Fig. 6: KF9/C12-specific memory T cells are sustained for 6-12 months in COVID-19 convalescents.

a Frequency of magnitude of T cells response (upper graphs) and responders (lower graphs) to individual overlapping peptide (OLP) pairs in SARS-CoV-2 N proteins (2 μg/ml) or the KF9 peptide (100 nM) in seven COVID-19 convalescents (n = 7) with a haplotype HLA-A24-B52-C12 6 months after infection are shown. Data are expressed as median (upper panel). b Detection of HLA-C*12:02-restricted KF9/C12-specific T cells in PBMCs from seven COVID-19 convalescents (n = 7) by KF9/C12 tetramer. A statistically significant difference is determined by unpaired Mann–Whitney t-test (*p = 0.0167; versus HLA-C*12:02-negative convalescents). c Distribution of naïve, central memory (TCM), effector memory (TEM), and terminally differentiated effector memory cells (TEMRA) among the KF9/C12-specific CD8⁺ T cells in PBMCs from seven COVID-19 convalescents (n = 7) and data are expressed as median. A statistically significant difference is determined by unpaired Mann–Whitney t-test (*p = 0.0312; versus TEM). See Supplementary Fig. 5f (Gating strategy). d, e Celltrace dye-labelled PBMCs were stimulated with N overlapping or KF9 peptides. The frequency (d) and differentiation status (e) of tetramer⁺ KF9/C12-specific CD8⁺ T cells were analysed by flow cytometry after 7 days of stimulation and data are expressed as median. Statistically significant differences are determined by Wilcoxon matched-pairs signed rank test in proliferating KF9/C12-specific T cells (*p = 0.0156; versus no peptide) (d) and the TEM phenotypes at 6 months (*p = 0.0496; versus TEMEA) (e).