Interaction of pigeon cytochrome c-(43–58) peptide analogs with either T cell antigen receptor or I-A^b molecule

Abstract

We determined that a pigeon cytochrome c-derived peptide, p43–58, possesses two anchor residues, 46 and 54, for binding with the I-A^b molecule that are compatible to the position 1 (P1) and position 9 (P9) of the core region in the major histocompatibility complex (MHC) class II binding peptides, respectively. In the present study to analyze each binding site between P1 and P9 of p43–58 to either I-A^b or T cell antigen receptor (TCR), we investigated T cell responses to a series of peptides (P2K, P3K, P4K, P5K, P6K, P7K, and P8E) that sequentially substituted charged amino acid residues for the residues at P2 to P8 of p43–58. T cells from C57BL/10 (I-A^b) mice immunized with P4K or P6K did not mount appreciable proliferative responses to the immunogens, but those primed with other peptides (P2K, P3K, P5K, P7K, and P8E) showed substantial responses in an immunogen-specific manner. It was demonstrated by binding studies that P1 and P9 functioned as main anchors and P4 and P6 functioned as secondary anchors to I-A^b. Analyses of Vβ usage of T cell lines specific for these analogs suggested that P8 interacts with the complementarity-determining region 1 (CDR1)/CDR2 of the TCR β chain. Furthermore, sequencing of the TCR on T cell hybridomas specific for these analogs indicated that P5 interacts with the CDR3 of the TCR β chain. The present findings are consistent with the three-dimensional structure of the trimolecular complex that has been reported for TCR/peptide/MHC class I molecules.

Formation of a trimolecular complex among T cell antigen receptor (TCR), major histocompatibility complex (MHC) molecule, and peptide antigen (Ag) is essential for T cell activation (1, 2). Recently, Garcia et al. (3) and Garboczi et al. (4) crystallized 2C TCR–K^b–dEV8 and TCR–HLA-A2–Tax peptide, respectively. These crystallized trimolecular complexes were shown to possess similar structures. The long dimensions of the TCR and MHC class I peptide binding surfaces are tilted approximately 20° to 30° toward the diagonal. These restrictions may place the CDR1 and CDR2 of the TCR β chain over the β1 and α1 helices of the MHC class I molecule, respectively. This orientation of TCR to MHC class I peptide resembles the model of TCR–MHC class II complex proposed by Sant’Angelo et al. (5), in which the complementarity-determining region (CDR)1 and CDR2 of the TCR α chain lie over the N-terminal region of the peptide, CDR3 of the TCR α chain interacts with the central position (P) 5 of the peptide, and CDR3 of the TCR β chain has interactions extending from P5 to P8. However, this orientation appears to differ from the model proposed by Jorgensen et al. (6), in which the TCR is rotated approximately 90° relative to the peptide–MHC surface. Thus far, no crystal structure of trimolecular complex, including class II molecules, has been determined, and it is unclear whether formation of the trimolecular complex is predicted by a general model demonstrated with a MHC class I molecule or is determined individually by each of the components.

In prior studies (7, 8), it has been demonstrated that P1 of a pigeon cytochrome c-(43–58) peptide (p43–58) is a main anchor to the I-A molecule and a particular amino acid at the P1 determines allele-specific peptide binding. Hammer et al. (9) and Jardetzky et al. (10) also showed that the P1 residue of HLA-DR1-binding peptides represents the most important anchor. The functional role of P1 was confirmed by crystallographic analysis of peptide Ag anchoring within the HLA-DR1 groove (11). In our subsequent study (12), we found that p43–58 analogs with alanine at P9 possessed the highest affinity among the I-A^b-binding peptides with phenylalanine at P1 examined. The importance of alanine at P9 was shown in other allele-specific bindings. Thus, we considered that both P9 and P1 might function as main anchors in a variety of MHC class II molecules. However, the p43–58 analogs with amino acids at P1 that are unable to interact with certain I-A molecules were no longer immunogenic in mice bearing the I-A molecules, even after changing the P9 residue to alanine (12).

In the present study, I-A^b and/or TCR contact sites of p43–58 and its analogs were reevaluated using functional assays of T cell responses. Peptides (P2K, P3K, P4K, P5K, P6K, P7K, and P8E) with substitutions of lysine (K) or glutamic acid (E) (the original amino acid at P8 is lysine) in the core region of p43–58 were used as immunogens. Because lysine possesses a long side chain with a positive charge, we considered that these substitutions might substantially influence interaction between the peptides and the TCR. Jorgensen et al. (6) showed that substitution of charged amino acids at the TCR contact sites of the peptide Ag resulted in induction of the response of T cells possessing corresponding oppositely charged amino acids in the V(D)J junctional regions of the TCR α or β chain. We show herein that P2, P3, P5, P7, and P8 function as TCR contact sites and P4 and P6 function as secondary anchors to MHC class II molecules. Furthermore, it is demonstrated that P8 and P5 interact with CDR1/CDR2 and CDR3 of TCR β chain, respectively.

MATERIALS AND METHODS

Peptides.

Preparation, purification, and sequence confirmation of the peptides used in this study were performed as previously described (12). Amino acid sequences of p43–58 and its analogs are shown in Fig. 1.

Figure 1.

Summary of amino acid sequences of peptides used for analysis of the functional site mapping of p43–58. The peptides are named by position and the substituted amino acid. Residue 46 corresponds to P1 as shown in crystallography of HLA-DR1.

T Cell Proliferation Assay.

T cell proliferation assays were performed as described elsewhere (12). Briefly, C57BL/10 (B10) mice were immunized with a synthetic peptide (20 nmol) in complete Freund’s adjuvant. After 7–10 days, T cell-enriched fractions (4 × 10⁵ per well) from the draining lymph nodes were cultured with Ag-presenting cells (APC) (1 × 10⁵ per well) and peptides in 96-well plates for 3 days. T cell responses were determined by [³H]thymidine incorporation (12).

T Cell Lines and Proliferation Assay.

Lymph node T cells (4 × 10⁶ per well) from B10 mice primed with each peptide were cultured with APC (1 × 10⁶ per well) plus immunogen (20 μM) in 24-well plates. After 14 days of culture, live cells were restimulated with the immunogen peptide and APC for 4 days and rested for 10 days. This stimulation and resting cycle was repeated several times (more than five cycles), and the peptide-specific T cell lines were established. For proliferation assay, T cell lines (3 × 10⁴ per well) were cultured with the peptides and APC (5 × 10⁵ per well) in 96-well microtiter plates for 2 days. Thereafter, [³H]thymidine incorporation was measured as described above.

T Cell Hybridomas.

T cells primed with p43–58 analogs were stimulated with the analogs for 4 days and then fused with BW1100. After hypoxanthine/aminopterin/thymidine (HAT) selection and cloning, T cell hybridomas were obtained. Responsiveness of hybridomas to the immunogens was determined by using a CTLL-2 assay (12).

Competitive Inhibition Assay for Peptide Binding.

This assay was performed as described previously (13). Biotinylated 46F50E54A peptide was used as a positive indicator (13). The mean fluorescence (MF) of viable cells was measured after excluding propidium iodide-positive cells. Inhibition rate was calculated by the following formula: (MF in the absence of inhibitor − MF in the presence of inhibitor)/(MF in the absence of inhibitor − MF of negative control).

Flow Cytometry Analysis.

Stimulated T cell lines were incubated with each Vβ- or Vα-specific antibody for 30 min at 4°C. Antibodies used in this study were B20.6 (anti-Vβ2) (14), KJ25 (anti-Vβ3) (15), KT4–10 (anti-Vβ4) (16), MR9–4 (anti-Vβ5) (17), 44–22-1 (anti-Vβ6) (18), TR310 (anti-Vβ7) (19), F23.1 (anti-Vβ8) (20), F23.2 (anti-Vβ8.2) (21), KJ16–133 (anti-Vβ8.1 and -8.2) (22), MR10–2 (anti-Vβ9) (23), KT10b (anti-Vβ10) (24), RR3–15 (anti-Vβ11) (25), MR11–1 (anti-Vβ12) (26), MR12–4 (anti-Vβ13) (27), 14.2 (anti-Vβ14) (28), and 2C11 (anti-CD3ɛ) (29). The secondary antibody was fluorescein-conjugated goat anti-mouse IgG (Organon Teknika–Cappel) or goat anti-rat Ig (Southern Biotechnology Associates). Proportions of each TCR-positive cell were calculated as follows: positive cell percent (%) = percent of TCR V region-positive cells/CD3-positive cells.

DNA Sequencing.

Total RNA was extracted from hybridomas by using the phenol/guanidine method. cDNAs were synthesized by using reverse transcriptase and TCR Cα (5′-CTCGTCGACGTCGGTGAACAGGCAGAGGG-3′) or Cβ (5′-CTCGGATCCTTTTGTTTGTTTGCAAT-3′) specific primer. DNA fragments were amplified by PCR with Vβ8 (5′-CTCGAATTCATGGGCTCCAGACTCTTCTT-3′), Vβ14 (5′-CTCGAATTCGCTCAGACTATCCATCAATGG-3′), Vα11 (5′-CTCGAATTCTCTGCTCTGAGATGCAATTTT-3′) specific or Vα consensus (Operon Technologies) primers. These PCR products were digested by EcoRI and BamHI or NotI and SalI following ligation with pBluescript II (Stratagene Cloning System) and subcloning. Inserted plasmid DNAs were used as templates for sequence reactions. Sequencing was performed by a dye/dideoxynucleotide terminator method using an automatic DNA sequencer 373A (Applied Biosystems).

RESULTS

Influence of Single Replacements of the Core Region Residues of p43–58 by Charged Amino Acids on the T Cell Responses.

On the basis of the prospect that substitution(s) in the core region of peptide Ag by charged amino acids alters interaction between the peptide Ag and MHC and/or TCR (6), we prepared peptides that serially replaced residues in the core region of p43–58 with lysine (K) or glutamic acid (E) (P2K, P3K, P4K, P5K, P6K, P7K, and P8E) (Fig. 1). Because it had been shown that P1 and P9 are critical sites for interaction with I-A^b, analogs with substitutions of charged amino acids at P1 or P9 were not examined in the present experiments (7, 8, 12, 13).

Fig. 2 shows T cell proliferative responses in I-A^b mice primed with p43–58, P2K, P3K, P4K, P5K, P6K, P7K, or P8E. As reported earlier (7), P5K and P7K induced immunogen-specific T cell responses (Fig. 2 e and g). Thus, it was again revealed that P5K and P7K functioned as TCR epitopes. When the I-A^b mice had been immunized with p43–58, p43–58 was the most potent stimulator of the primed T cells in vitro (Fig. 2a). P8E, P2K, and P4K evoked medium to low responses of p43–58-primed T cells. T cells primed with P3K or P8E responded to the P3K or P8E in vitro, respectively, but not to p43–58 (Fig. 2 c and h). This finding suggests that P3 and P8 interact with TCR. P2K generated immunogen-specific T cell responses. However, the P2K-primed T cells cross-reacted slightly with p43–58 (Fig. 2b). Thus, it seems that P2 also is a TCR contact site, although the interaction of P2 with TCR may be minimal. In contrast to these analogs, P4K generated T cell responses specific for both P4K and p43–58 in I-A^b mice (Fig. 2d). However, the proliferative responses were detected only at high concentrations of the P4K or p43–58. P6K was incapable of inducing T cell responses (Fig. 2f).

Figure 2.

T cell proliferative responses to the peptides with serial substitutions to lysine or glutamic acid in the core region of p43–58. B10 mice were immunized with 20–100 nmol of p43–58 (a), P2K (b), P3K (c), P4K (d), P5K (e), P6K (f), P7K (g), or P8E (h), and 7–10 days later T cell responses were quantitated by [³H]thymidine incorporation. The data are indicated as Δcpm in which values with peptides are compared with values without peptides.

We then examined T cell responses against peptides carrying the oppositely charged amino acid, glutamic acid, at P4 or P6. P4E- or P6E-primed T cells showed essentially the same response pattern as that of P4K- or P6K-primed T cells, respectively (Fig. 3 a and c). In contrast, P4F, which carries a noncharged amino acid, phenylalanine, at P4, generated T cell responses 100-fold greater than those generated with P4K and P4E (Fig. 3b). The P4F-primed T cells exhibited negligible cross-reactivity to p43–58. P6L, with the noncharged amino acid leucine at P6, induced immunogen-specific T cell responses with weak cross-reactivity to p43–58 at high Ag doses (Fig. 3d). Thus, p43–58 analogs with a noncharged amino acid at P4 or P6 generated immunogen-specific T cell responses. These response patterns suggest that P4 and P6 also interact with TCR, as has been reported previously (7). The weak or nonimmunogenicity of P4K, P4E, P6K, and P6E may be attributable to the charged amino acids located at P4 or P6.

Figure 3.

T cell proliferative responses to P4E, P4F, P6E, P6L, P4E/P9A, and P6E/P9A. B10 mice were immunized with 20–100 nmol of P4E (a), P4F (b), P6E (c), P6L (d), P4E/P9A (e), or P6E/P9A (f), and T cell responses were measured as described in the legend of Fig. 2.

Recovery of Immunogenicity of Weakly or Nonimmunogenic Peptide Ag by Substitution of Alanine Residue at P9.

It has been demonstrated that changing the P9 residue of p43–58 to alanine enhances markedly the binding capacity to I-A^b (12). Indeed, P4E/P9A, which had been prepared by substituting alanine for asparagine at P9 in P4E, elicited T cell responses 100-fold more potently than did P4E (Fig. 3e). The P4E/P9A-primed T cells responded considerably to P4E/P9A but weakly to P9A. The magnitude of the T cell response to P4E/P9A was comparable to that of P4F-primed T cell responses against P4F (Fig. 3b). Similarly, asparagine to alanine substitution in P6E at P9 (P6E/P9A) endowed the peptide Ag with substantial immunogenicity in I-A^b mice (Fig. 3f). However, the P6E/P9A-primed T cells responded to P9A more strongly than to the immunogen P6E/P9A (heteroclitic). In any case, it seems that P9 anchor works epistatically for P4 and P6 anchors.

P4 and P6 Work as Minor Anchors and P9 as a Major Anchor.

To examine whether the weak or nonimmunogenicity of analog peptides with substitutions of charged amino acids at P4 or P6 were attributable to their ineffective interaction with TCR or MHC, we performed semiquantitative binding assays (13). Table 1 shows that P6L considerably inhibited binding of biotinylated 46F50E54A to I-A^b transfectants, although the inhibition by P6L was low compared with that by p43–58. It should be noted that inhibition by P6K or P6E was significantly lower than that by P6L. Table 1 also shows that P4E and P4K exhibit very low inhibitory activities to the binding between biotinylated 46F50E54A and I-A^b compared with that by p43–58 or P4F. These results demonstrate that the weak or nonimmunogenicity seen with P6K, P6E, P4E, and P4K is attributable to the low binding capacity of these peptides to the I-A^b molecules. Thus, it appears that P4 and P6 work as either MHC- or TCR-binding sites.

Table 1.

Competitive inhibition assay for binding between the peptides and I-A^b

Inhibitor	% inhibition
p43–58	53.8  ± 9.8
P4E	6.7  ± 6.5
P4K	12.0  ± 7.0
P4F	49.5  ± 5.0
P6E	8.9  ± 0.4
P6K	6.1  ± 3.7
P6L	18.3  ± 1.8
P9A	73.1  ± 14.8
P4E/P9A	65.7  ± 9.5
P6E/P9A	22.3  ± 5.2

Percent inhibition was calculated by using the formula shown in Materials and Methods. The means of the percent inhibition are shown with SD. 

P4E/P9A and P6E/P9A exhibit high inhibitory activities against binding between biotinylated 46F50E54A and I-A^b molecules compared with those of P4E and P6E (Table 1). In addition, P4E/P9A as well as P9A showed greater inhibition than P6E/P9A. These findings are quite consistent with those seen in the T cell proliferative responses (Figs. 2 and 3). Thus, it seems that alanine at P9 endows these peptide Ag with high affinity for I-A^b molecules.

Vβ Usage of the T Cell Lines Specific for p43–58 Analogs.

Since it was shown that P2-P8 residues of p43–58 interacted with TCR, we considered that the TCR carrying negatively charged amino acids corresponding to lysine might be selected after repetitive stimulations with peptide Ag carrying lysine. According to this prospect, T cell lines specific for each analog were established after stimulation plus a resting interval more than five times. The responding patterns of these T cell lines were essentially the same as those of bulk T cell responses prior to restimulations (data not shown). However, a P4K-specific T cell line showed negligible reactivity to p43–58, although P4K-primed T cells cross-reacted with p43–58 (Fig. 2d). Because P6K did not elicit T cell responses, we established P6L-specific T cell lines. The P6L-specific T cell lines reacted with P6L in an immunogen-specific manner.

When Vβ usage of T cell blasts specific for the p43–58 analogs was analyzed after secondary stimulation in vitro, the T cell blasts exhibited a wide variety of Vβ usage (data not shown). However, after multiple stimulations, each T cell line specific for the p43–58 analogs showed a characteristic Vβ usage (Table 2). Because the total proportion of each of the Vβ-positive cell population analyzed attained almost 100%, we considered that the profile of Vβ usage of these T cell lines reflected the Ag specificity of the responding T cell repertoires. Quite high proportions of Vβ8.1- and Vβ14-positive cells were seen in the T cell lines specific for p43–58, P2K, P3K, P4K, P5K, P5E, P5V, P6L, and P7K, and the ratios between Vβ8.1- plus Vβ14-positive cells and cells expressing the other Vβs fell between 55.5 and 82.0 (Table 2). These proportions were consistent in each of the experiments, which were carried out separately (compare P5E-specific T cell lines 1 and 2; 62.1 and 55.5, respectively, and P5V-specific lines 1 and 2; 73.4 and 76.8, respectively) (Table 2). The profiles of the Vβ usage did not change after several additional stimulations (data not shown). It appeared that the specific pattern of Vβ usages was not altered by a particular amino acid on the TCR binding position of p43–58 analogs (compare P5K, P5E line 1, P5V line 1, and p43–58; 66.1, 62.1, 73.4 and 71.7, respectively). In contrast, P8E-specific T cell lines expressed different Vβs. The ratios of Vβ8.1- plus Vβ14-positive cells vs. other Vβ-positive cells in the P8E-specific T cell lines 1 and 2 were markedly low compared with those seen in other T cell lines (31.1 and 21.8, repectively).

Table 2.

V_β usage (%) of the TCR specific for p43–58 analogs

Vβ	p43–58	P2K	P3K	P4K	P5K	P5E		P5V		P6L	P7K	P8E
						1	2	1	2			1	2
2	0.0	3.3	0.4	0.0	0.2	0.0	0.1	0.2	0.1	0.0	0.2	11.0	5.6
3	0.2	0.1	0.5	0.0	2.1	0.1	0.1	0.2	0.1	1.5	0.1	2.3	0.0
4	0.2	0.7	0.7	0.1	0.0	0.0	0.0	0.1	0.1	0.5	0.3	25.6	27.0
5	0.1	5.6	0.3	0.2	0.1	0.0	0.1	0.2	0.1	0.3	0.1	0.3	0.0
6	5.7	12.6	1.9	2.7	0.6	0.2	0.3	6.0	5.0	0.5	0.4	4.3	11.5
7	1.6	5.6	0.2	0.2	6.2	4.8	3.0	0.4	0.4	0.4	0.3	4.1	9.5
8.1	36.4	43.7	39.6	49.5	29.6	28.8	22.1	29.9	32.7	4.0	42.8	13.8	4.8
8.2	15.6	1.6	29.0	13.0	12.6	1.5	1.2	11.4	10.6	13.6	18.4	8.1	27.4
8.3	0.4	0.0	0.3	5.9	17.6	0.7	0.5	6.8	6.7	0.0	0.7	0.0	0.4
9	0.4	1.6	0.1	0.0	0.2	0.0	0.0	0.5	0.4	0.4	0.3	1.3	0.7
10	0.8	0.7	0.1	0.4	0.2	0.1	0.3	1.7	1.4	0.6	0.1	1.9	2.0
11	0.1	0.0	0.5	0.0	0.1	0.1	0.2	0.4	0.1	0.3	0.1	18.9	4.2
12	0.2	0.0	0.2	0.0	0.2	0.0	0.1	0.9	0.3	0.9	0.1	0.0	0.0
13	9.4	3.2	0.2	24.4	0.6	35.9	42.3	3.8	2.8	0.5	0.0	2.2	3.7
14	51.4	39.6	36.8	28.1	49.9	42.3	38.0	60.0	60.5	84.6	44.4	22.3	20.9
Vβ 8.1 + 14/other Vβs	71.7	70.4	69.0	62.3	66.1	62.1	55.5	73.4	76.8	82.0	80.5	31.1	21.8

Amino Acid Sequences of TCR of T Cell Hybridomas Specific for p43–58 Analogs.

We then determined amino acid sequences of the V(D)J junctional regions of the TCR α and β chains of various T cell hybridomas specific for p43–58 analogs. It was shown that Vβ8.1-positive T cell hybridomas (BD series) or T cell clones [DB14 (7) and DB16] specific for p43–58 possessed glycine or threonine at β97 in the V–J junction that might be involved in making hydrogen bonds with aspartic acid at P5 of p43–58 (Fig. 4a). Garcia et al. (3) observed that glycine and many other small residues are at the apices of the CDRs of the TCR, and these authors considered that the absence of protruding side chains might allow the TCR to access the peptide–MHC surface more closely. Thus, a smaller subset of side chains might be able to supply the main binding energy for the specific TCR–peptide–MHC interaction.

Figure 4.

Sequences of CDR3 of the TCRs specific for p43–58 analogs. T cell hybridomas—BD, B50K, B52K, or B53E—were specific for p43–58, P5K, P7K, or P8E, respectively. T cell clones—DB14 and DB16—were specific for p43–58. Amino acid sequences of CDR3 of the TCR β chain (a) and α chain (b) are shown.

In contrast to the BD series hybridomas, Vβ8.1-positive T cell hybridomas specific for P5K (B50K series) possessed tryptophan, glutamine, or aspartic acid at β97 that might interact with lysine at P5. A similar finding was obtained with Vβ14-positive B50K series hybridomas that possessed negatively charged amino acids at β101 in the V–J junction (Fig. 4a). Thus, lysine at P5 of the P5K peptide seems to interact with the glutamic acid or aspartic acid at position 101 of TCR β chain in a salt-bridge or charge-charge pair manner. In contrast, none of Vβ14-positive B52K and B53E series hybridomas specific for P7K and P8E, respectively, possessed reciprocally charged amino acids at the same position. Since the amino acid at P5 of P7K and P8E is aspartic acid, tyrosine and serine at the β101 of B52K series hybridomas and glutamine and asparagine at the β101 of B53E series hybridomas may be involved in making hydrogen bonds (Fig. 4a).

However, we could not detect specific residues on the TCR α chain that interact with the charged amino acid of substituted p43–58 analogs (Fig. 4b). Nevertheless, recent structural and functional studies of TCR–peptide interactions demonstrated that CDR3 of TCR α chain also interacts with P5 of the peptides (5). Altogether, the present findings permit the conclusion that residue at P5 of p43–58 analogs, a main TCR-contacting site, interacts with the CDR3 of TCR β chains.

DISCUSSION

It has been reported that a network of hydrogen bonds between asparagine (N) at α62, α69, and β82 of HLA-DR1, DR3, and I-E^k molecules and the peptide backbones appears to assume a polyproline II-like conformation (11, 30, 31). It seems that P1, P4, P6, and P9 of these class II binding peptides work as anchors. Because the key residues for making the hydrogen bonds are preserved among I-A, HLA-DR1, DR3, and I-E^k, we considered that the I-A binding peptides also showed a similar conformation. Recently, Sant’Angelo et al. (5) reported a similar conformation in another peptide Ag system that binds to I-A^k. Thus, it seems that peptides in the MHC class II groove are generally forced into a similar conformation.

Aside from the hydrogen bond network, side chains of amino acids at the anchors P1, P4, P6, and P9 contribute to interaction between the peptides and MHC class II molecules. In the P1 pocket of HLA-DR1, a small β chain residue at position 86 allows for a larger aromatic P1 side chain on the peptides, whereas Phe^β86 in the P1 pocket of I-E^k selects smaller hydrophobic residues (11, 31). Pro^β86 of I-A^b possesses a small nonpolar side chain. It thus seems that the P1 pocket of I-A^b permits phenylalanine at P1 of p43–58 to enter. However, since Ala^α56 of I-A^b also selects phenylalanine at P1 of p43–58 analogs, the P1 pocket of I-A^b may be formed in a slightly different manner from that of HLA-DR1 (11). The P9 pocket of I-E^k forms a narrow hydrophobic tunnel with Glu^β9 at its base (31). On the other hand, because Tyr^β9 may occupy the P9 pocket of I-A^b, it appears to be too shallow to fit well with alanine at P9 of the peptide Ag. The P4 and P6 pockets appear to be less restrictive. The present and previous studies (7, 32) showed that either aromatic phenylalanine or small polar serine at P4 and either large nonpolar proline or small polar threonine at P6 of p43–58 analogs did not significantly affect the interaction between the p43–58 analogs and I-A^b. However, change of a single amino acid to a charged amino acid at P4 or P6 of the analog peptides resulted in inefficient interaction between the analogs and the putative P4 or P6 pockets of I-A^b, respectively. The precise mechanism underlying this phenomenon remains unclear.

It was revealed that P3 and P8 as well as P5 and P7 (7, 32) functioned as TCR contact sites. These findings were largely compatible with the crystallographic studies (11, 30, 31) in which P2, P5, and P8 of the peptide Ag were shown to be the major TCR contact sites. However, P3 but not the P2 of p43–58 was defined as a major TCR contact site in the present study. Thus, it remains elusive whether the N-terminal half of p43–58 forms a polyproline II-like helix.

In the present report, we showed that P5 and P8 of p43–58 interacted with CDR3 and CDR2/CDR1 of TCR β chain, respectively. This structure of p43–58 in the groove of I-A^b may be similar to that of hemagglutinin-(306–318) in the groove of HLA-DR1 (11). The side chain of lysine at P8 of hemagglutinin-(306–318) protrudes over the α-helix of DR1 α chain and thus appears to be accessible to TCR. However, the possibility remains that substitution at the P8 position also induces a conformational change in the α-helix of A^b α, which then interacts with CDR1/2 of TCR β. In any case, our findings appear to favor the orientation proposed by Sant’Angelo et al. (5). Altogether, it seems to us that the orientation between TCR and the MHC–peptide ligand can be generalized among the trimolecular complexes irrespective of whether the MHC class I or class II molecules are involved, as has been discussed by Sant’Angelo et al. (5).

ABBREVIATIONS

TCR

T cell antigen receptor

CDR

complementarity-determining region

MHC

major histocompatibility complex

Ag

antigen(s)

APC

antigen-presenting cell(s)

V

variable

C

constant

P

position

p43–58

pigeon cytochrome c-(43–58) peptide

B10

C57BL/10