Peptide length determines the outcome of TCR/peptide-MHCI engagement

Abstract

αβ-TCRs expressed at the CD8(+) T-cell surface interact with short peptide fragments (p) bound to MHC class I molecules (pMHCI). The TCR/pMHCI interaction is pivotal in all aspects of CD8(+) T-cell immunity. However, the rules that govern the outcome of TCR/pMHCI engagement are not entirely understood, and this is a major barrier to understanding the requirements for both effective immunity and vaccination. In the present study, we discovered an unexpected feature of the TCR/pMHCI interaction by showing that any given TCR exhibits an explicit preference for a single MHCI-peptide length. Agonists of nonpreferred length were extremely rare, suboptimal, and often entirely distinct in sequence. Structural analysis indicated that alterations in peptide length have a major impact on antigenic complexity, to which individual TCRs are unable to adapt. This novel finding demonstrates that the outcome of TCR/pMHCI engagement is determined by peptide length in addition to the sequence identity of the MHCI-bound peptide. Accordingly, the effective recognition of pMHCI Ag, which is a prerequisite for successful CD8(+) T-cell immunity and protective vaccination, can only be achieved by length-matched Ag-specific CD8(+) T-cell clonotypes.

Figure 1

MHCI-peptide length preference reflects the length of the “index” peptide. Target cells (6 × 10⁴) expressing either HLA A*0201 or HLA B*3508 were pulsed in duplicate with the following “sizing scan” mixtures (1mM) for 2 hours at 37°C: X⁸, X⁹, X¹⁰, X¹¹, X¹², and X¹³ (where X is any of the 19 proteogenic L-amino acids excluding cysteine). Subsequently, 3 × 10⁴ ILA1, 1E6, SB14, SB10, or SB27 CD8⁺ T cells (Table 1) were added and incubated overnight. The supernatant was then harvested and assayed for MIP1β by ELISA. Values are expressed as a percentage of the highest response, where black represents 100% of the highest response and white represents 0% of the highest response. The raw data used to construct this figure can be viewed in supplemental Figure 1.

Figure 2

MHCI-restricted CD8⁺ T-cell clones exhibit explicit peptide length recognition footprints. Target cells (6 × 10⁴) expressing either HLA A*0201 or HLA B*3508 were pulsed in duplicate with mixtures from 8mer, 9mer, 10mer, 11mer, 12mer, or 13mer CPL scans (100 μM) at 37°C. After 2 hours, 3 × 10⁴ ILA1 (A), 1E6 (B), MEL5 (C), or SB27 (D) CD8⁺ T cells were added and incubated overnight. The supernatant was then harvested and assayed for MIP1β by ELISA. For ILA1 and MEL5 CD8⁺ T cells, red, yellow, green, dark blue, mid blue, and light blue depict > 50%, 40%-50%, 30%-40%, 20%-30%, 10%-20%, and 0%-10% MIP1β production (pg/mL) in response to 10⁻⁵ M index peptide, respectively. For 1E6 and SB27 CD8⁺ T cells, red, orange, yellow, green, dark blue, and light blue depict > 100%, 80%-100%, 60%-80%, 40%-60%, 20%-40% and 0%-20% MIP1β production (pg/mL) in response to 10⁻⁵ M index peptide, respectively. Each heat map is representative of a minimum of 3 different replicate experiments, each performed in duplicate.

Figure 3

MHCI-peptide length preference is not determined by MHCI binding. T2 cells (0.5 × 10⁶) were incubated in RPMI 1640 medium with either 100μM HPVGEADYFEY (non-HLA A*0201 binder), 100μM GILGFVFTL (HLA A*0201 binder) or 1mM of the indicated sizing scan mixture (X⁸, X⁹, X¹⁰, X¹¹, X¹², or X¹³) at 26°C for 14-16 hours, then at 37°C for 2 hours, before staining for HLA A*0201 surface expression. The conditions used here are representative of those that were used for the experiments in Figure 1 and supplemental Figure 1. Duplicate samples were acquired for each condition using a FACSCanto II flow cytometer. Data were analyzed with FlowJo software. NP indicates no peptide. Error bars represent SDs.

Figure 4

The 1E6 TCR exhibits a strict preference for 10mer peptides. C1R-HLA A*0201 cells (6 × 10⁴) were pulsed with the indicated 9mer (A-B) or 10mer (C-D) peptides at the concentrations depicted for 2 hours at 37°C. Subsequently, 3 × 10⁴ 1E6 CD8⁺ T cells were added and incubated overnight. The supernatant was then harvested and assayed for MIP1β by ELISA. Error bars represent SDs.

Figure 5

The MEL5 TCR exhibits a strict preference for 10mer peptides. C1R-HLA A*0201 cells (6 × 10⁴) were pulsed with the indicated 10mer (A-B), 9mer (C), 11mer (C), or 8mer (E) peptides at the concentrations depicted for 2 hours at 37°C. Subsequently, 3 × 10⁴ MEL5 CD8⁺ T cells were added and incubated overnight. The supernatant was then harvested and assayed for MIP1β by ELISA. Error bars represent SDs. (D,F) MEL5 CD8⁺ T cells (5 × 10⁴) were incubated with PE-conjugated HLA A*0201 tetramer (25 μg/mL) folded around ELAGIGILTV (black), FWLLPAWAL (red), FWLLGAWAL (blue), FFAGGIGIRTI (cyan), FLAGGIGIRTL (green), WLLPAWGV (yellow), or WLLPTWGV (pink) for 15 minutes at 37°C and then stained with 5 μL of 7-aminoactinomycin D for 30 minutes at 4°C, washed twice, and resuspended in PBS. Negative control staining is shown in dark purple.

Figure 6

Poor recognition of shorter peptides by the SB27 TCR. T2-HLA B*3508 cells (6 × 10⁴) were pulsed with the indicated 9mer (A-B) or 13mer (C-D) peptides at the concentrations depicted for 2 hours at 37°C. Subsequently, 3 × 10⁴ SB27 CD8⁺ T cells were added and incubated overnight. The supernatant was then harvested and assayed for MIP1β by ELISA. Error bars represent SDs.

Figure 7

Peptide length has a major impact on TCR degeneracy. (A) 1E6 CD8⁺ T-cell recognition of a set of 30 11mer peptides sampled by CPL-based importance. (B) 1E6 CD8⁺ T-cell recognition of a set of 30 12mer peptides sampled by CPL-based importance. Assays were conducted as described in Figure 3. (C) CPL-based importance sampling was used to construct a degeneracy curve that describes the number of 11mer and 12mer peptides that can be recognized by the 1E6 TCR. The ordinate is the estimated number of peptides with a functional sensitivity of at least the value indicated on the abscissa. Error bars represent SDs.

Figure 8

The degree of MHCI-peptide backbone “bulging” is proportional to peptide length. Supposition of pMHCI structures with increasing epitope length. The peptides have been superimposed on HLA A*0201, which is shown in cartoon representation. MHCIs were aligned using the MHC α1 and β2 domains. Peptide backbones are shown with ribbon representation and include the 8mer B*3501-VPLRPMTY (orange), 9mer A*0201-ILAKFLHRL (red), 10mer1 A*0201-ELAGIGILTV (green), 10mer A*0201-ALWGPDPAAA (dark green), 11mer B*3508-HPVGEADYFEY (blue), 12mer B*3508-CPSQEPMSIYVY (grey), 13mer B*3508-LPEPLPQGQLTAY (yellow), and 14mer B*3501-LPAVVGLSPGEQEY (purple). (A) Side view demonstrating the different peptide bulges out the MHCI groove depending on the length of the peptide. The scale on the left was calculated using the conserved position of peptide residue 2 from the Cα atom as 0 Å. The MHC α2 helix has been removed for clarity. (B) “Bird's eye view” demonstrating the different sideways displacement of peptides toward MHCI α1 or α2 helices depending on the length of the peptide. The scale on the left was calculated using the average central position of all of the peptides as 0 Å. The MHC α1 (above) and α2 helix (below) are shown. PDB codes are 1A1N, 10GA, 2GT9, 2FZ3, 3BW9, 1ZHL, 1XH3, and 414W.