Report

Positive Selection Through a Motif in the αβ T Cell Receptor

See allHide authors and affiliations

Science  07 Aug 1998:
Vol. 281, Issue 5378, pp. 835-838
DOI: 10.1126/science.281.5378.835

Abstract

The two lineages of T cells, αβ and γδ, differ in their developmental requirements: only αβ T cells require major histocompatibility complex recognition, a process known as positive selection. The αβ T cell receptor (TCR), but not its γδ counterpart, contains a motif within the α-chain connecting peptide domain (α-CPM) that has been conserved over the last 500 million years. In transgenic mice expressing an αβ TCR lacking the α-CPM, thymocytes were blocked in positive selection but could undergo negative selection. Thus, the α-CPM seems to participate in the generation of signals required for positive selection.

Positive selection of αβ thymocytes generates a T cell repertoire that is self–major histocompatibility complex (MHC) restricted (1–4), whereas negative selection ensures that the immune system is self-tolerant (5–7). Although the αβ TCR mediates both forms of thymic selection, the distinction between positive and negative selection signals has been difficult to define. The TCR α-chain has an amino acid motif in its connecting peptide domain (α-CPM) that is conserved in all α-chains from bony fish to humans (8). Although the α-CPM has been conserved over the last 500 million years (9), it is not found within the γδ TCR. αβ TCRs containing a defective α-CPM are unresponsive to antigens and are aberrantly associated to the CD3 complex (8). To test whether the α-CPM imparts a specific signaling function to αβ T cells during thymic selection, we generated transgenic mice expressing a wild-type αβ TCR or a TCR that lacked the α-CPM and followed the selective events during T cell ontogeny. These experiments showed that thymocytes expressing TCRs without a complete α-CPM were not efficiently positively selected, but nevertheless could be negatively selected.

The wild-type 3BBM74 TCR (10) is positively selected in mice expressing I-Ab and negatively selected in mice expressing I-Abm12. Transgenic mice expressing a Vα2.1/Vβ8.1 wild-type 3BBM74 receptor (encoded by the α wild-type and β wild-type cDNAs) or an α-CPM defective 3BBM74 TCR [encoded by the αIII and βIII cDNAs; described in (11)] were generated by injecting DNA into C57BL/6 (B6) zygotes. Recombination deficient, B6.RAG-2–/– mice expressing either the wild-type or mutant TCR transgenes were used in all experiments.

The mutant and wild-type TCRs were comparably expressed on the surface of thymocytes (Fig. 1A); however, the mutant mice contained fewer TCRhi thymocytes. The mean of TCR expression on mutant peripheral T cells was slightly decreased (∼30%), but the range of TCR expression on mutant and wild-type T cells was similar (Fig. 1A). Wild-type or mutant TCRs were immunoprecipitated from thymocytes or lymph node T cells, and the presence of the CD3 γ, δ, and ɛ subunits and ζ chains was detected by protein immunoblotting (Fig. 1B). The CD3δ subunit was absent from immunoprecipitates of mutant TCRs, even in the presence of the mild detergent digitonin. Furthermore, the ζ chain did not coprecipitate with the mutant TCR expressed by peripheral T cells. In contrast, all CD3 subunits and the ζ chain were coimmunoprecipitated with the wild-type TCR (Fig. 1B).

Figure 1

Surface expression and composition of wild-type and mutant TCR/CD3 complexes. (A) Thymocytes and lymph node cells from wild-type and mutant B6.Rag-2–/– transgenic mice were harvested and stained with the mAb to Vα2, B20.1 (34), which is specific for the transgenic α-chain. (B) Thymocytes and lymph node cells from wild-type and mutant transgenic mice were harvested and lysed in a buffer containing 1% digitonin. The TCR complex was immunoprecipitated with B20.1 mAb and resolved by SDS–polyacrylamide gel electrophoresis. The CD3 γ, δ, ɛ, and ζ proteins were identified as previously described (35).

Analysis of the CD48 (double-negative; DN), CD4+8+ (double-positive; DP), and CD4+8 or CD48+(single-positive; SP) thymocyte populations showed a reduction in the percentage of CD4+8 SP cells in the mutant animals (Fig. 2A). The percentage (Fig. 2B) and total number (12) of CD4+8SP thymocytes could differ by a factor of 30. The thymi of mice expressing the mutant TCR maintained a low number of CD4+8 thymocytes for >10 weeks (Fig. 2B); thus, the developmental defect was not relieved with time. Thymi from mutant TCR mice contained twice the number of cells as thymi from transgenic mice expressing the wild-type TCR (12), and the percentage of DP thymocytes was increased in the mutant mice (Fig. 2A). To study the efficiency of allelic exclusion, a parameter of positive selection (13), wild-type and mutant animals were crossed to B6 mice to restore recombination. Expression of the wild-type TCR excluded the rearrangement of endogenous α-chains, whereas expression of the mutant receptor did not (12). Taken together, these data suggest a severe block in positive selection.

Figure 2

Analysis of positive selection in the thymus. B6.Rag-2–/– mice expressing either the wild-type or the mutant TCR were used for these experiments. (A) Thymocytes were collected, stained with mAbs to CD4 and CD8 (34), and analyzed by flow cytometry. The numbers indicate the percentage of cells in each quadrant. (B) After staining with mAbs to CD4 and CD8 (34) and flow cytometric analysis, the percentage of CD4+8 SP thymocytes was calculated. Each data point represents the mean percentage of CD4+8 SP thymocytes from two or three mice. Wild type (open circles); α-CPM mutant (filled squares).

The appearance of transgenic T cells in the periphery was analyzed in recombination-defective (RAG-2–/–) mice (Fig. 3A). Within the first 4 weeks, T cells expressing the mutant receptor increased slowly, reflecting the inefficient positive selection in the thymus (Fig. 2). By 7 weeks, the number of splenic T cells in wild-type and mutant TCR transgenic mice were roughly equivalent (Fig. 3A). These mutant lymphocytes were thymus-derived CD4+8 T cells (Fig. 3B) (14). Even though CD4+ T cells expressing the mutant receptor accumulated slowly in the periphery, they were unresponsive to the I-Abm12 alloantigen (Fig. 3, C and D), probably because of inadequate ζ chain coupling to the mutant TCR (Fig. 1B). Thus, peripheral T cells in the mutant mice may have been selected by an escape or a default pathway (15). Alternatively, the mutant T cells may have been refractory to antigen stimulation subsequent to an expansion in the periphery.

Figure 3

Analysis of peripheral cells in mice bearing a positive selection ligand (I-Ab). (A) Splenocytes were collected, counted, stained with mAbs to Vα2 and Vβ8 (34), and analyzed by flow cytometry. After determining the percentage of TCR+ cells and the total number of splenocytes, the total number of T cells per spleen was calculated. Each data point represents the mean number of T cells per spleen from two or three mice. Splenocytes were used because it is difficult to obtain lymph nodes from extremely young Rag-2–/– animals. Wild type (open circles); α-CPM mutant (filled squares). (B) Lymph node cells from 7-week-old B6.Rag-2–/– mice expressing either the wild-type or the mutant TCR were collected, stained with mAbs to CD4 and Vα2 (34), and analyzed by flow cytometry. The numbers indicate the percentage of cells in each quadrant. (C) Mixed leukocyte cultures were initiated between 2 × 105responder lymph node cells from 7-week-old mice expressing either the wild-type (open circles) or the mutant (filled squares) TCR and titrated numbers of irradiated stimulator spleen cells from B6.C.H-2-bm12 (I-Abm12) mice. After 4 days, the cultures were pulsed with 0.5 μCi of [3H]thymidine overnight, and the incorporation was determined. There was no response to stimulator cells from B6 (I-Ab) mice (12). (D) Mixed leukocyte cultures were carried out as described in (C), and the concentration of interleukin-3 was determined as described (8). Wild type (open circles); α-CPM mutant (filled squares). There was no response to stimulator cells from B6 (I-Ab) mice (12).

Mice expressing the wild-type or the mutant TCR were crossed to B6.C.H-2-bm12, Rag-2–/– animals to introduce a ligand (I-Abm12) that induces negative selection. Thymocytes and lymph node cells from offspring expressing I-Abm12 and the transgenic TCR were analyzed. There were few CD4+8 SP thymocytes in mice expressing either the wild-type TCR (Fig. 4A) or the mutant TCR, and the expression of I-Abm12 reduced thymocyte number by ∼30% in both strains (12). In the periphery of both wild-type and mutant mice, there were few TCR+CD4+ cells (Fig. 4A), indicating that both the wild-type and mutant T cells were negatively selected.

Figure 4

Analysis of negative selection induced by the alloantigen I-Abm12. (A) Seven-week-old B6.Rag-2–/– I-Ab/bm12 mice expressing either the wild-type or the mutant TCR were used for these experiments. Thymocytes were stained with mAbs to CD4 and CD8 (34) and analyzed by flow cytometry. Lymph node cells were stained with mAbs to CD4 and Vα2 (34) and analyzed by flow cytometry. The numbers indicate the percentage of cells in each quadrant. (B) B6.Rag-2–/– (I-Ab) mice expressing either the wild-type (open squares) or mutant (filled squares) TCR were used in an in vitro assay (16–19) to determine the responsiveness of DP thymocytes to I-Abm12. Thymocytes were cocultured for 16 hours with splenocytes from B6 (I-Ab) or B6.C.H-2-bm12 (I-Abm12) animals (36) in the presence of titrated amounts of the I-Abm12 blocking mAb, 3JP (34). Cells were then stained with mAbs to Vα2, CD4, and CD8 (34). DP thymocytes were gated and analyzed for intensity of CD4 and CD8 staining. (C) Thymocytes were cocultured for 12 to 16 hours with I-Abm12 splenocytes (36). Cells were harvested, stained with Hoechst 33342, and subsequently stained with mAbs to Vα2, CD4, and CD8 (34). DP thymocytes were gated and analyzed for intensity of Hoechst 33342 staining. Thymocytes undergoing apoptosis stain more brightly with Hoechst 33342; the percentage of these cells is indicated in the figure.

To directly examine the susceptibility of DP thymocytes to negative selection, an in vitro assay was used (16–19). DP thymocytes respond to negative selection ligands by down-regulating CD4 and CD8. Therefore, thymocytes from B6.Rag-2–/–(I-Ab) mice expressing either the wild-type or the mutant TCR were cultured with antigen presenting cells (APCs) from B6.C.H-2-bm12 (I-Abm12) mice in the presence of varying amounts of an I-Abm12 blocking antibody (Fig. 4B). In this way, the I-Abm12 alloantigen available on the APC surface could be titrated. Both wild-type and mutant DP thymocytes responded similarly to the I-Abm12 antigen over an equivalent concentration range of blocking antibody by down-regulating CD4 and CD8 (Fig. 4B). Thus, DP thymocytes bearing the α-CPM mutant TCR were as responsive to their ligand as wild-type DP thymocytes. Staining the thymocytes in these cultures with Hoechst 33342 revealed that both wild-type and mutant thymocytes became apoptotic as a consequence of antigen (I-Abm12) recognition (Fig. 4C). Thus, DP thymocytes bearing a TCR lacking the α-CPM were able to undergo negative selection that resulted in apoptosis.

We show that positive selection is more profoundly affected than negative selection in mice expressing a TCR with a mutant α-CPM, which suggests that positive selection is regulated by a distinct structure within the αβ heterodimer. Although the mutant TCR contains alterations in its connecting peptide, transmembrane, and cytoplasmic domains (11), the α-CPM is likely to be the critical element. Positive selection is normal in transgenic animals coexpressing a wild-type α chain and the βIII chain (12). Furthermore, only TCRs with a mutant α-CPM exhibit a poor association to CD3δ (20). That positive selection is defective in CD3δ–/– mice (21) is consistent with our observation that the α-CPM mutant TCR is aberrantly associated to the CD3δ chain (Fig. 1B).

These data support a molecular affinity model for thymic selection, which proposes that the TCR/CD3 complex decides between positive and negative selection by initiating qualitatively distinct signaling pathways. A low-affinity ligand would initiate positive selection through the α-CPM and CD3δ. A high-affinity ligand would activate additional elements of the TCR/CD3 complex, generating distinct negative selection signals. That independent signals mediate positive and negative selection has been postulated (22), and other experimental evidence is consistent with this idea (23–26).

The conserved structural differences between αβ and γδ TCRs may be related to the differences in the ontogeny of αβ and γδ T cells (21, 2729). Unlike γδ cells, the development of αβ T cells depends on an MHC- driven, intrathymic, positive selection and the presence of the α-CPM and CD3δ (21) within the TCR complex. Given that the α-CPM has been conserved in a T cell lineage for which positive selection is obligatory, we suggest that the α-CPM evolved to facilitate the type of signals specifically required for positive selection.

  • * Present address: Malaghan Institute of Medical Research, Post Office Box 7060, Wellington South, New Zealand.

  • To whom correspondence should be addressed. E-mail: palmer{at}bii.ch

REFERENCES AND NOTES

View Abstract

Stay Connected to Science

Navigate This Article