Cytotoxic T lymphocytes form a critical component of SARS-CoV-2 immunity by recognizing viral peptides bound to HLA class I molecules. Here, we identified the spike-derived peptide NYNYLYRLF448-456 (NF9) as the immunodominant HLA-A*24:02–restricted epitope in both convalescent and vaccinated donors. Across cohorts, A24/NF9-specific responses were dominated by public TCR motifs featuring TRAV12-1 paired with TRBJ2-7 and a conserved CDR3β sequence (CASSXXXGYEQYF). Using a panel of 13 TCRs, we mapped recognition of single amino acid substitutions within NF9 and identified residue 5 (L452) as the principal determinant of escape. The L452R substitution, characteristic of the Delta variant, abolished recognition across all tested TCRs despite preserved HLA binding. Crystallography of a representative public TCR (P1-15) revealed that mutation at position-5 reoriented the peptide within HLA-A*24:02, flipping the adjacent Y453 side chain into the peptide-binding groove and eliminating the dominant TCR contact. This position-5–driven conformational switch provided a structural mechanism for universal loss of NF9 recognition by HLA-A*24:02–restricted T cells. Consistent with this, Delta-infected convalescents failed to mount de novo NF9-5R–specific responses while retaining responses to the conserved A24/QI9 spike epitope. Together, these findings defined the basis of A24/NF9 recognition and showed how 1 mutation remodeled peptide presentation to abrogate TCR responses.
Takeshi Nakama, Aaron Wall, Garry Dolton, Li Rong Tan, Hannah Thomas, Hiroshi Hamana, Yoshiki Aritsu, Toong Seng Tan, Mako Toyoda, Yoshihiko Goto, Huanyu Li, Mizuki Kitamatsu, Keiko Udaka, Yusuke Miyashita, Hiroyuki Oshiumi, Kimitoshi Nakamura, Yoji Nagasaki, Rumi Minami, Hirotomo Nakata, Pierre J. Rizkallah, Hiroyuki Kishi, Takamasa Ueno, Andrew K. Sewell, Chihiro Motozono
Usage data is cumulative from March 2026 through May 2026.
| Usage | JCI | PMC |
|---|---|---|
| Text version | 979 | 0 |
| 361 | 0 | |
| Figure | 149 | 0 |
| Supplemental data | 77 | 0 |
| Citation downloads | 94 | 0 |
| Totals | 1,660 | 0 |
| Total Views | 1,660 | |
Usage information is collected from two different sources: this site (JCI) and Pubmed Central (PMC). JCI information (compiled daily) shows human readership based on methods we employ to screen out robotic usage. PMC information (aggregated monthly) is also similarly screened of robotic usage.
Various methods are used to distinguish robotic usage. For example, Google automatically scans articles to add to its search index and identifies itself as robotic; other services might not clearly identify themselves as robotic, or they are new or unknown as robotic. Because this activity can be misinterpreted as human readership, data may be re-processed periodically to reflect an improved understanding of robotic activity. Because of these factors, readers should consider usage information illustrative but subject to change.