Ag-specific T cell recognition is mediated through direct interaction of clonotypic TCRs with complexes formed between Ag-presenting molecules and their bound ligands. Although characterized in substantial detail for class I and class II MHC encoded molecules, the molecular interactions responsible for TCR recognition of the CD1 lipid and glycolipid Ag-presenting molecules are not yet well understood. Using a panel of epitope-specific Abs and site-specific mutants of the CD1b molecule, we showed that TCR interactions occur on the membrane distal aspects of the CD1b molecule over the α1 and α2 domain helices. The location of residues on CD1b important for this interaction suggested that TCRs bind in a diagonal orientation relative to the longitudinal axes of the α helices. The data point to a model in which TCR interaction extends over the opening of the putative Ag-binding groove, making multiple direct contacts with both α helices and bound Ag. Although reminiscent of TCR interaction with MHC class I, our data also pointed to significant differences between the TCR interactions with CD1 and MHC encoded Ag-presenting molecules, indicating that Ag receptor binding must be modified to accommodate the unique molecular structure of the CD1b molecule and the unusual Ags it presents.

T cells are a key component of the vertebrate immune system, playing a critical role in host defense against many microbial pathogens. The molecular mechanisms involved in T cell recognition of conventional peptide Ags presented by MHC molecules are now understood in great detail (1, 2). More recently, another pathway for Ag presentation has been identified which is independent of MHC-encoded Ag-presenting molecules. This pathway activates T cells by presenting foreign and, potentially, also self Ags through MHC-like CD1 proteins (3, 4). This mechanism for Ag recognition has also been shown to be mediated by specific T cell Ag receptors (5, 6) but appears to present predominantly or exclusively lipid and glycolipid Ags rather than peptides to T cells (7, 8, 9, 10, 11, 12).

Human CD1 is a family of five nonpolymorphic β2-microglobulin-associated proteins that are encoded outside the MHC locus but have structural similarity to MHC class I proteins (3, 13, 14). Substantial insight into the structure of CD1 proteins was provided by the crystal structure of the mouse CD1d1 protein, the mouse homologue of the human CD1d molecule (15). This work showed striking overall structural similarity between CD1 and classical MHC class I molecules, despite the relatively limited amino acid sequence homology between these two protein families. The α1 and α2 domains in both families are organized to form a similar Ag-binding superdomain structure composed of two regions containing anti-parallel α helices supported by an underlying β-pleated sheet. In addition, MHC class I and CD1 each contain a membrane proximal Ig-like α3 domain, which in both cases is responsible for the majority of noncovalent interactions with β2-microglobulin.

Despite the overall similarities between MHC class I and CD1 molecules, these molecules show significant structural differences consistent with their roles in the binding and presentation of separate chemical classes of antigenic ligands. The most significant structural differences are in the membrane distal α1 and α2 domains, where the putative CD1 ligand-binding site is deeper and substantially more hydrophobic over its surface than are the peptide-binding grooves of MHC class I and II molecules (15, 16). The differences in the shape and physicochemical characteristics of the CD1 molecule compared with MHC-encoded Ag-presenting molecules are consistent with the proposed function of CD1, making the molecule particularly well suited for binding and presentation of lipid and glycolipid Ags.

Previous studies with TCR reconstitution in T cell tumor lines to confer CD1-restricted Ag recognition provide strong direct evidence that such responses are mediated by the clonotypic Ag receptors (TCRs) of T cells (6). TCR interactions with MHC class I and class II molecules have been extensively analyzed (1, 2), and the crystal structures of trimolecular complexes formed among αβ TCRs, MHC class I/peptide complexes (17, 18, 19, 20, 21), and MHC class II/peptide complexes (22) have recently been solved. These structures reveal that the TCR recognizes a surface on the MHC class I and class II molecules formed by the two membrane distal domains of each molecule (α1 and α2 for MHC class I and α1 and β1 for MHC class II). As was correctly predicted by mutagenesis studies (23), TCR interaction with MHC class I molecules occurs such that the long axis of the TCR footprint forms an angle of ∼45–50 degrees relative to the longitudinal axis of the bound peptide Ag. In this orientation, the TCR α-chain is positioned such that it lies mainly over the α2 helix, with the TCR β-chain predominantly over the α1 helix and the CDR3 loops of each chain approaching the center of the molecule in close apposition to exposed portions of a bound peptide. Mutagenesis studies suggested that TCR interactions with MHC class II molecules might display a similar orientation (24, 25). However, recent data from x-ray crystallographic analysis of the trimolecular complex formed between the D10 TCR, a 16-residue foreign peptide Ag, and the I-Ak molecule indicate that the TCR orientation on MHC class II molecules is rotated counterclockwise relative to its position on MHC class I. This creates what has been described as an orthogonal rather than a diagonal orientation of the TCR on the face of the Ag-presenting molecule (22).

Largely on the basis of studies of the influence of Ag structure on recognition by CD1-restricted T cell lines, we have previously proposed a model for Ag recognition in this system that involves the direct contact of TCRs with the hydrophilic head groups of glycolipid and lipid Ags bound in the CD1 groove (4, 11). However, it has not been determined whether this recognition involves contact of the TCR with only the head group of the Ag or the formation of a more extensive interface among the TCR, the bound Ag, and the surface of the CD1 molecule. To address this question, we have studied the effects of mutations on the surface of the membrane distal domains of CD1b on TCR reactivity. Here we describe the effects of point mutations of exposed residues on the membrane distal face of CD1b on Ag-specific responses by CD1b-restricted human T cell lines. Our results indicated that multiple molecular interactions occur between the TCRs of CD1b-restricted T cells and the surface formed by the helices of the α1 and α2 domains of CD1b. Furthermore, the results were consistent with a common molecular footprint for the interaction of CD1b with the TCRs of T cells specific for a variety of different lipid and glycolipid Ags and suggested more extensive contacts between the TCR and the α1 domain of CD1b than has been found for TCR/MHC class I interactions. These findings support a model for TCR/CD1 interactions in which many of the general features of TCR/MHC complexes have been maintained, while also indicating unique features that may be related to the ability of CD1 proteins to present bound lipid or glycolipid Ags to T cells.

The cell lines 293T and T2 were provided by Drs. Lloyd Klickstein and Peter Cresswell, respectively, and have been previously described (26, 27). The derivation and characterization of the CD1b-restricted, mycobacterial lipid Ag-specific T cell lines DN1 and LDN5, and of the CD1b-restricted ganglioside-reactive T cell clones GG123B and GG33A, have been previously described in detail (11, 12, 28). The CD1b-specific autoreactive Mt2.21 T cell clone was derived from the CD4-depleted PBMC of a randomly selected normal blood donor by in vitro stimulation with CD1+ monocyte-derived dendritic cells (DC4; produced as described below) in T cell medium (TCM) (8) containing 10 μg/ml of a detergent extract from Mycobacterium tuberculosis (29). These cells were subsequently cloned by limiting dilution and expanded by PHA stimulation. The phenotype of the Mt2.21 clone was TCR αβ+ TCR γδ, CD4, CD8α+, and CD8β, and it proliferated in the absence of any exogenous foreign Ag when cocultured with monocyte-derived DCs that expressed CD1b. Further characterization, which will be presented in detail elsewhere (M. Vincent and M. Brenner, manuscript in preparation), showed that the responses of Mt2.21 were strongly inhibited by mAbs specific for the CD1b protein and that Mt2.21 was specific for CD1b+ target cells in cytolytic assays.

mAbs were either used as mouse ascites fluids or purified from culture supernatants by protein G affinity column chromatography (Pharmacia, Uppsala, Sweden). Previously unpublished mAbs BCD1b1, BCD1b5, and BCD1b6 were produced from mice immunized with CD1+ human monocyte-derived DCs as previously described (30). The specificities of anti-CD1 Abs were confirmed by FACS analysis of CD1a-, CD1b-, CD1c-, and CD1d-transfected C1R cells (8, 28). Abs tested included BCD1b1 (IgG1, anti-CD1b; S. M. Behar and S. A. Porcelli, unpublished experiments), BCD1b2 (IgG1, anti-CD1b (31)), BCD1b3 (IgG1, anti-CD1b (30)), BCD1b5 (IgG1, anti-CD1b; S. M. Behar and S. A. Porcelli, unpublished experiments), BCD1b6 (IgG1, anti-CD1b; S. M. Behar and S. A. Porcelli, unpublished experiments), WM-25 (IgG1, anti-CD1b (32)), 7C4 (IgG1, anti-CD1b, (33)), NuT2 (IgG1, anti-CD1b (34)), and 4A7.6.5 (IgG2a, anti-CD1b (35)). Nonbinding isotype-matched controls were ascites fluids or purified Igs generated from mouse myeloma cell lines P3X63Ag8 (P3; IgG1) and RPC5.4 (IgG2a), both obtained from the American Type Culture Collection (ATCC; Manassas, VA). The anti-transferrin receptor mAb 5E9 (IgG1; ATCC) was used as a cell-binding control Ab. Alexa-488 and CY2-labeled Abs were produced from protein G column-purified culture supernatants (BCD1b2, BCD1b3, and NuT2) or IgG purified from ascites (BCD1b5) according to the manufacturers’ instructions (Alexa-488, Molecular Probes, Eugene, OR; CY2, Amersham Life Sciences, Pittsburgh, PA). Flow cytometry was performed as described (36), with gating on viable cells according to forward and side scatter and exclusion of propidium iodide-stained cells.

Mutant CD1b cDNAs were generated with the pAlterMax mutagenesis kit (Promega, Madison, WI.) following the manufacturer’s instructions. CD1b cDNA cloned into the XbaI site of the pAltermax plasmid (Promega) was used as the template for mutagenesis reactions. Variations from the standard protocol included the use of the R408 helper phage to produce a single stranded DNA template (as described in the appendix to the manufacturer’s instructions (Promega)) and the addition of 0.01 U/μl of DpnI (New England Biolabs, Beverly, MA) to mutagenesis reactions to degrade residual native double-stranded templates. The following oligonucleotides were used, each encoding a mutation of a single codon as indicated by the standard system of nomenclature consisting of the single letter code for the amino acid of the wild-type CD1b, followed by the residue position number in the mature protein, followed by the single letter code for the amino acid that was substituted by the mutation (e.g., alanine in most cases): S59A: AGCAACCTCCTTATCAGCAAAGTTACCTTTAGACC; K61A: CTCAGCAACCTCCGCATCACTAAAGTTACC; E65A: CTCCTCTAACGCAGCAACCTCC; E68A: CTCGGAATATCGCCTCTAACTC; R71A: CCAAAGATGTAGACTGCGAATATCTCCTC; Y73F: GAATCCAAAGATGAAGACTCGGAATATC; R79A: CAAAGTCTTGTACTTCTGCAGCGAATCCAAAG; E80A: AAAGTCTTGTACTGCTCGAGC; E80K: AAGTCTTGTACTTTTCGAGCGA; D83A: ACCGGCAAAGGCTTGTACTTC; D83K: ATCACCGGCAAATTTTTGTACTTCTCG; D87A: CATCTGGAAAGCACCGGCAAA; D87K: TTTCATCTGGAATTTACCGGCAAAGTC; K143A: AGTGCACAGAATGCCTGTGCCCTGC; Q150A: GATACCTTGATATGCTATGATTAGTGC; Y151F: GATACCTTGAAATTGTATGATTA; Q152A: TTCCATGATACCTGCATATTGTATGATTAG; E156A: TCTCACAGTTGCCATGATACC; T157A: GAATTCTCACAGCTTCCATGATACC; E164A: GGGGCAGGTTGCATAGAGGAG; T165A: TCGGGGGCAGGCTTCATAGAG; R168A: GCCCAAGAGATATGCGG GGCAGGTTTC.

Mutant constructs were screened by DNA sequencing both for the presence of the desired mutations and for the absence of random secondary mutations.

Transient transfectants of 293T cells or stable transfectants of T2 cells were generated by electroporation of wild-type (WT) or mutated CD1b cDNA constructs. Cultures of 293T cells were grown at 0.5 × 106 cells/plate in 175-cm2 plates in RPMI 1640 supplemented with 10% FCS (RPMI/FCS). After 48 h in culture cells were released using trypsin-EDTA solution (Life Technologies, Gaithersburg, MD) and washed twice in RPMI 1640. Electroporations were conducted with 107 cells in 400 μl RPMI 1640 in 0.4-cm electroporation cuvettes (Bio-Rad Laboratories, Hercules, CA), using a single discharge (250 V, 960 μF) delivered by a Gene Pulser Apparatus (Bio-Rad). Mutant CD1b constructs were used at 35 μg/ml for electroporations, and WT transfectants were produced in parallel using WT CD1b pAltermax plasmids in amounts varying between 7.5 and 75 μg/ml per transfection to yield a range of different expression levels.

T2 cell transfectants were produced by the same protocol except that the pGKpuro plasmid, containing a puromycin resistance gene, was cotransfected at a 1:50 molar ratio with the CD1b pAlterMax plasmid. After transfection, cells were incubated in RPMI/FCS for 48 h before the addition of 250 ng/ml puromycin (Sigma, St. Louis, MO) for an additional 48 h, collected by centrifugation, and cultured in RPMI/FCS without puromycin. After 1 wk, the cells were stained for 1 h at 4°C with 1 μg/ml of the anti-CD1b mAb 4A7.6.5 and positively selected using goat anti-mouse IgG Dynabeads M450 (Dynal, Olso, Norway) according to the manufacturer’s instructions. Positive selection was repeated at weekly intervals for two additional rounds of selection before cloning. T2 cell transfectants were cloned using the autoclone module of an EPICS flow cytometer (Coulter Epics Elite, Hialeah, FL) per the manufacturer’s instructions. In brief, cells were stained with CY2-conjugated mAb BCD1b3 at 1 μg/ml for 1 h at 4°C. They were then analyzed by FACS, selected by level of CD1b expression, and sorted at 1 cell/well into 96-well round-bottom-well microtiter plates. Each clone was then expanded in the presence of 2000 irradiated (7500 rad) T2 feeder cells per well for 2 wk and reanalyzed for surface CD1b expression by FACS. Those clones expressing CD1b molecules at levels with mean fluorescent intensities (MFIs) between channels 100 and 200 were retained for functional studies in Ag presentation assays.

Monocytes were isolated from leukocyte concentrates of normal donors by plastic adherence (37) and incubated in RPMI/FCS medium containing 100 U/ml GM-CSF (Immunex, Seattle, WA) and 200 U/ml IL-4 (Peprotech, Norwood, MA) to induce differentiation into DCs and the expression of CD1 molecules. Specific activities of cytokine preparations were determined using the GM-CSF- and IL-4-responsive M07e cell line (kindly provided by Genetics Institute, Cambridge, MA). Units reported here are M07e units (i.e., 1 U was defined as the amount of cytokine giving 50% maximal proliferation at 48 h of 20,000 M07e cells in a volume of 0.2 ml RPMI/FCS medium). Monocyte-derived DCs were harvested after 48 h culture in medium with GM-CSF plus IL-4, analyzed by FACS to confirm expression of CD1b, and subsequently used in T cell assays.

Proliferation assays were conducted in 96-well flat-bottom microtiter plates in 200 μl/well TCM. The CD48 T cell line DN1 (CD1b restricted and M. tuberculosis mycolic acid specific) was cultured with irradiated (5000 rad) monocyte-derived DCs (5 × 104 of each cell type) in 0.2 ml TCM with or without 25 μg/ml M. tuberculosis mycolic acids (Sigma). T cell proliferation assays were performed in the presence or absence of CD1b-specific mAbs (25 μg/ml purified IgG or a 1/100 dilution of ascites fluid) as noted in the figure legends. Control Abs included P3 (IgG1; nonbinding negative control), RPC5.4 (IgG2a; nonbinding negative control) and mAb 5E9 (IgG1; cell binding negative control). All mAb preparations used had no significant inhibitory effects on the proliferation of PHA (Difco/Becton Dickinson, Sparks, MD)-stimulated PBMC cultures, thus ruling out nonspecific inhibitory effects.

T cell activation by CD1b WT and mutant transfectants was assessed by ELISA for IFN-γ release. Assays were performed using 50,000 each of APCs (T2 or 293T transfectants) and T cells per well in 96-well flat-bottom plates. All transfectants were analyzed by FACS within 24 h of each assay to determine surface levels of CD1b expression using mAb BCD1b3, which was shown in preliminary studies to bind to an epitope of CD1b that was not disrupted by any of the mutations studied (Table I and unpublished data). Transient transfectants with MFI values of <40, which were shown in pilot studies using WT CD1b constructs to stimulate weak or inconsistent T cell responses when used as APCs, were not included in functional studies. Because FACS analysis of transient transfectants showed that CD1b expression was detectable within 24 h of electroporation and peaked within 48 to 96 h posttransfection, T cell assays with transient transfectants were initiated 48 h after transfection.

Table I.

Staining patterns of Absa

MutantAb
BCD1b1BCD1b2BCD1b3BCD1b5BCD1b6WM25NuT27C44A7.6.5
WT 
S59A − 
K61A − 
E65A − 
E68A 
R71A 
Y73F 
R79A 
E80A 
E80K 
D83A 
D83K 
D87A 
D87K 
K143A 
L147A 
Q150A 
Y151F 
Q152A 
E156A 
T157A − 
T165A 
R168A 
MutantAb
BCD1b1BCD1b2BCD1b3BCD1b5BCD1b6WM25NuT27C44A7.6.5
WT 
S59A − 
K61A − 
E65A − 
E68A 
R71A 
Y73F 
R79A 
E80A 
E80K 
D83A 
D83K 
D87A 
D87K 
K143A 
L147A 
Q150A 
Y151F 
Q152A 
E156A 
T157A − 
T165A 
R168A 
a

Plasmids containing cDNA constructs were expressed in 293T and T2 cell lines, and surface expression of CD1b was determined by FACS analysis of mAb staining with each of the Abs listed. Abs P3 (IgG1, nonbinding), RPC5.4 (IgG2a, nonbinding), and 5E9 (IgG1, anti-transferrin receptor) were included as negative and positive controls (data not shown). Staining levels similar to WT expression (log MFI >50% of log MFI for WT CD1b) are indicated as “+”. Those Abs that demonstrated no appreciable staining by FACS are listed as “−”.

Optimal conditions for T cell activation in the IFN-γ release assay were established in preliminary experiments for each T cell line studied. For some T cell lines, addition of PMA was required. When used, PMA was titrated between 1 and 4 ng/ml to give maximal Ag-specific responses without an increase in background IFN-γ production. For T cell lines reactive to exogenous lipid Ags, Ags were used at a concentration that was determined in dose titration experiments to be the minimal concentration required to stimulate peak IFN-γ production. This was 50 ng/ml glucose monomycolate (GMM, purified from Mycobacterium phlei as previously described (11)) for LDN5, 30 μg/ml mycolic acids (Sigma) for DN1, and 30 μg/ml GM1 (Sigma) for GG33A and GG123B. T cell assays were incubated at 37°C with 5% CO2 for 48 h, at which time culture supernatants were harvested and analyzed by ELISA for IFN-γ production as previously described (38).

As previously described, a three-dimensional model of the CD1b molecule based on the known murine CD1d crystal structure has been generated using a standard homology model building approach (6). Neither insertions nor deletions in either α helix were required to align the sequences in the homology match supporting this model, thus suggesting a high likelihood of accuracy within these regions. Using this model, we identified the positions of polar amino acids on the α1 and α2 helices with side chains that were predicted to be solvent exposed and oriented away from the Ag-binding groove (Fig. 1). Selected amino acids were individually mutated using in vitro site-directed mutagenesis, producing a total of 22 mutations distributed over 19 different positions. Nineteen of these were conservative mutations, substituting an alanine for a hydrophilic amino acid or phenylalanine for tyrosine, and three were charge reversal mutations, substituting a basic amino acid for an acidic amino acid (Table I).

FIGURE 1.

Positions of residues mutated on CD1b and location of mutants affecting mAb binding. Amino acid residues selected for mutation are indicated by the circles superimposed on the ribbon diagram (a list of all individual mutations produced is included in Table I). Circles labeled “N” and “B” indicate the locations of amino acid mutations that caused loss of binding of the NuT2 or BCD1b6 Abs, respectively. CD1b expression on cells transfected by mutants containing an alanine substitution at position 164 (∗) were not well detected by any anti-CD1b Abs tested, most likely reflecting a significant defect in folding or expression of this particular mutant. The remaining open circles indicate the location of mutations that did not effect binding of any mAb tested.

FIGURE 1.

Positions of residues mutated on CD1b and location of mutants affecting mAb binding. Amino acid residues selected for mutation are indicated by the circles superimposed on the ribbon diagram (a list of all individual mutations produced is included in Table I). Circles labeled “N” and “B” indicate the locations of amino acid mutations that caused loss of binding of the NuT2 or BCD1b6 Abs, respectively. CD1b expression on cells transfected by mutants containing an alanine substitution at position 164 (∗) were not well detected by any anti-CD1b Abs tested, most likely reflecting a significant defect in folding or expression of this particular mutant. The remaining open circles indicate the location of mutations that did not effect binding of any mAb tested.

Close modal

Point mutations were produced in cDNA constructs and subsequently expressed in either epithelial (293T) or lymphoblastoid (T2) cell lines. To ensure that mutant CD1b proteins were properly folded and expressed, we analyzed each mutant for binding to a panel of nine anti-CD1b mAbs. Of the 22 mutants, all but 1 (E164A) were expressed on the cell surface at levels comparable with those obtained with transfection of equivalent amounts of WT CD1b cDNA, based on binding of this panel of mAbs. In multiple experiments, the relative expression levels of E164A were markedly lower than those of WT CD1b (data not shown), and this mutant was not further analyzed. All of the other mutants were recognized by at least eight of the nine mAbs, suggesting that the local and global conformational properties were preserved in the mutated CD1b molecules (Table I).

Four of the 22 site-specific mutations inhibited Ab binding by 1 of the 9 mAbs tested. Three of these (S59A, K61A, and E65A) were predicted to lie on solvent-exposed areas of adjacent loops on the amino-terminal side of the α1 helix (Fig. 1). Mutations at these positions markedly reduced binding of the NuT2 mAb, potentially identifying the site at which this mAb bound to CD1b. A fourth mutation on the carboxy-terminal side of the α2 helix (T157A) inhibited the binding of a different Ab, BCD1b6, but not any of the other Abs tested (Table I and Fig. 1). Thus, the Abs in our panel recognized at least three distinct molecular features of the CD1b molecule, and at least two of these mapped to the helices in the α1 and α2 domains.

To confirm and extend the mapping of mAb-binding sites on CD1b, cross-blocking experiments were done to determine which mAbs bound to sterically nonoverlapping sites. Cell lines stably transfected with WT CD1b were exposed overnight to an excess of unlabeled anti-CD1b mAb and then assessed for the binding of a second fluorochrome-labeled anti-CD1b mAb. Using this approach, three distinct and nonoverlapping patterns of Ab binding were identified (Table II). mAbs 4A7.6.5, BCD1b3, and BCD1b2 all cross-blocked each other, and the binding of these three mAbs was inhibited by preincubation with mAb WM-25. mAbs BCD1b5, BCD1b6, and NuT2 shared a second pattern, and mAb BCD1b1 was not cross-blocked by any other Ab. These results showed that this panel of anti-CD1b Abs recognized at least three distinct serologic epitopes. However, further structural diversity could also be appreciated, because the three distinct Ab-binding interactions defined for mAbs BCD1b5, BCD1b6, and NuT2 using the CD1b mutants all fell within one serologic epitope defined by Ab cross-blocking.

Table II.

Cross-blocking with anti-CD1b Absa

BCD1b1BCD1b2BCD1b34A7.6.5WM25BCD1b5BCD1b6NuT2
BCD1b1 − − − − − − − 
BCD1b2 − +/− − − − 
BCD1b3 − − − − 
4A7.6.5 − − − − 
BCD1b5 − − − − − 
BCD1b1BCD1b2BCD1b34A7.6.5WM25BCD1b5BCD1b6NuT2
BCD1b1 − − − − − − − 
BCD1b2 − +/− − − − 
BCD1b3 − − − − 
4A7.6.5 − − − − 
BCD1b5 − − − − − 
a

Transfected C1R B lymphoblastoid cells expressing WT CD1b were first incubated overnight with the unlabeled blocking Ab shown across the top row, followed by addition of the Alexa-488 fluorochrome-conjugated Ab shown in the first column, and FACs analysis was done 1 h later. Abs that completely inhibited the subsequent binding of Alexa-488 labeled Ab are labeled “+”; partial inhibition (∼75% reduction of MFI) is indicated by “+/−”; no inhibition is indicated by “−”. mAb 7C4 stained weakly and could not be reliably evaluated for cross-blocking in this study.

On the basis of our model of the human CD1b structure and analogies with the MHC-encoded Ag-presenting molecules, we hypothesized that the membrane distal α1 and α2 domains plus determinants from Ag bound in the groove would form the target for TCR recognition by CD1b-restricted T cells. Thus, Abs that bind to the most membrane distal portions of the CD1 molecule (i.e., near the surface formed by the α1 and α2 helices) would be expected to efficiently block T cell reactivity. To assess this, we tested the ability of each mAb in our panel to block the CD1b-restricted proliferative response of T cell line DN1 to its specific mycobacterial lipid Ag, mycolic acid (Fig. 2). The NuT2 and BCD1b6 mAbs, which were predicted to bind near the membrane distal α-helical surface of the molecule based on their reactivity with the panel of CD1b mutants, blocked T cell reactivity completely. This was also the case for mAb BCD1b5, which belonged to the same serologically defined epitope group. The mAbs that mapped to the other two defined epitope groups (which could not be structurally mapped on the CD1b protein by the panel of mutants analyzed here) also blocked T cell reactivity, although in general less potently. These data were consistent with the hypothesis that the molecular surface formed by the α1 and α2 helices was a critically important site for TCR interactions, because steric hindrance resulting from Ab binding to this region of CD1b strongly inhibited T cell responses.

FIGURE 2.

Inhibition of T cell responses by CD1b-specific mAbs. T cell proliferation assays were performed with the CD1b-restricted T cell line DN1, using monocyte-derived DC as APCs. Saturating amounts of the indicated mAbs were added before the addition of the T cells (25 μg/ml for purified Abs or a 1/100 dilution for ascites), and these remained in the cultures throughout the assay. Cultures were incubated for 5 days in the presence or absence of Ag (mycolic acid at 10 μg/ml), and T cell proliferation was determined by [3H]thymidine incorporation. Results are shown as the percent inhibition of thymidine incorporation compared with replica cultures containing no mAbs. Isotype-matched binding and nonbinding control mAbs were also tested, including RPC5.4 (nonbinding IgG2a), P3 (nonbinding IgG1), and 5E9 (anti-transferrin receptor, IgG1) in both ascites and purified forms. Control mAbs gave no significant inhibition of proliferation (data not shown).

FIGURE 2.

Inhibition of T cell responses by CD1b-specific mAbs. T cell proliferation assays were performed with the CD1b-restricted T cell line DN1, using monocyte-derived DC as APCs. Saturating amounts of the indicated mAbs were added before the addition of the T cells (25 μg/ml for purified Abs or a 1/100 dilution for ascites), and these remained in the cultures throughout the assay. Cultures were incubated for 5 days in the presence or absence of Ag (mycolic acid at 10 μg/ml), and T cell proliferation was determined by [3H]thymidine incorporation. Results are shown as the percent inhibition of thymidine incorporation compared with replica cultures containing no mAbs. Isotype-matched binding and nonbinding control mAbs were also tested, including RPC5.4 (nonbinding IgG2a), P3 (nonbinding IgG1), and 5E9 (anti-transferrin receptor, IgG1) in both ascites and purified forms. Control mAbs gave no significant inhibition of proliferation (data not shown).

Close modal

To more precisely define the molecular determinants responsible for TCR interactions with CD1b, T cell recognition of Ags presented by CD1b mutants was tested using a panel of previously established CD1b-restricted human T cell lines. The T cells tested recognized a range of different foreign or self lipid and glycolipid Ags, representing most of the spectrum of currently known CD1b-presented Ags. These Ags differ significantly with respect to their hydrophilic head groups, which have been postulated to be the part of the Ag that interacts most directly with the TCRs of CD1b-restricted T cells (4, 10, 11). Among the five T cell lines studied, the TCR gene segment usage by three (DN1, LDN5, and Mt2.21) has been established, and this revealed no sharing of V or J gene segment usage (Ref. 6 ; M. Vincent, unpublished data). Thus, unlike the human and murine CD1d-restricted NK T cell population that expresses a canonical TCRα chain paired with a restricted repertoire of TCRβ chains (10), these CD1b-restricted T cells appear to use diverse TCRs for Ag recognition.

As a preliminary screen, all 21 of the CD1b mutations were expressed transiently in 293T cells and assessed for their Ag presenting capacity relative to WT CD1b as determined by IFN-γ release from the mycobacterial lipid Ag-reactive T cell lines DN1 and LDN5 (Fig. 3). To ensure that differences in reactivity were not related to artifacts of transient transfection, a subset of 12 of the CD1b mutants were also stably transfected into the T2 cell line. These included all of the alanine substitution mutants that showed a >50% reduction of IFN-γ production by either DN1 or LDN5 T cells when expressed transiently in 293T cells (E65A, R71A, R79A, E80A, D83A, D87A, T157A, and T165A), as well as four additional mutants (Y73F, Y151F, Q152A, E156A). Responses of the mycobacterial Ag-specific T cell lines revealed significant and reproducible effects of several of the point mutations on CD1b Ag presenting function, with similarities as well as several notable differences being observed between the two T cells (Fig. 3). In general, the patterns produced by 293T and T2 cell transfectants were similar, given that those mutants producing a >1-log reduction in IFN-γ production in the transient transfection system showed similar effects in all cases when used as stable transfectants (Fig. 3). Marked decreases in IFN-γ production (>1 log) were seen for T cell line DN1 when mutants R79A, E80A, D83A, or T165A were expressed in either APC line (Fig. 3). Smaller reductions in Ag presenting activity were also reproducibly noted for R71A, Y73F, and D87A.

FIGURE 3.

Effect of CD1b mutations on Ag recognition by mycolic acid and GMM-specific T cell lines. Left (a and b), Transient transfections of CD1b WT and mutant constructs were done in 293T cells. These were evaluated by flow cytometry to determine the level of CD1b surface expression relative to WT and used as APCs for either the mycolic acid-specific DN1 T cell line (a) or the GMM-specific line LDN5 (b). The degree of activation was determined by measuring the relative amount of IFN-γ production after 48 h of incubation of T cells with transfectants and Ag (either mycolic acid, 30 μg/ml, or GMM, 50 ng/ml). For each assay, a panel of WT transient transfectants expressing a range of different surface levels of CD1b was included. Relative activity indicates the IFN-γ release for each mutant as a percentage of that induced by WT-expressing cells with similar CD1b surface levels (MFI ±30%). To ensure reproducibility, experiments were repeated with either stable transfectants (as described below) or with a minimum of three separate transient transfections of 293T cells. Mean values of data generated from separate transfections are shown with N representing the number of experiments averaged to produce the mean. Each separate experiment was performed in triplicate, and the SEs of the means are shown by the error bars. Right (a and b), Stable transfections of CD1b WT and mutant constructs were done in T2 cells, and subclones selected for CD1b surface expression were used as APCs in IFN-γ release assays. Surface levels of CD1b on all T2 transfectants were determined by FACS within 48 h of assaying Ag-presenting function, and these are plotted on the x-axis vs the level of IFN-γ release on the y-axis. A panel of WT CD1 clones with a range of MFI values for CD1b staining was included to control for the effect of CD1b surface expression on the strength of Ag presentation. IFN-γ production by DN1 in response to mycolic acid (30 μg/ml) (a) or GMM (50 ng/ml) (b) presentation varied with MFI as shown for WT (▪). A first order polynomial equation was used to generate a nonlinear regression with the WT transfectant data, and the dashed straight line shows the best fit line through these points. Symbols show the means of triplicate values and 1 SD for each mutant as indicated, and the results are representative of at least three separate experiments for each mutant.

FIGURE 3.

Effect of CD1b mutations on Ag recognition by mycolic acid and GMM-specific T cell lines. Left (a and b), Transient transfections of CD1b WT and mutant constructs were done in 293T cells. These were evaluated by flow cytometry to determine the level of CD1b surface expression relative to WT and used as APCs for either the mycolic acid-specific DN1 T cell line (a) or the GMM-specific line LDN5 (b). The degree of activation was determined by measuring the relative amount of IFN-γ production after 48 h of incubation of T cells with transfectants and Ag (either mycolic acid, 30 μg/ml, or GMM, 50 ng/ml). For each assay, a panel of WT transient transfectants expressing a range of different surface levels of CD1b was included. Relative activity indicates the IFN-γ release for each mutant as a percentage of that induced by WT-expressing cells with similar CD1b surface levels (MFI ±30%). To ensure reproducibility, experiments were repeated with either stable transfectants (as described below) or with a minimum of three separate transient transfections of 293T cells. Mean values of data generated from separate transfections are shown with N representing the number of experiments averaged to produce the mean. Each separate experiment was performed in triplicate, and the SEs of the means are shown by the error bars. Right (a and b), Stable transfections of CD1b WT and mutant constructs were done in T2 cells, and subclones selected for CD1b surface expression were used as APCs in IFN-γ release assays. Surface levels of CD1b on all T2 transfectants were determined by FACS within 48 h of assaying Ag-presenting function, and these are plotted on the x-axis vs the level of IFN-γ release on the y-axis. A panel of WT CD1 clones with a range of MFI values for CD1b staining was included to control for the effect of CD1b surface expression on the strength of Ag presentation. IFN-γ production by DN1 in response to mycolic acid (30 μg/ml) (a) or GMM (50 ng/ml) (b) presentation varied with MFI as shown for WT (▪). A first order polynomial equation was used to generate a nonlinear regression with the WT transfectant data, and the dashed straight line shows the best fit line through these points. Symbols show the means of triplicate values and 1 SD for each mutant as indicated, and the results are representative of at least three separate experiments for each mutant.

Close modal

In contrast, only mutation E80A produced a >1-log reduction for LDN5 in both the transient and stable expression systems (Fig. 3), with one additional mutation (T157A) showing a reproducible but smaller reduction in Ag presenting activity when expressed in either 293T or T2 cells. Only a single mutant (E156A) that produced a reduction in Ag presenting function when stably expressed in T2 cells did not show a comparable reduction when expressed transiently in 293T cells. The differential recognition of this CD1b mutant likely reflected a greater sensitivity of the T2-stable transfectant system to changes in the efficiency of Ag presentation and not unrelated variations between T2 cell clones. This conclusion was supported by the observation that this same T2 transfectant presented Ag to the DN1 T cell line as efficiently as WT CD1b, indicating that the reduction observed for the LDN5 response was not related to nonspecific changes in the particular stable transfectant clone used. In summary, there was excellent agreement in the results obtained using either transient or stable transfection systems. Consistent with their diverse TCR structure previously shown by DNA sequencing (6), each T cell line was characterized by a distinct pattern of CD1b reactivity with only the mutations at position E80 affecting both cell lines.

To extend this analysis to T cells recognizing a different class of glycolipid Ags, Ag presentation to two additional CD1b-restricted T cell clones recognizing glycolipid Ags with larger and more complex oligosaccharide structures was studied using the stable T2 transfectants. T cell clones GG33A and GG123B both respond to GM1, a ceramide with a five-sugar polar cap which is thus four or five sugars larger, respectively, than the polar cap of GMM or mycolic acid (12). Although both GM1-reactive clones were strongly affected by mutations R71A and D83A, only GG33A showed significantly reduced responses to GM1 presented by six additional mutants (R79A, E80A, D87A, E156A, T157A, and T165A) (Fig. 4). Thus, as for the mycobacterial Ag-reactive T cell clones, the GM1-reactive clones showed distinct but overlapping patterns of recognition of the panel of CD1b mutants.

FIGURE 4.

Presentation of ganglioside Ags by CD1b mutants. T2 cell-stable transfectants were used as APCs for the CD1b-restricted GG33A (a) and GG123B (b) T cell clones specific for GM1. The IFN-γ production in response to 30 μg/ml GM1 vs MFI for CD1b expression from FACS analyses is graphed. The IFN-γ production from T cells reacting to CD1b WT transfectants of various MFIs are shown for comparison (▪). A nonlinear regression was used to generate the dashed line as in Fig. 3. The results shown are representative of at least three separate experiments for each mutant analyzed.

FIGURE 4.

Presentation of ganglioside Ags by CD1b mutants. T2 cell-stable transfectants were used as APCs for the CD1b-restricted GG33A (a) and GG123B (b) T cell clones specific for GM1. The IFN-γ production in response to 30 μg/ml GM1 vs MFI for CD1b expression from FACS analyses is graphed. The IFN-γ production from T cells reacting to CD1b WT transfectants of various MFIs are shown for comparison (▪). A nonlinear regression was used to generate the dashed line as in Fig. 3. The results shown are representative of at least three separate experiments for each mutant analyzed.

Close modal

We also examined the responses of a CD1b-autoreactive T cell clone (Mt2.21), which responded to CD1b-expressing APCs in the absence of a defined lipid Ag. It is currently not known whether this T cell clone recognizes CD1b independent of any bound ligand or whether a bound self lipid Ag is required for recognition. Clone Mt2.21 was tested against either WT or a total of 21 mutant CD1b transfectants of 293T and T2 cells and showed a unique pattern of reactivity against the panel of CD1b mutants (Fig. 5). In this case, multiple mutations expressed in the transient system (E68A, E80K, D83K) and in the stable system (R71A, Y73F, E80A, D83A, and T157A) produced reductions in CD1b recognition by Mt2.21 of >1 log. In addition, several mutants tested in the transient expression system showed a significant but less marked effect on recognition of CD1b by Mt2.21 (D87K, Q150A, and Q152A).

FIGURE 5.

Effects of CD1b mutations on responses of a CD1b-autoreactive T cell clone. Left, Transient transfections of CD1b WT and mutant constructs were done in 293T cells and evaluated for their ability to stimulate IFN-γ release by CD1b-autoreactive T cell clone Mt2.21 without addition of any exogenous Ag. Methods used for the analysis and data presentation are identical with those described for analysis of Ag presentation by transient transfectants in the legend to Fig. 3. In all cases, the means of data generated from three separate transfections with experiments from each transfection performed in triplicate are shown. Error bars demonstrate SEs of the means for these data. Right, Recognition of CD1b mutants expressed in stable T2 transfectants by T cell clone Mt2.21. The methods used and analysis of data were identical with those described for analysis of Ag presentation by stable T2 transfectants in the legend to Fig. 3. The results shown are representative of at least three separate experiments for each mutant analyzed.

FIGURE 5.

Effects of CD1b mutations on responses of a CD1b-autoreactive T cell clone. Left, Transient transfections of CD1b WT and mutant constructs were done in 293T cells and evaluated for their ability to stimulate IFN-γ release by CD1b-autoreactive T cell clone Mt2.21 without addition of any exogenous Ag. Methods used for the analysis and data presentation are identical with those described for analysis of Ag presentation by transient transfectants in the legend to Fig. 3. In all cases, the means of data generated from three separate transfections with experiments from each transfection performed in triplicate are shown. Error bars demonstrate SEs of the means for these data. Right, Recognition of CD1b mutants expressed in stable T2 transfectants by T cell clone Mt2.21. The methods used and analysis of data were identical with those described for analysis of Ag presentation by stable T2 transfectants in the legend to Fig. 3. The results shown are representative of at least three separate experiments for each mutant analyzed.

Close modal

The results of the functional analyses of the CD1b mutants demonstrated that each of the five CD1b-restricted T cell lines studied showed different patterns of sensitivity to mutations introduced into the α1 and α2 helices of CD1b. To provide greater insight into the relation between the effects of individual mutations on interactions of the TCR with CD1b, the predicted locations of the functionally significant mutants on the CD1b proteins for each T cell line analyzed are shown in Fig. 6. When viewed in this way, the data convey several general conclusions. First, it is apparent that the various T cells analyzed vary substantially in terms of the number of single point mutations that could be shown to have an impact on Ag recognition. At one extreme were DN1 and Mt2.21, which were sensitive to mutations at approximately one-half of the positions analyzed (Fig. 6, a and e). In contrast, LDN5 was affected by mutations at only 3 of 18 positions (Fig. 6 b). The difference observed in the number of functionally significant mutations could have resulted from differences in the sites of interaction for the different TCRs tested or could have reflected differences in relative avidity of each different TCR for its cognate CD1/Ag complex.

FIGURE 6.

Predicted locations of CD1b mutations affecting T cell recognition. The locations of mutations having an effect on T cell reactivity are shown on a ribbon diagram of the three-dimensional model of CD1b for each T cell line examined in this study (a through e). The numbers indicate the position of each residue in the linear amino acid sequence of CD1b. •, Those positions that were tested in both transient and stable transfection systems and were found in both cases to show a >1-log reduction in Ag presenting function when mutated. All other residues that were found when mutated to show at least a 0.5-log reduction in Ag presenting function in one or both assay systems are indicated as hatched circles. ○, Positions that were tested but found to produce a <0.5-log reduction in Ag presenting function in all assays performed. A composite figure that includes all of the mutations having an effect on any of these T cells is also shown (f).

FIGURE 6.

Predicted locations of CD1b mutations affecting T cell recognition. The locations of mutations having an effect on T cell reactivity are shown on a ribbon diagram of the three-dimensional model of CD1b for each T cell line examined in this study (a through e). The numbers indicate the position of each residue in the linear amino acid sequence of CD1b. •, Those positions that were tested in both transient and stable transfection systems and were found in both cases to show a >1-log reduction in Ag presenting function when mutated. All other residues that were found when mutated to show at least a 0.5-log reduction in Ag presenting function in one or both assay systems are indicated as hatched circles. ○, Positions that were tested but found to produce a <0.5-log reduction in Ag presenting function in all assays performed. A composite figure that includes all of the mutations having an effect on any of these T cells is also shown (f).

Close modal

In fact, the spatial distribution of the active mutants for the different T cell lines suggested that a single footprint of the TCR on the surface of the CD1b molecule could have accounted for the pattern observed with all of these T cell lines. Thus, nearly all of the functionally active mutations in the α2 domain were clustered in the proximal C-terminal half of the α2 helix (residues 156–165). In the α1 domain, the functionally active mutants were generally spread over a longer stretch encompassing the central third and proximal part of the C-terminal third of the α1 helix. This distribution of active mutations appeared to be significantly different from what has previously been demonstrated in mutagenesis studies on TCR interactions with MHC class I molecules (23), which predicted an orientation of binding that has been recently confirmed by x-ray crystallography of this complex (17, 18, 19, 20). For example, whereas mutations affecting the interaction of TCRs with MHC class I molecules tend to center around the distal one-third of the α1 and α2 helices (23), the functionally significant mutations on CD1b tended to be shifted more toward the center of each helix. Furthermore, there was an unequal effect of mutation of the α1 helix when compared with the α2 helix, with mutations at only three positions in the α2 helix producing a >1-log decrease in T cell responses, compared with six positions on the α1 helix.

Overall, our data indicated that extensive interactions occurred between the TCR and discrete areas formed by the side chain atoms of amino acids in the α1 and α2 helices, at least for the panel of T cells tested in this study. In addition, although the number of critical contact points appeared to vary between the different T cells analyzed, the data were consistent with a single molecular footprint for the TCR that was oriented diagonally with respect to the long axes of the α1 and α2 helices (Figs. 6 f and 7). This diagonal pattern of interaction was also predicted by the modeling of the TCR/CD1b interaction previously proposed by Grant et al. (6), in which the binding interface was modeled primarily on the basis of the known TCR/MHC class I crystal structures.

We have analyzed the interactions of the TCRs of five CD1b-restricted T cell lines using site directed mutagenesis of the CD1b molecule, providing to our knowledge the first available molecular detail of this specific immune interaction. Previous studies have not addressed the question of whether the TCR makes contacts primarily or exclusively with the exposed portion of the Ag bound to CD1, or whether it also makes contacts with portions of the CD1 protein during Ag recognition. The data presented here showed that Ag recognition by CD1b-restricted T cells required access of the TCR to the surface of CD1b formed by the helices of the α1 and α2 domains and that this recognition was significantly affected by multiple conservative substitutions of amino acids contributing to this surface. The positions that were selected for mutagenesis in this study represented polar amino acids on the α1- and α2-helical surfaces that were predicted by molecular modeling of the CD1b structure to be solvent exposed and oriented away from the Ag-binding groove. Thus, although not yet directly assessed experimentally, we judge it to be unlikely that these mutations will have a significant impact on lipid Ag binding by CD1b. These mutations fell on either side of the opening to the CD1 groove, suggesting that the same surface of the TCR that contacts CD1b also interacts with the solvent-exposed portions of bound Ag. These data provide strong support for the hypothesis that recognition of Ags presented by CD1b involves simultaneous contact by the TCR of determinants contributed from both the bound antigenic ligand and amino acid residues on the exposed face of the CD1b protein.

Another finding to emerge from our analysis was the distribution of the mutations affecting Ag recognition by CD1b-restricted T cells, which led us to propose a footprint for TCR interactions with CD1b that was oriented diagonally to the horizontal axes of the α helices. The apparent area of interaction defined for TCR/CD1b interactions in the current study bore important similarities to TCR interactions with MHC class I, which have been shown by mutagenesis studies and by x-ray crystallography to occur predominantly between the high points or “peaks” along the surface of the MHC class I α helices (18, 19, 23). Similar peaks also exist on the mouse CD1 structure as determined by x-ray crystallography (15) and are predicted to exist on CD1b based on homology model building (Fig. 7 and Ref. 6). Although the horizontal distance between the α helices is not conserved between CD1 and MHC molecules (14.4 for muCD1d vs 18.4 Å for MHC class I), the distance between these high points on the α1 and α2 helices is similar for the two proteins (between 23 and 25 Å for each; see Fig. 7, a and b). Thus, a cleft is formed between the high points of the CD1b α helices which is approximately the correct width to allow a close interaction by the TCR. Similar to what was shown previously for MHC class I molecules, the mutations of CD1b that affected TCR recognition in our study fell within a topographical canyon formed between these two peaks, thus strengthening this general principle for TCR interactions with MHC class I molecules and extending it to at least one group of MHC class I-like molecules.

FIGURE 7.

Comparison of the TCR interaction surfaces of CD1 and MHC class I molecules. Representations of the α carbon backbones determined by x-ray crystallography of H-2Kb (a) and mouse CD1d (b) reveal significant differences in the distance between the α1 and α2 helices in these molecules. In contrast, the distances between the high points or peaks of the two helices (arrows) are more closely conserved at ∼23 Å for H-2Kb and 25 Å for CD1 (distances are measured between α carbons). The locations of mutations effecting recognition of H-2Kb and of positions corresponding to residues that when mutated in CD1b showed the greatest effects on Ag presenting function (i.e., a reduction in Ag presenting function of >1 log for at least one of the T cell lines analyzed), are shown as filled circles. A box in a demonstrates the relative orientation of the TCR on H-2Kb as originally proposed by Sun et al. (23 ) and later confirmed by the crystal structure of the 2C TCR/H-2Kb complex (18 ). Representations of the surface on which TCR interaction occurs for H-2Kb (c) and for the human CD1b model structure (d) were generated using the GRASP program, with locations of mutations affecting TCR interactions shown in blue. The opening of the groove of CD1b is shown in yellow, and the groove of the H-2Kb molecule is shown with a bound VSV nonamer peptide Ag. Both molecules contain a diagonal depression that follows the general angle of the β-pleated strands, which is known for H-2Kb to comprise the major interface of MHC class I/TCR interactions. In the case of H-2Kb, the solvent-exposed area of the VSV peptide is part of this depression. Similarly, for CD1b, the groove opening is within the diagonal depression, suggesting that a bound lipid or glycolipid Ag protruding from this opening would contribute to the interface with the TCR. a and b are modified from the work of Zeng et al. (15 ) and c is reproduced with permission from Sun et al. (23 ).

FIGURE 7.

Comparison of the TCR interaction surfaces of CD1 and MHC class I molecules. Representations of the α carbon backbones determined by x-ray crystallography of H-2Kb (a) and mouse CD1d (b) reveal significant differences in the distance between the α1 and α2 helices in these molecules. In contrast, the distances between the high points or peaks of the two helices (arrows) are more closely conserved at ∼23 Å for H-2Kb and 25 Å for CD1 (distances are measured between α carbons). The locations of mutations effecting recognition of H-2Kb and of positions corresponding to residues that when mutated in CD1b showed the greatest effects on Ag presenting function (i.e., a reduction in Ag presenting function of >1 log for at least one of the T cell lines analyzed), are shown as filled circles. A box in a demonstrates the relative orientation of the TCR on H-2Kb as originally proposed by Sun et al. (23 ) and later confirmed by the crystal structure of the 2C TCR/H-2Kb complex (18 ). Representations of the surface on which TCR interaction occurs for H-2Kb (c) and for the human CD1b model structure (d) were generated using the GRASP program, with locations of mutations affecting TCR interactions shown in blue. The opening of the groove of CD1b is shown in yellow, and the groove of the H-2Kb molecule is shown with a bound VSV nonamer peptide Ag. Both molecules contain a diagonal depression that follows the general angle of the β-pleated strands, which is known for H-2Kb to comprise the major interface of MHC class I/TCR interactions. In the case of H-2Kb, the solvent-exposed area of the VSV peptide is part of this depression. Similarly, for CD1b, the groove opening is within the diagonal depression, suggesting that a bound lipid or glycolipid Ag protruding from this opening would contribute to the interface with the TCR. a and b are modified from the work of Zeng et al. (15 ) and c is reproduced with permission from Sun et al. (23 ).

Close modal

Despite this general similarity predicted between CD1 and MHC class I interactions with TCRs, significant differences were also evident in the distribution of mutations affecting recognition of CD1b compared with those previously found to affect MHC class I recognition (Fig. 7). This probably indicates differences in the orientation and docking of TCRs with the MHC class I and CD1b molecules that relate to the unique structural features of CD1b that allow it to bind and present lipid and glycolipid Ags. Rather than having a relatively shallow Ag-binding groove from which the amino acid side chains of an extended linear antigenic peptide can protrude, CD1 molecules have a larger and deeper cavity lined with hydrophobic residues that appears to be specialized for binding and presenting Ags with a strongly hydrophobic moiety, such as lipids and glycolipids. The putative Ag-binding pocket of CD1b is deepened by elevation of the helices relative to the base of the α1 and α2 domains and by an increase in the vertical tilt of the α1 and α2 helices relative to the horizontal axis of the β-pleated sheet. Thus, the height of the peaks relative to the troughs of the α1 and α2 helices are accentuated when compared with MHC class I molecules. Also distinct from MHC class I molecules is the opening of the groove in CD1b, which is smaller than the opening in MHC class I and is shifted laterally such that the CDR3 loops of the TCR would be unlikely to contact a bound Ag if positioned as they are in TCR/MHC class I complexes (6, 15, 39).

These differences in the structure of the TCR contact surface of CD1b as compared with MHC class I suggest that the docking of TCRs must occur in a slightly different way for the two types of Ag-presenting molecules. Our data are consistent with an orientation of the TCR on CD1b that is slightly rotated relative to its position on MHC class I. For MHC class I molecules, TCR interaction occurs in a cleft that is formed between the peaks of the α1 and α2 helices (Fig. 7, arrows) (18, 19). This places the CDR3 loops of the α- and β-chains of the TCR near the center of the class I molecule over positions 4 or 5 of a bound antigenic peptide, with the CDR1 and -2 loops making additional contacts with either peptide or the α helices of MHC class I. Although mutations affecting TCR recognition of CD1b-presented Ags fell consistently between the maximal elevations of the α1 and α2 helices as predicted by molecular modeling of CD1b, these mutants extended further toward the amino-terminal ends of the α1 and to a lesser extent α2 helices than would have been predicted by direct analogy to MHC class I/TCR interactions (Figs. 6 and 7). Thus, the long axes of TCRs bound to CD1 molecules may be closer to the perpendicular of the α1 and α2-helical axes than in the TCR/MHC class I complexes.

This proposed difference in the orientation of the TCR relative to the α1 and α2 helices of CD1b as compared with MHC class I molecules is consistent with the molecular modeling of the TCR/CD1b interaction, as previously reported by Grant et al. (6). This model makes the assumption that the TCR is positioned such that the TCR α-chain lies primarily over the α2 domain of CD1b, whereas TCRβ is positioned primarily over the α1 domain (6). Although not yet demonstrated by direct experimental evidence, this orientation appears reasonable based on the modeling and also on the fact that all TCR/MHC complexes analyzed by x-ray crystallography thus far show this orientation (17, 18, 19, 20, 21, 22). In docking the two proteins during the generation of that model, it was noted that the increased elevation of the peaks of the CD1b α1 and α2 helices and the smaller distance between the two α helices in comparison with MHC class I molecules prevented placement of the TCR in the same orientation as observed for TCR/H-2Kb interactions. When the same such docking is attempted with CD1b, the amino-terminal end of the α1 helix sterically hinders the binding of the TCR α-chain onto the carboxy-terminal side of the α2 helix of CD1b. This unfavorable contact can be relieved by a counterclockwise rotation of the TCR relative to the exposed face of the CD1b molecule (M. Degano, unpublished observations). Thus, in this respect, the orientation of the TCR on CD1b may more closely resemble that recently shown for the interaction of the D10 TCR on the I-Ak MHC class II molecule (22), in which case the TCR is rotated in a counterclockwise direction to avoid steric hindrance by the amino-terminal portion of the bound peptide.

Another important effect of this proposed counterclockwise rotation of the TCR on CD1b is that, based on the molecular modeling of these proteins, it places the highly variable CDR3 loops directly over the opening of the putative CD1b Ag-binding cavity where they could make contacts with the protruding hydrophilic head group of a bound lipid or glycolipid Ag (6). Our mutational analysis also suggested a greater contribution of individual residues in the α1 helix to the TCR-binding interface compared with residues in the α2 helix, suggesting that the CD1b interaction with the TCR β-chain may dominate the molecular interactions responsible for TCR binding for at least some CD1-restricted T cells. This is in contrast to data available for TCR/MHC class I interactions, which suggest a predominant role of the TCR α-chain (17, 20). Ding et al. (20) have suggested that a tilt of the vertical axis of the TCR toward the α2 helix may contribute to the preferential interactions of the class I molecule with TCR α-chains. Similarly, a tilt of the vertical axis of the TCRs of CD1b-restricted T cells toward the α1 helix could explain the proposed extensive interaction of the TCR β-chain with this helix. This orientation would also position much of the TCR α-chain further from the CD1b α2 helix, accounting for the relative paucity of interactions with this helix detected by our analysis. Nevertheless, mutations on the CD1b α2 helix do have an effect on T cell recognition (Fig. 6) and, at least for CD1d-restricted NKT cells, an important role for interaction with the TCR α-chain is implied by the preferential expression of an invariant TCR α-chain by the majority of these T cells (40, 41, 42).

The molecular interactions of TCRs with CD1b molecules examined in this study reveal many features that are conserved with MHC class I and suggest general mechanisms for the interactions of αβ TCRs with their various targets. Our results show that, as for MHC class I/TCR interactions, the membrane distal aspects of the α1 and α2 domains of CD1b are crucially important for interaction with the TCR. In combination with previous findings on the fine specificity of glycolipid Ag recognition (11), the current study supports the hypothesis that Ag-specific recognition by the TCRs of CD1-restricted T cells involves multiple contacts of the TCR with the CD1 protein as well as with portions of the bound antigenic ligand. Although the topology of this interface is different in detail, the substantial similarities between the MHC class I and CD1 molecules, particularly in the α1 and α2 helices that form peaks and valleys contributing to the binding interface, may dictate the relative orientation of these molecules in the complex with the TCR. Conservation of this general architecture among Ag-presenting molecules may explain why TCRs assembled from the same sets of V, D, and J gene segments can be used to recognize both MHC class I and II molecules as well as the evolutionarily distant CD1 family members. The ability of the TCR to dock effectively with CD1 proteins allows the T cell system to incorporate into its repertoire a distinct universe of nonpeptide Ags, providing an alternative mechanism of immune recognition that may contribute to host defense against microbial pathogens.

We thank Lloyd Klickstein for generously providing 293T cell lines; Peter Cresswell for the T2 cell line; Daniel Olive for making available the hybridoma line for production of the 4A7.6.5 mAb; Dr. K. Sagawa for providing the NuT2 mAb; Massimo Degano, Thilo Stehle, Ian Wilson, Ethan Grant, and Victor Hsu for their many helpful discussions; and Jonathan Higgins, Thilo Stehle, and Massimo Degano for critically reading the manuscript.

1

This work was supported by National Institutes of Health Grants K11AI013858 to A.M., AI40135 and AI45889 to S.A.P., AI22553 to R.L.M., and AR01996 to M.V. Additional support was provided by a grant from the American Cancer Society to S.A.P. and by Swiss Science National Foundation Grant 31.45518.95 to G.D.L.

4

Abbreviations used in this paper: DC, dendritic cell; TCM, T cell medium; WT, wild type; MFI, mean fluorescence intensity; GMM, glucose monomycolate; GM1, monosialoganglioside GM1.

1
Davis, M. M., J. J. Boniface, Z. Reich, D. Lyons, J. Hamp, B. Arden, Y. Chien.
1998
. Ligand recognition by αβ T cell receptors.
Annu. Rev. Immunol.
16
:
523
2
Garcia, K. C., L. Teyton, I. A. Wilson.
1999
. Structural basis of T cell recognition.
Annu. Rev. Immunol.
17
:
369
3
Porcelli, S. A..
1995
. The CD1 family: a third lineage of antigen-presenting molecules.
Adv. Immunol.
59
:
1
4
Porcelli, S. A., R. L. Modlin.
1999
. The CD1 system: antigen-presenting molecules for T cell recognition of lipids and glycolipids.
Annu. Rev. Immunol.
17
:
297
5
Porcelli, S., M. B. Brenner, J. L. Greenstein, S. P. Balk, C. Terhorst, P. A. Bleicher.
1989
. Recognition of cluster of differentiation 1 antigens by human CD4CD8 cytolytic T lymphocytes.
Nature
341
:
447
6
Grant, E. P., M. Degano, J. P. Rosat, S. Stenger, R. L. Modlin, I. A. Wilson, S. A. Porcelli, M. B. Brenner.
1999
. Molecular recognition of lipid antigens by T cell receptors.
J. Exp. Med.
189
:
195
7
Beckman, E. M., S. A. Porcelli, C. T. Morita, S. M. Behar, S. T. Furlong, M. B. Brenner.
1994
. Recognition of a lipid antigen by CD1-restricted αβ+ T cells.
Nature
372
:
691
8
Beckman, E. M., A. Melian, S. M. Behar, P. A. Sieling, D. Chatterjee, S. T. Furlong, R. Matsumoto, J. P. Rosat, R. L. Modlin, S. A. Porcelli.
1996
. CD1c restricts responses of mycobacteria-specific T cells. Evidence for antigen presentation by a second member of the human CD1 family.
J. Immunol.
157
:
2795
9
Sieling, P. A., D. Chatterjee, S. A. Porcelli, T. I. Prigozy, R. J. Mazzaccaro, T. Soriano, B. R. Bloom, M. B. Brenner, M. Kronenberg, P. J. Brennan.
1995
. CD1-restricted T cell recognition of microbial lipoglycan antigens.
Science
269
:
227
10
Kawano, T., J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, K. Motoki, H. Ueno, R. Nakagawa, H. Sato, E. Kondo, et al
1997
. CD1d-restricted and TCR-mediated activation of Vα14 NKT cells by glycosylceramides.
Science
278
:
1626
11
Moody, D. B., B. B. Reinhold, M. R. Guy, E. M. Beckman, D. E. Frederique, S. T. Furlong, S. Ye, V. N. Reinhold, P. A. Sieling, R. L. Modlin, et al
1997
. Structural requirements for glycolipid antigen recognition by CD1b-restricted T cells.
Science
278
:
283
12
Shamshiev, A., A. Donda, I. Carena, L. Mori, L. Kappos, G. de Libero.
1999
. Self glycolipids as T-cell autoantigens.
Eur. J. Immunol.
29
:
1667
13
Calabi, F., C. Milstein.
1986
. A novel family of human major histocompatibility complex-related genes not mapping to chromosome 6.
Nature
323
:
540
14
Martin, L. H., F. Calabi, C. Milstein.
1986
. Isolation of CD1 genes: a family of major histocompatibility complex- related differentiation antigens.
Proc. Natl. Acad. Sci. USA
83
:
9154
15
Zeng, Z., A. R. Castano, B. W. Segelke, E. A. Stura, P. A. Peterson, I. A. Wilson.
1997
. Crystal structure of mouse CD1: an MHC-like fold with a large hydrophobic binding groove.
Science
277
:
339
16
Porcelli, S. A., B. W. Segelke, M. Sugita, I. A. Wilson, M. B. Brenner.
1998
. The CD1 family of lipid antigen presenting molecules.
Immunol. Today
19
:
362
17
Garcia, K. C., M. Degano, L. R. Pease, M. Huang, P. A. Peterson, L. Teyton, I. A. Wilson.
1998
. Structural basis of plasticity in T cell receptor recognition of a self peptide-MHC antigen.
Science
279
:
1166
18
Garcia, K. C., M. Degano, R. L. Stanfield, A. Brunmark, M. R. Jackson, P. A. Peterson, L. Teyton, I. A. Wilson.
1996
. An αβ T cell receptor structure at 2.5 Å and its orientation in the TCR-MHC complex.
Science
274
:
209
19
Garboczi, D. N., P. Ghosh, U. Utz, Q. R. Fan, W. E. Biddison, D. C. Wiley.
1996
. Structure of the complex between human T-cell receptor, viral peptide and HLA-A2.
Nature
384
:
134
20
Ding, Y. H., K. J. Smith, D. N. Garboczi, U. Utz, W. E. Biddison, D. C. Wiley.
1998
. Two human T cell receptors bind in a similar diagonal mode to the HLA- A2/Tax peptide complex using different TCR amino acids.
Immunity
8
:
403
21
Teng, M. K., A. Smolyar, A. G. Tse, J. H. Liu, J. Liu, R. E. Hussey, S. G. Nathenson, H. C. Chang, E. L. Reinherz, J. H. Wang.
1998
. Identification of a common docking topology with substantial variation among different TCR-peptide-MHC complexes.
Curr. Biol.
8
:
409
22
Reinherz, E. L., K. Tan, L. Tang, P. Kern, J. Liu, Y. Xiong, R. E. Hussey, A. Smolyar, B. Hare, R. Zhang, et al
1999
. The crystal structure of a T cell receptor in complex with peptide and MHC class II.
Science
286
:
1913
23
Sun, R., S. E. Shepherd, S. S. Geier, C. T. Thomson, J. M. Sheil, S. G. Nathenson.
1995
. Evidence that the antigen receptors of cytotoxic T lymphocytes interact with a common recognition pattern on the H-2Kb molecule.
Immunity
3
:
573
24
Jorgensen, J. L., U. Esser, B. Fazekas de St. Groth, P. A. Reay, M. M. Davis.
1992
. Mapping T-cell receptor-peptide contacts by variant peptide immunization of single-chain transgenics.
Nature
355
:
224
25
Sant’Angelo, D. B., G. Waterbury, P. Preston-Hurlburt, S. T. Yoon, R. Medzhitov, S. C. Hong, C. A. Janeway, Jr.
1996
. The specificity and orientation of a TCR to its peptide-MHC class II ligands.
Immunity
4
:
367
26
Klickstein, L. B., M. R. York, E. Luther, T. A. Springer.
1993
. Identification of novel GPI-anchored antigens by analysis of GPI-anchor deficient cells with mAb in the blind panel. S. F. Schlossman, Jr, and L. Boumsell, Jr, and W. Gilks, Jr, and J. M. Harlan, Jr, and T. Kishimoto, Jr, and C. Morimoto, Jr, and J. Ritz, Jr, and S. Shaw, Jr, and R. Silverstein, Jr, and T. Springer, Jr, et al eds. In
Leukocyte Typing V
Vol. 2
:
1478
Oxford University Press, Oxford.
27
Salter, R. D., P. Cresswell.
1986
. Impaired assembly and transport of HLA-A and -B antigens in a mutant TxB cell hybrid.
EMBO J.
5
:
943
28
Porcelli, S., C. T. Morita, M. B. Brenner.
1992
. CD1b restricts the response of human CD48 T lymphocytes to a microbial antigen.
Nature
360
:
593
29
Brusca, J. S., J. D. Radolf.
1994
. Isolation of integral membrane proteins by phase partitioning with Triton X-114.
Methods Enzymol.
228
:
182
30
Behar, S. M., S. A. Porcelli, E. M. Beckman, M. B. Brenner.
1995
. A pathway of costimulation that prevents anergy in CD28 T cells: B7-independent costimulation of CD1-restricted T cells.
J. Exp. Med.
182
:
2007
31
Melián, A., Y. J. Geng, G. K. Sukhova, P. Libby, S. A. Porcelli.
1999
. CD1 expression in human atherosclerosis: a potential mechanism for T cell activation by foam cells.
Am. J. Pathol.
155
:
775
32
Favaloro, E. J., K. F. Bradstock, S. Kamath, G. Dowden, D. Gillis, V. George.
1986
. Characterization of a p43 human thymocyte antigen.
Disease Markers
4
:
261
33
Cattoretti, G., E. Berti, C. Parravicini, M. Buscaglia, F. Cappio, R. Caputo, A. Cerri, L. Crosti, D. Delia, G. Gaiera, et al
1989
. Expression of CD1 molecules on dendritic cells: otogeny, epitope analysis on normal and malignant cells, and tissue distribution. W. Knapp, Jr, and B. Dorkin, Jr, and W. R. Guilks, Jr, and E. P. Reiber, Jr, and R. E. Schmitt, Jr, and H. Stein, Jr, and A. E. G. von dem Borne, Jr, eds.
Leukocyte Typing IV
262
Oxford University Press, Oxford.
34
Boumsell, L., M. Amiot, H. Dastot, M. Schmid.
1989
. Analysis of the CD1 battery of the Vienna Workshop in the T cell lineage. W. Knapp, Jr, and B. Dorken, Jr, and W. R. Gilks, Jr, and E. P. Rieber, Jr, and R. E. Schmidt, Jr, and H. Stein, Jr, and A. E. G. von dem Borne, Jr, eds.
Leukocyte Typing IV
259
Oxford University Press, Oxford.
35
Olive, D., P. Dubreuil, C. Mawas.
1984
. Two distinct TL-like molecular subsets defined by monoclonal antibodies on the surface of human thymocytes with different expression on leukemia lines.
Immunogenetics
20
:
253
36
Panchamoorthy, G., J. McClean, R. L. Modlin, C. T. Morita, S. Ishikawa, M. B. Brenner, H. Band.
1991
. A predominance of the T cell receptor Vγ2/Vδ2 subset in human mycobacteria-responsive T cells suggests germline encoded recognition.
J. Immunol.
147
:
3360
37
Anegon, I., D. Grolleau, J. P. Soulillou.
1991
. Regulation of HILDA/LIF gene expression in activated human monocytic cells.
J. Immunol.
147
:
3973
38
Jackman, R. M., S. Stenger, A. Lee, D. B. Moody, R. A. Rogers, K. R. Niazi, M. Sugita, R. L. Modlin, P. J. Peters, S. A. Porcelli.
1998
. The tyrosine-containing cytoplasmic tail of CD1b is essential for its efficient presentation of bacterial lipid antigens.
Immunity
8
:
341
39
Wilson, I. A., P. J. Bjorkman.
1998
. Unusual MHC-like molecules: CD1, Fc receptor, the hemochromatosis gene product, and viral homologs.
Curr. Opin. Immunol.
10
:
67
40
Bendelac, A., O. Lantz, M. E. Quimby, J. W. Yewdell, J. R. Bennink, R. R. Brutkiewicz.
1995
. CD1 recognition by mouse NK1+ T lymphocytes.
Science
268
:
863
41
Lantz, O., A. Bendelac.
1994
. An invariant T cell receptor α chain is used by a unique subset of major histocompatibility complex class I-specific CD4+ and CD48 T cells in mice and humans.
J. Exp. Med.
180
:
1097
42
Bendelac, A., M. N. Rivera, S. H. Park, J. H. Roark.
1997
. Mouse CD1-specific NK1 T cells: development, specificity, and function.
Annu. Rev. Immunol.
15
:
535