Two conformationally distinct and stable forms of Qa-1b, one strongly associated with β2-microglobulin (β2m) and the other associated with a novel molecule, gp44, were observed during immunochemical studies on the expression of Qa-1b molecules in mouse spleen cells. Both forms are efficiently processed and expressed at the cell surface. However, a large proportion of Qa-1b was found to be disulfide linked to gp44 without any detectable β2m. In TAP1-deficient mice, both forms undergo carbohydrate processing and are expressed on the cell surface, suggesting that they may traffic using a pathway not requiring a TAP association step. Consistent with this, size exclusion chromatography of newly synthesized class I molecules shows that high molecular mass complexes containing H-2Kk do not contain Qa-1b. Although Qa-1b can be stably expressed without β2m, there was no maturation of either form in cells from β2m-deficient mice where heavy chains were rapidly degraded. These results suggest that Qa-1b, like most other class I molecules, requires β2m for an initial folding step. However, β2m is not essential for subsequent processing of Qa-1b molecules.

Class I histocompatibility Ags alert the immune system to viruses by displaying peptides derived from virally encoded proteins on the surface of infected cells. Cells expressing class I molecules bearing viral epitopes are efficiently destroyed by host CTL. Newly synthesized class I heavy chains (Hcs) first associate with calnexin, an endoplasmic reticulum (ER)3-resident chaperone, and then bind β2-microglobulin (β2m) (1, 2). Stabilization of class I/β2m heterodimers is usually dependent upon acquisition of a suitable peptide, delivered by the TAP peptide transporter located in the ER (3, 4). In TAP-deficient cells where the peptide supply to class I molecules is severely reduced, many Hcs are unable to form stable structures and are targeted for degradation (5).

Peptides that enter the ER independently of TAPs, such as cleaved leader sequences, can induce assembly of class I molecules in TAP-deficient cells (6). It has also been shown that the expression of some class I molecules is intrinsically TAP independent (7, 8). The class I-like molecules CD1 and TL can both be expressed on the surface of cells lacking a functional TAP transporter, and there is increasing evidence that they may acquire antigenic epitopes in intracellular compartments other than the ER (9, 10).

Mouse Qa-1 is one of a group of class I molecules designated as nonclassical or class Ib histocompatibility Ags, and that elicits strong CTL responses (11). At least two other members of this group, CD1 and H-2 M3, are involved in the presentation of bacterial and mycobacterial Ags to T cells (12, 13). Qa-1b is expressed in many mouse tissues, associates with β2m, and can present peptide Ags to CTL (11). Previous studies have shown that some CTL can recognize Qa-1b on cells defective in TAP function (14).

Maturation of class I molecules has usually been studied using alloantibodies that recognize correctly folded heterotrimers of Hcs, β2m, and peptide. More recently, alternative conformations of class I molecules, including precursor forms as well as cell surface molecules, have been described using other Hc-specific Abs (15). We have produced rabbit antisera against unique peptide sequences present in the cytoplasmic domain of Qa-1b. These antisera do not cross-react with other mouse class I molecules, and recognize both correctly folded and conformationally altered Qa-1b molecules. By comparing the reactivity of alloantiserum with peptide-specific Abs, we have identified two forms of Qa-1b that differ in their association with β2m and with other components of the maturation pathway.

Alloantiserum specific for Qa-1b was produced by immunization of B6.Tlaa × A strain mice with cells from A.Tlab. Anti-KSFQ (anti-Qa-1b polyclonal) was produced by immunizing New Zealand White rabbits with the peptide KSFQKDAMLMF. Hybridoma cells producing monoclonal anti-Kb (34.2.12) and anti-Kk (16.3.1) were obtained from the American Type Culture Collection (Rockville, MD).

A genomic clone of H-2Ld carrying the regular H-2 promoter region was kindly provided by Dr. A. Mellor. The Qa-1b/Ld construct was made by introducing the α1 and α2 domains of Qa-1b into the Ld gene as described previously for the Qa-1b/Dd construct (16). The construct was cloned into pBR327 and the linearized DNA was used to transfect L cells. Cells were grown in RPMI 1640 (Life Technologies, Grand Island, NY) containing 10% FCS, 50 IU/ml penicillin, 50 IU/ml streptomycin, and 2 mM glutamine (complete medium) at 37°C in 5% CO2.

CBA mouse spleen cells were cultured for 40 to 50 h with 2 μg/ml of Con A in complete medium containing 1 mM 2-ME. For labeling, 2 × 107 cells were washed three times at room temperature in RPMI medium lacking methionine and cysteine and containing no additives. A total of 200 μCi of Pro-Mix, [35S](methionine + cysteine) (Amersham, Little Chalfont, U.K.) was added and cells were incubated for 10 min at 37°C. After labeling, cells were either placed on ice, or chased in complete medium for appropriate times.

Labeled cells were washed once in ice-cold isotonic TBS to remove radioisotopes, and lysed in 1 ml of PBS containing 1% Triton X-100 (Pierce, Chester, U.K.), proteinase inhibitors, and 10 mM iodoacetamide. After 10 min on ice, lysates were centrifuged at 20,000 × g in a Sigma (St. Louis, MO) refrigerated bench microfuge. Cleared lysates were then treated with protein A beads and unimmunized serum (mouse or rabbit) at 4°C overnight. Protein A beads preloaded with specific antiserum were rotated at 4°C with precleared lysates for 2 h, and then harvested at 100 × g. After extensive washing, immunoprecipitates were treated with SDS containing sample buffer at 95°C with or without 10 mM DTT. Samples were then subjected to electrophoresis on 12% polyacrylamide gels, treated with autoradiographic enhancers (Amplify, Amersham, U.K.), dried, and autoradiographed on Kodak BioMax MR film.

Cells, 2 × 107, were washed thoroughly in cold PBS and suspended in 200 μl of PBS on ice. A total of 5 μl of 1 mg/ml of lactoperoxidase solution was added followed by 5 μl of a dilute hydrogen peroxide solution (30 vol diluted 1:104). Labeling was started by adding 10 μl (1 mCi) of [125I]Na solution (Amersham) and removing the tube from the ice. The reaction was continued for 30 min, adding 5 μl of dilute hydrogen peroxide solution at 10-min intervals. Cells were then washed in PBS containing 1 mM sodium iodide and lysed. 125I-labeled samples were visualized using Kodak BioMax MS film using an appropriate intensifying screen.

Cells labeled with [35S]methionine for 5 min were diluted to 7 ml with warm complete RPMI medium in a 10-ml polypropylene tube, and 1 ml was removed and placed immediately in ice. This chase suspension was maintained at 37°C with occasional mixing, removing samples at 30-min intervals. Lysis and immunoprecipitation were performed as described above, and then immunoprecipitates were treated with Endo (endoglycosidase) H (Boehringer Mannheim, Mannheim, Germany) according to the manufacturer’s instructions before gel electrophoresis.

First, immunoprecipitates were treated with 20 μl of 1.0% SDS at 95°C for 5 min with or without 1 mM DTT. Then, 1 ml of PBS containing 1% Triton X-100 (Pierce) was added and the beads removed. The supernatant containing the released Ag was then subjected to a second round of immunoprecipitation using protein A beads preloaded with Ab. Samples were then subjected to gel electrophoresis under reducing conditions.

CBA spleen cell blasts were metabolically labeled for 20 min as described above and lysed in PBS containing proteinase inhibitors and 1% digitonin. Lysates were centrifuged at 20,000 × g for 15 min and precleared with unimmunized serum. Precleared lysates were loaded onto a Superose 6 size exclusion gel filtration column (Pharmacia, St. Albans, U.K.) equilibrated with PBS containing 1% digitonin and run at a flow rate of 0.25 ml/min. The column was calibrated using gel filtration protein standards (Bio-Rad, Hemel Hempstead, U.K.), thyroglobulin (670 kDa), bovine gamma globulin (158 kDa), and chicken OVA (44 kDa). One-milliliter fractions were collected and each fraction subjected to immunoprecipitation with protein A beads preloaded with specific antiserum/Ab at 4°C for 4 h. After three washes in digitonin-containing PBS buffer, samples were analyzed by gel electrophoresis under reducing conditions.

To visualize Qa-1b molecules, Con A-activated CBA mouse spleen cells (H-2k, Qa-1b) were either metabolically labeled with [35S]methionine or surface labeled with 125I. Cell lysates were treated either with an anti-Qa-1b alloantiserum or with a rabbit antiserum prepared against a Qa-1b cytoplasmic peptide motif (anti-KSFQ) and protein A beads. Immunoprecipitates were subjected to gel electrophoresis under reducing conditions and the results shown in Figure 1, A and B. Both reagents detect Qa-1b Hcs, but in very different amounts. This was not due to differences in Ab titer, since Ab was always present in excess. Interestingly, both reagents coprecipitated equal amounts of β2m. Similar results were obtained with metabolically labeled (Fig. 1,A) and surface-labeled (Fig. 1,B) cells. When lysates were extensively precleared with the alloantiserum, all of the β2m-associated Hcs were removed (Fig. 1,C). However, Hcs devoid of β2m were still detectable using the anti-KSFQ Ab (Fig. 1 C). To investigate whether anti-KSFQ Ab might cross-react with proteins other than Qa-1b, it was tested on lysates of [35S]methionine-labeled B10.BR (H-2k, Qa-1a) spleen cells. No 40- to 50-kDa material was immunoprecipitated from these lysates (data not shown), indicating that all of the labeled 40- to 50-kDa material detected with anti-KSFQ was associated with Qa-1b. These results suggested that CBA lysates contain two serologically distinguishable forms of Qa-1b Hcs. One is a stable Qa-1b2m heterodimer recognized both by the alloantiserum and by anti-KSFQ. There are also Qa-1b Hcs present that contain little or no β2m and are recognized only by anti-KSFQ. It also shows that a significant proportion of the Qa-1b does not form a stable association with β2m.

FIGURE 1.

Alloantiserum and anti-KSFQ define different forms of Qa-1b. A, [35S]methionine-labeled CBA lysates immunoprecipitated using alloantiserum or anti-KSFQ; B, 125I surface-labeled lysates treated similarly; and C, [35S]methionine-labeled lysates precleared three times sequentially with alloantiserum, then treated with anti-KSFQ. Hc denotes class I heavy chain.

FIGURE 1.

Alloantiserum and anti-KSFQ define different forms of Qa-1b. A, [35S]methionine-labeled CBA lysates immunoprecipitated using alloantiserum or anti-KSFQ; B, 125I surface-labeled lysates treated similarly; and C, [35S]methionine-labeled lysates precleared three times sequentially with alloantiserum, then treated with anti-KSFQ. Hc denotes class I heavy chain.

Close modal

To examine the intracellular trafficking of Qa-1b molecules, CBA spleen cell blasts were labeled for 10 min with [35S]methionine and chased for various times in medium containing excess unlabeled methionine. Precleared lysates were treated with anti-KSFQ and maturation of Qa-1b monitored by treatment with the glycosidase Endo H to monitor the conversion of immature, high-mannose, carbohydrates to complex forms. As shown in Figure 2,A, Endo H-sensitive Qa-1b molecules have a half-life of about 60 min before becoming resistant. The half-life of immature carbohydrates on H-2Kk molecules in the same cells was less than 30 min (Fig. 2,B). Samples from the pulse-chase analysis were also investigated by electrophoresis under nonreducing conditions. Without reduction of the samples, high m.w. forms of Qa-1b were found immediately after the labeling pulse, and were increased in quantity during the chase period (Fig. 2 C). Potential artifacts due to oxidation occurring in the lysate were minimized by including iodoacetamide in the lysis buffer. No high m.w. forms of H-2Kk molecules were found in the same lysate (data not shown). These results show that some Qa-1b molecules form disulfide-linked complexes early in the maturation pathway.

FIGURE 2.

Pulse-chase analysis of Qa-1b and H-2Kk molecules in CBA cells. A, Cells were labeled for 10 min with [35S]methionine, chased for the times indicated, and immunoprecipitated with anti-KSFQ. B, Lysates from A were immunoprecipitated with anti-H-2Kk. In A and B, samples were treated with Endo H before SDS-PAGE. Hc denotes class I heavy chain. EHR Hc denotes running position of Endo H-resistant Hcs. C, Samples from A were run unreduced. **, Denotes high molecular mass aggregates.

FIGURE 2.

Pulse-chase analysis of Qa-1b and H-2Kk molecules in CBA cells. A, Cells were labeled for 10 min with [35S]methionine, chased for the times indicated, and immunoprecipitated with anti-KSFQ. B, Lysates from A were immunoprecipitated with anti-H-2Kk. In A and B, samples were treated with Endo H before SDS-PAGE. Hc denotes class I heavy chain. EHR Hc denotes running position of Endo H-resistant Hcs. C, Samples from A were run unreduced. **, Denotes high molecular mass aggregates.

Close modal

CBA spleen cells were surface-labeled and lysates were treated with either the alloantiserum or with anti-KSFQ. To reduce the complexity of the gel patterns, N-linked carbohydrate side chains were removed by treatment of immunoprecipitates with endoglycosidase F (Endo F). Endo F treatment of labeled material isolated using the alloantiserum reduced the size of Qa-1b Hcs to a single band of 35 kDa, consistent with removal of the two N-linked carbohydrate chains. However, Endo F treatment of anti-KSFQ immunoprecipitates revealed the 35-kDa Hc band and an additional component of 37 kDa (Fig. 3,A). The 37-kDa component was not eliminated by extended treatment with excess Endo F and is not, therefore, a consequence of incomplete carbohydrate removal. To determine whether the 37-kDa polypeptide was synthesized internally, cells were labeled with [35S]methionine for 1 h and a similar immunoprecipitation conducted. Very little 37-kDa material was detectable after a 1-h label, but if cells were labeled continuously for 5 h with [35S]methionine, 37-kDa material was clearly detectable using anti-KSFQ (Fig. 3 B). This indicates that the 37-kDa Qa-1b-associated polypeptide is indeed synthesized by the cells but either has a slow turnover rate or a large intracellular pool size. Since the fully glycosylated cell surface form of the 37-kDa polypeptide is approximately 44 kDa, it has been called gp44.

FIGURE 3.

Coisolation of Qa-1b with gp44 using conformation-independent Abs. A, Immunoprecipitates of 125I-labeled Qa-1b molecules treated with Endo F. B, CBA cells were [35S]methionine labeled for 1 h (lane2) or 5 h (lanes1 and 3) and immunoprecipitated with anti-KSFQ. Immunoprecipitates in lanes2 and 3 were treated with Endo F. C, Immunoprecipitates of 125I-labeled Qa-1b/Ld molecules not treated with Endo F. Hc denotes class I heavy chain. Hc(EF) denotes Endo F-treated Hc. gp44(EF) denotes Endo F-treated gp44.

FIGURE 3.

Coisolation of Qa-1b with gp44 using conformation-independent Abs. A, Immunoprecipitates of 125I-labeled Qa-1b molecules treated with Endo F. B, CBA cells were [35S]methionine labeled for 1 h (lane2) or 5 h (lanes1 and 3) and immunoprecipitated with anti-KSFQ. Immunoprecipitates in lanes2 and 3 were treated with Endo F. C, Immunoprecipitates of 125I-labeled Qa-1b/Ld molecules not treated with Endo F. Hc denotes class I heavy chain. Hc(EF) denotes Endo F-treated Hc. gp44(EF) denotes Endo F-treated gp44.

Close modal

A hybrid class I molecule consisting of the α1 and α2 domains of Qa-1b linked to the α3 domain of H-2Ld was also examined. Transfected L cells expressing the hybrid molecule were surface labeled with 125I and lysates treated with either anti-Qa-1b alloantiserum or with the 28.14.8 Ab, which is specific for the α3 domain of H-2Ld. The hybrid molecule showed a weak β2m association similar to native Qa-1b. However, since the hybrid Qa-1b/Ld Hc carries an additional N-linked carbohydrate chain, it is 2 to 3 kDa larger than Qa-1b. As shown in Figure 3 C, the α3-specific Ab 28.14.8 reveals the hybrid class I Hc and gp44, which are clearly distinguishable without Endo F treatment. Consistent with previous data, the anti-Qa-1b alloantiserum detect only the hybrid Hc, indicating that Hcs associated with gp44 are not recognized by alloantibodies. This result shows that gp44 is detectable using Abs other than anti-KSFQ.

To investigate further the nature of the association between Qa-1b and gp44, a two-stage immunoprecipitation experiment was performed. Lysates of surface-labeled cells were treated with anti-KSFQ and immunoprecipitates washed as before. Labeled polypeptides were then released from the immunosorbent with a hot SDS solution, fresh lysis buffer added, and the eluted material treated a second time with anti-KSFQ on protein A beads. Finally, the samples were treated with Endo F and subjected to gel electrophoresis. As shown in Figure 4, when immunosorbents from the first immunoprecipitation were treated with SDS alone, Qa-1b and gp44 remain associated in the second round (Fig. 4, lane1). If, however, the reducing agent DTT is included in the first SDS retrieval step, gp44 does not reappear in the second immunoprecipitate (Fig. 4, lane2). This indicates that Qa-1b and gp44 are disulfide linked. It also shows that anti-KSFQ does not recognize gp44 directly but does so only by virtue of its association with Qa-1b.

FIGURE 4.

Qa-1b and gp44 are covalently linked. 125I-labeled CBA lysates were immunoprecipitated with alloantiserum (lane3) or anti-KSFQ (lane4). Alternatively, anti-KSFQ immunoprecipitates were treated with either SDS alone (lane1) or SDS and DTT (lane2) and subjected to a second anti-KSFQ immunoprecipitation step as described in Materials and Methods. All samples were Endo F treated. Hc denotes class I heavy chain.

FIGURE 4.

Qa-1b and gp44 are covalently linked. 125I-labeled CBA lysates were immunoprecipitated with alloantiserum (lane3) or anti-KSFQ (lane4). Alternatively, anti-KSFQ immunoprecipitates were treated with either SDS alone (lane1) or SDS and DTT (lane2) and subjected to a second anti-KSFQ immunoprecipitation step as described in Materials and Methods. All samples were Endo F treated. Hc denotes class I heavy chain.

Close modal

It has been previously reported that T cells can recognize Qa-1b in a TAP-dependent or a TAP-independent way (14). This implies that Qa-1b molecules can reach the cell surface in the absence of TAP molecules. To test this, activated spleen cells from TAP1-deficient mice (H-2b, Qa-1b) were surface labeled and lysates prepared as before. Using anti-KSFQ, Qa-1b was easily detectable by immunoprecipitation (Fig. 5,A, lane2). Surface forms were also detected using the alloantiserum (data not shown). In contrast, only low levels of H-2Kb were found in the same lysates (Fig. 5,A, lane4). Pulse-chase analysis showed that a significant proportion of Qa-1b molecules are processed to Endo H-resistant forms in TAP1-deficient cells, indicating that intracellular transport of Qa-1b occurs in the absence of TAP function (Fig. 5 B).

FIGURE 5.

Expression of Qa-1b in TAP1-deficient mice. A, Surface expression of Qa-1b and H-2Kb in 125I-labeled B10 and TAP1-deficient mice. B, Pulse-chase analysis of Qa-1b in TAP1-deficient mice. Lysates were immunoprecipitated with anti-KSFQ and samples were Endo H treated. Hc denotes class I heavy chain. EHR Hc denotes running position of Endo H-resistant Hcs.

FIGURE 5.

Expression of Qa-1b in TAP1-deficient mice. A, Surface expression of Qa-1b and H-2Kb in 125I-labeled B10 and TAP1-deficient mice. B, Pulse-chase analysis of Qa-1b in TAP1-deficient mice. Lysates were immunoprecipitated with anti-KSFQ and samples were Endo H treated. Hc denotes class I heavy chain. EHR Hc denotes running position of Endo H-resistant Hcs.

Close modal

It has been shown recently that class I molecules associate transiently in the ER with a protein complex consisting of the molecular chaperons calnexin and calreticulin, another molecule called tapasin, and the TAP1/TAP2 heterodimer, as a prelude to peptide loading (17). To investigate a possible physical association between Qa-1b and the TAP machinery, lysates of metabolically labeled CBA blasts were fractionated by size-exclusion chromatography in digitonin-containing buffer to preserve weak molecular associations. Fractions were then treated sequentially with anti-KSFQ and 16.3.1 to detect Qa-1b and H-2Kk, respectively. The results are shown in Figure 6. Using this methodology, we were able to detect class I molecules associated with different molecular mass ranges. Some H-2Kk was found in complexes with molecular masses above the 670-kDa marker, possibly representing molecules associated with the TAP machinery (Fig. 6,A). H-2Kk molecules were also found in complexes with molecular masses below the 158-kDa standard marker, probably representing assembled heterotrimeric complexes of Hc, β2m, and peptide. In contrast, Qa-1b was found in complexes with molecular masses in the range of 100- to 300-kDa only, and not in the high molecular mass complexes that contain H-2Kk (Fig. 6,B). The presence of material in the 300- to 400-kDa range in Figure 6 B indicates some variability in the composition of Qa-1b-containing complexes. In a separate experiment, anti-TAP1 immunoprecipitates, in digitonin-containing buffer, of metabolically labeled CBA lysates were eluted and reprecipitated with anti-KSFQ. No Qa-1b molecules were found (data not shown), consistent with their failure to strongly associate with TAPs.

FIGURE 6.

Gel filtration analysis of Qa-1b- and H-2Kk-containing complexes. Fractions 7 to 18 were immunoprecipitated using anti-H-2Kk (A) or anti-KSFQ (B) as described in Materials and Methods. Running positions of size markers are shown above each panel. Hc denotes class I heavy chain.

FIGURE 6.

Gel filtration analysis of Qa-1b- and H-2Kk-containing complexes. Fractions 7 to 18 were immunoprecipitated using anti-H-2Kk (A) or anti-KSFQ (B) as described in Materials and Methods. Running positions of size markers are shown above each panel. Hc denotes class I heavy chain.

Close modal

Our results show that β2m is not associated with a large proportion of Qa-1b Hcs in CBA cells. This suggested that Qa-1b expression might be independent of β2m. We therefore examined the maturation and cell surface expression of Qa-1b in β2m-deficient mice by pulse-chase analysis. We were unable to detect any Endo H-resistant forms of Qa-1b in β2m-deficient mice, and Endo H-sensitive Hcs disappear during the chase period (Fig. 7). Furthermore, no cell surface Qa-1b was detectable by surface labeling (data not shown). These results show that, without β2m, Qa-1b is rapidly degraded. They also show that, once the Hcs have undergone their initial folding step, β2m is no longer essential for subsequent maturation and stable cell surface expression.

FIGURE 7.

Pulse-chase analysis of Qa-1b in cells from β2m-deficient mice. [35S]methionine-labeled cells were chased for the indicated times and immunoprecipitated with anti-KSFQ. Samples were Endo H treated. Hc denotes class I heavy chain. EHR Hc denotes running position of Endo H-resistant Hcs.

FIGURE 7.

Pulse-chase analysis of Qa-1b in cells from β2m-deficient mice. [35S]methionine-labeled cells were chased for the indicated times and immunoprecipitated with anti-KSFQ. Samples were Endo H treated. Hc denotes class I heavy chain. EHR Hc denotes running position of Endo H-resistant Hcs.

Close modal

The stability of cell surface heterotrimeric complexes of class I Hcs, β2m, and peptide is critically dependent upon the affinity of the peptide bound. Release of peptide either during intracellular transport or at the cell surface results in loss of β2m and alteration of the Hc conformation. Most class I molecules bind a range of peptides with a spectrum of affinities. Inevitably, therefore, some cell surface Hcs will acquire an aberrant conformation. These forms are not readily detectable, since the majority of mAbs recognize only stable heterotrimeric forms. The availability of conformation-independent Abs, which recognize class I Hcs irrespective of their association with β2m and peptide, has allowed detailed studies on the trafficking and cell surface presentation of several class I molecules, including H-2Kb (15), H-2Db (18), H-2Ld (19), HLA-B27 (20), HLA-Cw1 (21), and Qa-1b (this study). The emerging picture is that conformational variants of class I molecules do exist and are stably expressed at the cell surface. The function of these variants remains unclear.

An interesting and special feature of Qa-1b is its ability to bind the nonamer peptide, AMAPRTLLL, derived from the leader sequences of H-2D and H-2L class I molecules, in preference to other peptides (11). Recently, peptide elution studies show AMAPRTLLL indeed to be the major peptide bound, with very few others detected (16, 22). In this study, we have shown that, as expected, association of Qa-1b with β2m is poor in CBA (H-2k) mice that lack AMAPRTLLL. We were, however, surprised to find that, despite poor β2m association, Qa-1b molecules do not undergo rapid proteolysis, but rather traffic normally and are stably expressed on the cell surface. These cell surface Qa-1b molecules, which lack β2m and did not react with our conformation-dependent alloantiserum, were detected with the conformation-independent peptide-specific antiserum (anti-KSFQ). This suggests that, if AMAPRTLLL is not present, Qa-1b adopts an alternative stable conformation.

It is widely accepted that trafficking and maturation of class I Hcs involves the recruitment of several accessory molecules. One of these is calnexin, an ER resident lectin-like chaperone, to which nascent Hcs bind cotranslationally and remain bound while they undergo correct folding upon acquisition of β2m. On subsequent binding to the TAP machinery and on binding peptides they are then released and bound for the cell surface. It should be noted that recent studies have shown that, in calnexin-deficient cells, rates of class I Hc maturation were not affected, suggesting that other ER molecular chaperons can perform the same job (23, 24). One possible explanation for the weak association between Qa-1b and β2m observed here is that Hcs are capable of folding correctly independently of β2m. This is not clearly the case, since Qa-1b Hcs are rapidly degraded in β2m-deficient mice, and cell surface labeling experiments showed them to be absent. Therefore, Qa-1b molecules appear to share the same β2m-dependent initial folding steps as other class I molecules, but subsequent β2m association appears not to be essential for their trafficking and maturation. An exception to the β2m-dependence rule is H-2Db, which is expressed on the surface of cells from β2m-deficient cells, although at low levels (18).

Previous studies have shown that, in cells defective in TAP expression and function, the rate of intracellular transport of class I molecules is reduced, possibly because calnexin remains associated for longer with Hc-β2m heterodimers (25). Our pulse-chase experiments show that Qa-1b becomes Endo H resistant 60 to 90 min after synthesis, much slower than H-2Kk molecules, which become resistant within 30 min. Similar rates of maturation and cell surface expression of Qa-1b were observed in both normal and TAP1-deficient mice. Cell surface expression of most class I molecules in TAP-deficient cells is generally low, since Hc-β2m dimers without peptide are unstable. However, Qa-1b molecules are abundant and stable on the cell surface in TAP1-deficient mice. In this respect, Qa-1b resembles the class Ib molecules TL and CD1 (7, 8). Earlier work has shown that CTL recognition of Qa-1b is not dependent on TAPs (14, 26). Paradoxically, many Qa-1-specific CTL recognize their targets in a TAP-dependent manner, possibly reflecting a requirement for the transporter in delivering peptide epitopes into the lumen of the ER. Our results, therefore, clearly distinguish the dual roles of TAP molecules in the assembly of class I molecules and in the delivery of peptides.

It is not unusual for free class I Hcs to form either homodimers or associate with additional proteins. Indeed, this may be a way to stabilize free Hcs that would otherwise be targeted for degradation. Wolf and Cook reported that Qa-1b Hcs associate with free H-2Ld Hc and with another unidentified molecule, Qsm, and in doing so exclude β2m (27). In their experiments, no additional molecules were detected in H-2k strains. In contrast, our results demonstrate another polypeptide, gp44, which associates with Qa-1b in H-2k and H-2b strains. Using biosynthetic labeling, gp44 labels poorly, and is therefore difficult to detect. Also, if carbohydrates are not removed, gp44 was difficult to distinguish from Qa-1b on SDS-PAGE due to the closeness of their polypeptide chain sizes. It was, in contrast, easy to distinguish Qa-1b/Ld Hc from gp44 since the extra carbohydrate on the chimeric class I Hc reduces substantially its SDS-PAGE mobility (Fig. 3 C). This experiment also demonstrates that the α3 domain of Qa-1b is unimportant for gp44 association.

Evidence that gp44 is a novel molecule comes from the following observations. First, gp44 associates with Qa-1b in mice deficient in the class II-associated invariant chain (data not shown). Second, other known accessory molecules in the TAP pathway, e.g., calreticulin (28), ectocalreticulin (28), and tapasin (29, 30) have higher molecular masses: 52 kDa, 62 kDa, and 47.5 kDa, respectively. And third, preliminary N-terminal sequence analysis of gp44 appears to exclude other known class I Hcs (H. Djaballah and P. J. Robinson, unpublished observations). It is not yet clear whether gp44 is membrane bound; further characterization is under way.

All class I molecules have four conserved cysteine residues that form two intrachain disulfide bonds, Cys101-Cys164 and Cys203-Cys259. The additional cysteine residue at position 114 of Qa-1b is located on one of the β strands at the base of the peptide-binding cleft, making it solvent accessible and able to form a disulfide bridge with free cysteine residues on gp44 (Fig. 8). A cysteine residue at this position of class I Hcs is unusual. Although some mouse class I molecules have a free cysteine residue at position 121, this is less likely to form disulfide bridges because of poor solvent accessibility. However, this may account for the formation of disulfide-linked homodimers observed previously (31). Cysteine residues are also present in the cytoplasmic domains of several class I Hcs, including Qa-1b. However, the reducing environment of the cytoplasm effectively rules out formation of disulfide bridges in vivo involving cytoplasmic cysteine residues. In addition, we find no evidence for dimers of Qa-1b, but detect larger complexes that form within minutes of Hc synthesis and contain gp44. One might predict that formation of such complexes may not allow a good contact between the Hcs and the TAP molecular complex. Our gel filtration studies showing that Qa-1b and H-2Kk are present in different molecular mass fractions is consistent with this hypothesis.

FIGURE 8.

Molecular images of the peptide-combining site of Qa-1b highlighting the position of Cys114. The main part of the figure shows a cartoon view of the α2 domain of the model of Qa-1b (this region is indicated by the gray shading in the inset, where the molecule is seen from the top). The positions of the serine residues at positions 143 and 147, which are well conserved as threonine and tryptophan in other class I molecules, are shown, as is the conserved disulfide bond between Cys101 and Cys164. A free cysteine residue is present at position 114, on the second β strand of the α2 domain and points up into the peptide binding groove. The figure was produced using MOLSCRIPT (32).

FIGURE 8.

Molecular images of the peptide-combining site of Qa-1b highlighting the position of Cys114. The main part of the figure shows a cartoon view of the α2 domain of the model of Qa-1b (this region is indicated by the gray shading in the inset, where the molecule is seen from the top). The positions of the serine residues at positions 143 and 147, which are well conserved as threonine and tryptophan in other class I molecules, are shown, as is the conserved disulfide bond between Cys101 and Cys164. A free cysteine residue is present at position 114, on the second β strand of the α2 domain and points up into the peptide binding groove. The figure was produced using MOLSCRIPT (32).

Close modal

The scheme we propose for the maturation of Qa-1b molecules is shown in Figure 9. It is based upon a recent maturation scheme for class I molecules proposed by Sadasivan et al. (17). Hcs associate cotranslationally with calnexin in the ER, where they bind β2m and commence folding. Next, calnexin is replaced by calreticulin, as is the case for most class Ia molecules, forming a quasi-stable complex with the chaperone. Here, similarity with the class Ia maturation pathway ends. Rather than associating with the TAP machinery, a decision is now made as to whether Hcs bind peptide or gp44. If peptide binds, β2m association is stabilized and the heterotrimeric complex is detectable by alloantiserum. If, instead, gp44 forms a disulfide bridge with Qa-1b, possibly via cysteine 114, β2m is released and the complex is no longer detectable using our alloantiserum. In either case, calreticulin is recycled while the Hcs undergo further maturation and transport to the cell surface. Additional experiments will be required to determine the precise order of events, and establish whether other class Ib molecules may follow a similar maturation pathway.

FIGURE 9.

Proposed scheme for Qa-1b maturation and possible origins of the two cell surface forms.

FIGURE 9.

Proposed scheme for Qa-1b maturation and possible origins of the two cell surface forms.

Close modal

The results described above, obtained using a variety of immunochemical and biochemical techniques, shed some light on the trafficking and maturation of Qa-1b and suggest that stable cell surface expression of Qa-1b Hcs lacking β2m can be accounted for by covalent association with a novel molecule gp44.

We thank Drs. O. Smithies and M. Merkenschlager for β2m-deficient mice; Drs. D. Kioussis, O. Williams, and S. Tonegawa for TAP1-deficient mice; Drs. C. Benoist and D. Mathis for Ii-deficient mice; and John Trowsdale for anti-TAP1 Abs. We also thank Drs. Jim Kaufman and Danny Altman for critically reading the manuscript.

1

H.D. is supported by the Leukaemia Research Fund.

3

Abbreviations used in this paper: ER, endoplasmic reticulum; Hc, heavy chain; β2m, β2-microglobulin; Endo, endoglycosidase.

1
York, I., K. L. Rock.
1996
. Antigen processing and presentation by the class I major histocompatibility complex.
Annu. Rev. Immunol.
14
:
369
2
Jackson, M. R., P. A. Peterson.
1993
. Assembly and intracellular transport of MHC class I molecules.
Annu. Rev. Cell Biol.
9
:
207
3
Suh, W. K., E. K. Mitchell, Y. Yang, P. A. Peterson, G. L. Waneck, D. B. Williams.
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
4
Ortmann, B., M. J. Androlewicz, P. Cresswell.
1994
. MHC class I/β2-microglobulin complexes associate with TAP transporters before peptide binding.
Nature
368
:
864
5
Van Kaer, L., P. G. Ashton Rickardt, H. L. Ploegh, S. Tonegawa.
1992
. TAP1 mutant mice are deficient in antigen presentation, surface class I molecules, and CD48+ T cells.
Cell
71
:
1205
6
Wei, M. L., P. Cresswell.
1992
. HLA-A2 molecules in an antigen processing mutant cell contain signal sequence derived peptides.
Nature
356
:
443
7
Holcombe, H. R., A. R. Castaño, H. Cheroutre, M. Teitell, J. K. Maher, P. A. Peterson, M. Kronenberg.
1995
. Nonclassical behavior of the thymus leukemia antigen: peptide transporter independent expression of a nonclassical class I molecule.
J. Exp. Med.
181
:
1433
8
Brutkiewicz, R. R., J. R. Bennink, J. W. Yewdell, A. Bendelac.
1995
. TAP independent, β2-microglobulin dependent surface expression of functional mouse CD1.1.
J. Exp. Med.
182
:
1913
9
Prigozy, T. I., P. A. Sieling, D. Clemens, P. L. Stewart, S. M. Behar, S. A. Porcelli, M. B. Brenner, R. L. Modlin, M. Kronenberg.
1997
. The mannose receptor delivers lipoglycan antigens to endosomes for presentation to T cells by CD1b molecules.
Immunity
6
:
187
10
Sugita, M., R. M. Jackman, E. Van Donselaar, S. M. Behar, R. A. Rogers, P. J. Peters, M. B. Brenner, S. A. Porcelli.
1996
. Cytoplasmic tail dependent localization of CD1b antigen presenting molecules to MIICs.
Science
273
:
349
11
Soloski, M. J., A. DeCloux, C. J. Aldrich, J. Forman.
1995
. Structural and functional characteristics of the class Ib molecule, Qa-1.
Immunol. Rev.
147
:
67
12
Moody, D. B., M. Sugita, P. J. Peters, M. B. Brenner, S. A. Porcelli.
1996
. The CD1 restricted T-cell response to mycobacteria.
Res. Immunol.
147
:
550
13
Bouwer, H. G., K. F. Lindahl, J. R. Baldridge, C. R. Wagner, R. A. Barry, D. J. Hinrichs.
1994
. An H2-T MHC class Ib molecule presents Listeriamonocytogenes-derived antigen to immune CD8+ cytotoxic T cells.
J. Immunol.
152
:
5352
14
Aldrich, C. J., R. Waltrip, E. Hermel, M. Attaya, K. F. Lindahl, J. J. Monaco, J. Forman.
1992
. T cell recognition of Qa-1b antigens on cells lacking a functional TAP2 transporter.
J. Immunol.
149
:
3773
15
Machold, R. P., H. L. Ploegh.
1996
. Intermediates in the assembly and degradation of class I major histocompatibility complex (MHC) molecules probed with free heavy chain specific monoclonal antibodies.
J. Exp. Med.
184
:
2251
16
Cotterill, L. A., H. J. Stauss, M. M. Millrain, D. J. Pappin, D. Rahman, B. Canas, P. Chandler, A. Stackpoole, E. Simpson, P. J. Robinson, P. J. Dyson.
1997
. Qa-1 interaction and T cell recognition of the Qa-1 determinant modifier peptide.
Eur. J. Immunol.
27
:
2123
17
Sadasivan, B., P. J. Lehner, B. Ortmann, T. Spies, P. Cresswell.
1996
. Roles for calreticulin and a novel glycoprotein, tapasin, in the interaction of MHC class I molecules with TAP.
Immunity
5
:
103
18
Allen, H., J. Fraser, D. Flyer, S. Calvin, R. A. Flavell.
1986
. β2-microglobulin is not required for cell surface expression of the murine class I histocompatibility antigen H-2Db or of a truncated H-2Db.
Proc. Natl. Acad. Sci. USA
83
:
7447
19
Smith, J. D., N. B. Myers, J. Gorka, T. H. Hansen.
1993
. Model for the in vivo assembly of nascent Ld class I molecules and for the expression of unfolded Ld molecules at the cell surface.
J. Exp. Med.
178
:
2035
20
Carreno, B. M., T. H. Hansen.
1994
. Exogenous peptide ligand influences the expression and half-life of free HLA class I heavy chains ubiquitously detected at the cell surface.
Eur. J. Immunol.
24
:
1285
21
Setini, A., A. Beretta, C. De-Santis, R. Meneveri, A. Martayan, M. C. Mazzilli, C. Appella, A. G. Siccardi, P. G. Natali, P. Giacomini.
1996
. Distinctive features of the alpha 1-domain alpha helix of HLA-C heavy chains free of β2-microglobulin.
Hum. Immunol.
46
:
69
22
DeCloux, A., A. S. Woods, R. J. Cotter, M. J. Soloski, J. Forman.
1997
. Dominance of a single peptide bound to the class Ib molecule, Qa-1b.
J. Immunol.
158
:
2183
23
Sadasivan, B. K., A. Cariappa, G. L. Waneck, P. Cresswell.
1995
. Assembly, peptide loading, and transport of MHC class I molecules in a calnexin negative cell line.
Cold Spring Harbor Symp. Quant. Biol.
60
:
267
24
Scott, J. E., J. R. Dawson.
1995
. MHC class I expression and transport in a calnexin-deficient cell line.
J. Immunol.
155
:
143
25
Degen, E., M. F. Cohen Doyle, D. B. Williams.
1992
. Efficient dissociation of the p88 chaperone from major histocompatibility complex class I molecules requires both β2-microglobulin and peptide.
J. Exp. Med.
175
:
1653
26
Hermel, E., E. Grigorenko, K. F. Lindahl.
1991
. Expression of medial class I histocompatibility antigens on RMA-S mutant cells.
Int. Immunol.
3
:
407
27
Wolf, P. R., R. G. Cook.
1995
. The class Ib molecule Qa-1b forms heterodimers with H-2Ld and a novel 50 kDa glycoprotein encoded centromeric to I-E beta.
J. Exp. Med.
181
:
657
28
Zhu, Q., P. Zelinka, T. White, M. L. Tanzer.
1997
. Calreticulin-integrin bidirectional signaling complex.
Biochem. Biophys. Res. Commun.
232
:
354
29
Li, S., H-O. Sjogren, U. Hellman, R. F. Pettersson, P. Wang.
1997
. Cloning and functional characterization of a subunit of the transporter associated with antigen processing.
Proc. Natl. Acad. Sci. USA
94
:
8708
30
Ortmann, B., J. Coperman, P. J. Lehner, B. Sadasivan, J. A. Herberg, A. G. Grandea, S. R. Ridell, R. Tampe, T. Spies, J. Trowsdale, P. Cresswell.
1997
. A critical role for tapasin in the assembly and function of multimeric MHC class I-TAP complexes.
Science
277
:
1306
31
Capps, G. G., B. E. Robinson, K. D. Lewis, M. C. Zuniga.
1993
. In vivo dimeric association of class I MHC heavy chains: possible relevance to class I-β2-microglobulin dissociation.
J. Immunol.
151
:
159
32
Kraulis, P. J..
1991
. MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures.
J. Appl. Crystal.
24
:
946