Fig 1.
Cohort demographics, HLA-A*11:01 population frequencies and HLA-I profiles of Indigenous Australians.
(A) Frequency of HLA-A*11:01 according to geographic region (left panel) and within representative Asian (China, n = 150; Vietnam n = 170; Oman n = 118; Pakistan n = 104; Taiwan n = 55), Indigenous (Papua New Guinea (PNG) Madang, n = 65; Australia Cape York Peninsula Aborigine, n = 103), and Caucasian (Australia Caucasian, n = 134; USA Caucasian, n = 265) populations (right panel). Allele frequencies were obtained from the Allele Frequency Net Database (allelefrequencies.net) HLA classical allele frequency search tool. (B) Circos plot displaying co-expression of HLA-A alleles within the Indigenous LIFT cohort (n = 127, alleles = 2n) and non-Indigenous cohort (n = 223, alleles = 2n). HLA-A*11:01 is depicted with red chords. (C) Frequency of HLA-B (top) and HLA-C (bottom) alleles co-expressed with HLA-A*11:01 in Indigenous LIFT cohort and non-Indigenous individuals (n = 38, alleles = 2n).
Fig 2.
CD8+ T cell responses towards known HLA-A*11:01-restricted IAV epitopes.
(A) Frequency of IFN-γ+ of CD8+ T cells stimulated with pooled and individual previously published HLA-A*11:01-restricted IAV epitopes. CD8+ T cells were expanded with either pool 1a or 1b. Median and IQR are shown (n = 6–7). Representative concatenated FACS plots of one donor are shown for IFN-γ responses to the negative control (DMSO), peptide pool 1a and 1b, M113-21, NP342-351 and M1125-134. (B) Frequency of IFN-γ+ of CD8+ T cells after stimulation with M1121-138 and overlapping peptides within this region (n = 6–7). Representative concatenated FACS plots of one donor are shown of the IFN-γ response to each condition. (C) Frequency of IFN-γ+ of CD8+ T cells after expansion and re-stimulation with variants of M1126-134, with M1121-133 as a negative control. Median and IQR are shown (n = 4). Representative FACS plots are shown for CD8+ T cell responses to M1125-134 and M1126-134 after expansion with either peptide. (D) Proportions of activated CD8+ T cells with 1–4 functions after expansion and re-stimulation with M1125-134 or M1126-134 (n = 4). A permutations test was used to determine statistical significance. (A-B) Friedman test with Dunn’s correction for multiple comparisons was used to determine statistical significance.
Fig 3.
Identification of novel HLA-A*11:01-restricted peptides for influenza A and influenza B viruses.
(A) Length distribution of HLA-A*11:01 ligands mapping to the human proteome (1% FDR, non-redundant by sequence), after filtering of peptides identified as ligands of endogenous HLA of the C1R cell line. Bars show the mean number of peptides of each length assigned in 3, 2 and 2 experiments for uninfected, A/X31-infected and B/Malaysia/2506/04-infected conditions, respectively. Points show peptide numbers for individual replicates. (B) Binding motif for 9–11 amino acid peptides generated from combined non-redundant human sequences (1% FDR) across the seven data sets, where n is the total number of sequences for each length. Sequence logos were generated using Seq2logo 2.0 [46]. (C, D) Length and (E, F) proteome distributions of (C, E) A/X31- and (D, F) B/Malaysia/2506/04-derived ligands (non-redundant by sequence, combined from 2 data sets) identified as strong (SB, % Rank ≤ 0.5), weak (WB, 0.5 < % Rank ≤ 2) or potential binders (PB) of HLA-A*11:01 at a 1% FDR (solid bars) and at scores below the 1% FDR threshold (checked bars). %Rank binding calculated using NetMHC 4.0 [29,30]. Influenza peptide sequences are available in S1 Dataset.
Fig 4.
CD8+ T cell reactivity towards IAV peptides.
(A) Frequency of IFN-γ+CD8+ T cells following stimulation with DMSO (negative control) and IAV peptide pools. Wilcoxon signed-rank test was used to determine statistical significance. Median and IQR are shown (n = 7). Representative concatenated FACS plots from one donor are shown. (B-C) Frequency of IFN-γ+CD8+ T cells after stimulation with individual IAV peptides from (B) Pool A and (C) Pool B. Friedman test with Dunn’s correction for multiple comparisons was used to determine statistical significance. Median and IQR are shown (n = 6–7). Representative FACS plots showing CD8+ T cell responses to (B) PB2320-331 and PB2323-331, and (C) PB1659-669. (D) Polyfunctionality of CD8+ T cells responsive towards PB2323-331, PB2320-331 or PB1659-669, and the associated DMSO negative control. A permutations test was used to determine statistical significance.
Fig 5.
CD8+ T cell responses to novel IBV epitopes.
(A) Frequency of IFN-γ+CD8+ T cells after stimulation with individual IBV peptide pools. Median and IQR are shown (n = 10). Representative FACS plots are shown for CD8+ T cell responses from one donor towards peptide pool 1–3 and the negative control. (B) Frequency of IFN-γ+CD8+ T cells in response to stimulation with individual peptides from Pool 1. Median and IQR are shown (n = 2–9). A representative FACS plot of one donor with a CD8+ T cell response to NP511-520 or NS1186-195. (C) Frequency of IFN-γ+CD8+ T cells in response to stimulation with individual peptides from Pool 2. Median and IQR are shown (n = 2–9). Representative FACS plot of one donor with a CD8+ T cell response to M141-49. (D) Proportions of CD8+ T cells with 1–4 functions after stimulation with NP511-520, NS1186-195 or M141-49 (n = 8 or 9, respectively). A permutations test was used to determine statistical significance. (A-C) Friedman test with Dunn’s correction for multiple comparisons was used to determine statistical significance.
Fig 6.
Structures of IBV peptides complexed with HLA-A*11:01.
(A-C) NP511-520 (green), M141-49 (sand) and NS1186-195 (pink) peptides shown as sticks in complex with HLA-A*11:01 (grey). (D-F) Overlay of (D) NP156-166 (green) and M141-49 (sand) and (E) NP511-520 (green) and NS1186-195 (pink), depicted as sticks, or (F) NP511-520 (green), M141-49 (sand) and NS1186-195 (pink) peptides with P1, P2 and PΩ depicted as sticks and remaining residues as cartoons, in complex with HLA-A*11:01 (grey).
Fig 7.
Immunodominance of HLA-A*02:01 and HLA-A*11:01 IAV and IBV epitopes.
(A) PBMCs harvested from the peripheral whole blood of HLA-A*02:01+/-A*11:01+ donors were infected with either A/X31 or B/Malaysia/2506/04. PBMCs were expanded for 10 days and re-stimulated with C1R cells transduced with HLA-A*02:01 or -A*11:01 peptide-pulsed with HLA-A*02:01 or -A*11:01-restricted influenza peptides, respectively, to determine their immunodominance. (B, C) Frequency of IFN-γ+CD8+ T cells after re-stimulation with IAV- or IBV-infected C1R cells, or A*02:01- or A*11:01-restricted (B) IAV or (C) IBV peptides. Median and IQR are shown (n = 3, one experiment, n = 6, two independent experiments). (B, C) Representative concatenated FACS plots are shown for one donor depicting IFN-γ responses within CD8+ T cells for each stimulation condition. Friedman test with Dunn’s multiple comparisons was used to determine significant differences between peptide stimulation conditions.
Table 1.
Conservation analysis of immunogenic influenza peptides.
Table 2.
Comparison of experimentally identified and predicted HLA-A*11:01-restricted influenza epitopes.