The CD8 coreceptor contributes to the recognition of peptide-MHC (pMHC) ligands by stabilizing the TCR-pMHC interaction and enabling efficient signaling initiation. It is unclear though, which structural elements of the TCR ensure a productive association of the coreceptor. The α-chain connecting peptide motif (α-CPM) is a highly conserved sequence of eight amino acids in the membrane proximal region of the TCR α-chain. TCRs lacking the α-CPM respond poorly to low-affinity pMHC ligands and are unable to induce positive thymic selection. In this study we show that CD8 participation in ligand binding is compromised in T lineage cells expressing mutant α-CPM TCRs, leading to a slight reduction in apparent affinity; however, this by itself does not explain the thymic selection defect. By fluorescence resonance energy transfer microscopy, we found that TCR-CD8 association was compromised for TCRs lacking the α-CPM. Although high-affinity (negative-selecting) pMHC ligands showed reduced TCR-CD8 interaction, low-affinity (positive-selecting) ligands completely failed to induce molecular approximation of the TCR and its coreceptor. Therefore, the α-CPM of a TCR is an important element in mediating CD8 approximation and signal initiation.

Upon binding a peptide-MHC (pMHC)4 ligand, the αβ TCR complex initiates an intracellular signal; when a T cell or thymocyte accumulates a sufficient number of TCR and other signals, a cellular response is induced. An important element in TCR signal initiation is the coreceptor, either CD4 or CD8. These coreceptors bind to MHC class II or class I molecules, respectively (1, 2, 3, 4). The coreceptor binding site on an MHC molecule is distinct from its peptide-binding domain and therefore does not interfere with TCR binding (5). By binding to the same MHC molecule as the TCR (6), the coreceptor increases the overall affinity of TCR-pMHC binding (7, 8). Moreover, the CD4 and CD8 coreceptors are important for signal initiation due to their association with the tyrosine kinase Lck (9), which is required for critical early events in TCR signaling (10, 11, 12). Lck phosphorylates the ITAMs of the CD3 molecules within the TCR complex, especially those of CD3ζ (13). Moreover, Lck phosphorylates tyrosine residues on the Syk family kinase ZAP70 (14), thereby increasing the enzymatic activity of ZAP70 (15). Therefore, the coreceptor has an important function in coupling ligand binding with the initiation of signal transduction.

Although high-affinity TCR-pMHC interactions exhibit some degree of CD8 independence, lower affinity interactions (KD values ≥ 3 μM) require CD8 for Ag recognition (16). In the absence of CD8 binding to pMHC, primary CD8 T cells fail to form conjugates with APCs even in the presence of high concentrations of antigenic peptide (17). In this context, the CD8αβ heterodimer, but not the CD8αα homodimer, significantly increases the avidity of TCR-mediated ligand recognition as measured by binding to soluble monomeric pMHC (18). The same study suggests that CD8β facilitates TCR signal induction, not only by increasing the avidity of TCR-ligand binding but also by docking TCR/CD3 to glycolipid-enriched microdomains.

During thymocyte development, the ability of the TCR to read ligand affinity plays an important role in establishing a self-tolerant T cell repertoire (19, 20). Preselection CD4+CD8+ double-positive (DP) thymocytes expressing a MHC I-restricted TCR respond to self-Ags by differentiating into mature CD8+ single-positive (SP) thymocytes, if the apparent affinity of TCR-CD8/pMHC binding is below a discrete threshold (KD > 6 μM) (21). In contrast, preselection DP thymocytes, whose TCRs bind self-pMHC Ag with an apparent affinity above the selection threshold (KD < 6 μM), initiate apoptosis and fail to develop into mature T lineage cells (21). In this way, overtly self-reactive T cells are removed (negatively selected) from the developing T cell repertoire. The coreceptors play an important role in the selection outcome (22, 23). For example, CD8β-deficient mice select severely reduced numbers of MHC I-restricted cytotoxic Τ cells (24, 25).

Exactly how the coreceptor and TCR interact to initiate distinct signals with defined cellular consequences is not completely understood. The TCR α-chain contains an evolutionarily conserved motif in its constant region termed the α-chain connecting peptide motif (α-CPM). Mutations in this conserved membrane proximal motif (FETDXNLN) promote unresponsiveness to antigenic stimuli (26) and defects in positive selection; interestingly, negative selection with an agonist ligand was unaffected (27, 28). TCR/CD3 complexes where the α-CPM has been substituted with amino acids from the TCRδ membrane proximal domain exhibit a reduced association with the CD3δ subunit (27). Interestingly, thymocytes from CD3δ-deficient mice are defective in undergoing positive selection as well, implying a role for CD3δ in linking TCR and coreceptor molecules (29). Finally, α-CPM-deficient TCRs bind agonist ligands but cannot cooperate with CD8 to increase ligand binding (30). This further supports the idea that the α-CPM may mediate a molecular interaction between CD8 and the TCR, either directly or via CD3δ.

To directly measure the role of the α-CPM in the TCR-CD8 interaction, we measured fluorescence resonance energy transfer (FRET) between fluorescently tagged CD3ζ (of the TCR complex) and CD8β (31, 32). We used a well-defined set of peptides inducing positive or negative selection of thymocytes expressing the OT-I TCR (33) to assess the role of the α-CPM in mediating an interaction between CD3ζ and CD8β. In T cell hybridomas expressing either a wild-type (wt) or α-CPM mutant OT-I TCR, CD3ζ-CD8β interactions within the immunological synapse (34) varied in both intensity and time. Although negative-selecting peptides induced fast and sustained FRET signals in hybridomas expressing the wt OT-I TCR, positive-selecting peptides induced markedly delayed and weaker FRET signals. T cell hybridomas expressing α-CPM-deficient TCRs exhibited diminished and delayed CD3ζ-CD8β interactions in response to negative-selecting peptides. Strikingly, positive-selecting peptides did not induce detectable CD3ζ-CD8β interactions, even though both TCR/CD3 complexes and CD8 molecules were efficiently recruited to the synapse. Because the α-CPM is located in the constant domain of every TCR and does not interfere with TCR specificity (30), we suggest a role for the α-CPM in translating low affinity TCR-pMHC binding across the plasma membrane by ensuring a close association of the coreceptor to the TCR.

OVA variant peptides were synthesized and purified as described (20, 35) and had the following affinity hierarchy for the OT-I TCR: OVA > Q4 > Q4R7 > T4 > Q4H7 > G4 > E1 ≫ VSV (33). Anti-CD8β (clone 53-5.8), anti-H-2Kb (clone AF6-88.5), anti-TCRβ (H57-597), anti-Vα2 (B20.1), capturing anti-mIL-2 (JES6-1A12), and detecting anti-mIL-2-biotin (JES6-5H4) were from BD Biosciences. Rabbit monoclonal anti-GFP was from Epitomics, mouse monoclonal anti-GFP was from Santa Cruz Biotechnology, and anti-phosphotyrosine (4G10) was from Upstate Biotechnology.

The OT-I TCR comprises rearranged TCRα (Vα2-Jα26) and TCRβ (Vβ5-Dβ2-Jβ2.6) chains and is derived from the Kb-restricted, OVA257–264-specific CTL clone 149.42 (36). The cDNAs encoding the OT-Iα and OT-Iβ wt chains were recovered by standard PCR techniques from retroviral vectors described earlier (37) and ligated into the BamHI and XhoI sites of the lentiviral expression vector pTRIPΔU3EF1αEGFP (N. Taylor, Centre National de la Recherche Scientifique Unité Mixte de Recherche 5535, Montpellier, France), replacing the BamHI-EGFP-XhoI fragment. Resulting plasmids were referred to as pTRIPΔU3EF1α-OTIα and pTRIPΔU3EF1α-OTIβ, respectively. cDNAs encoding the OT-I TCR αδ chimeras were constructed using the previously described II and IV chimeric cDNAs (26) encoding the 3BBM74 TCR (38). The SpeI-XhoI fragment of pTRIPΔU3EF1α-OTIα was replaced by the corresponding fragment from the αII or αIV chimeric α-chain. Similarly, the OT-Iβγ chimera III was constructed by replacing the XbaI-XhoI fragment of pTRIPΔU3EF1α-OTIβ with the corresponding chimeric βIII fragment (26). The CD8β-YFP (32) and CD3ζ-CFP (31) constructs (where YFP is yellow fluorescent protein and CFP is cyan fluorescent protein) have been described.

Fetal thymic organ culture (FTOC) was performed as described (19). In brief, thymic lobes were excised from OT-I Rag−/−β2m−/− transgenic mice (where β2m is β2-microglobulin) at a gestational age of day 15.5. FTOCs included exogenous β2m (5 μg/ml) and peptide (2 μM for negative-selecting peptides and 20 μM for positive-selecting ligands). After 7 days of culture, thymocytes were subjected to flow cytometry analysis.

Tetramer-binding studies were performed as described (33). In brief, DP thymocytes from OT-I Rag−/−β2m−/− mice were stained with fluorescently labeled tetramers and KD values were determined by nonlinear regression analysis of the mean fluorescence intensities (MFI). Differences related to the replacement of the α-CPM (ΔKD) were calculated as ratios between the α-CPM and the OT-I wt KD values. Monomer binding studies were performed as described (21). In brief, the ratio (r) of free TCR was determined by two-step photoaffinity labeling of DP thymocytes from T1 Rag−/−β2m−/− mice. Ligand concentration-dependent binding curves were analyzed by nonlinear regression and KD values were determined. ΔKD values were calculated between the α-CPM and the transmembrane (TM) control KD values.

Supernatants of Phoenix packaging cells (G. Nolan, Stanford University, Stanford, CA) were used to transduce the TCR/CD8-deficient hybridoma 58 (39) sequentially with constructs for CD8α, CD8β-YFP, and CD3ζ-CFP. Resulting hybridomas were sorted for high CFP and YFP expression. To introduce the wt or chimeric OT-I α-chains and the wt or chimeric OT-I β-chains, CaCl2-based transfection of 293T cells was used to produce lentiviral supernatants for subsequent transduction.

Transduced hybridomas were FACS sorted for similar TCR expression among wt, mutant α-CPM and TM control hybridomas. All cells were grown in RPMI 1640 supplemented with 4% FCS, 2 mM l-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, and 50 μM 2-ME.

An Olympus IX81 inverted microscope was used in this study. A triggered F-View II camera was attached to the microscope and the MT20 illuminator with a 150-watt xenon lamp (Olympus). The system was run by CellR software (Olympus). Optical filter cubes on an automated turret were used to change excitation and emission corresponding to the various channels. The filter cubes had the following spectral alignments (center/bandpass): YFP excitation, 500/20 nm; YFP emission, 535/30 nm; CFP excitation, 436/20 nm; CFP emission, 480/40 nm. The FRET filter cube was a combination of the CFP excitation filter and the YFP emission filter. Beam splitting was achieved with a 455-nm-long pass dichroic mirror in the CFP filter cube and a 515-nm-long pass dichroic mirror in the YFP and FRET filter cubes. A 60×, 1.4 numerical aperture oil objective was used.

For CFP and YFP, the efficiency of energy transfer is >50% within the Förster radius (R0) of 4.9–5.2 nm and approximates zero beyond 2× R0 (∼10 nm). Cross-talk coefficients of CFP fluorescence into the FRET channel (equal to d) and YFP into the FRET channel (equal to a) were calibrated using hybridomas transduced with either CD3ζ-CFP or CD8β-YFP alone. For the specified filter cubes the cross-talk coefficients were d = 36.8% and a = 18.1%. For each image, the CFP, YFP, and FRET channels were acquired using the three specified filter cubes. Background was subtracted as the averaged signal from a user-specified, cell-free region of each image. To calculate FRET efficiency, the formula E = [FRET − (a × CFP) − (d × YFP)]/[FRET − (a × CFP) − (d × YFP) + (G × CFP)], which was described earlier (31, 40), was used. G is the independently calibrated ratio of sensitized emission in the FRET channel before photobleaching to donor recovery in the CFP channel after acceptor photobleaching (41). For the imaging system used here, G = 1.7 ± 0.1. FRET efficiencies were calculated from hybridoma:APC interfaces. For illustration, the NIH ImageJ software was used and the calculated FRET efficiency images were illustrated in pseudo-colors. Statistical analysis of E was performed using Student’s two-sample t test assuming two-tailed distribution and unequal variance. The recruitment of fluorescently tagged proteins to the synapse was calculated as background-corrected ratio of the MFI of the synapse region divided by the MFI of a user-defined membrane region outside of the synapse.

Tap2-deficient RMA-S cells (42) were loaded with the exogenous OVA peptide variants described above. We varied the loading concentration of each peptide to generate similar pKb expressions. RMA-S cells were incubated at 29°C overnight followed by peptide addition for 30 min at 29°C and incubation for 3 h at 37°C. pMHC surface expression was assessed by anti-H-2Kb-PE staining and flow cytometry analysis.

For IL-2 detection, 96-well flat-bottom ELISA plates (Nunc MaxiSorp) were coated with 60 μl of anti-IL-2 at 2 μg/ml in PBS with 0.02% NaN3 overnight at 4°C. For Ag stimulation, 105 OT-I hybridomas and 105 peptide-pulsed RMA-S cells were incubated in flat-bottom wells for 24 h at 37 °C. One hundred microliters of the supernatant was added to the washed and blocked ELISA plates for 1 h at room temperature. Anti-mouse IL-2-biotin secondary Ab was applied at 2 μg/ml for 1 h at room temperature. Streptavidin-HRP (Zymed Laboratories) was applied at 0.3 μg/ml for 1 h at room temperature. Conversion of o-phenylenediamine dihydrochloride (Sigma-Aldrich) was allowed for 5 min in the dark followed by stopping the reaction with 50 μl of 10% H2SO4 per well. OD490 nm was detected and normalized to a standard dilution series by Softmax Pro software (Molecular Devices).

OT-I hybridomas (105) and peptide-pulsed RMA-S cells (105) were incubated in flat-bottom wells at 37°C. Cells were fixed with 2% formaldehyde, washed, and stained for Vβ5 and analyzed by flow cytometry. Data were presented as the percentage of TCRs remaining on the cell surface.

OT-I hybridomas (2 × 105) and peptide-pulsed RMA-S cells (2 × 105) were incubated in flat-bottom wells at 37°C. Stimulation was stopped by fixation with 2% formaldehyde. Cells were washed with PBS, 10 mM Tris, and finally with buffer C from the SlowFade antifade kit (Invitrogen). Cells were mounted on standard glass slides in SlowFade antifade mounting medium.

Hybridomas (2 × 107) and peptide-pulsed RMA-S cells (2 × 107) were incubated in round-bottom tubes at 37°C. Cells were lysed in 1% Triton X-100, 50 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 20 mM β-glycerophosphate, 1 mM Na3VO4, 10 mM NaF, 1 mM PMSF, 2 μg/ml aprotinin, 5 μg/ml leupeptin, and 1 mg/ml pefabloc. Lysates were precleared by centrifugation and with protein G-Sepharose beads (GE Healthcare) and were immunoprecipitated overnight with rabbit monoclonal anti-GFP (Epitomics). After washing, samples were separated by electrophoresis with 10% SDS-polyacrylamide gels under reducing conditions, blotted, and probed with anti-phosphotyrosine (clone 4G10; Upstate Biotechnology) followed by anti-mouse-HRP (Cell Signaling) and ECL detection (GE Healthcare). For a loading control, membranes were stripped and probed with mouse monoclonal anti-GFP (Santa Cruz Biotechnology).

The wt and chimeric TCRs used in this study are schematically represented in Fig. 1,A. As point mutations in the α-CPM led to abrogation of surface TCR expression (26), we generated chimeric TCRs that contain corresponding domains from the γδ TCR, which lacks the α-CPM. Fig. 1,A shows the membrane proximal connecting peptide (CP), TM, and cytoplasmic (Cyto) domains of the constant regions of the receptors used in this study. The α-CPM is indicated with an asterisk (∗), TCR δ-sequences that replace α-sequences are shown in light gray, and TCR γ-sequences that replace β-sequences are shown in dark gray in Fig. 1,A. The α-CPM mutant chain (αIV in Fig. 1,A) is a chimera encoding V, J, and parts of the C domain from the TCR α-chain followed by a segment of Cδ, which lacks the α-CPM. The βIII chain (Fig. 1,A) encodes V, D, J, and part of the C domain from the TCR β-chain followed by the Cγ TM and cytoplasmic regions. The TM control receptor is composed of the βIII chain and the αII chain, which is identical to the αIV chain except that it contains the α-CPM (Fig. 1,A). Therefore, the TM control receptor and the α-CPM mutant receptor are distinguished by the presence or absence of the α-CPM. The V (variable) regions and the bulk of the C (constant) regions among the various receptors were identical and were not affected by the replacements of the domains shown in Fig. 1 A.

FIGURE 1.

Schematic representation of the TCR constant regions used in this study and characterization of the peptide variants. A, Schematic representation of the connecting peptide (CP), TM, and cytoplasmic (Cyt) domains of the wt and chimeric TCRs used. The α-CPM sequence is indicated with an asterisk (∗). Open bars, TCRα and TCRβ sequences; light gray bars, TCRδ sequences; dark gray bars, TCRγ sequences. The α-CPM mutant receptor (αIV/βIII) is similar to the TM control receptor (αII/βIII) except that the α-CPM has been removed and replaced by the corresponding TCRδ sequence in the αIV chain (26 ). B, Flow cytometric analysis of thymocytes expressing wt or α-CPM mutant (mut) OT-I TCR (as previously shown; Ref. 28 ). Thymocytes from C57BL/6 Rag2−/− mice, transgenic for the wt or α-CPM mutant TCR, were stained with mAbs against CD4 and CD8β. The numbers in the dot plots indicate the percentage of CD8αβ+CD4 SP cells. C, Flow cytometric analysis of thymocytes expressing wt, α-CPM mutant, or TM control T1 TCR. Thymocytes from BALB/c Rag2−/− mice, which were transgenic for the wt, α-CPM mutant, or TM control TCR, were stained and analyzed as in B. D, and E, FTOC analysis. Thymi from day 15.5 embryos of either OT-I wt (D) or α-CPM mutant (E) Rag−/−β2m−/− mice were incubated with exogenously added individual peptide and β2m. The ability of a peptide to induce the differentiation of CD8αβ+CD4 SP cells was analyzed by flow cytometry after 7 days of culture. Cells were electronically gated for TCRβ expression and the CD4/CD8β expression profiles are shown with the percentage of CD8αβ+CD4 SP cells indicated. In wt FTOCs the pattern of negative selection (neg) seen with 2 μM peptide is also seen over a broad concentration range of negative-selecting peptides (data not shown). Positive selection (pos) in FTOC is more efficient using higher peptide concentrations; FTOCs stimulated with 20 μM peptide are shown in the figure. Lack of positive or negative selection is indicated by “null.”

FIGURE 1.

Schematic representation of the TCR constant regions used in this study and characterization of the peptide variants. A, Schematic representation of the connecting peptide (CP), TM, and cytoplasmic (Cyt) domains of the wt and chimeric TCRs used. The α-CPM sequence is indicated with an asterisk (∗). Open bars, TCRα and TCRβ sequences; light gray bars, TCRδ sequences; dark gray bars, TCRγ sequences. The α-CPM mutant receptor (αIV/βIII) is similar to the TM control receptor (αII/βIII) except that the α-CPM has been removed and replaced by the corresponding TCRδ sequence in the αIV chain (26 ). B, Flow cytometric analysis of thymocytes expressing wt or α-CPM mutant (mut) OT-I TCR (as previously shown; Ref. 28 ). Thymocytes from C57BL/6 Rag2−/− mice, transgenic for the wt or α-CPM mutant TCR, were stained with mAbs against CD4 and CD8β. The numbers in the dot plots indicate the percentage of CD8αβ+CD4 SP cells. C, Flow cytometric analysis of thymocytes expressing wt, α-CPM mutant, or TM control T1 TCR. Thymocytes from BALB/c Rag2−/− mice, which were transgenic for the wt, α-CPM mutant, or TM control TCR, were stained and analyzed as in B. D, and E, FTOC analysis. Thymi from day 15.5 embryos of either OT-I wt (D) or α-CPM mutant (E) Rag−/−β2m−/− mice were incubated with exogenously added individual peptide and β2m. The ability of a peptide to induce the differentiation of CD8αβ+CD4 SP cells was analyzed by flow cytometry after 7 days of culture. Cells were electronically gated for TCRβ expression and the CD4/CD8β expression profiles are shown with the percentage of CD8αβ+CD4 SP cells indicated. In wt FTOCs the pattern of negative selection (neg) seen with 2 μM peptide is also seen over a broad concentration range of negative-selecting peptides (data not shown). Positive selection (pos) in FTOC is more efficient using higher peptide concentrations; FTOCs stimulated with 20 μM peptide are shown in the figure. Lack of positive or negative selection is indicated by “null.”

Close modal

Thymocytes from Rag2−/− mice expressing transgenic wt or α-CPM mutant TCRs were analyzed to determine the influence of the α-CPM on thymic selection. As previously shown, wt OT-I transgenic thymocytes generated substantial numbers of CD8αβ+CD4 SP cells (19, 43), whereas only very limited numbers of α-CPM mutant thymocytes were positively selected (Fig. 1,B) (28). Similarly, in the T1 TCR transgenic system (21) CD8αβ SP cells were generated in the wt and TM control transgenic mice but not in α-CPM mutant mice, clearly indicating a positive selection defect (Fig. 1 C).

The ability of different OVA variant peptides to induce positive or negative selection (see Materials and Methods for affinity hierarchy) was tested in FTOC (19). As shown here (Fig. 1,D) and previously (33), the strong OVA, Q4, and Q4R7 ligands induced negative selection in FTOC as indicated by a low percentage of CD8αβ+CD4 SP thymocytes. In contrast, the weak ligands Q4H7 and E1 generated substantial numbers of CD8αβ+CD4 SP cells, which is indicative of positive selection (Fig. 1,D and Ref. 33). The noncognate VSV ligand induced neither positive nor negative selection as reflected by high numbers of CD8αβ+CD4+ DP thymocytes. In contrast with wt OT-I thymocytes, α-CPM mutant thymocytes exhibited a substantially different selection outcome in response to the various peptides (Fig. 1,E). Only OVA was identified as negative selector, given the paucity of CD8αβ+CD4 SP cells. All other ligands were nonselectors of α-CPM mutant thymocytes; the presence of high numbers of DP thymocytes and relatively few SP thymocytes indicate the lack of positive or negative selection (Fig. 1 E).

The lack of selection in α-CPM mutant FTOC might indicate that the α-CPM mutant receptor cannot bind cognate ligands. To determine the influence of the α-CPM mutation on ligand binding, we performed pMHC tetramer and monomer binding studies. Both binding assays involved concomitant TCR and CD8 coreceptor binding to the MHC and were performed on preselection DP thymocytes at 37°C. α-CPM mutant thymocytes exhibited decreased ligand binding compared with OT-I wt thymocytes when incubated with fluorescently labeled pMHC tetramers (Fig. 2,A). Dissociation constants, KD, were calculated by nonlinear regression analysis of the MFI values of bound tetramers. Relative to wt OT-I thymocytes, α-CPM mutant thymocytes bound all tetramers tested less avidly. The reduction of tetramer avidity (ΔKD) was 3- to 4-fold (Table I).

FIGURE 2.

Ligand binding of thymocytes with wt or mutant (mut) α-CPM TCRs. A, pMHC tetramer binding to thymocytes expressing OT-I wt or α-CPM mutant TCRs. Thymocytes from C57BL/6 Rag−/−β2m−/− mice expressing either the wt or α-CPM mutant OT-I TCR were incubated with various PE-labeled tetramers at indicated concentrations at 37°C. Binding was quantified by flow cytometry and nonlinear regression analysis. A representative experiment of at least three independent experiments is shown. B, pMHC monomer binding to thymocytes expressing the T1 wt, α-CPM mutant, or TM control TCR. Thymocytes from BALB/c Rag−/−β2m−/− mice expressing the T1 TCR variants were subjected to two-step labeling experiments as described (21 ). pMHC Δ223/227 monomers that significantly reduce CD8 binding were used with the 4P ligand (right panel). A representative experiment from a total of two or more independent experiments is shown.

FIGURE 2.

Ligand binding of thymocytes with wt or mutant (mut) α-CPM TCRs. A, pMHC tetramer binding to thymocytes expressing OT-I wt or α-CPM mutant TCRs. Thymocytes from C57BL/6 Rag−/−β2m−/− mice expressing either the wt or α-CPM mutant OT-I TCR were incubated with various PE-labeled tetramers at indicated concentrations at 37°C. Binding was quantified by flow cytometry and nonlinear regression analysis. A representative experiment of at least three independent experiments is shown. B, pMHC monomer binding to thymocytes expressing the T1 wt, α-CPM mutant, or TM control TCR. Thymocytes from BALB/c Rag−/−β2m−/− mice expressing the T1 TCR variants were subjected to two-step labeling experiments as described (21 ). pMHC Δ223/227 monomers that significantly reduce CD8 binding were used with the 4P ligand (right panel). A representative experiment from a total of two or more independent experiments is shown.

Close modal
Table I.

KD values for pMHC tetramer bindinga

StrainTetramer Ligand
OVA-KbQ4R7-KbT4-KbQ4H7-Kb
OT-I wt 4.4 × 10−9 4.5 × 10−8 5.5 × 10−8 5.0 × 10−8 
α-CPM mutant 1.7 × 10−8 1.4 × 10−7 2.1 × 10−7 2.0 × 10−7 
ΔKD (KD α-CPM/KD wt3.9-fold 3.1-fold 3.8-fold 4.0-fold 
StrainTetramer Ligand
OVA-KbQ4R7-KbT4-KbQ4H7-Kb
OT-I wt 4.4 × 10−9 4.5 × 10−8 5.5 × 10−8 5.0 × 10−8 
α-CPM mutant 1.7 × 10−8 1.4 × 10−7 2.1 × 10−7 2.0 × 10−7 
ΔKD (KD α-CPM/KD wt3.9-fold 3.1-fold 3.8-fold 4.0-fold 
a

KD values [M] determined for pMHC tetramer binding to OT-I wt or α-CPM mutant CD4+CD8+ DP thymocytes. ΔKD represents KD α-CPM mutant/KD wt.

Binding studies using monomeric pMHC ligands (21, 44) involve peptides carrying a photoreactive azidobenzoic acid (ABA) that can be photochemically cross-linked to the TCR upon specific ligand binding. The T1 TCR is specific for the pMHC complex SYIPSAEK(ABA)I/H-2Kd, with the photoreactive ABA attached to the lysine at position 8 of the peptide. The ligand variant 4L was created by replacing proline at position 4 of the agonist peptide (referred to as 4P) (21). Concentration-dependent binding curves were analyzed by nonlinear regression and KD values were calculated. For low affinity (positive-selecting) ligands, half-maximal TCR saturation was not reached (data not shown) and therefore KD values could not be determined. Similar to the tetramer-binding experiments, DP thymocytes from α-CPM mutant T1 mice bound pMHC ligands with ∼4-fold lower apparent affinity (Table II) compared with wt or TM control thymocytes. The reduced apparent affinity of α-CPM mutant DP thymocytes to bind pMHC is most likely due to defective cooperativity between TCR and CD8, as ligand binding was similar when the interaction between CD8 and MHC was disrupted by point mutations in the MHC (4P-Kd Δ223/227; Fig. 2,B and Ref. 21). In both the tetramer and monomer binding experiments, the 3-to 4-fold decrease in ligand binding displayed by α-CPM mutant thymocytes was similar for all ligands (Tables I and II) and was independent of TCR specificity (i.e., similar affinity decreases were observed for OT-I TCR- and T1 TCR-expressing thymocytes). Because the CD8 coreceptor is also involved in TCR signaling initiation, we wondered whether the α-CPM mutation might affect signal initiation as well. To test this hypothesis, we set up a system to observe TCR-CD8 interaction directly by FRET microscopy.

Table II.

KD values for pMHC monomer bindinga

StrainMonomer Ligand
4L-Kd4P-Kd4P-Kd Δ223/227
T1 wt 2.8 × 10−8 8.8 × 10−8 1.4 × 10−6 
α-CPM mutant 1.5 × 10−7 3.9 × 10−7 1.4 × 10−6 
TM control 3.5 × 10−8 9.5 × 10−8 1.7 × 10−6 
ΔKD (KD α-CPM/KD TM control4.3-fold 4.1-fold 0.8-fold 
StrainMonomer Ligand
4L-Kd4P-Kd4P-Kd Δ223/227
T1 wt 2.8 × 10−8 8.8 × 10−8 1.4 × 10−6 
α-CPM mutant 1.5 × 10−7 3.9 × 10−7 1.4 × 10−6 
TM control 3.5 × 10−8 9.5 × 10−8 1.7 × 10−6 
ΔKD (KD α-CPM/KD TM control4.3-fold 4.1-fold 0.8-fold 
a

KD values determined for pMHC monomer binding to T1 wt, α-CPM mutant, or TM control CD4+CD8+ DP thymocytes. Mean values of three experiments are shown.

The TCR/CD8-deficient hybridoma 58 was successively transfected with constructs for CD8α, CD8β-YFP, CD3ζ-CFP, and TCR chains for the wt, mutant α-CPM, or TM control receptors. Resulting OT-I wt, α-CPM mutant, and TM control T cell hybridomas were FACS sorted for similar expression of TCR-Vα2, CD8α (data not shown), CD8β-YFP, and CD3ζ-CFP (Fig. 3,A). CD3ζ-CFP surface expression was assessed by fluorescence microscopy of T cell hybridomas before and after transduction of the OT-I wt TCR (Fig. 3,B). Plasma membrane expression of CD3ζ-CFP was only observed in TCR-positive cells. The localization of CD3ζ-CFP at the plasma membrane was observed for the wt and for both chimeric receptors (data not shown), indicating that CD3ζ-CFP was assembled in surface TCR/CD3 complexes. Moreover, the CD3ζ-CFP molecules were tyrosine phosphorylated upon pervanadate treatment, pointing to its functionality in wt, TM control, and mutant hybridomas (Fig. 3 C).

FIGURE 3.

Characterization of T cell hybridomas. A, OT-I TCR, CD3ζ-CFP (ζ-CFP), and CD8β-YFP expressed on wt, α-CPM mutant, and TM control T cell hybridomas were analyzed by flow cytometry. CD3ζ-CFP expression was measured by CFP detection. CD8β-YFP expression was measured by a fluorescent mAb to CD8β and YFP detection. B, Expression and localization of CD3ζ-CFP. When OT-I wt TCR is expressed on the cell surface, CD3ζ-CFP localizes to the plasma membrane. TCR expression is shown upon staining with anti-Vα2 and Alexa Fluor 546. C, Phosphorylation of the CD3ζ-CFP (pζ-CFP) construct. OT-I wt or α-CPM mutant hybridomas (20 × 106) were treated with 100 μM pervanadate. Cell lysates were immunoprecipitated with rabbit anti-GFP (Epitomics) and then probed with anti-phosphotyrosine (α-pTyr; clone 4G10) mAb. Stripping and reprobing with mouse anti-GFP (α-GFP; Santa Cruz Biotechnology) revealed the presence of CD3ζ-CFP in all samples.

FIGURE 3.

Characterization of T cell hybridomas. A, OT-I TCR, CD3ζ-CFP (ζ-CFP), and CD8β-YFP expressed on wt, α-CPM mutant, and TM control T cell hybridomas were analyzed by flow cytometry. CD3ζ-CFP expression was measured by CFP detection. CD8β-YFP expression was measured by a fluorescent mAb to CD8β and YFP detection. B, Expression and localization of CD3ζ-CFP. When OT-I wt TCR is expressed on the cell surface, CD3ζ-CFP localizes to the plasma membrane. TCR expression is shown upon staining with anti-Vα2 and Alexa Fluor 546. C, Phosphorylation of the CD3ζ-CFP (pζ-CFP) construct. OT-I wt or α-CPM mutant hybridomas (20 × 106) were treated with 100 μM pervanadate. Cell lysates were immunoprecipitated with rabbit anti-GFP (Epitomics) and then probed with anti-phosphotyrosine (α-pTyr; clone 4G10) mAb. Stripping and reprobing with mouse anti-GFP (α-GFP; Santa Cruz Biotechnology) revealed the presence of CD3ζ-CFP in all samples.

Close modal

For APCs, we used RMA-S murine lymphoma cells expressing H-2Kb loaded with exogenously added individual peptides (42). RMA-S-H-2Kb cells were pulsed with the indicated peptides at concentrations yielding similar surface expressions of pKb (Fig. 4,A). To assess functional responses, we tested hybridomas for their ability to internalize the TCR and release IL-2 upon stimulation with peptide-loaded APCs. The high-affinity/negative-selecting ligands OVA and Q4R7 induced ∼35% TCR endocytosis in OT-I wt or TM control hybridomas after 60 min of incubation (Fig. 4,B). The positive-selecting ligand Q4H7 showed slightly decreased and delayed kinetics of TCR endocytosis. In contrast, the lower affinity, positive-selecting ligand G4 failed to induce detectable loss of TCR from the cell surface, similarly as the noncognate ligand VSV. In the α-CPM mutant hybridomas, TCR endocytosis was substantially reduced, implicating a role for the α-CPM in the early cellular response to Ag. As readout for a late cellular response, we examined IL-2 release. The negative-selecting ligands OVA and Q4R7 induced IL-2 secretion from wt and TM control hybridomas (Fig. 4,C). Neither the positive-selecting Q4H7 ligand nor the noncognate VSV ligand produced detectable amounts of IL-2 (Fig. 4 C). The α-CPM mutant hybridomas failed to produce IL-2 in response to any of the ligands, excluding the possibility that the α-CPM mutant cells could accumulate weak signals over time that would not be detected by measuring TCR internalization.

FIGURE 4.

Stimulation of T cell hybridomas with peptide-loaded APCs. A, Peptide-loading of RMA-S cells. RMA-S cells were loaded with the indicated concentrations (conc.) of peptide as described in Materials and Methods. Staining with a fluorescently labeled H2-Kb Ab and flow cytometric analysis revealed similar MFI values (MedFI) and therefore similar peptide-Kb expression with all peptides. B, TCR internalization. T cell hybridomas were stimulated with peptide-loaded RMA-S cells (loading concentration as indicated in A). At the indicated time points TCR expression was determined by flow cytometry. TCR expression was set as 100% at 0 min and MFI signals were normalized to those values. C, IL-2 production. T cell hybridomas were stimulated with peptide-loaded RMA-S cells (loading concentration as indicated in A). After 24 h, supernatants were harvested and the amounts of IL-2 determined by ELISA. Calcium ionophore/PHA/PMA stimulation was used as positive control (+++). Ø, Not detected.

FIGURE 4.

Stimulation of T cell hybridomas with peptide-loaded APCs. A, Peptide-loading of RMA-S cells. RMA-S cells were loaded with the indicated concentrations (conc.) of peptide as described in Materials and Methods. Staining with a fluorescently labeled H2-Kb Ab and flow cytometric analysis revealed similar MFI values (MedFI) and therefore similar peptide-Kb expression with all peptides. B, TCR internalization. T cell hybridomas were stimulated with peptide-loaded RMA-S cells (loading concentration as indicated in A). At the indicated time points TCR expression was determined by flow cytometry. TCR expression was set as 100% at 0 min and MFI signals were normalized to those values. C, IL-2 production. T cell hybridomas were stimulated with peptide-loaded RMA-S cells (loading concentration as indicated in A). After 24 h, supernatants were harvested and the amounts of IL-2 determined by ELISA. Calcium ionophore/PHA/PMA stimulation was used as positive control (+++). Ø, Not detected.

Close modal

The approximation of two molecules within 10 nm of each other is an indirect measure of their interaction and can be assessed by FRET imaging (40). For CFP and YFP, the efficiency of energy transfer is >50% within a radius of ∼5 nm and approximates zero beyond ∼10 nm. Because the αβ TCR has a diameter of ∼10 nm (7), significant FRET between CFP and YFP strongly suggests physical interaction between the carrier molecules CD3ζ and CD8β used in these studies. We compared the FRET signals induced between CD3ζ-CFP and CD8β-YFP in the T cell hybridomas upon stimulation with peptide-loaded APCs. Following stimulation of OT-I wt hybridomas with OVA-loaded RMA-S cells, CD3ζ-CFP and CD8β-YFP were recruited to the synapse, where FRET efficiency increased significantly (Fig. 5,A, top panel). Stimulation with the noncognate VSV ligand did not recruit CD3ζ-CFP to the synapse and consequently no significant FRET was detected (Fig. 5,A, bottom panel). CD8β-YFP was recruited to the APC contact area even with the noncognate VSV ligand, as previously observed (32). In contrast, α-CPM mutant hybridomas efficiently recruited both CD3ζ-CFP and CD8β-YFP to the synapse but failed to induce substantial FRET (Fig. 5,A, middle panel). In time course experiments (Fig. 5,B) the high-affinity, negative-selecting ligands OVA and Q4R7 induced a fast and sustained FRET between CD3ζ-CFP and CD8β-YFP in OT-I wt and TM control hybridomas, whereas the low-affinity, positive-selecting ligands Q4H7 and G4 induced a delayed and weaker FRET signal. It is noteworthy that with the positive-selecting Q4H7 ligand, whose affinity for the OT-I receptor is only 2- to 3-fold lower than that of the negative selector Q4R7 (33), the FRET signal developed significantly more slowly. However, by 60 min the FRET signal induced by Q4H7 reached a similar intensity as that induced by negative-selecting ligands (Fig. 5 B).

FIGURE 5.

Ligand-induced CD3ζ-CD8β interaction. A, Fluorescence images of conjugates between peptide-loaded APCs and OT-I wt or α-CPM mutant (mut) hybridomas (loading concentration as indicated in Fig. 4 A). Conjugates are shown as phase contrast images; CD3ζ-CFP in cyan, CD8β-YFP in yellow, and FRET efficiency images as donor-ratioed, compensated color gradients according to the algorithm described in Materials and Methods. The color code is depicted in the scale bar. White arrows indicate synapses. B, CD3ζ-CFP and CD8β-YFP FRET measurements at the interface of T cell hybridoma:APC conjugates. The indicated average values ± SEM originate from n ≥ 20 conjugates for OVA, Q4R7, Q4H7, and G4 and n ≥ 10 for VSV. Negative-selecting ligands (OVA and Q4R7) are depicted in red and positive-selecting ligands (Q4H7 and G4) in blue. The noncognate VSV ligand is shown in black. For OT-I wt and TM control hybridomas, the differences in FRET efficiency induced by the weakest negative selector, Q4R7, or the strongest positive selector, Q4H7, were significant (p < 0.05) at 10, 15, 20, and 30 min. For the α-CPM mutant hybridomas, the differences in FRET signals induced by Q4R7 or Q4H7 were significant (p < 0.05) at 20, 30, 45, and 60 min. To compare the same ligands between the T cell hybridomas, we compared FRET signals between the α-CPM mutant and the TM control hybridoma. For Q4R7, the FRET signals were significantly different for all time points (p < 0.05). For Q4H7, the FRET signals in these two cell lines differed at 10, 15, 20, 30, 45, and 60 min. C and D, Recruitment (recr.) of CD3ζ-CFP (C) and CD8β-YFP (D) to the synapse. Fold recruitment of CD3ζ-CFP and CD8β-YFP to the synapse was calculated as described in Materials and Methods. The values are indicated as mean ± SEM and originate from n ≥ 20 conjugates for OVA, Q4R7, Q4H7, and G4 and n ≥ 10 conjugates for VSV.

FIGURE 5.

Ligand-induced CD3ζ-CD8β interaction. A, Fluorescence images of conjugates between peptide-loaded APCs and OT-I wt or α-CPM mutant (mut) hybridomas (loading concentration as indicated in Fig. 4 A). Conjugates are shown as phase contrast images; CD3ζ-CFP in cyan, CD8β-YFP in yellow, and FRET efficiency images as donor-ratioed, compensated color gradients according to the algorithm described in Materials and Methods. The color code is depicted in the scale bar. White arrows indicate synapses. B, CD3ζ-CFP and CD8β-YFP FRET measurements at the interface of T cell hybridoma:APC conjugates. The indicated average values ± SEM originate from n ≥ 20 conjugates for OVA, Q4R7, Q4H7, and G4 and n ≥ 10 for VSV. Negative-selecting ligands (OVA and Q4R7) are depicted in red and positive-selecting ligands (Q4H7 and G4) in blue. The noncognate VSV ligand is shown in black. For OT-I wt and TM control hybridomas, the differences in FRET efficiency induced by the weakest negative selector, Q4R7, or the strongest positive selector, Q4H7, were significant (p < 0.05) at 10, 15, 20, and 30 min. For the α-CPM mutant hybridomas, the differences in FRET signals induced by Q4R7 or Q4H7 were significant (p < 0.05) at 20, 30, 45, and 60 min. To compare the same ligands between the T cell hybridomas, we compared FRET signals between the α-CPM mutant and the TM control hybridoma. For Q4R7, the FRET signals were significantly different for all time points (p < 0.05). For Q4H7, the FRET signals in these two cell lines differed at 10, 15, 20, 30, 45, and 60 min. C and D, Recruitment (recr.) of CD3ζ-CFP (C) and CD8β-YFP (D) to the synapse. Fold recruitment of CD3ζ-CFP and CD8β-YFP to the synapse was calculated as described in Materials and Methods. The values are indicated as mean ± SEM and originate from n ≥ 20 conjugates for OVA, Q4R7, Q4H7, and G4 and n ≥ 10 conjugates for VSV.

Close modal

In the mutant α-CPM hybridomas FRET signals were markedly reduced, pointing toward a defect in CD3ζ-CD8β approximation (Fig. 5,B, middle panel). Strikingly, in absence of the α-CPM the low-affinity ligands Q4H7 and G4 caused no detectable CD3ζ-CD8β interaction over background (Fig. 5,B). The reduced FRET signals in the α-CPM mutant hybridomas could not be attributed to reduced TCR or CD8 recruitment to the synapse, because CD3ζ-CFP and CD8β-YFP are recruited similarly as in OT-I wt or TM control hybridomas (Fig. 5, C and D). Therefore, an intact α-CPM is required for cytoplasmic approximation of the CD8 coreceptor and the TCR for low-affinity ligands, explaining the defect in positive selection in α-CPM deficient mice.

To test whether TCR-CD8 association at the plasma membrane correlates with the initiation of intracellular signaling, we directly examined the phosphorylation of CD3ζ-CFP, an early event in TCR signal transduction (24, 45). For the Western blot analysis the same time points were chosen as in the microscopy-based FRET experiments. The OT-I wt and TM control hybridomas exhibited similar patterns of CD3ζ-CFP phosphorylation (Fig. 6, A and C). The high-affinity/negative-selecting ligands OVA and Q4R7 induced rapid and sustained phosphorylation of CD3ζ-CFP, whereas the low-affinity/positive-selecting ligand Q4H7 triggered delayed and weaker phosphorylation of CD3ζ-CFP. In the α-CPM mutant hybridomas only moderate phosphorylation was observed at late time points in response to the high-affinity ligands OVA and Q4R7, whereas markedly reduced phosphorylation was induced by the positive-selecting ligand Q4H7 (Fig. 6,B). The noncognate VSV ligand failed to induce phosphorylation in any of the hybridomas. Therefore, the extent and kinetics of TCR-CD8 interaction as measured by FRET correlates with the extent and kinetics of CD3ζ-CFP phosphorylation (Figs. 5,B and Fig. 6 D).

FIGURE 6.

Ligand induced CD3ζ-CFP phosphorylation. A, Tyrosine phosphorylation of CD3ζ-CFP. OT-I wt hybridomas were stimulated with peptide-loaded APCs. CD3ζ-CFP phosphorylation was assessed by immunoprecipitation and Western blotting. Anti-phosphotyrosine mAb (clone 4G10) indicates phosphorylated CD3ζ-CFP (pζ-CFP). Only one species of phosphorylated CD3ζ-CFP could be observed (∼50 kDa) in contrast to endogenous CD3ζ, which exhibits phosphorylated forms of increased molecular sizes (45 ). For signal normalization, membranes were stripped and probed with anti-GFP mAb. Representative blots are shown (n ≥ 2). B and C, α-CPM mutant (B) and TM control (C) T cell hybridomas were probed for CD3ζ-CFP phosphorylation as described in A. D, Densitometric evaluation of CD3ζ-CFP phosphorylation. Densitometry of the Western blots from A–C was performed using a phosphorimaging device (ChemiImager 5500; Alpha Innotech). Mean intensity values were normalized to the values obtained at 0 min and to the anti-GFP signal using the Alpha Ease FC software (Alpha Innotech).

FIGURE 6.

Ligand induced CD3ζ-CFP phosphorylation. A, Tyrosine phosphorylation of CD3ζ-CFP. OT-I wt hybridomas were stimulated with peptide-loaded APCs. CD3ζ-CFP phosphorylation was assessed by immunoprecipitation and Western blotting. Anti-phosphotyrosine mAb (clone 4G10) indicates phosphorylated CD3ζ-CFP (pζ-CFP). Only one species of phosphorylated CD3ζ-CFP could be observed (∼50 kDa) in contrast to endogenous CD3ζ, which exhibits phosphorylated forms of increased molecular sizes (45 ). For signal normalization, membranes were stripped and probed with anti-GFP mAb. Representative blots are shown (n ≥ 2). B and C, α-CPM mutant (B) and TM control (C) T cell hybridomas were probed for CD3ζ-CFP phosphorylation as described in A. D, Densitometric evaluation of CD3ζ-CFP phosphorylation. Densitometry of the Western blots from A–C was performed using a phosphorimaging device (ChemiImager 5500; Alpha Innotech). Mean intensity values were normalized to the values obtained at 0 min and to the anti-GFP signal using the Alpha Ease FC software (Alpha Innotech).

Close modal

TCRs lacking the α-CPM are recruited to the synapse but do not efficiently interact with the CD8 coreceptor. Our FRET experiments indicate that close approximation of the two molecules is delayed and reduced when α-CPM mutant T cell hybridomas encounter a strong ligand and completely absent if they encounter a weak ligand. Therefore, the defective transmission of weak signals in the α-CPM mutant TCR (28) is a consequence of deficient TCR-CD8 interaction and the subsequently decreased CD3ζ phosphorylation.

We previously showed that thymocytes expressing an α-CPM mutant TCR fail to be positively selected independently of the TCR specificities tested (27, 28). In this study, we showed that positive selectors and weak negative selectors fail to induce any form of thymic selection; only the strong agonist OVA induces negative selection in α-CPM mutant FTOCs. Previous studies performed with hybridoma cells showed a defect of TCR-CD8 cooperation in binding to pMHC (30). To quantify this binding defect, we measured TCR ligand binding on thymocytes of OT-I and T1 TCR transgenic mice expressing wt or mutant α-CPM constant regions. A ∼4-fold reduction in TCR binding affinity (Fig. 2 and Tables I and II) could be measured between wt and α-CPM mutant TCRs; this difference in ligand binding was not evident with pMHC ligands that fail to engage CD8 (4P-Kd Δ223/227; Fig. 2,B and Table II). Therefore, reduced pMHC binding of α-CPM mutant TCRs can plausibly be attributed to defective TCR-CD8 cooperation and not to altered TCR affinity per se. Moreover, this demonstrates that putative structural alterations in the α-CPM mutant TCR do not affect TCR affinity. Nevertheless, the slight reduction in apparent affinity of α-CPM deficient TCRs did not fully explain their thymic selection defects (Fig. 1). For example, the α-CPM mutant TCR binds the Q4H7 ligand with a tetramer KD of 200 nM (2 × 10−7 M) but is a nonselector for this TCR. Because the positive-selecting complex, G4-Kb binds the wt OT-I receptor with a similar affinity (KD of 300 nM (3 × 10−7 M); M. A. Daniels and E. Palmer, unpublished results), a KD of 200 nM is well within the range of positive selection. Therefore, the slight reduction in binding affinity does not completely account for the selection defect.

To more fully understand the mechanism, by which α-CPM mutant receptors fail during thymic selection, we used a T cell hybridoma expressing labeled CD8β and CD3ζ molecules (32). Our experiments confirm previous work (32) showing that the CD8 coreceptor is recruited to the synapse similarly for all pMHC ligands (Fig. 5,D), whereas the extent of TCR/CD3 recruitment correlated with ligand affinity (Fig. 5,C and Ref. 46). Measuring FRET between CD8β-YFP and CD3ζ-CFP in the synapse, which reflects CD8β-CD3ζ interaction, we observed different interaction kinetics for negative- and positive-selecting ligands. Negative-selecting ligands induce a rapid and sustained FRET signal, whereas positive-selecting ligands induce slow and delayed FRET (Fig. 5,B), extending a previous study (46). In the present study we show that the transition between the FRET patterns induced by negative- or positive-selecting ligands (Fig. 5,A) correlates with the affinity selection threshold previously described (33). These data demonstrate that positive-selecting pMHC ligands exhibit reduced TCR-CD8 interaction in the synapse compared with negative-selecting ligands, which correlates with decreased CD3ζ phosphorylation during signal initiation in response to positive-selecting ligands (Fig. 6 A and Ref. 33).

Our data further show that the absence of the α-CPM has no discernible effect on TCR or CD8 recruitment to the immunological synapse (Fig. 5, C and D). Both CD8 and TCR/CD3 were similarly recruited to the synapse in all cell lines studied (OT-I wt, α-CPM mutant, and TM control; Fig. 5, C and D). Strikingly, CD8β-CD3ζ interaction was significantly reduced in T cell hybridomas expressing α-CPM mutant receptors (Fig. 5,B). This emphasizes that TCR and coreceptor can be colocalized without any interaction as detected by FRET (31) (32). This reduction of molecular interaction under the plasma membrane had a clear functional consequence. Phosphorylation of CD3ζ was markedly reduced in α-CPM mutant hybridomas (Fig. 6 D). These data suggest that an important role of the α-CPM is to facilitate the physical approximation of the intracellular domain of CD8 to the TCR/CD3 complex, a requirement for ITAM phosphorylation.

FRET signals observed in α-CPM mutant hybridomas were not always predictive of selection outcome. The high affinity ligands OVA and Q4R7 induced similarly weak FRET signals in α-CPM mutant cells, similar to those observed in wt or TM control cells stimulated by positive-selecting ligands (Fig. 5,B). In α-CPM mutant FTOCs OVA is a negative selector, which might be expected because OVA can induce negative selection in wt OT-I FTOCs even in the absence of CD8 expression (47). Thus, negative selection in OVA-stimulated α-CPM mutant FTOCs might occur even without the induction of FRET. In contrast, Q4R7 is a nonselector in α-CPM mutant FTOCs (Fig. 1 E), even though this ligand induces a similar FRET signal as does OVA. It is possible that the fluorescent protein labels exaggerate the TCR-CD8 interaction induced by Q4R7, although we consider this less likely because mutant forms of CFP and YFP were used that minimize spurious dimerization (32). Nevertheless, these data clearly show that the α-CPM mutation leads to a diminution in TCR-CD8 interaction as evidenced by a reduction in the contribution of CD8 to pMHC binding, TCR coreceptor approximation, and signal initiation. The corresponding FTOCs show the consequences of this weakened interaction on thymic selection.

We propose a model in which the lateral approximation of CD8 and the TCR is similar to the approximation of two sides of a zipper (Fig. 7). The molecules are first brought together by the pMHC ligand and, while the ligand remains bound to the receptor/coreceptor pair, their close apposition continues downward in the direction of the T cell. The α-CPM functions as a zipper tooth on the TCR side of the CD8/TCR complex, stabilizing the two sides of the zipper. If the membrane proximal domains of CD8 and TCR become “zipped,” the CD8-associated Lck can tyrosine phosphorylate the ITAMs of the TCR/CD3 complex. Once the positive-selecting ligand dissociates, CD8 and the TCR disengage and Lck-mediated phosphorylation is terminated. TCRs lacking the α-CPM might not stabilize the TCR-CD8 zipper, compromising ITAM phosphorylation. Signaling by low affinity ligands, which occupy the TCR for a relatively short time, is severely affected by the lack of the α-CPM. Agonist ligands such as OVA could occupy the α-CPM mutant TCR for a sufficiently long time to compensate for the lack of an important tooth in the receptor/coreceptor zipper. Therefore, only the highest affinity ligands are capable of mediating negative selection in mutant α-CPM thymocytes (Fig. 1).

FIGURE 7.

The zipper model of TCR-CD8 approximation. In absence of an intact α-CPM, the CD8 coreceptor binds to the pMHC but only inefficiently engages the TCR at the level of the TM and cytoplasmic domains. A candidate for linking the TCR and the CD8 coreceptor is the CD3δ molecule, which has been previously shown to be poorly associated with the TCR/CD3 complex in α-CPM-deficient cells (2629 ) is indicated in the left panel. In analogy to a zipper, the α-CPM receptor is missing a tooth that hinders further zippering in the direction of the T cell plasma membrane (left image). As a consequence, the ITAMs of the CD3 complex, especially CD3ζ, are inadequately phosphorylated. Only high-affinity ligands with a sufficiently long half-life of TCR binding can compensate for the α-CPM mutation. An intact α-CPM therefore allows tight CD8 association and closure of the zipper. In our model, this corresponds to a fully “zipped” approximation (right panel) where the coreceptor-associated Lck (lck) can fully access the ITAMs of the CD3 complex.

FIGURE 7.

The zipper model of TCR-CD8 approximation. In absence of an intact α-CPM, the CD8 coreceptor binds to the pMHC but only inefficiently engages the TCR at the level of the TM and cytoplasmic domains. A candidate for linking the TCR and the CD8 coreceptor is the CD3δ molecule, which has been previously shown to be poorly associated with the TCR/CD3 complex in α-CPM-deficient cells (2629 ) is indicated in the left panel. In analogy to a zipper, the α-CPM receptor is missing a tooth that hinders further zippering in the direction of the T cell plasma membrane (left image). As a consequence, the ITAMs of the CD3 complex, especially CD3ζ, are inadequately phosphorylated. Only high-affinity ligands with a sufficiently long half-life of TCR binding can compensate for the α-CPM mutation. An intact α-CPM therefore allows tight CD8 association and closure of the zipper. In our model, this corresponds to a fully “zipped” approximation (right panel) where the coreceptor-associated Lck (lck) can fully access the ITAMs of the CD3 complex.

Close modal

The α-CPM is present in the TCR α-chain throughout vertebrate evolution (26). α-CPM-deficient receptors associate poorly with CD3δ, and related work has previously shown that the CD3δ molecule is also required to establish a link between the TCR and the CD8 coreceptor (29). Therefore, the α-CPM and CD3δ seem to be required for an optimal functional interaction with the CD8 coreceptor and for recognition of low affinity ligands that mediate positive selection and homeostatic expansion of peripheral T cells. The α-CPM may serve an equivalent function for the CD4 coreceptor, because a class II MHC restricted α-CPM mutant TCR also fails to generate positive selection signals (27).

We thank Dr. Naomi Taylor (Centre National de la Recherche Scientifique Unité Mixte de Recherche 5535, Montpellier, France) for viral vectors and packaging cell lines, M. Cavallari for help with ELISA experiments, and E. Wagner for animal care. The animal experiments were conducted in accordance with the Federal and Cantonal laws of Switzerland.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by grants from the Swiss National Science Foundation, European Association of Plastic Surgeons, Novartis, and Hoffmann La Roche (to E.P.), National Institutes of Health Grant R01AI074074 (to N.R.J.G.), and U.S. Cancer Research Institute (to M.A.D.), and T32HL07195-30 (to P.P.Y.).

4

Abbreviations used in this paper: pMHC, peptide-MHC; ABA, azidobenzoic acid; α-CPM, α-chain connecting peptide motif; β2m, β2-microglobulin; CFP, cyan fluorescent protein; DP, CD4+CD8+ double positive; FRET, fluorescence resonance energy transfer; FTOC, fetal thymic organ culture; MFI, mean fluorescence intensity; SP, CD4CD8+ single positive; TM, transmembrane; YFP, yellow fluorescent protein; wt, wild type.

1
Cammarota, G., A. Schierle, B. Takacs, D. Doran, R. Knorr, W. Bannworth, J. Guardiola, F. Sinigaglia.
1992
. Identification of a CD4 binding site on the b2 domain of HLA-DR molecules.
Nature
356
:
799
-781.
2
Koenig, R., L. Y. Huang, R. N. Germain.
1992
. MHC class II interaction with CD4 mediated by a region analogous to the MHC class I binding site for CD8.
Nature
356
:
796
-798.
3
Norment, A. M., R. D. Salter, P. Parham, V. H. Engelhard, V. H. Littman.
1988
. Cell-cell adhesion mediated by CD8 and MHC class I molecules.
Nature
336
:
79
-81.
4
Salter, R. D., H. Benjamin, P. K. Wesley, S. E. Buxton, T. P. J. Garrett, C. Clayberger, A. M. Krensky, A. M. Norment, D. R. Littman, P. Parham.
1990
. A binding site for the T-cell co-receptor CD8 on the α3 domain of HLA-A2.
Nature
345
:
41
-46.
5
Huang, J., L. J. Edwards, B. D. Evavold, Z. Cheng.
2007
. Kinetics of MHC-CD8 interaction at the T cell membrane.
J. Immunol.
179
:
7653
-7662.
6
Gallagher, P. F., F. de St. Groth, J. F. A. P. Miller.
1989
. CD4 and CD8 molecules can physically associate with the same T-cell receptor.
Proc. Natl. Acad. Sci. USA
86
:
10044
-10048.
7
Garcia, K. C., M. Degano, R. L. Stanfield, A. Brunmark, M. R. Jackson, P. A. Peterson, L. Teyton, I. A. Wilson.
1996
. An α-β T cell receptor structure at 2.5 A and its orientation in the TCR-MHC complex.
Science
274
:
209
-219.
8
Wooldridge, L., H. A. van den Berg, M. Glick, E. Gostick, B. Laugel, S. L. Hutchinson, A. Milicic, J. M. Brenchley, D. C. Douek, D. A. Price, A. K. Sewell.
2005
. Interaction between the CD8 coreceptor and MHC class I stabilizes TCR-antigen complexes at the cell surface.
J. Biol. Chem.
280
:
27491
-27502.
9
Veillette, A., M. A. Bookman, E. M. Horak, J. B. Bolen.
1988
. The CD4 and CD8 T cell surface antigens are associated with the internal membrane tyrosine-protein kinase p56lck.
Cell
55
:
301
-308.
10
Zamoyska, R..
1998
. CD4 and CD8: modulators of T-cell receptor recognition of antigen and of immune responses?.
Curr. Opin. Immunol.
10
:
82
-87.
11
Trobridge, P. A., K. A. Forbush, S. D. Levin.
2001
. Positive and negative selection of thymocytes depends on Lck interaction with the CD4 and CD8 coreceptors.
J. Immunol.
166
:
809
-818.
12
Molina, T. J., K. Kishihara, D. P. Siderovskid, W. van Ewijk, A. Narendran, E. Timms, A. Wakeham, C. J. Paige, K. U. Hartmann, A. Veillette, et al
1992
. Profound block in thymocyte development in mice lacking p56lck.
Nature
357
:
161
-165.
13
Lewis, L. A., C. D. Chung, J. Chen, J. R. Parnes, M. Moran, V. P. Patel, M. C. Miceli.
1997
. The Lck SH2 phosphotyrosine binding site is critical for efficient TCR-induced processive tyrosine phosphorylation of the ζ-chain and IL-2 production.
J. Immunol.
159
:
2292
-2300.
14
van Oers, N. S. C., N. Killeen, A. Weiss.
1996
. Lck regulates the tyrosine phosphorylation of the T cell receptor subunits and ZAP-70 in murine thymocytes.
J. Exp. Med.
183
:
1053
-1062.
15
LoGrasso, P. V., J. Hawkins, L. J. Frank, D. Wisniewski, A. Marcy.
1996
. Mechanism of activation for Zap-70 catalytic activity.
Proc. Natl. Acad. Sci. USA
93
:
12165
-12170.
16
Holler, P. D., D. M. Kranz.
2003
. Quantitative analysis of the contribution of TCR/pepMHC affinity and CD8 to T cell activation.
Immunity
18
:
155
-164.
17
Potter, T. A., K. Grebe, B. Freiberg, A. Kupfer.
2001
. Formation of supramolecular activation clusters on fresh ex vivo CD8+ T cells after engagement of the T cell antigen receptor and CD8 by antigen-presenting cells.
Proc. Natl. Acad. Sci. USA
98
:
12624
-12629.
18
Arcaro, A., C. Gregoire, T. R. Bakker, L. Baldi, M. Jordan, L. Goffin, N. Boucheron, F. Wurm, P. A. van der Merwe, B. Malissen, I. F. Luescher.
2001
. CD8β endows CD8 with efficient coreceptor function by coupling T cell receptor/CD3 to raft-associated CD8/p56Lck complexes.
J. Exp. Med.
194
:
1485
-1495.
19
Hogquist, K. A., S. C. Jameson, W. R. Heath, J. L. Howard, M. J. Bevan, F. R. Carbone.
1994
. T cell receptor antagonist peptides induce positive selection.
Cell
76
:
17
-28.
20
Alam, S. M., P. J. Travers, J. L. Wung, W. Nasholds, S. Redpath, S. C. Jameson, N. R. J. Gascoigne.
1996
. T-cell-receptor affinity and thymocyte positive selection.
Nature
381
:
616
-620.
21
Naeher, D., M. A. Daniels, B. Hausmann, P. Guillaume, I. F. Luescher, E. Palmer.
2007
. A constant affinity threshold for T cell tolerance.
J. Exp. Med.
204
:
2553
-2559.
22
Fung-Leung, W. P., M. W. Schilham, A. Rahemtulla, T. M. Kündig, M. Vollenweider, J. Potter, W. van Ewijk, T. W. Mak.
1991
. CD8 is needed for development of cytotoxic T but not helper T cells.
Cell
65
:
443
-450.
23
Rahemtulla, A., W. P. Fung-Leung, M. W. Schilham, T. M. Kündig, S. R. Sambhara, A. Narendran, A. Arabian, A. Wakeham, C. J. Paige, R. M. Zinkernagel, et al
1991
. Normal development and function of CD8+ cells but markedly decreased helper cell activity in mice lacking CD4.
Nature
353
:
181
-184.
24
Nakayama, T., A. Singer, E. D. Hsi, L. E. Samelson.
1989
. Intrathymic signalling in immature CD4+CD8+ thymocytes results in tyrosine phosphorylation of the T-cell receptor ζ-chain.
Nature
341
:
651
-655.
25
Fung-Leung, W. P., T. M. Kündig, K. Ngo, J. Panakos, J. De Sousa-Hitzler, E. Wang, P. S. Ohashi, T. W. Mak, C. Y. Lau.
1994
. Reduced thymic maturation but normal effector function of CD8+ T cells in CD8β gene-targeted mice.
J. Exp. Med.
180
:
959
-967.
26
Baeckstroem, B. T., E. Milia, A. Peter, B. Jaureguiberry, C. T. Barldari, E. Palmer.
1996
. A motif within the T cell receptor a chain constant region connecting peptide domain controls antigen responsiveness.
Immunity
5
:
437
-447.
27
Baeckstroem, B. T., U. Mueller, B. Hausmann, E. Palmer.
1998
. Positive selection through a motif in the αβ T cell receptor.
Science
281
:
835
-838.
28
Werlen, G., B. Hausmann, E. Palmer.
2000
. A motif in the αβ T-cell receptor controls positive selection by modulating ERK activity.
Nature
406
:
422
-426.
29
Doucey, M. A., L. Goffin, D. Naeher, O. Michielin, P. Baumgärtner, P. Guillaume, E. Palmer, I. F. Luescher.
2003
. CD3δ establishes a functional link between the T cell receptor and CD8.
J. Biol. Chem.
278
:
3257
-3263.
30
Naeher, D., I. F. Luescher, E. Palmer.
2002
. A role for the α-chain connecting peptide motif in mediating TCR-CD8 cooperation.
J. Immunol.
169
:
2964
-2972.
31
Zal, T., A. Zal, N. R. J. Gascoigne.
2002
. Inhibition of T cell receptor-coreceptor interactions by antagonist ligands visualized by live FRET imaging of the T-hybridoma immunological synapse.
Immunity
16
:
521
-535.
32
Yachi, P. P., J. Ampudia, N. R. J. Gascoigne, T. Zal.
2005
. Nonstimulatory peptides contribute to antigen-induced CD8-T cell receptor interaction at the immunological synapse.
Nat. Immunol.
6
:
785
-792.
33
Daniels, M. A., E. Teixeiro, J. Gill, B. Hausmann, D. Roubaty, K. Holmberg, G. Werlen, G. A. Hollaender, N. R. J. Gascoigne, E. Palmer.
2006
. Thymic selection threshold defined by compartmentalization of Ras/MAPK signalling.
Nature
444
:
724
-729.
34
Grakoui, A., S. K. Bromley, C. Sumen, M. M. Davis, A. S. Shaw, P. M. Allen, M. L. Dustin.
1999
. The immunological synapse: a molecular machine controlling T cell activation.
Science
285
:
221
-227.
35
Santori, F. R., W. C. Kieper, S. M. Brown, Y. Lu, T. A. Neubert, K. L. Johnson, S. Naylor, S. Vukmanovic, K. A. Hogquist, S. C. Jameson.
2002
. Rare, structurally homologous self-peptides promote thymocyte positive selection.
Immunity
17
:
131
-140.
36
Kelly, J. M., S. J. Sterry, S. Cose, S. J. Turner, J. Fecondo, S. Rodda, P. J. Fink, F. R. Carbone.
1993
. Identification of conserved T cell receptor CDR3 residues contacting known exposed peptide side chains from a major histocompatibility complex class I-bound determinant.
Eur. J. Immunol.
23
:
3318
-3326.
37
Stotz, S. H., L. Bollinger, F. R. Carbone, E. Palmer.
1999
. T cell receptor (TCR) antagonism without a negative signal: evidence from t cell hybridomas expressing two independent TCRs.
J. Exp. Med.
189
:
253
-263.
38
DiGiusto, D. L., E. Palmer.
1994
. An analysis of sequence variation in the β-chain framework and complementarity determining regions of an allo-reactive T cell receptor.
Mol. Immunol.
31
:
693
-670.
39
Letourneur, F., B. Malissen.
1989
. Derivation of a T cell hybridoma variant deprived of functional T cell receptor αβ chain transcripts reveals a nonfunctional α-mRNA of BW5147 origin.
Eur. J. Immunol.
19
:
2269
-2274.
40
Zal, T., N. R. J. Gascoigne.
2004
. Using live FRET imaging to reveal early protein-protein interactions during T cell activation.
Curr. Opin. Immunol.
18
:
418
-427.
41
Zal, T., N. R. J. Gascoigne.
2004
. Photobleaching-corrected FRET efficiency imaging of live cells.
Biophys. J.
86
:
3923
-3939.
42
Ljunggren, H. G., N. J. Stam, C. Oehlen, J. J. Neefjes, P. Hoeglund, M. T. Heemels, J. Bastin, T. N. M. Schumacher, A. Townsend, K. Kaerre, H. L. Ploegh.
1990
. Empty MHC class I molecules come out in the cold.
Nature
346
:
476
-481.
43
Clarke, S. R., M. Barnden, C. Kurts, F. R. Carbone, J. F. Miller, W. R. Heath.
2000
. Characterization of the ovalbumin-specific TCR transgenic line OT-I: MHC elements for positive and negative selection.
Immunol. Cell Biol.
78
:
110
-117.
44
Luescher, I. F., J. C. Cerottini, P. Romero.
1994
. Photoaffinity labeling of the T cell receptor on cloned cytotoxic T lymphocytes by covalent photoreactive ligand.
J. Biol. Chem.
269
:
5574
-5582.
45
Love, P. E., E. W. Shores.
2000
. ITAM multiplicity and thymocyte selection: how low can you go?.
Immunity
12
:
591
-597.
46
Yachi, P. P., J. Ampudia, T. Zal, N. R. J. Gascoigne.
2006
. Altered peptide ligands induce delayed CD8-T cell receptor interaction: a role for CD8 in distinguishing antigen quality.
Immunity
25
:
1046
-1054.
47
Goldrath, A. W., K. A. Hogquist, M. J. Bevan.
1997
. CD8 lineage commitment in the absence of CD8.
Immunity
6
:
633
-642.