely essential. Cells from tapasin‐deficient mice could be recognized by and could stimulate alloreactive CTL (11,12), and furthermore such mice were tolerant to immunization with cells expressing syngeneic class I molecules (12). Thus, there appears to be substantial overlap between the self peptides expressed in tapasin‐containing versus tapasin‐deficient cells. In addition, cell‐surface class I molecules in the absence of tapasin interaction are less stable than in wild‐type cells, but are more stable than class I molecules that completely lack peptides. This suggests the presence of an altered peptide repertoire in class I molecules loaded in the absence of tapasin rather than a complete lack of peptides (8,12). Direct analysis of peptides bound to mouse H‐2Kb and human HLA‐B*2705 molecules in the absence or presence of tapasin revealed a reduction in the overall yield of peptides from tapasin‐deficient cells (8,10). However, those peptides that could be recovered exhibited considerable (but not complete) overlap with the repertoire of peptides recovered from class I molecules expressed in tapasin‐containing cells. Therefore, there appear to be both quantitative and qualitative influences of tapasin on the loading of peptides onto class I molecules. How tapasin effects its functions in peptide loading remains unclear. Several recent studies have suggested functions that include the ER retention of empty class I molecules (8,11,16), modulation of TAP expression levels (15), promotion of peptide binding and translocation by TAP (11,17), and stabilization of peptide‐deficient H chain–β2m heterodimers (15). Furthermore, some of tapasin’s effects may be indirect since tapasin binding to class I molecules is believed to occur in collaboration with CRT. It has been reported that class I molecules that do not bind tapasin also do not associate with CRT, leading to a model of cooperative binding between tapasin and CRT (18,19). A similar model has been suggested for the association of ERp57 with class I (19,20). To add to the complexity of tapasin functions, various human class I allotypes have been shown to possess different levels of tapasin dependence. For example, HLA‐B*2705 expressed in tapasin‐negative 721.220 cells is expressed on the cell surface at levels similar to that observed in tapasin‐expressing cells and can present viral antigens to CTL. In contrast, HLA‐B*4402 surface expression and antigen presentation are highly dependent on the presence of tapasin, whereas HLA‐B8 exhibits an intermediate dependence (9). In the murine system as well, Myers et al. showed that on the surface of tapasin‐deficient cells the H‐2Kb, ‐Kd and ‐Ld molecules differ in their ratios of peptide‐receptive versus conformationally mature forms (7). Given the diversity of functions attributed to tapasin and its apparent cooperation with CRT and ERp57 in associating with the peptide‐loading complex, there is considerable interest in understanding how the components of this complex are organized. Cresswell and co‐workers demonstrated that up to four class I–tapasin complexes are associated with each TAP heterodimer (14). Several subsequent studies focused on defining the nature of the class I–tapasin association. Using truncation mutants, the N‐terminal 50 residues of tapasin were shown to mediate its association with class I (21). Conversely, site‐directed mutagenesis identified several residues within the class I H chain that are important for binding to tapasin (18,22–25). These include the glycosylation site at residue 86 in the α1 domain of the H‐2Ld molecule (18), residues 115, 122 and 134 in the α2 domain of HLA‐A2 (23–25) or residues 133 and 134 in the α2 domain of H‐2Ld (18) and, within the α3 domain, residue 222 in H‐2Dd (22) or 227 in H‐2Ld (18). For some of these mutants, peptide deficiency phenotypes similar to those observed in tapasin‐deficient cells were observed (18,22–24), whereas others did not appear to be impaired in peptide binding (25). Since previous studies on the nature of the tapasin–class I interaction focused almost entirely on the H‐2Ld and HLA‐A2 molecules, and various phenotypes were observed with different non‐tapasin‐binding mutants, we decided to undertake a systematic examination of a third molecule, H‐2Dd, in an effort to identify residues involved in tapasin association and to characterize phenotypes associated with a loss of tapasin binding. Significant differences were observed in the sites of tapasin interaction with H‐2Dd and in the association of Dd mutants with CRT indicating isotype‐specific differences in the organization of the peptide‐loading complex. Methods Mutagenesis and transfection The pSV2‐neo plasmid containing the H‐2Dd gene was provided by Dr T. Potter (National Jewish Center for Immunology and Respiratory Diseases, Denver, CO). DNA segments encoding the α2 and α3 domains of Dd were subcloned separately into pBluescript and site‐directed mutations were generated using the QuikChange mutagenesis kit (Stratagene, La Jolla, CA). The following oligonucleotides were used for mutagenesis (lower case letters denote mutated bases): R111→A: 5′‐CTG CCA GTA CCC Ggc GAG GAG GCG CCC‐3′ D122→A: 5′‐CAG GGC GAT GTA tgC GCA GCC GTC GTA GG‐3′ N127→A: 5′‐CGT TTT CAG GTC cTG Ggc CAG GGC GAT GTA ATC‐3′ E128→A: 5′‐CCA CGT TTT CAG GTC TgC GTT CAG GGC G‐3′ D129→A: 5′‐CGT CCA CGT TTT CAG GgC TTC GTT CAG G‐3′ T132→A: 5′‐CCG CCG TCC ACG ccT TaA GGT CTT CGT TCA GG‐3′ N127→A, E128→A: 5′‐CCA CGT TTT CAG GTC TgC Ggc CAG GGC GAT GTA ATC‐3′ D129→A, T132→A: 5′‐CCG CCG TCC ACG cTT TCA GGg CTT CGT TCA GG‐3′ A135→Q, A136→Q: 5′‐GCG CCG CCA TGT CCt gCt gCG TCC ACG TTT TCA GG‐3′ N220→A: 5′‐GGT CAG CTC CTC gcc GGC CAA CTG CCA GGT CAG‐3′ D212→A, T214→A: 5′‐CTG CCA GGT CAG ctg GAT GgC AGC AGG GTA GAA GC‐3′ T225→A, Q226→A: 5′‐CAA GCT CCA TTT CCg cGG cCA GCT CCT CCC C‐3′ All mutations were confirmed by DNA sequencing before the mutated fragment was inserted back into pSV2‐neo containing the rest of the H‐2Dd gene. PSV2‐neo plasmids containing mutant Dd genes were transfected into mouse L cells (H‐2k) using the Superfect reagent (Qiagen, Hilden, Germany). Stable transfectants were obtained by geneticin selection (Life Technologies, Gaithersburg, MD) and cells positive for surface Dd expression were separated from negative cells by FACS. Mutant cells expressing D129A/T132A could not be sorted due to low expression, thus clones were obtained by limiting dilution. Cells, antibodies and other reagents Transfected mouse L cells were grown in DMEM (Life Technologies) supplemented with 2 mM glutamine, 10% FCS, antibiotics and 500 µg/ml geneticin. For detection of H‐2Dd, mAb 34‐2‐12S and 34‐5‐8S were used, the latter being specific for β2m‐associated Dd (26). mAb 16‐3‐1N, which recognizes β2m‐associated H‐2Kk (27), was used as a positive control in flow cytometry experiments. Rabbit anti‐TAP2 antiserum was provided by Drs Y. Yang and P. Peterson (R. W. Johnson, La Jolla CA). Anti‐tapasin antiserum was raised in rabbits against the C‐terminal 20 amino acids of murine tapasin (28). Rabbit anti‐CNX antiserum directed against the ER luminal domain of CNX has been described previously (29). Anti‐CRT antiserum, SPA‐600, was purchased from StressGen (Victoria, BC, Canada). The H‐2Dd‐binding peptide, designated Tum– (sequence NGPPHSNNF), was synthesized by the Alberta Peptide Institute (Edmonton) and was used in peptide occupancy experiments. Flow cytometry To determine the cell‐surface levels of the various H‐2Dd mutants, transfected L cells (3.5 × 105) were removed from plates by trypsinization and incubated on ice for 15–20 min in 0.1 ml of FACS buffer (HBSS with 1% BSA and 0.01% NaN3) containing 1.5 µg of either mAb 34‐5‐8S or mAb 16‐3‐1N. Cells were washed once and incubated for 15–20 min on ice with 0.4 µg of fluorescein‐conjugated goat anti‐mouse IgG (Jackson ImmunoResearch, West Grove, PA) in 0.1 ml FACS buffer. Cells were washed twice and resuspended in 0.3 ml of FACS buffer containing 20 µg/ml propidium iodide for the detection of live cells. Analysis followed within 30 min using an Epics Elite flow cytometer (Beckman Coulter, Fullerton, CA). For experiments measuring the turnover of cell‐surface Dd molecules, cells were removed from plates by treatment with 2.5 mM EDTA in PBS and, following antibody incubations, were fixed in 0.5% paraformaldehyde prior to analysis. Metabolic radiolabeling and immunoisolation Typically, 5 × 106 transfected L cells in a single 100‐mm plate were starved for 30 min with Met‐free RPMI and radiolabeled for 30 min in 1 ml of medium containing 0.3 mCi of [35S]Met (>1000 Ci/mmol; Amersham, Little Chalfont, UK). Cells were lysed at 4°C for 30 min in PBS, pH 7.4, containing 1% digitonin, 10 mM iodoacetamide, 60 µg/ml Pefabloc (Roche, Basel, Switzerland), and 10 µg/ml each of leupeptin, antipain and pepstatin. Lysates were centrifuged at 11,000 g to pellet nuclei and cell debris. To isolate Dd molecules associated with TAP, tapasin or CRT, supernatant fractions were subjected to a first round of immunoisolation by incubating with anti‐TAP, anti‐tapasin or anti‐CRT antisera for 2 h. Immune complexes were recovered over a period of 1 h with Protein A–agarose beads. The beads were then heated at 40°C in PBS containing 0.2% SDS for 1 h to disrupt protein complexes. Eluted material was adjusted to 5% skim milk and 2% Nonidet P‐40, and a second round of immunoisolation was performed at 4°C for 2 h with the anti‐Dd mAb 34‐2‐12S or an isotype‐matched control mAb, MKD6. Immune complexes recovered with Protein A beads were analyzed by SDS–PAGE (10% gels). For pulse–chase experiments, 1 × 106 transfected L cells were radiolabeled for 10 min as above except that 0.1 mCi of [35S]Met was used. Cells were washed and then chased for various periods in DMEM containing 1 mM Met. Lysis was conducted in PBS containing 1% Nonidet P‐40, 2 mM iodoacetamide and protease inhibitors, and then lysates were centrifuged as above. In the case of peptide occupancy experiments, supernatant fractions were incubated for 1 h at 4°C in the presence or absence of 50 µM Tum– peptide and then incubated at 37°C for an additional h. Peptide‐containing Dd molecules were isolated with mAb 34‐5‐8S and immune complexes were collected with Protein A–agarose beads. Dd molecules were eluted from the beads with 0.1 M citrate buffer, pH 6, containing 0.1% SDS and digested with endo‐β‐N‐acetylglucosaminidase H (Endo H; New England Biolabs, Beverly, MA) prior to analysis by SDS–PAGE. Radioactive bands were quantified using a STORM Phosphorimager and ImageQuant software (Molecular Dynamics, Sunnyvale, CA). In all cases, band intensities were corrected by subtracting a background value determined by quantifying a blank area of the gel corresponding in size to the radioactive band of interest. Turnover kinetics of cell‐surface H‐2Dd L cell transfectants (∼1.5 × 106) were incubated for 18 h at 26°C in serum‐free medium containing 10 µg/ml of human β2m (Sigma, St Louis, MO). Cells were then washed with cold medium and resuspended in 4 ml of pre‐warmed DMEM containing 10 µg/ml of Brefeldin A (Sigma). A 1‐ml aliquot of cells was immediately transferred to a tube containing 2 ml of FACS buffer while the rest was transferred to a 37°C water bath. Aliquots were collected at the indicated time points and analyzed by FACS as described above using the conformation‐dependent H‐2Dd mAb 34‐5‐8S. Results Cell‐surface expression of H‐2Dd mutants We previously demonstrated that residue E222 within the α3 domain of H‐2Dd is involved in association with tapasin/TAP, but residues E223 and E232 are not (22). To identify additional segments of the H‐2Dd H chain that participate in association with tapasin/TAP, we selected 13 residues of the H chain α2 and α3 domains to mutate (Fig. 1). These residues were chosen based on the following criteria: conservation among class I molecules from diverse vertebrates, solvent accessibility, and their location on a face of the H chain defined by residues T134, E222 and E227, previously implicated in tapasin/TAP association with different class I molecules (18,22–24). Residues involved in H chain interaction with either peptide or β2m were avoided. Where possible, mutants were generated in pairs to allow screening of a large number of residues in a minimum amount of time. Single mutants were obtained subsequently if the double mutants displayed a deficiency in tapasin/TAP association. All residues were mutated to alanine except for A135 and A136, which were mutated to glutamine. Each of the mutants was stably expressed in mouse L cells and cells positive for cell‐surface H‐2Dd were sorted from negative cells. To avoid cloning artifacts, these Dd‐positive cells were used in subsequent experiments without further selection. Initially, mutants were compared in terms of their relative synthetic rates of Dd H chains. In most cases, rates were within 10% of wild‐type Dd (data not shown). For those mutants that differed by a greater percentage (50–70% of wild‐type for E222 and D122 mutants), cell numbers were adjusted accordingly to facilitate comparisons in subsequent experiments. As an initial screen for mutants that may be impaired in tapasin/TAP association, the levels of cell‐surface Dd expression were compared using the conformation‐sensitive mAb 34‐5‐8S. It is well established that in the absence of association with tapasin/TAP, many class I molecules are expressed at the surface of cells at substantially reduced levels as a consequence of inefficient peptide loading (4,7–14). As shown in Fig. 2, several of the mutants (R111A, D212A/T214A, N220A and T225A/Q226A) exhibited surface expression levels comparable to wild‐type Dd. In contrast, mutants D122A and N127A/E128A had reduced surface expression that was ∼50% of wild‐type, and mutants D129A/T132A and A135Q/A136Q were expressed at low levels, ∼20% of wild‐type. This trend did not change significantly when expression levels were adjusted to compensate for different Dd H chain synthetic rates (Fig. 2). These results suggested that mutants D122A and N127A/E128A, D129A/T132A and A135Q/A136Q may be candidates for proteins with altered tapasin/TAP interactions. To ascertain which residue(s) within the double mutants were responsible for the reduced surface expression, we generated the single mutants N127A, E128A, D129A and T132A. Unfortunately, we were unable to obtain the single mutants A135Q or A136Q. Mutants N127A and E128A exhibited levels of surface expression that were very close to wild‐type, indicating that simultaneous mutation of both residues was required to create the low surface expression phenotype. Mutant D129A retained very low expression at the cell surface and T132A was expressed at ∼50% of wild‐type, indicating that these residues are individually capable of influencing surface expression. Mutant Dd molecules with reduced surface expression are deficient in tapasin/TAP interactions To determine if the mutations introduced in the Dd molecule had an effect on its interaction with tapasin/TAP, we assayed for mutant association with tapasin or TAP by sequential immunoisolation. Initially, TAP or tapasin was isolated from digitonin lysates of radiolabeled cells and then immune complexes were disrupted in SDS and any associated Dd molecules were recovered in a second round of immune isolation with the anti‐Dd mAb, 34‐2‐12S (Fig. 3A). As expected, based on our previous work (22), the wild‐type Dd H chain could be recovered from TAP or tapasin immune isolates indicating that the wild‐type Dd molecule associates with both tapasin and TAP (Fig. 3A, lanes 4 and 5). In contrast, mutant E222K, which we previously showed was incapable of tapasin/TAP association (22), could not be recovered from tapasin or TAP immune complexes (Fig. 3A, lanes 6 and 7). Note that in these experiments, TAP and tapasin could be detected in the second round of isolation with the anti‐Dd mAb (Fig. 3A, lanes 4–7). Since these proteins were also present in control experiments wherein an isotype‐matched mAb (MKD6) was used in place of anti‐Dd mAb (Fig. 3A, lanes 2 and 3), this indicates either incomplete dissociation of initial anti‐tapasin or anti‐TAP immune complexes in SDS or renaturation of some of the anti‐tapasin or anti‐TAP antibodies during the second round of isolation. This is a consequence of the low temperature employed in the first round SDS dissociation step that was required to preserve the mAb 34‐2‐12S epitope for isolation of Dd in the second round. Examination of the remaining Dd mutants revealed that mutants R111A, D212A/T214A, N220A and T225A/Q226A remain capable of binding TAP and tapasin (Fig. 3A). This correlates well with their wild‐type levels of expression at the cell surface. In contrast, mutants D122A, N127A/E128A, D129A/T132A and A135Q/A136Q failed to associate with TAP and tapasin (Fig. 3A), again consistent with their reduced cell‐surface expression. To ensure that the latter group of mutants did not interact weakly with tapasin/TAP, we repeated the experiment in the presence of the proteasome inhibitor Z‐L3V5. We and others have shown that a reduction in peptide supply prolongs the association of empty class I molecules with the peptide‐loading complex (30,31). No residual interaction of the mutants with tapasin or TAP was detected in this experiment (data not shown). The sequential immunoisolation technique was also used to test the capability of the single Dd mutants to interact with tapasin. As shown in Fig. 3(B), mutants N127A, E128A and T132A exhibited very low but detectable association with tapasin (∼10% relative to wild‐type Dd), whereas mutant D129A failed to bind to tapasin. To ensure that the mutations introduced in Dd did not cause substantial misfolding of the molecule, we isolated the various mutants and showed that they all are capable of association with β2m (Fig. 3C). Furthermore, all mutants retained the ability to bind exogenous peptide in cell lysates (see Fig. 5 below). Therefore, in addition to our previous identification of residue 222 in the α3 domain (22), we have identified a cluster of residues (residues 122–136) within the α2 domain of H‐2Dd involved in its interaction with tapasin and TAP. Intracellular transport of non‐tapasin‐binding Dd mutants Conflicting ER to Golgi transport phenotypes have been reported as a consequence of either tapasin deficiency or mutations in class I H chains that prevent association with tapasin/TAP. In some instances, a lack of interaction with tapasin results in accelerated class I transport out of the ER, suggesting that tapasin is involved in ER retention of class I molecules (8,16,23–25). Lack of interaction with the peptide‐loading complex can, however, cause inefficient peptide loading and result in ER retention of empty class I molecules by other ER components such as CNX (22,32). We followed the ER to Golgi transport of wild‐type Dd and non‐tapasin‐binding Dd mutants by pulse–chase analysis and Endo H digestion. Conversion of a class I H chain from an Endo H‐sensitive form to a resistant form is due to processing of its N‐linked oligosaccharide chains and coincides with the passage of the class I molecule from the ER into medial and trans Golgi compartments. Consequently, it may be used as a measure of ER to Golgi transport kinetics. Figure 4 shows that wild‐type Dd molecules were exported from the ER to Golgi with t1/2 = 52 min, consistent with previous observations (22). The mutants fell into three categories: D122A and E128A were found to be transported faster than wild‐type Dd with t1/2 = 39 and 45 min respectively; mutants N127A/E128A, N127A and T132A were all efficiently transported out of the ER but at rates substantially slower than wild‐type (t1/2 = 90, 98 and 84 min respectively); and mutants D129A/T132A, A135Q/A136Q and D129A were exported extremely slowly with t1/2 ≫ 3 h suggesting that they were largely retained in the ER as stable proteins. These findings revealed that there is no clear correlation between the degree of tapasin interaction of the various Dd mutants and their ER to Golgi transport phenotypes. Molecules that associated to a low degree with tapasin exhibited rates either faster (E128A) or slower (N127A and T132A) than normal transport rates. Likewise, mutants that failed to bind to tapasin were transported either faster than normal (D122A), somewhat slower than normal (N127A/E128A) or were largely retained in the ER (D129A/T132A, A135Q/A136Q and D129A). Most H‐2Dd mutants are occupied with peptides Tapasin has been shown to influence the amount and sequence of peptides binding to class I molecules (7–14). To determine if an impaired tapasin interaction affects the peptide occupancy status of our mutants, we performed two independent assays. In the first experiment, we compared the thermal stability of different mutants in cell lysates to that of the wild‐type molecule. It is well established that class I molecules that lack peptides or contain low‐affinity peptides will dissociate upon incubation at 37°C and will not be recognized by conformation‐sensitive mAb (33). Transfected cells were radiolabeled for 10 min with [35S]Met and chased for 1 h. Following lysis, they were incubated at 4°C for 1 h in the absence or presence of a stabilizing Dd‐binding peptide and then shifted to 37°C for an additional 1 h prior to immune isolation with the conformation‐sensitive mAb 34‐5‐8S. As shown in Fig. 5(A), the bulk (60%) of Endo H‐resistant, wild‐type Dd molecules could be recovered whether or not stabilizing exogenous peptides were present, indicating that most wild‐type Dd molecules had been loaded with peptides and acquired thermal stability. In our hands, Endo H‐sensitive class I molecules rarely exhibited thermal stability suggesting that peptide loading is followed rapidly by export to the Golgi. As a control for the peptide‐deficient phenotype, we examined the E222K mutant that does not associate with tapasin/TAP and acquires peptides inefficiently (22). Accordingly, only 29% of Endo H‐resistant E222K molecules exhibited thermal stability. None of the other Dd mutants that were converted to Endo H‐resistant forms exhibited the same degree of thermal sensitivity as E222K. Rather, E128A and D122A exhibited an intermediate level of thermal stability (44 and 48% respectively), and mutants N127A/E128A, N127A and T132A were indistinguishable from wild‐type Dd (ranging from 60 to 70% thermal stability) (Fig. 5A). Thus, all of these mutants had been loaded to a substantial degree with stabilizing peptides. In the second approach, we monitored the stability of mutant molecules that were expressed at the cell surface. To ensure that the entire spectrum of surface‐expressed molecules was examined, cells were incubated overnight at 26°C in the presence of exogenous β2m to preserve any labile molecules that might lack peptide or contain suboptimal peptides. The temperature was then shifted to 37°C and the turnover of surface Dd molecules was monitored over time by flow cytometry. As expected, wild‐type Dd molecules remained largely stable over the 2 h duration of the experiment (Fig. 5B). Conversely, empty Dd molecules expressed in TAP‐deficient LKD8c cells (34) were extremely unstable with most mAb reactivity disappearing within 30 min at 37°C. Of the Dd mutants created in this study, only those that were severely impaired in their export from the ER exhibited thermal stabilities substantially less than wild‐type Dd. As shown in Fig. 5(B), mutants D129A/T132A, A135Q/A136Q and D129A lost roughly half of their reactivity with mAb 34‐5‐8S over the 2 h period, indicating the presence of empty molecules or molecules loaded with readily dissociable peptides in addition to peptide‐stabilized molecules. The results for the remaining mutants were consistent with the findings of the thermal stability assays in cell lysates (compare Fig. 5A and B). Mutants D122A, N127A/E128A, N127A, E128A and T132A all exhibited surface decay kinetics that were not significantly different from wild‐type Dd, indicating that they were loaded with high‐affinity, stabilizing peptides despite their impaired associations with tapasin/TAP. H‐2Dd mutants bind CRT and CNX Several previous studies using various mouse or human class I molecules have correlated a lack of association with tapasin with a loss of CRT binding, suggesting that tapasin and CRT associate with the peptide‐loading complex in a cooperative fashion (18,19). To determine if this observation also holds true for the H‐2Dd molecule, we immunoisolated CRT from cells transfected with wild‐type or mutant Dd molecules and assessed the presence of Dd among the CRT‐associated proteins by subsequent immunoisolation with mAb 34‐2‐12S (Fig. 6A, right panel). A direct immunoisolation with mAb 34‐2‐12S was also performed to evaluate the synthetic rate of each mutant. The latter experiment revealed that mutants E222K and D122A were synthesized at lower rates than the other mutants (Fig. 6A, left panel). Remarkably, despite the complete absence of tapasin association for each of the mutants tested (see Fig. 3A), all could be recovered from anti‐CRT immune precipitates. This was not simply due to non‐specific adsorption of Dd to immune complexes since none could be recovered when mAb 34‐2‐12S was replaced by the isotype‐matched control mAb, MKD6. Furthermore, with the exception of E222K, the various mutants were recovered from the pool of CRT‐associated proteins as readily as wild‐type Dd. Given this unexpected finding, it was important to determine if association of the mutant molecules with CNX was also maintained since it is conceivable that CRT could be substituting for some loss of CNX interaction. Using the same type of sequential immunoisolation approach used to characterize the Dd–CRT interaction, we established that all of the mutant proteins tested retained the ability to associate with CNX (Fig. 6B). The finding that wild‐type Dd and various Dd mutants that fail to associate with tapasin remain capable of binding CRT suggests that the cooperative binding model cannot be extended to all class I molecules. Discussion This study was undertaken to define more precisely the sites of interaction between class I H chains and tapasin, and to assess the phenotypes of mutant class I molecules that fail to bind to tapasin. Previous studies addressing the nature of the H chain–tapasin or –TAP interaction focused mainly on the mouse H‐2Ld (18) and human HLA‐A2 (23–25) molecules. In the case of H‐2Ld, most of the mutants were assessed for binding to TAP, but not tapasin and the mutations were non‐conservative in nature. Non‐conservative mutations, by causing steric or charge perturbations, may identify residues in proximity to contact sites rather than the interacting residues themselves. In the HLA‐A2 study, conservative mutations were employed, but a relatively small set of mutants was generated (25). In an effort to assess H chain isotype differences in tapasin interaction and to produce a more comprehensive picture of H chain–tapasin contact sites, we created 13 conservative mutants in the H‐2Dd molecule. Candidate residues for mutagenesis were rejected if they were not highly conserved among class I molecules from different species, were not solvent exposed or were located distant from residues previously implicated in tapasin binding (Fig. 1). As summarized in Fig. 7, we found that of the residues mutated within the α3 domain CD8 binding loop (residues 222–229), only E222 that we studied previously (22) was involved in Dd binding to tapasin. Mutations at N220, E223 [studied previously (22)], T225 or Q226 had no effect. We did not mutate residue E227 or residue E229 since previous mutagenesis studies at these positions in the Dd molecule failed to reveal the reduction in cell‐surface expression that would be expected if association with tapasin was impaired (35). Additional α3 domain residues located on the same face of the Dd molecule, D212 and T214, were also not involved in interactions with tapasin. Consistent with the involvement of the CD8 binding loop in tapasin interaction, mutation of residue E227 in the Ld molecule has been shown to prevent tapasin binding (18) (Fig. 7). Analogous mutations in the HLA‐A2 molecule have not been tested for tapasin interaction. However, in a recent report, HLA‐B27 possessing the double D227K/E229K mutation was shown to exhibit impaired association with tapasin (36). We also identified a primary area of interaction with tapasin located in the α2 domain of the Dd H chain. Residues D122, N127, E128, D129 and T132 all affected the binding of Dd to tapasin. Residues A135/A136, only tested in combination, also abrogated interaction with tapasin (Fig. 7 and Table 1). Yu et al. identified the same region of the Ld molecule although interaction with TAP rather than tapasin was assayed for most mutants (18). If one assumes that tapasin association is equally affected, then the segment from E128 to A136 is implicated in Ld binding to tapasin (Fig. 7). Some of the identified Ld residues are unlikely to be in direct contact with tapasin, such as L130 and W133, which are hydrophobic and directed away from solvent. Their mutation to hydrophilic residues likely caused perturbation in neighboring residues that do contact tapasin. Our study extends the interacting segment identified in Ld to include N127 and D122 (Fig. 7). Beissbarth et al. also identified D122 as a tapasin‐interacting residue in the HLA‐A2 molecule along with Q115 (Fig. 7) (25). Interestingly, residue T134, which was previously implicated in binding to tapasin when mutated to lysine in the A2 and Ld molecules (23,24), showed no effect when mutated conservatively to alanine in A2 (25). This could be due to T134 making only a weak contact with tapasin but, when mutated to lysine, the presence of the bulky side chain prevents other key contact residues from binding to tapasin. Surprisingly, mutation of residues D129 and T132 to alanine in the A2 molecule did not affect tapasin binding (25), but our analogous mutations in Dd largely prevented the interaction. This suggests that there are some differences in the manner in which different class I molecules associate with tapasin. However, despite these differences, the overall patterns of tapasin interaction sites on the H‐2Dd, H‐2Ld and HLA‐A2 molecules appear to be fairly well conserved (Fig. 7). The suggestion that tapasin forms a bridge between class I molecules and TAP (4) predicts that any H chain mutant that fails to bind tapasin should also fail to bind to TAP. It has been observed by Bangia et al. that a weak interaction between class I and TAP may exist in the absence of tapasin (21). However, in our hands, no mutants defective in tapasin binding were found that retained binding to TAP. Association with TAP/tapasin was always lost simultaneously, supporting the view that tapasin acts as an obligatory bridge between class I H chain–β2m heterodimers and the TAP transporter. Examination of the H‐2Dd mutants deficient in tapasin interaction revealed a complex array of phenotypes. With respect to export from the ER, a lack of tapasin interaction has been documented to either increase (8,16,21,23–25) or decrease (22,28) the rate of class I transport from the ER to Golgi. Both phenotypes were observed for the non‐tapasin binding Dd mutants (Table 1). Mutants D122A and E128A were shown to leave the ER faster than wild‐type Dd but most of the Dd mutants displayed retarded transport. Mutants N127A/E128A, N127A and T132A were slightly delayed, whereas mutants D129A, D129A/T132A and A135Q/A136Q were extremely slow to leave the ER. Thus there was no correlation between tapasin interaction and ER to Golgi transport rate. Furthermore, although all non‐tapasin binding mutants exhibited reduced surface expression compared to wild‐type Dd, there was also no clear correlation between tapasin interaction and degree of peptide occupancy as assessed by thermal stability (Table 1). Most non‐tapasin‐binding mutants were indistinguishable from wild‐type Dd in terms of thermal stability. Only the D129A, D129A/T132A and A135Q/A136Q mutants exhibited decreased peptide occupancy. Given the well‐documented effects of tapasin on increasing peptide‐binding efficiency of most class I molecules including Dd (4,7–14), it seems likely that the diverse phenotypes we observe are the result of more complex influences than a simple binding or non‐binding to tapasin. It is possible that mutants with near‐normal ER export rates and thermal stabilities retain some interaction with tapasin in living cells that is disrupted upon immune isolation, even when the ability to detect association is enhanced through the use of a proteasome inhibitor (data not shown). Such residual interaction may be sufficient to provide near normal peptide loading. For the very slowly transported mutants that do exhibit peptide deficiency, it may be that in addition to a loss of tapasin association, the mutations induce some degree of misfolding that is detected by ER chaperones, thereby leading to retention. Such misfolding would likely be quite localized since these mutants retained the ability to associate with β2m and bind exogenous peptide. Despite these caveats, the mutants remain informative as to the residues of the Dd H chain that contribute to tapasin binding. In the early steps of class I H chain folding, before any association with the peptide‐loading complex, the dominant interaction of human and mouse H chains is with the chaperone CNX (37–39). CNX promotes H chain folding and retains free H chains in the ER (1–3). CNX also binds ERp57 and the TAP/tapasin complex independently of class I, where it is thought to stabilize the proteins until the recruitment of H chain–β2m heterodimers (38). Upon H chain association with β2m and subsequent formation of the peptide‐loading complex, CRT largely replaces CNX, although in mouse CNX can still be a major component (4,5,30). Investigations into the interplay between constituents of the peptide‐loading complex, which consists of H chain–β2m, CRT, ERp57, tapasin and TAP, have indicated that the presence of tapasin is required for the inclusion of CRT and ERp57 in the complex. Neither of the latter components has been found associated with class I molecules in the absence of tapasin (18–20). This raises the question of whether CRT binds directly to class I molecules that are somehow made receptive by their simultaneous association with tapasin (CRT–class I–tapasin) or if CRT binds to tapasin which in turn associates with class I (CRT–tapasin–class I). In tapasin‐negative insect cells, it has been shown that CRT can substitute for CNX and promote the folding of free mouse Db H chains (39), which implies a physical contact between CRT and the H chain. Furthermore, in the current study, Dd mutants that failed to bind detectably to tapasin retained the ability to bind to CRT. Thus, at least for H‐2Dd, CRT appears to bind directly to Dd–β2m heterodimers. Unfortunately, we were unable to extend our examination of the interactions of non‐tapasin‐binding Dd mutants to include ERp57. Extensive co‐immunoisolation experiments with anti‐ERp57 antiserum or anti‐Dd mAb failed to detect an interaction of this enzyme even with the wild‐type Dd molecule (data not shown). It is not clear why we can readily detect CRT interactions with Dd–β2m heterodimers in the absence of apparent tapasin association in contrast to several other studies using different human or mouse class I molecules. We cannot exclude the possibility of a weak tapasin interaction for certain Dd mutants prior to cell lysis despite their lack of detectable association even under conditions of proteasome inhibition. However, given the strong co‐isolation of CRT we observed with the various mutants, it seems unlikely that the binding to CRT can be explained solely by the existence of a putative residual tapasin interaction. It is well established that CRT and CNX can interact with non‐native protein substrates via protein–protein interactions as well as via a lectin site (40,41). Therefore, one possibility is that polypeptide‐based binding of CRT to Dd in the absence of tapasin is stronger than with other human or mouse class I molecules due to differences in H chain sequences. Another possibility may be that the organization of CRT in the peptide‐loading complex may not be the same for all class I molecules. Indeed, while this manuscript was under review, Hansen and co‐workers described another example of this unusual behavior in which an S132K mutation in the HLA‐B27 molecule disrupted tapasin/TAP interactions but association with CRT was maintained (36). The functions of CRT remain one of the more enigmatic aspects of the peptide‐loading complex. The determination of whether CRT acts as a chaperone directly for class I in this complex or indirectly via tapasin will fundamentally influence subsequent experimentation on the functional aspects of CRT in class I biogenesis. Acknowledgements We thank Drs Tery Potter, Young Yang and Per Peterson for their kind gifts of reagents, and the Alberta Peptide Institute for peptide synthesis. M.‐E. P. was the recipient of a Medical Research Council of Canada Studentship and an Ontario Government Studentship. This work was supported by the National Cancer Institute of Canada with funds from the Canadian Cancer Society. Abbreviations CNX—calnexin CRT—calreticulin CTL—cytotoxic T lymphocyte Endo H—endo‐β‐N‐acetylglucosaminidase H ER—endoplasmic reticulum TAP—transporter associated with antigen processing View largeDownload slide Fig. 1. Locations of residues mutated in the H‐2Dd molecule. The polypeptide backbones of the Dd H chain and associated β2m are depicted in ribbon representation, and the α1, α2 and α3 domains of the H chain are indicated. Residues selected for mutation are shown in single‐letter amino acid code followed by their positions in the H chain primary sequence. View largeDownload slide Fig. 1. Locations of residues mutated in the H‐2Dd molecule. The polypeptide backbones of the Dd H chain and associated β2m are depicted in ribbon representation, and the α1, α2 and α3 domains of the H chain are indicated. Residues selected for mutation are shown in single‐letter amino acid code followed by their positions in the H chain primary sequence. View largeDownload slide Fig. 2. Cell‐surface expression of wild‐type and mutant Dd molecules. Mouse L cells transfected with wild‐type or mutant Dd genes were incubated with the β2m‐dependent mAb 34‐5‐8S followed by incubation with FITC‐labeled goat anti‐mouse IgG and analysis by flow cytometry. The mean fluorescence obtained for each mutant was plotted relative to wild‐type Dd which was set at 100% (white bars). In addition, surface Dd levels were adjusted based on variation in H chain synthetic rates (gray bars). View largeDownload slide Fig. 2. Cell‐surface expression of wild‐type and mutant Dd molecules. Mouse L cells transfected with wild‐type or mutant Dd genes were incubated with the β2m‐dependent mAb 34‐5‐8S followed by incubation with FITC‐labeled goat anti‐mouse IgG and analysis by flow cytometry. The mean fluorescence obtained for each mutant was plotted relative to wild‐type Dd which was set at 100% (white bars). In addition, surface Dd levels were adjusted based on variation in H chain synthetic rates (gray bars). View largeDownload slide Fig. 3. Association of Dd mutants with tapasin and TAP. (A) Transfected L cells were radiolabeled for 30 min with [35S]Met, lysed in digitonin lysis buffer and then subjected to a first round of immunoisolation with either anti‐tapasin (lanes 1) or anti‐TAP2 (lanes 2) antisera. Immune complexes adsorbed on Protein A beads were dissociated and subjected to a second round of immunoisolation with the β2m‐independent Dd mAb 34‐2‐12S. Control isolations included the isotype‐matched control mAb MKD6 in the second isolation. In addition, wild‐type Dd was isolated directly with mAb 34‐2‐12S to serve as a H chain mobility standard. The mobilities of the Dd heavy chain (Dd HC), TAP and tapasin are indicated. (B) The experiment in (A) was repeated with an additional set of mutants with the exception that only anti‐tapasin antiserum was used in the first round of sequential immunoisolation. (C) Transfected mouse L cells expressing various Dd mutants were radiolabeled and lysed as in (A) and then subjected to immunoisolation with mAb 34‐2‐12S. Recovered proteins were analyzed on a 15% SDS–polyacrylamide gel. Note that both the Dd H chain and β2m can be detected in each lane. View largeDownload slide Fig. 3. Association of Dd mutants with tapasin and TAP. (A) Transfected L cells were radiolabeled for 30 min with [35S]Met, lysed in digitonin lysis buffer and then subjected to a first round of immunoisolation with either anti‐tapasin (lanes 1) or anti‐TAP2 (lanes 2) antisera. Immune complexes adsorbed on Protein A beads were dissociated and subjected to a second round of immunoisolation with the β2m‐independent Dd mAb 34‐2‐12S. Control isolations included the isotype‐matched control mAb MKD6 in the second isolation. In addition, wild‐type Dd was isolated directly with mAb 34‐2‐12S to serve as a H chain mobility standard. The mobilities of the Dd heavy chain (Dd HC), TAP and tapasin are indicated. (B) The experiment in (A) was repeated with an additional set of mutants with the exception that only anti‐tapasin antiserum was used in the first round of sequential immunoisolation. (C) Transfected mouse L cells expressing various Dd mutants were radiolabeled and lysed as in (A) and then subjected to immunoisolation with mAb 34‐2‐12S. Recovered proteins were analyzed on a 15% SDS–polyacrylamide gel. Note that both the Dd H chain and β2m can be detected in each lane. View largeDownload slide Fig. 4. Intracellular transport of Dd mutants. Cells expressing either wild‐type or mutant Dd molecules were radiolabeled for 10 min with [35S]Met and chased in medium containing unlabeled Met for periods of up to 3 h. Aliquots of cells were removed at the indicated times and Dd molecules were recovered from cell lysates with mAb 34‐2‐12S. Isolated proteins were digested with Endo H prior to analysis by SDS–PAGE. The mobilities of Endo H‐sensitive (Endo Hs) and Endo H‐resistant (Endo Hr) H chains are indicated. View largeDownload slide Fig. 4. Intracellular transport of Dd mutants. Cells expressing either wild‐type or mutant Dd molecules were radiolabeled for 10 min with [35S]Met and chased in medium containing unlabeled Met for periods of up to 3 h. Aliquots of cells were removed at the indicated times and Dd molecules were recovered from cell lysates with mAb 34‐2‐12S. Isolated proteins were digested with Endo H prior to analysis by SDS–PAGE. The mobilities of Endo H‐sensitive (Endo Hs) and Endo H‐resistant (Endo Hr) H chains are indicated. View largeDownload slide Fig. 5. Dd mutants that are impaired in tapasin binding are occupied with peptides. (A) L cells expressing wild‐type or mutant Dd were radiolabeled with [35S]Met for 10 min, chased with unlabeled medium for 1 h and lysed in NP‐40 lysis buffer. Lysates were incubated for 1 h at 4°C in the presence (+) or absence (–) of a Dd‐binding peptide and subsequently transferred to 37°C for 1 h before being subjected to immunoisolation with the β2m‐dependent mAb 34‐5‐8S. The recovered material was digested with Endo H and analyzed by SDS–PAGE. The number below each pair of lanes indicates the percentage of thermostable, Endo H‐resistant H chain recovered in the absence of exogenous peptide relative to that recovered in the presence of exogenous peptide, as assessed by densitometric analysis. (B) Transfected L cells were incubated for 18 h at 26°C in the presence of exogenous human β2m to allow for the stabilization of class I molecules at the cell surface. Brefeldin A (10 µg/ml) was then added to prevent the surface appearance of new Dd molecules and cells were transferred to 37°C for the indicated times. The relative amounts of β2m‐containing Dd molecules at the cell surface were assessed by flow cytometry at each time point using mAb 34‐5‐8S. View largeDownload slide Fig. 5. Dd mutants that are impaired in tapasin binding are occupied with peptides. (A) L cells expressing wild‐type or mutant Dd were radiolabeled with [35S]Met for 10 min, chased with unlabeled medium for 1 h and lysed in NP‐40 lysis buffer. Lysates were incubated for 1 h at 4°C in the presence (+) or absence (–) of a Dd‐binding peptide and subsequently transferred to 37°C for 1 h before being subjected to immunoisolation with the β2m‐dependent mAb 34‐5‐8S. The recovered material was digested with Endo H and analyzed by SDS–PAGE. The number below each pair of lanes indicates the percentage of thermostable, Endo H‐resistant H chain recovered in the absence of exogenous peptide relative to that recovered in the presence of exogenous peptide, as assessed by densitometric analysis. (B) Transfected L cells were incubated for 18 h at 26°C in the presence of exogenous human β2m to allow for the stabilization of class I molecules at the cell surface. Brefeldin A (10 µg/ml) was then added to prevent the surface appearance of new Dd molecules and cells were transferred to 37°C for the indicated times. The relative amounts of β2m‐containing Dd molecules at the cell surface were assessed by flow cytometry at each time point using mAb 34‐5‐8S. View largeDownload slide Fig. 6. Dd mutants that do not bind tapasin remain associated with CRT. Cells were radiolabeled for 30 min with [35S]Met and lysed in digitonin lysis buffer. Lysates were subjected to immunoisolation with anti‐Dd mAb 34‐2‐12S (A and B, left panels) or to sequential immunoisolation with either anti‐CRT or anti‐CNX antisera followed by mAb 34‐2‐12S antibody to recover CRT‐associated Dd molecules (A, right panel) or CNX‐associated Dd molecules (B, right panel) respectively. View largeDownload slide Fig. 6. Dd mutants that do not bind tapasin remain associated with CRT. Cells were radiolabeled for 30 min with [35S]Met and lysed in digitonin lysis buffer. Lysates were subjected to immunoisolation with anti‐Dd mAb 34‐2‐12S (A and B, left panels) or to sequential immunoisolation with either anti‐CRT or anti‐CNX antisera followed by mAb 34‐2‐12S antibody to recover CRT‐associated Dd molecules (A, right panel) or CNX‐associated Dd molecules (B, right panel) respectively. View largeDownload slide Fig. 7. Locations of mutations in the H‐2Dd, H‐2Ld and HLA‐A2 molecules that result in loss of association with tapasin. The polypeptide backbone is represented by the thin line and the side chains of the mutated residues are represented in stick form. (A) The entire H chain is depicted; β2m has been omitted for clarity. The arrow indicates the viewpoint for the detailed images shown in (B). (B) A detailed view of α2 domain residues 111–150. Images were created in SwissPdbViewer and rendered with POV‐Ray. View largeDownload slide Fig. 7. Locations of mutations in the H‐2Dd, H‐2Ld and HLA‐A2 molecules that result in loss of association with tapasin. The polypeptide backbone is represented by the thin line and the side chains of the mutated residues are represented in stick form. (A) The entire H chain is depicted; β2m has been omitted for clarity. The arrow indicates the viewpoint for the detailed images shown in (B). (B) A detailed view of α2 domain residues 111–150. Images were created in SwissPdbViewer and rendered with POV‐Ray. Table 1. Summary of mutant phenotypes Mutant  Surface expression (% wild‐type)  Tapasin interaction  ER to Golgi transport rate t1/2 (min)  Percent thermostable Endo Hr molecules  Turnover rate of cell‐surface molecules  Wild‐type  100  +  52  60  wild‐type (slow)  R111A  80  NT  NT  NT  NT  D122A  50  –  39  48  like wild‐type  N127A             E128A  50  –  100  67  like wild‐type  N127A  80  +/–  127  70  like wild‐type  E128A  70  +/–  45  44  like wild‐type  D129A             T132A  20  –  >>180  NA  moderate  D129A  15  –  >>180  NA  moderate  T132A  50  +/–  84  60  like wild‐type  A135Q             A136Q  20  –  >>180  NA  moderate  D212A             T214A  80  NT  NT  NT  NT  N220A  85  NT  NT  NT  NT  T225A             Q226A  110  NT  NT  NT  NT  LKD8  10  NT  NT  NT  rapid  Mutant  Surface expression (% wild‐type)  Tapasin interaction  ER to Golgi transport rate t1/2 (min)  Percent thermostable Endo Hr molecules  Turnover rate of cell‐surface molecules  Wild‐type  100  +  52  60  wild‐type (slow)  R111A  80  NT  NT  NT  NT  D122A  50  –  39  48  like wild‐type  N127A             E128A  50  –  100  67  like wild‐type  N127A  80  +/–  127  70  like wild‐type  E128A  70  +/–  45  44  like wild‐type  D129A             T132A  20  –  >>180  NA  moderate  D129A  15  –  >>180  NA  moderate  T132A  50  +/–  84  60  like wild‐type  A135Q             A136Q  20  –  >>180  NA  moderate  D212A             T214A  80  NT  NT  NT  NT  N220A  85  NT  NT  NT  NT  T225A             Q226A  110  NT  NT  NT  NT  LKD8  10  NT  NT  NT  rapid      NT, not tested; NA, not applicable. View Large References 1 Vassilakos, A., Cohen‐Doyle, M. F., Peterson, P. A., Jackson, M. R. and Williams, D. B. 1996. The molecular chaperone calnexin facilitates folding and assembly of class I histocompatibility molecules. EMBO J.  15: 1495. Google Scholar 2 Jackson, M. R., Cohen‐Doyle, M. F., Peterson, P. A. and Williams, D. B. 1994. Regulation of MHC class I transport by the molecular chaperone, calnexin (p88, IP90). Science  263: 384. Google Scholar 3 Tector, M. and Salter, R. D. 1995. Calnexin influences folding of human class I histocompatibility proteins but not their assembly with beta 2‐microglobulin. J. Biol. Chem.  270: 19638. Google Scholar 4 Sadasivan, B., Lehner, P. J., Ortmann, B., Spies, T. and Cresswell, P. 1996. Roles for calreticulin and a novel glycoprotein, tapasin, in the interaction of MHC class I molecules with TAP. Immunity  5: 103. Google Scholar 5 Solheim, J. C., Harris, M. R., Kindle, C. S. and Hansen, T. H. 1997. Prominence of beta 2‐microglobulin, class I heavy chain conformation, and tapasin in the interactions of class I heavy chain with calreticulin and the transporter associated with antigen processing. J. Immunol.  158: 2236. Google Scholar 6 Cresswell, P., Bangia, N., Dick, T. and Diedrich, G. 1999. The nature of the MHC class I peptide loading complex. Immunol. Rev.  172: 21. Google Scholar 7 Myers, N. B., Harris, M. R., Connolly, J. M., Lybarger, L., Yu, Y. Y. and Hansen, T. H. 2000. K(b), k(d), and L(d) molecules share common tapasin dependencies as determined using a novel epitope Tag [In process citation]. J. Immunol.  165: 5656. Google Scholar 8 Barnden, M. J., Purcell, A. W., Gorman, J. J. and McCluskey, J. 2000. Tapasin‐mediated retention and optimization of peptide ligands during the assembly of class I molecules. J. Immunol.  165: 322. Google Scholar 9 Peh, C. A., Burrows, S. R., Barnden, M., Khanna, R., Cresswell, P., Moss, D. J. and McCluskey, J. 1998. HLA‐B27‐restricted antigen presentation in the absence of tapasin reveals polymorphism in mechanisms of HLA class I peptide loading. Immunity  8: 531. Google Scholar 10 Purcell, A. W., Gorman, J. J., Garcia‐Peydro, M., Paradela, A., Burrows, S. R., Talbo, G. H., Laham, N., Peh, C. A., Reynolds, E. C., Lopez De Castro, J. A. and McCluskey, J. 2001. Quantitative and qualitative influences of tapasin on the class I peptide repertoire. J. Immunol.  166: 1016. Google Scholar 11 Grandea, A. G., III, Golovina, T. N., Hamilton, S. E., Sriram, V., Spies, T., Brutkiewicz, R. R., Harty, J. T., Eisenlohr, L. C. and Van Kaer, L. 2000. Impaired assembly yet normal trafficking of MHC class I molecules in Tapasin mutant mice. Immunity  13: 213. Google Scholar 12 Garbi, N., Tan, P., Diehl, A. D., Chambers, B. J., Ljunggren, H. G., Momburg, F. and Hammerling, G. J. 2000. Impaired immune responses and altered peptide repertoire in tapasin‐deficient mice. Nat. Immunol.  1: 234. Google Scholar 13 Grandea, A. G., III, Androlewicz, M. J., Athwal, R. S., Geraghty, D. E. and Spies, T. 1995. Dependence of peptide binding by MHC class I molecules on their interaction with TAP. Science  270: 105. Google Scholar 14 Ortmann, B., Copeman, J., Lehner, P. J., Sadasivan, B., Herberg, J. A., Grandea, A. G., Riddell, S. R., Tampe, R., Spies, T., Trowsdale, J. and Cresswell, P. 1997. A critical role for tapasin in the assembly and function of multimeric MHC class I–TAP complexes. Science  277: 1306. Google Scholar 15 Lehner, P. J., Surman, M. J. and Cresswell, P. 1998. Soluble tapasin restores MHC class I expression and function in the tapasin‐negative cell line .220. Immunity  8: 221. Google Scholar 16 Schoenhals, G. J., Krishna, R. M., Grandea, A. G., III, Spies, T., Peterson, P. A., Yang, Y. and Fruh, K. 1999. Retention of empty MHC class I molecules by tapasin is essential to reconstitute antigen presentation in invertebrate cells. EMBO J.  18: 743. Google Scholar 17 Li, S., Paulsson, K. M., Chen, S., Sjogren, H. O. and Wang, P. 2000. Tapasin is required for efficient peptide binding to transporter associated with antigen processing. J. Biol. Chem.  275: 1581. Google Scholar 18 Yu, Y. Y., Turnquist, H. R., Myers, N. B., Balendiran, G. K., Hansen, T. H. and Solheim, J. C. 1999. An extensive region of an MHC class I alpha 2 domain loop influences interaction with the assembly complex. J. Immunol.  163: 4427. Google Scholar 19 Hughes, E. A. and Cresswell, P. 1998. The thiol oxidoreductase ERp57 is a component of the MHC class I peptide‐loading complex. Curr. Biol.  8: 709. Google Scholar 20 Harris, M. R., Lybarger, L., Yu, Y. Y., Myers, N. B. and Hansen, T. H. 2001. Association of ERp57 with mouse MHC class I molecules is tapasin dependent and mimics that of calreticulin and not calnexin. J. Immunol.  166: 6686. Google Scholar 21 Bangia, N., Lehner, P. J., Hughes, E. A., Surman, M. and Cresswell, P. 1999. The N‐terminal region of tapasin is required to stabilize the MHC class I loading complex. Eur. J. Immunol.  29: 1858. Google Scholar 22 Suh, W. K., Derby, M. A., Cohen‐Doyle, M. F., Schoenhals, G. J., Fruh, K., Berzofsky, J. A. and Williams, D. B. 1999. Interaction of murine MHC class I molecules with tapasin and TAP enhances peptide loading and involves the heavy chain alpha3 domain. J. Immunol.  162: 1530. Google Scholar 23 Lewis, J. W., Neisig, A., Neefjes, J. and Elliott, T. 1996. Point mutations in the alpha 2 domain of HLA‐A2.1 define a functionally relevant interaction with TAP. Curr. Biol.  6: 873. Google Scholar 24 Peace‐Brewer, A. L., Tussey, L. G., Matsui, M., Li, G., Quinn, D. G. and Frelinger, J. A. 1996. A point mutation in HLA‐A*0201 results in failure to bind the TAP complex and to present virus‐derived peptides to CTL. Immunity  4: 505. Google Scholar 25 Beissbarth, T., Sun, J., Kavathas, P. B. and Ortmann, B. 2000. Increased efficiency of folding and peptide loading of mutant MHC class I molecules. Eur. J. Immunol.  30: 1203. Google Scholar 26 Ozato, K., Mayer, N. M. and Sachs, D. H. 1982. Monoclonal antibodies to mouse major histocompatibility complex antigens IV. A series of hybridoma clones producing anti‐H‐2d antibodies and an examination of expression of H‐2d antigens on the surface of these cells. Transplantation  34: 113. Google Scholar 27 Ozato, K., Mayer, N. M. and Sachs, D. H. 1980. Hybridoma cell lines secreting monoclonal antibodies to mouse H‐2 and la antigens. J. Immunol.  137: 3881. Google Scholar 28 Grandea, A. G., III, Comber, P. G., Wenderfer, S. E., Schoenhals, G., Fruh, K., Monaco, J. J. and Spies, T. 1998. Sequence, linkage to H2‐K, and function of mouse tapasin in MHC class I assembly. Immunogenetics  48: 260. Google Scholar 29 Danilczyk, U. G. and Williams, D. B. 2001. The lectin‐chaperone calnexin utilizes polypeptide‐based interactions to associate with many of its substrates in vivo. J. Biol. Chem.  276: 25532. Google Scholar 30 Suh, W. K., Mitchell, E. K., Yang, Y., Peterson, P. A., Waneck, G. L. and Williams, D. B. 1996. MHC class I molecules form ternary complexes with calnexin and TAP and undergo peptide‐regulated interaction with TAP via their extracellular domains. J. Exp. Med.  184: 337. Google Scholar 31 Hughes, E. A., Ortmann, B., Surman, M. and Cresswell, P. 1996. The protease inhibitor, N‐acetyl‐l‐leucyl‐l‐leucyl‐leucyl‐l‐norleucinal, decreases the pool of major histocompatibility complex class I‐binding peptides and inhibits peptide trimming in the endoplasmic reticulum. J. Exp. Med.  183: 1569. Google Scholar 32 Grandea, A. G., III, Lehner, P. J., Cresswell, P. and Spies, T. 1997. Regulation of MHC class I heterodimer stability and interaction with TAP by tapasin. Immunogenetics  46: 477. Google Scholar 33 Ljunggren, H. G., Stam, N. J., Ohlen, C., Neefjes, J. J., Hoglund, P., Heemels, M. T., Bastin, J., Schumacher, T. N., Townsend, A., Karre, K., et al. 1990. Empty MHC class I molecules come out in the cold. Nature  346: 476. Google Scholar 34 Otten, G. R., Bikoff, E., Ribaudo, R. K., Kozlowski, S., Margulies, D. H. and Germain, R. N. 1992. Peptide and beta 2‐microglobulin regulation of cell surface MHC class I conformation and expression. J. Immunol.  148: 3723. Google Scholar 35 Connolly, J. M., Hansen, T. H., Ingold, A. L. and Potter, T. A. 1990. Recognition by CD8 on cytotoxic T lymphocytes is ablated by several substitutions in the class I alpha 3 domain: CD8 and the T‐cell receptor recognize the same class I molecule. Proc. Natl Acad. Sci. USA  87: 2137. Google Scholar 36 Harris, M. R., Lybarger, L., Myers, N. B., Hilbert, C., Solheim, J. C., Hansen, T. H. and Yu, Y. Y. 2001. Interactions of HLA‐B27 with the peptide loading complex as revealed by heavy chain mutations. Int. Immunol.  13: 1275. Google Scholar 37 Carreno, B. M., Solheim, J. C., Harris, M., Stroynowski, I., Connolly, J. M. and Hansen, T. H. 1995. TAP associates with a unique class I conformation, whereas calnexin associates with multiple class I forms in mouse and man. J. Immunol.  155: 4726. Google Scholar 38 Diedrich, G., Bangia, N., Pan, M. and Cresswell, P. 2001. A role for calnexin in the assembly of the MHC class I loading complex in the endoplasmic reticulum. J. Immunol.  166: 1703. Google Scholar 39 Danilczyk, U. G., Cohen‐Doyle, M. F. and Williams, D. B. 2000. Functional relationship between calreticulin, calnexin, and the endoplasmic reticulum luminal domain of calnexin. J. Biol. Chem.  275: 13089. Google Scholar 40 Ihara, Y., Cohen‐Doyle, M. F., Saito, Y. and Williams, D. B. 1999. Calnexin discriminates between protein conformational states and functions as a molecular chaperone in vitro. Mol. Cell  4: 331. Google Scholar 41 Saito, Y., Ihara, Y., Leach, M. R., Cohen‐Doyle, M. F. and Williams, D. B. 1999. Calreticulin functions in vitro as a molecular chaperone for both glycosylated and non‐glycosylated proteins. EMBO J.  18: 6718. Google Scholar Author notes Departments of 1Immunology and 2Biochemistry, University of Toronto, Toronto, Ontario M5S 1A8, Canada
TI - Mutant MHC class I molecules define interactions between components of the peptide‐loading complex
JO - International Immunology
DO - 10.1093/intimm/14.4.347
DA - 2002-04-01
UR - https://www.deepdyve.com/lp/oxford-university-press/mutant-mhc-class-i-molecules-define-interactions-between-components-of-sQOXOA70Ep
SP - 347
EP - 358
VL - 14
IS - 4
DP - DeepDyve
ER -