Identifying SARS-CoV-2–specific T cell epitope–derived peptides is critical for the development of effective vaccines and measuring the duration of specific SARS-CoV-2 cellular immunity. In this regard, we previously identified T cell epitope–derived peptides within topologically and structurally essential regions of SARS-CoV-2 spike and nucleocapsid proteins by applying an immunoinformatics pipeline. In this study, we selected 30 spike- and nucleocapsid-derived peptides and assessed whether these peptides induce T cell responses and avoid major mutations found in SARS-CoV-2 variants of concern. Our peptide pool was highly specific, with only a single peptide driving cross-reactivity in people unexposed to SARS-COV-2, and immunogenic, inducing a polyfunctional response in CD4+ and CD8+ T cells from COVID-19 recovered individuals. All peptides were immunogenic and individuals recognized broad and diverse peptide repertoires. Moreover, our peptides avoided most mutations/deletions associated with all four SARS-CoV-2 variants of concern while retaining their physicochemical properties even when genetic changes are introduced. This study contributes to an evolving definition of individual CD4+ and CD8+ T cell epitopes that can be used for specific diagnostic tools for SARS-CoV-2 T cell responses and is relevant to the development of variant-resistant and durable T cell–stimulating vaccines.

In most cases of SARS-CoV-2 infection, both neutralizing Abs and effector CD4+ and CD8+ T cells are induced, followed by the establishment of T cell and B cell memory (1). Immune control of COVID-19 has been shown to correlate with early induction of innate cytokines, including IFNs (2–4), and CD4+ and CD8+ T cells (5, 6). Vaccination strategies have focused on inducing adaptive immune responses against the SARS-CoV-2 spike (S) protein, which mediates viral entry into cells and is the primary target for neutralizing Abs (7–9). Despite the clear importance of high levels of neutralizing Abs in preventing infection, only low levels are needed to protect against severe disease (10), and there is evidence to suggest that specific CD8+ T cells, supported by CD4+ T cells, are also important for controlling disease (5, 6, 11, 12). Preclinical studies in primates demonstrated a role for CD8+ T cells along with neutralizing Ab in protection from SARS-CoV-2 infection (Ref. 13 and M. Gagne, K.S. Corbett, B.J. Flynn, K.E. Foulds, D.A. Wagner, S.F. Andrew, J.M. Todd, C.C. Honeycutt, L. McCormick, S.T. Nurmukhambetova, manuscript posted on bioRxiv, DOI: 10.1101/2021.10.23.465542). Of note, the efficient development of neutralizing Abs also depends on the induction of S-specific CD4+ Th cells (14). Moreover, SARS-CoV-2–specific memory T and B cells and follicular helper T cells have been detected in lung tissue and lung-associated lymph nodes in humans after recovery from COVID, suggesting at least local coordination of cellular and humoral immune memory (15). This is consistent with the important role for CD4+ and CD8+ T cells in the control and eradication of viral respiratory infections in general (16–18).

Regarding the importance of T cells in immune memory, SARS-CoV-2–specific CD4+ and CD8+ T cells have now been shown to persist for up to 12 mo after mild to moderate disease (12, 19–23). Pre-existing CD4+ and CD8+ T cells specific for seasonal coronaviruses (or SARS-CoV-1) may cross-react with SARS-CoV-2 and could lead to enhanced T cell immunity during COVID, limiting disease (24, 25). Furthermore, T cell immunity may also play a role in broadening immunity to reinfection with new variant strains (26, 27).

The phylogenetic analyses of SARS-CoV-2 sequences have shown that the virus continues to evolve. This viral evolution has contributed to the establishment of B.1.1.7 (Alpha), B.1.351 (Beta), and B.1.617.2 (Delta) variants of concern (VOCs), as well as the currently globally prevalent B.1.1.529 (Omicron) (28–33). These VOCs contain mutations identified within the S glycoprotein (28, 34–44). However, the impact of viral mutations within other protein sequences, such as nucleocapsid (NC), needs to be considered, as cellular immune responses to other viral proteins have been measured in recovered COVID-19 patients (45). Due to this continual viral evolution, identifying T cell epitopes within the mutation-intolerant regions of SARS-CoV-2 proteins is important for the development of new and effective vaccines against circulating and emerging VOCs (46–51). The mutation-intolerant SARS-CoV-2 protein regions can be determined by using protein network analysis (52). This protein network analysis identifies amino acid residues and their locations within a protein that have a large number of physicochemical interactions and spatial interconnections with other residues, contributing to the maintenance of protein structure and function (52–54). Also, the protein regions containing many of these highly networked amino acid residues are genetically conserved across different variants (52). Therefore, determining S and NC protein regions comprised of many of these highly networked amino acid residues can allow for selecting genetically conserved T cell epitopes that can circumvent T cell immune escape mutations associated with currently circulating and emerging VOCs (54).

Testing of CD4+ and CD8+ T cell immunity is commonly done with peptide pools for greater sensitivity rather than recombinant proteins, as peptides do not require extensive proteolytic processing and more efficiently access cross-presentation pathways to stimulate CD4+ and CD8+ T cells, respectively (55). They also allow for precise definition of individual epitopes. Peptide pools are often used as overlapping peptides covering the individual proteins or the whole viral proteome, or can be more selective. To avoid false-negative responses, it is important to examine CD4+ and CD8+ T cell immunity with peptide pools that can avoid mutations in S and other SARS-CoV-2 proteins that will arise as the virus evolves. Peptide pools that define cross-reactivity with other coronaviruses will ensure SARS-CoV-2 specificity and avoid false positives. In this study, we have used a more selective approach utilizing our recently developed immunoinformatics analysis pipeline that identifies topologically and structurally essential regions of the SARS-CoV-2 S and NC proteins containing many of the highly networked amino acid residues that are likely to be genetically conserved (54).

In this study, we employed our immunoinformatics pipeline to produce a pool of SARS-CoV-2–specific peptides, similar in sensitivity to commercial overlapping peptide pools, but with broader recognition of all viral variants and greater specificity to avoid cross-reactivity with other coronaviruses. These peptides recalled broad CD4+ and CD8+ T cell responses in recovered COVID-19 patients, and peptides with cross-reactivity to seasonal coronaviruses were identified. This study contributes to an evolving definition of individual CD4+ and CD8+ T cell epitopes, their individual and pooled cross-reactivity with prior coronavirus infection, and how this cross-reactivity influences subsequent natural T cell responses to COVID. This study provides specific diagnostic tools for SARS-CoV-2 CD4+ and CD8+ T cell responses and is relevant to the selection of Ags and development of universal T cell–stimulating vaccines that cover most global HLA types.

An immunoinformatics pipeline was developed to identify immunogenic peptides from highly networked regions of the SARS-CoV-2 S and NC (54). In our previous study, 57 peptides in total were selected from a consensus sequence derived from 40 SARS-CoV-2 sequences in 2019, and this consensus sequence was 100% identical to the Wuhan-Hu-1 (GenBank accession no. NC_045512). For 9- and 12-mer peptides, we assessed the binding capacity of these peptides to globally prevalent HLA class I (HLA-I; n = 27) and HLA class II (HLA-II; n = 26) alleles, resulting in 80–100% worldwide population coverage. These peptides were identified as clusters containing overlapping peptides when they were mapped to the S and NC protein sequences derived from the 2019 SARS-CoV-2 isolate (GenBank accession no. NC_045512.2). Several clusters contained many hydrophobic amino acid residues that hindered the efficient synthesis of the peptides. Thus, we extended the terminal regions of these clusters to include hydrophilic and polar amino acid residues that were found within the viral sequences. A total of 30 peptides were selected for synthesis at 95% purity (see Table I; Mimotopes, Mulgrave, VIC, Australia) for testing in PBMCs from recovered COVID-19 patients.

Table I.
Peptide information
Protein RegionPeptide IDPeptide SequencesPosition (start)Position (end)LengthHLA
FQF FQFCNDPFL 133 141 HLA-I 
YLQ YLQPRTFLL 269 277 HLA-I 
RVV RVVVLSFEL 509 517 HLA-I 
SII9 SIIAYTMSL 691 699 HLA-I 
SVT SVTTEILPV 721 729 HLA-I 
MIA MIAQYTSAL 869 877 HLA-I 
VLY VLYENQKLI 915 923 HLA-I 
VLN VLNDILSRL 976 984 HLA-I 
VVF VVFLHVTYV 1060 1068 HLA-I 
QLI QLIRAAEIRASANLAATK 1011 1028 18 HLA-II 
QAL QALNTLVKQLSSNFGAI 957 973 17 HLA-II 
MAY MAYRFNGIGVTQNVLY 902 917 16 HLA-II 
IPF IPFAMQMAYRFNGI 896 909 14 HLA-II 
GAA GAALQIPFAMQMAYRF 891 906 16 HLA-II 
EMI EMIAQYTSALLA 868 879 12 HLA-II 
SII14 SIIAYTMSLGAENS 691 704 14 HLA-II 
QSI QSIIAYTMSLGA 690 701 12 HLA-II 
KGI KGIYQTSNFRVQ 310 321 12 HLA-II 
EKG EKGIYQTSNFRV 309 320 12 HLA-II 
YVG YVGYLQPRTFLL 266 277 12 HLA-II 
FSN FSNVTWFHAIHVSGT 59 73 15 HLA-II 
LPF LPFFSNVTWFHAIHV 56 70 15 HLA-II 
STQ STQDLFLPFFSNVTWFH 50 66 17 HLA-II 
NC ILL ILLNKHIDA 351 359 HLA-I 
NC RTA RTATKAYNV 262 270 HLA-I 
NC ALN ALNTPKDHI 138 146 HLA-I 
NC IIW IIWVATEGA 130 138 HLA-I 
NC GII GIIWVATEGALN 129 140 12 HLA-II 
NC GYY GYYRRATRRIRG 85 96 12 HLA-II 
NC QIG QIGYYRRATRRIR 83 95 13 HLA-II 
Protein RegionPeptide IDPeptide SequencesPosition (start)Position (end)LengthHLA
FQF FQFCNDPFL 133 141 HLA-I 
YLQ YLQPRTFLL 269 277 HLA-I 
RVV RVVVLSFEL 509 517 HLA-I 
SII9 SIIAYTMSL 691 699 HLA-I 
SVT SVTTEILPV 721 729 HLA-I 
MIA MIAQYTSAL 869 877 HLA-I 
VLY VLYENQKLI 915 923 HLA-I 
VLN VLNDILSRL 976 984 HLA-I 
VVF VVFLHVTYV 1060 1068 HLA-I 
QLI QLIRAAEIRASANLAATK 1011 1028 18 HLA-II 
QAL QALNTLVKQLSSNFGAI 957 973 17 HLA-II 
MAY MAYRFNGIGVTQNVLY 902 917 16 HLA-II 
IPF IPFAMQMAYRFNGI 896 909 14 HLA-II 
GAA GAALQIPFAMQMAYRF 891 906 16 HLA-II 
EMI EMIAQYTSALLA 868 879 12 HLA-II 
SII14 SIIAYTMSLGAENS 691 704 14 HLA-II 
QSI QSIIAYTMSLGA 690 701 12 HLA-II 
KGI KGIYQTSNFRVQ 310 321 12 HLA-II 
EKG EKGIYQTSNFRV 309 320 12 HLA-II 
YVG YVGYLQPRTFLL 266 277 12 HLA-II 
FSN FSNVTWFHAIHVSGT 59 73 15 HLA-II 
LPF LPFFSNVTWFHAIHV 56 70 15 HLA-II 
STQ STQDLFLPFFSNVTWFH 50 66 17 HLA-II 
NC ILL ILLNKHIDA 351 359 HLA-I 
NC RTA RTATKAYNV 262 270 HLA-I 
NC ALN ALNTPKDHI 138 146 HLA-I 
NC IIW IIWVATEGA 130 138 HLA-I 
NC GII GIIWVATEGALN 129 140 12 HLA-II 
NC GYY GYYRRATRRIRG 85 96 12 HLA-II 
NC QIG QIGYYRRATRRIR 83 95 13 HLA-II 

Previously hospitalized SARS-CoV-2 convalescent patients were recruited from Westmead Hospital (Westmead, Sydney, NSW, Australia) during 2020–2021. A total of 33 participants were included in this study at 1–4 mo after SARS-CoV-2 recovery (see Table II). Their level of disease ranged from mild to severe. All donors were presumed to be infected with the 2019 SARS-CoV-2 strain, as no other variants had been identified in Australia at the time of infection, apart from one donor who was confirmed as infected with the Delta variant during 2021. Samples from four additional donors who were 1 mo postrecovery from infection with the Delta variant were provided by the COSIN (Collection of COVID-19 Outbreak Samples in NSW) biobank.

Table II.
Cohort demographics
Recovered COVID-19 ParticipantsPre–COVID-19 Participants
Total participants n = 37 n = 30 
 Age (y, range) 56.6 (24–81) 41.1 (25–75) 
 Male (%) 21 (56.8) 16 (53.3) 
 Hospitalized (%) 15 (40.5) NA 
 Days since symptom onset (IQR) 55 (45–100) NA 
Total HLA-A2+ participants n = 9  
 Age (y, range) 57.0 (27–75)  
 Male (%) 4 (44.5)  
 Hospitalized (%) 5 (55.6)  
 Days since symptom onset (IQR) 34 (49–83)  
Recovered COVID-19 ParticipantsPre–COVID-19 Participants
Total participants n = 37 n = 30 
 Age (y, range) 56.6 (24–81) 41.1 (25–75) 
 Male (%) 21 (56.8) 16 (53.3) 
 Hospitalized (%) 15 (40.5) NA 
 Days since symptom onset (IQR) 55 (45–100) NA 
Total HLA-A2+ participants n = 9  
 Age (y, range) 57.0 (27–75)  
 Male (%) 4 (44.5)  
 Hospitalized (%) 5 (55.6)  
 Days since symptom onset (IQR) 34 (49–83)  

IQR, interquartile range; NA, not applicable.

Blood samples were collected by venipuncture into lithium heparin tubes to be processed within 1 h. PBMCs were then isolated by Ficoll density gradient centrifugation and cryopreserved. The HLA genotype for each donor was determined by next-generation sequencing on 11 loci, using genomic DNA extracted from whole blood or PBMCs using a DNA blood mini kit (QIAamp). Stored PBMCs from prepandemic blood samples collected prior to 2019 from 30 Australian Red Cross Lifeblood donors were also used (see Table II).

Cryopreserved PBMCs were thawed at 37°C in RF10 (RPMI 1640 medium supplemented with 10% FCS and 1× penicillin-streptomycin, Thermo Fisher Scientific) and rested for 18 h at 37°C and 5% CO2. Cell viability was assessed by trypan blue exclusion.

Dual-color FluoroSpot plates precoated with anti–IFN-γ and anti–IL-2 capture Abs (Mabtech) were blocked with RPMI 1640 medium for 30 min prior to the addition of 150,000–300,000 PBMCs per well. Triple-color FluoroSpot plates also including anti–TNF-α capture Abs were used on a subset of samples. Each well was stimulated with either 1) a customized highly networked (HN) SARS-CoV-2 peptide pool, 2) customized HN SARS-CoV-2 individual peptides, 3) SARS-CoV-2 S PepTivator (combination of Prot_S, Prot_S1, and Prot_S+, Miltenyi Biotec), 4) SARS-CoV-2 NC PepTivator (Prot_N, Miltenyi Biotec), 5) EBV PepTivator (EBV Consensus, Miltenyi Biotec), or 6) PHA-L, in addition to two unstimulated wells (RF10 only). The concentration of each peptide was 1 μg/ml in the corresponding pool and PHA-L was 10 μg/ml. The concentration of the individual peptides was 4 μg/ml. Plates were incubated at 37°C and 5% CO2 for 20 h. The cells and medium were then decanted from the plate and the assay was developed according to the manufacturer’s instructions.

Fluorescent spots indicating cells that secreted IFN-γ (FITC filter) and/or IL-2 (Cy3 filter) and TNF-α (Cy5 filter) were detected with an IRIS FluoroSpot ELISPOT reader (Mabtech). The mean numbers of responding cells in negative controls were subtracted from stimulated samples to account for background responses, and results were expressed as spot-forming units (SFU) per 106 cells or as average relative spot volume.

The PBMCs were thawed in RF10 and rested overnight in RF10. Rested PBMCs (1–5 × 106 cells) were incubated in the presence of 4 μg/ml HN SARS-CoV-2 peptides for 1 h in AIM-V medium (Thermo Fisher Scientific). After the incubation with these peptides, cells were washed with RF10. The stimulated cells were cultured in 48-well plates at a density of 2 × 106 cells/ml in RF10 medium with 1 μg/ml anti-CD28 Ab (clone L293; BD Biosciences) for costimulation. After 24 h cells were supplemented with 100 IU/ml IL-2 (Lonza) and cultured for 14 d. The medium was replaced every 72 h with freshly prepared RF10 supplemented with IL-2.

The effector cytokine production and polyfunctionality of T cells were evaluated using the peptide-expanded cells. The cells were restimulated with 4 μg/ml HN SARS-CoV-2 peptides in the presence of costimulatory Abs (1 μg/ml anti-CD28 and anti-CD49d; BD Biosciences), monensin (GolgiStop, 0.9 μl/ml; BD Biosciences) and brefeldin A (GolgiPlug 1.4 μl/ml; BD Biosciences) for 5 h at 37°C. Anti–CD107a/b-FITC Abs (BD Biosciences) were also added to identify degranulating cells. Following stimulation, the cells were stained with LIVE/DEAD fixable near-IR dead cell stain kit (Thermo Fisher Scientific) and the following conjugated Abs: anti–CD3-BUV496, anti–CD8-PerCP-Cy5.5 (BD Biosciences), and anti–CD4-allophycocyanin (BioLegend). Cells were then fixed, permeabilized (Cytofix/Cytoperm; BD Biosciences), and stained using anti–IL-2-BV421, anti–TNF-α-PE/Cy7, and anti–IFN-γ-PE Abs (BD Biosciences). The cells were acquired on an LSRFortessa X-20 flow cytometer (BD Biosciences), and the data were analyzed using FlowJo v10. For the analysis, we first gated cells based on their morphology using forward scatter height versus side scatter area, followed by the exclusion of doublets using forward scatter area versus forward scatter height gating. Viable cells were then selected using a LIVE/DEAD staining assay to exclude dead cells. Subsequently, we quantified the frequencies of degranulating and cytokine-producing cells among CD3+ and either CD4+ or CD8+ T cells. To study the polyfunctionality of the T cells, Boolean gating on IFN-γ, IL-2, TNF-α, and CD107a/b, the latter for CD8+ T cells only, was performed in FlowJo and analyzed using SPICE 6.0 software (https://niaid.github.io/spice/) following the technical considerations published by the software developers (56).

To assess the genetic variability of SARS-CoV-2 S and NC protein sequences derived from Alpha (B.1.1.7 lineage and Q sublineages), Beta (B.1.351 and its sublineages), and Delta (B.1.617.2 and all AY sublineages) VOCs, whole-genome viral sequences were downloaded from Global Initiative on Sharing All Influenza Data (GISAID; https://www.gisaid.org/) (57, 58). More than 1.7 million SARS-CoV-2 genomic sequences isolated from December 2019 to August 2021 were compiled. These genomic sequences were aligned to the SARS-CoV-2 reference sequence isolated in December 2019 (GenBank accession no. NC_045512.2) using MAFFT (59). Next, the regions encoding S and NC protein sequences were extracted and translated using Biostrings, stringr, and seqinr R packages, followed by refined sequence alignment (60–62). The sequences containing >5% ambiguous sites were removed during the sequence alignment. The sequences that contained >50% amino acid positions that do not align with other protein sequences were removed. In addition, the protein sequences that were shorter than 1273 aa for S and 419 aa for NC were removed. The genetically identical protein sequences were represented by one unique sequence. The resulting sequence alignments contained 66,518 and 14,949 genetically unique sequences for the S and NC proteins. For S protein sequences, the alignment contained data from Alpha (n = 58,964), Beta (n = 7,554), and Delta VOCs (n = 76,734). The alignment contained 3,094 Alpha, 826 Beta, and 11,029 Delta VOC sequences for the NC protein region.

To determine the distribution of mutations associated with the VOCs within our sequence alignments, we assessed the proportion of S protein sequences containing the mutations that define Alpha, Beta, and Delta SARS-CoV-2 variants (28–31). For each VOC, we also assessed the proportion of S protein sequences that contain previously characterized T cell escape mutations (47, 48, 50, 63, 64). Moreover, this study generated a comprehensive list of mutations associated with each VOC at individual sites for the NC protein region. This proportion of the sequences containing the mutations was calculated by using the epiDisplay R package (65). The resulting proportions of NC protein sequences were categorized into four different percent ranges (i.e., <1%, 1 to <10%, 10 to <80%, and 80–100%). We noted all of the positions that contained amino acid residues different from the SARS-CoV-2 2019 isolate. Any mutation that appears within ≥1% of S and NC protein sequences derived from each VOC was considered a significant contributor to the genetic diversity of the viral variants (66, 67). We assessed whether the 30 peptides contained mutations associated with SARS-CoV-2 VOCs by using Geneious version 8.1.9 (68).

In August 2022, the lineage comparison report was made available on GISAID, allowing for a comprehensive analysis of the mutations associated with the Omicron VOC and its sublineages (69). Therefore, we used this GISAID lineage comparison report for >4.5 million SARS-CoV-2 sequences to identify the mutations and deletions associated with the Omicron and its sublineages, BA.2, BA.5, and BA.2.75 (accessed in August 2022) (70–71). All mutations that were found within at least 1% of the S and NC protein sequences derived from the Omicron variants were compared with the 30 peptides by using Geneious version 8.1.9.

In this study, we determined the impact of the VOC-associated substitution mutations and/or deletions on the physicochemical properties of the S-derived peptides containing these genetic changes. To do this, we assessed whether these mutations/deletions changed the hydrophobicity, isoelectric point, and size of the peptides. The peptide sequences with and without the VOC-associated mutations/deletions were generated. The hydrophobicity of each amino acid residue within the peptide sequences was calculated using Geneious software (68). For each of the peptide sequences, the hydrophobicity values of individual amino acid residues were combined and divided by the length of the peptide sequence. This was calculated for the peptides with and without mutations/deletions associated with the VOCs. The isoelectric point for each individual peptide was computed by using EMBOSS Pepstats software (72). The isoelectric point of the peptide is the pH at which the peptide carries no net electrical charge. For the size of each peptide sequence, we calculated the following properties using ChemMine Tools: 1) molecular mass (unit: Da), and 2) geometric diameter (unit: Å) (73). To perform this calculation, the peptide sequences with or without the VOC mutations/deletions were converted to the simplified molecular-input line-entry system (SMILES) using PepSMI online software (https://www.novoprolabs.com/tools/convert-peptide-to-smiles-string). When this conversion is completed, the chemical structure of a peptide sequence is represented as a linear format (i.e., string). The linear format of each peptide sequence was used as the input for the ChemMine Tools software. The geometric diameter calculated by the ChemMine Tools represents the maximum distance from one atom to the other within the peptide sequence (74). Therefore, the geometric diameter indicates the maximum length of a peptide chemical structure.

To determine whether parametric tests are appropriate for the ex vivo data, we assessed the departure of the data from a Gaussian distribution by applying an Anderson–Darling test, D’Agostino–Pearson test, Shapiro–Wilk test, and Kolmogorov–Smirnov test. The ex vivo data significantly deviated from the Gaussian distribution; therefore, nonparametric alternatives were used for all of the comparisons throughout this study. Specifically, a Friedman’s test was used to compare three or more paired groups using Stata 15 (release 15; StataCorp, College Station, TX). A two-tailed Mann–Whitney or Wilcoxon signed rank tests were used to compare two unpaired or paired groups for a single variable using GraphPad Prism 9. For all comparisons, statistical significance was considered when p <0.05.

The study protocol was approved by the Human Research Ethics Committees of the Western Sydney Local Health District (North Parramatta, NSW, Australia; 2020/ETH00844) and was conducted according to the Declaration of Helsinki and International Conference on Harmonization Good Clinical Practice guidelines and local regulatory requirements. Written informed consent was obtained from all participants before study procedures.

We have previously identified HN T cell epitope–derived peptides within topologically and structurally important regions of SARS-CoV-2 S and NC proteins and selected 26 peptides for ex vivo assessment (54). For the HN peptides that were ≥12 mer in length (n = 15), we included the T cell epitope–derived peptides that were predicted to bind to more than two HLA-II alleles. For the HN peptides from the S protein that were 9 mer in length (n = 8), we selected peptides that had a wide range of predicted peptide:HLA complex stabilities (t1/2 range, 21–641 min) (54). For the NC protein, we included three highly networked 9-mer peptides that were the most promising for HLA-I Ag presentation and immunogenicity (54). Although we reported the predicted HLA-I restriction of the HN peptides previously (54), we have not analyzed this in detail for HLA-II. The latter requires extensive binding assays (75), as prediction algorithms are less accurate for HLA-II than for HLA-I (76).

In addition to the 26 peptides selected from our in silico analysis, we included four additional peptides that are known to elicit an immune response against SARS-CoV-1 and three of these peptides (S derived, QLIRAAEIRASANLAATK and VVFLHVTYV; NC derived, ALNTPKDHI) are listed as SARS-CoV-2 T cell epitopes in Immune Epitope Database (IEDB) (77–81). The fourth peptide (QALNTLVKQLSSNFGAI [QAL]), identified within the S protein, contained two HN 9-mer peptides that were predicted to be the most promising for HLA-I Ag presentation and immunogenicity by our in silico analysis pipeline (i.e., ALNTLVKQL and KQLSSNFGA) (54). Therefore, a total of 30 peptides were synthesized at 95% purity for ex vivo assessment (Table I).

To validate our pipeline, we collected PBMCs from 37 participants who had recovered from mild to severe COVID-19, 1–4 mo prior, during the wave of the original 2019 strain or the 2021 wave of the Delta VOC (Table II). The 30 selected HN SARS-CoV-2 T cell peptides were pooled and tested for T cell immunogenicity on PBMCs from 29 of these participants with various HLA types (Table III). The production of IFN-γ and IL-2 by effector T cells was assessed by dual-color FluoroSpot (Fig. 1A). Commercial overlapping peptide pools covering the complete S and NC proteins were used for comparison. IFN-γ and/or IL-2 production was observed in 22 of 29 (76%) participants in response to our HN SARS-CoV-2 peptide pool and 19 (66%) participants showed dual-positive cells producing both cytokines (Fig. 1B). There were no statistically significant differences between responses to the commercial S and NC pools and the HN peptide pool. The responses from participants infected with the Delta strain were indistinguishable from those infected with the 2019 strain.

Table III.
HLA-A, HLA-B, DRB1, and DQB1 allele frequency in the donor cohort (n = 32)
HLA Class IHLA Class II
AFrequency (%)BFrequency (%)DRB1Frequency (%)DQB1Frequency (%)
A*01 9.38 B*07 15.63 DRB1*01 15.63 DQB1*02 40.63 
A*02 31.25 B*08 9.38 DRB1*02 3.13 DQB1*03 50.00 
A*03 18.75 B*13 9.38 DRB1*03 15.63 DQB1*04 3.13 
A*11 31.25 B*14 3.13 DRB1*04 21.88 DQB1*05 34.38 
A*23 9.38 B*15 21.88 DRB1*07 31.25 DQB1*06 31.25 
A*24 28.13 B*18 12.50 DRB1*08 6.25   
A*25 3.13 B*27 9.38 DRB1*09 15.63   
A*26 3.13 B*35 18.75 DRB1*10 9.38   
A*29 3.13 B*37 6.25 DRB1*11 6.25   
A*30 6.25 B*39 6.25 DRB1*12 12.50   
A*31 3.13 B*40 9.38 DRB1*13 9.38   
A*32 6.25 B*44 21.88 DRB1*14 9.38   
A*33 12.50 B*46 6.25 DRB1*15 28.13   
A*34 6.25 B*47 3.13 DRB1*16 3.13   
A*68 9.38 B*48 3.13     
  B*51 12.50     
  B*54 3.13     
  B*56 3.13     
  B*58 3.13     
HLA Class IHLA Class II
AFrequency (%)BFrequency (%)DRB1Frequency (%)DQB1Frequency (%)
A*01 9.38 B*07 15.63 DRB1*01 15.63 DQB1*02 40.63 
A*02 31.25 B*08 9.38 DRB1*02 3.13 DQB1*03 50.00 
A*03 18.75 B*13 9.38 DRB1*03 15.63 DQB1*04 3.13 
A*11 31.25 B*14 3.13 DRB1*04 21.88 DQB1*05 34.38 
A*23 9.38 B*15 21.88 DRB1*07 31.25 DQB1*06 31.25 
A*24 28.13 B*18 12.50 DRB1*08 6.25   
A*25 3.13 B*27 9.38 DRB1*09 15.63   
A*26 3.13 B*35 18.75 DRB1*10 9.38   
A*29 3.13 B*37 6.25 DRB1*11 6.25   
A*30 6.25 B*39 6.25 DRB1*12 12.50   
A*31 3.13 B*40 9.38 DRB1*13 9.38   
A*32 6.25 B*44 21.88 DRB1*14 9.38   
A*33 12.50 B*46 6.25 DRB1*15 28.13   
A*34 6.25 B*47 3.13 DRB1*16 3.13   
A*68 9.38 B*48 3.13     
  B*51 12.50     
  B*54 3.13     
  B*56 3.13     
  B*58 3.13     

HLA typing was performed for 32 of the 37 donors in this study. Next-generation sequencing data for the DQB1 allele were not obtained for two donors.

FIGURE 1.

T cell response induced by highly networked SARS-CoV-2 peptide pool.

Selected highly networked (HN) SARS-CoV-2 peptides were combined in a single pool and tested in PBMCs from recovered COVID-19 participants. Alternatively, cells were stimulated with commercial spike (S) or nucleocapsid (NC) overlapping peptide pools or PHA. (AC) Cytokine production was assessed by FluoroSpot (A and B) and flow cytometry (C). (A) Representative FluoroSpot showing IFN-γ (green spots) and IL-2 (red spots) production by PBMCs from three recovered COVID-19 donors. (B) Quantification of spot-forming units (SFU) per 106 cells for IFN-γ, IL-2, or dual-secreting cells. Each data point represents a single donor (n = 29) with bars showing the median. Green dots represent values obtained from recovered COVID-19 participants infected with the Delta variant of SARS-CoV-2. All others were infected with the ancestral strain. Statistical significance was assessed using Friedman’s test, and p values were not significant (IFN-γ, p = 0.147; IL-2, p = 0.200; IFN-γ/IL-2, p = 0.132). (C) Representative intracellular flow cytometry data showing effector cytokine production and degranulation in CD4+ T cells and CD8+ T cells expanded in vitro with the HN SARS-CoV-2 peptide pool. Pie charts depict the average distribution of mono-, bi-, tri-, and tetrafunctional SARS-CoV-2–specific CD4+ and CD8+ T cells.

FIGURE 1.

T cell response induced by highly networked SARS-CoV-2 peptide pool.

Selected highly networked (HN) SARS-CoV-2 peptides were combined in a single pool and tested in PBMCs from recovered COVID-19 participants. Alternatively, cells were stimulated with commercial spike (S) or nucleocapsid (NC) overlapping peptide pools or PHA. (AC) Cytokine production was assessed by FluoroSpot (A and B) and flow cytometry (C). (A) Representative FluoroSpot showing IFN-γ (green spots) and IL-2 (red spots) production by PBMCs from three recovered COVID-19 donors. (B) Quantification of spot-forming units (SFU) per 106 cells for IFN-γ, IL-2, or dual-secreting cells. Each data point represents a single donor (n = 29) with bars showing the median. Green dots represent values obtained from recovered COVID-19 participants infected with the Delta variant of SARS-CoV-2. All others were infected with the ancestral strain. Statistical significance was assessed using Friedman’s test, and p values were not significant (IFN-γ, p = 0.147; IL-2, p = 0.200; IFN-γ/IL-2, p = 0.132). (C) Representative intracellular flow cytometry data showing effector cytokine production and degranulation in CD4+ T cells and CD8+ T cells expanded in vitro with the HN SARS-CoV-2 peptide pool. Pie charts depict the average distribution of mono-, bi-, tri-, and tetrafunctional SARS-CoV-2–specific CD4+ and CD8+ T cells.

Close modal

Polyfunctionality, defined as the induction of two or more functions, of CD4+ and CD8+ T cells was further assessed in PBMCs from 16 recovered COVID-19 participants by intracellular cytokine staining (ICS) for IFN-γ, IL-2 and TNF-α. In addition, the capacity to degranulate was evaluated on CD8+ T cells by measuring the mobilization of CD107a/b to the plasma membrane. These donors were selected based on having demonstrated a FluoroSpot response and their PBMCs were expanded in vitro with the peptide pool to increase the frequency of rare peptide-specific T cells. Thus, the assay measured recalled memory responses to the peptide pool, an important component of durable immunity. A polyfunctional CD4+ T cell response to the HN peptide pool was observed in all 16 donors tested and a CD8+ T cell response was observed in 9 of 16 (56%) donors tested. This response was dominated by IFN-γ and TNF-α production, a Th1 response, in CD4+ T cells and by the expression of IFN-γ, TNF-α and CD107a/b in CD8+ T cells (Fig. 1C). TNF-α expression by effector cells was also confirmed by FluoroSpot (Supplemental Fig. 1A), indicating in vivo relevance of this response.

To identify which peptides from our selected pool were responsible for the T cell responses observed, the 30 HN SARS-CoV-2 peptides were tested individually in PBMCs from recovered COVID-19 participants by FluoroSpot. The 17 HN SARS-CoV-2 class II peptides were tested in 37 participants (Table II); however, because our class I peptides were all predicted to bind to HLA-A*02 molecules, the 13 HN SARS-CoV-2 class I peptides were tested individually in 7 participants who were HLA-A*02 positive (Table III). The response to the HN peptides was broad with each participant responding to several peptides. All peptides, including S and NC peptides from both class I and class II, were immunogenic, eliciting IFN-γ and/or IL-2 responses in various participants (Fig. 2A, 2B). The repertoire of peptide responses varied between participants with 28 of 32 (87%) people recognizing one or more peptides evidenced by IFN-γ or IL-2 production, or both. The average number of peptides recognized by each person was 6.7 for class I and 6.9 for class II peptides. The median number of participants recognizing each class II peptide was 12 of 32 (interquartile range [IQR], 12–14), whereas the median number of participants recognizing each class I peptide was 3 of 7 (IQR, 3–6), but five of these class I peptides were recognized by at least 5 of the 7 HLA-A*02+ individuals tested.

FIGURE 2.

T cell response to individual NC and S-derived HN SARS-CoV-2 peptides.

Cytokine production by PBMCs from recovered COVID-19 participants exposed to HN SARS-CoV-2 individual peptides was assessed by FluoroSpot. (A and B) Quantification of spot-forming units (SFU) per 106 cells secreting IFN-γ and IL-2 after stimulation with MHC class II (A)– and MHC class I (B)–restricted peptides or commercial S and NC overlapping peptide pools. Each data point represents a single donor (n = 32 for MHC class II and n = 7 for MHC class I) and bars show the median. Green dots represent values obtained from a recovered COVID-19 participant infected with the Delta variant of SARS-CoV-2.

FIGURE 2.

T cell response to individual NC and S-derived HN SARS-CoV-2 peptides.

Cytokine production by PBMCs from recovered COVID-19 participants exposed to HN SARS-CoV-2 individual peptides was assessed by FluoroSpot. (A and B) Quantification of spot-forming units (SFU) per 106 cells secreting IFN-γ and IL-2 after stimulation with MHC class II (A)– and MHC class I (B)–restricted peptides or commercial S and NC overlapping peptide pools. Each data point represents a single donor (n = 32 for MHC class II and n = 7 for MHC class I) and bars show the median. Green dots represent values obtained from a recovered COVID-19 participant infected with the Delta variant of SARS-CoV-2.

Close modal

We extended the analysis of T cell polyfunctionality by ICS for the individual HN SARS-CoV-2 peptides on PBMCs from nine participants that were responsive by FluoroSpot and expanded in vitro with the HN peptide pool (Fig. 3A, 3B). The selection of HN peptides to be tested was defined by an individual’s responsiveness to each peptide detected by FluoroSpot. We observed a polyfunctional response in both CD4+ and CD8+ T cells to several peptides from both S and NC, shown in Fig. 3C, 3D, which included CD8+ T cell responses to embedded epitopes within class II peptides. Results are shown for peptides that elicited a CD4+ or CD8+ T cell response in more than one donor. Thus, CD8+ T cell responses are only shown for S-derived HN peptides in this study, but responses to NC-derived HN peptides were observed in individual donors (54). As for the peptide pools, the responses were dominated by IFN-γ and TNF-α production in CD4+ T cells and by the expression of IFN-γ, TNF-α, and CD107a/b in CD8+ T cells (Fig. 1C, 1D). High levels of serum TNF-α during acute illness are associated with increased disease severity (82). In line with this, we found a trend toward increased TNF-α production by CD4+ T cells in patients who had recovered from severe disease (i.e., were hospitalized with COVID-19) compared with patients who had recovered from mild symptomatic disease (i.e., were not hospitalized); however, the differences were not statistically significant (Supplemental Fig. 1B).

FIGURE 3.

Polyfunctionality of the T cell response to individual NC and S-derived HN SARS-CoV-2 peptides.

T cell polyfunctionality analysis after in vitro expansion with SARS CoV-2 peptides. (A and B) Representative dot plots showing intracellular cytokine production and degranulation level in CD4+ T cells (A) and CD8+ T cells (B) induced by individual HN SARS-CoV-2 peptides. (C and D) Pies depict the average distribution across n reactive donors of mono-, bi-, tri-, and tetrafunctional cells within SARS-CoV-2–specific CD4+ T cells (C) and CD8+ T cells (D).

FIGURE 3.

Polyfunctionality of the T cell response to individual NC and S-derived HN SARS-CoV-2 peptides.

T cell polyfunctionality analysis after in vitro expansion with SARS CoV-2 peptides. (A and B) Representative dot plots showing intracellular cytokine production and degranulation level in CD4+ T cells (A) and CD8+ T cells (B) induced by individual HN SARS-CoV-2 peptides. (C and D) Pies depict the average distribution across n reactive donors of mono-, bi-, tri-, and tetrafunctional cells within SARS-CoV-2–specific CD4+ T cells (C) and CD8+ T cells (D).

Close modal

Between 40 and 75% of individuals unexposed to SARS-CoV-2 have been reported to show a positive T cell response to SARS-CoV-2 epitopes due to the high homology between SARS-CoV-2 and seasonal coronaviruses (83–85). This cross-immunogenicity could lead to an overestimation of the specific response to SARS-CoV-2 epitopes, which could in turn give a false sense of protection over time against reinfection and severe disease. We thus questioned whether our selected HN SARS-CoV-2 peptides, with a predicted low homology to seasonal human coronaviruses, avoid a cross-reactive T cell response in individuals unexposed to SARS-CoV-2. To assess this, we exposed PBMCs collected from 20 healthy blood donors prior to 2019 to our HN SARS-CoV-2 peptide pool. In parallel we exposed the cells to commercial SARS-CoV-2 S and NC overlapping peptide pools. Whereas the commercial NC pool elicited cross-reactivity in only a handful of unexposed participants (5 of 20 [25%] IFN-γ+, and 2 of 20 [10%] IL-2+), the commercial S peptide pool was highly cross-reactive, inducing production of IFN-γ in 15 of 20 (75%) and IL-2 in 5 of 20 (25%) pre-COVID-19 participants. The number of unexposed people responding to S was approaching the number of recovered participants responding, although the magnitude of the response was still significantly lower in unexposed participants. In contrast, IFN-γ responses were only seen in 6 of 20 (30%) unexposed participants with our HN SARS-CoV-2 peptide pool (Fig. 4A). This response rate was significantly lower than the response rate in recovered participants. The relative amount of secreted IFN-γ (or mean relative spot volume) in response to either the commercial S peptide pool or our own HN SARS-CoV-2 peptide pool was also significantly lower in cells from pre–COVID-19 individuals when compared with recovered COVID-19 participants (Supplemental Fig. 2A).

FIGURE 4.

Cross-reactive T cell response to HN SARS-CoV-2 peptides.

Cytokine production by PBMCs from donors unexposed to SARS-CoV-2 (pre–COVID-19) stimulated with SARS-CoV-2 peptides was assessed by FluoroSpot. (A) Quantification of spot-forming units (SFU) per 106 cells secreting IFN-γ and IL-2 after stimulation with SARS-CoV-2 peptide pools. Data obtained from COVID-19 recovered donors are shown for comparison. Green dots represent values obtained from recovered COVID-19 participants infected with the Delta variant of SARS-CoV-2. All others were infected with the ancestral strain. (B) Quantification of SFU per 106 cells of IFN-γ and IL-2 and after stimulation with individual MHC class I (A)– and MHC class II (B)–restricted peptides. Each dot color represents a single donor. (C) Quantification of SFU per 106 cells of IFN-γ and IL-2 after stimulation with SARS-CoV-2 peptide pools. Each data point represents a single donor and bars show the median. Statistical significance was assessed using the Mann–Whitney test. * p ≤ 0.05, ** p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001; ns, not significant.

FIGURE 4.

Cross-reactive T cell response to HN SARS-CoV-2 peptides.

Cytokine production by PBMCs from donors unexposed to SARS-CoV-2 (pre–COVID-19) stimulated with SARS-CoV-2 peptides was assessed by FluoroSpot. (A) Quantification of spot-forming units (SFU) per 106 cells secreting IFN-γ and IL-2 after stimulation with SARS-CoV-2 peptide pools. Data obtained from COVID-19 recovered donors are shown for comparison. Green dots represent values obtained from recovered COVID-19 participants infected with the Delta variant of SARS-CoV-2. All others were infected with the ancestral strain. (B) Quantification of SFU per 106 cells of IFN-γ and IL-2 and after stimulation with individual MHC class I (A)– and MHC class II (B)–restricted peptides. Each dot color represents a single donor. (C) Quantification of SFU per 106 cells of IFN-γ and IL-2 after stimulation with SARS-CoV-2 peptide pools. Each data point represents a single donor and bars show the median. Statistical significance was assessed using the Mann–Whitney test. * p ≤ 0.05, ** p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001; ns, not significant.

Close modal

Next, to identify which HN peptides were responsible for inducing the cross-reactive T cell response, we tested our HN peptides individually (Fig. 4B). The HN S-derived peptide QAL showed the strongest response and most consistent cross-reactivity, indicating that it is likely the main driver of the cross-reactivity in the HN peptide pool for the donors tested in our study. To confirm this, we assessed the production of IFN-γ and IL-2 in PBMCs from a second group of 10 different individuals collected prior to the SARS-CoV-2 pandemic, treated with the HN SARS-CoV-2 peptide pool including or excluding the peptide QAL (Fig. 4C). Removing QAL significantly reduced the cross-immunogenicity of our HN SARS-CoV-2 peptide pool for IFN-γ in unexposed individuals (Fig. 4C) but did not significantly change the SARS-CoV-2–specific immunogenic response in the recovered COVID participants (Supplemental Fig. 2B). Indeed, several participants who did not respond to QAL still exhibited strong responses to the HN peptide pool (Supplemental Fig. 2C). Overall, our results indicate that the T cell epitopes derived from NC and S from SARS-CoV-2 identified by our immunoinformatics pipeline are immunogenic and highly specific when compared with commercial S-derived peptide pools.

To validate whether the peptides selected for ex vivo assessment avoided mutations identified within Alpha, Beta, and Delta VOCs, we compared mutations found within the extracted S protein region to the selected peptides to determine whether any of the peptides contain the mutations associated with VOCs. When conducting this analysis, genetically identical protein sequences were counted only once. This allowed our sequence alignments to contain the S protein sequences derived from all circulating SARS-CoV-2 variants. The genetic diversity of the sequences within our alignment was 2.4% for Alpha, 1.9% for Beta, and 2.3% for Delta VOCs. In addition, any mutation that appears within the S protein sequences at a frequency of <1% was considered a minor mutation, as it is less likely to contribute to the genetic diversity of the viral variants (66, 67). Some of the mutations that define VOCs are found at this low frequency in sequences of the VOC (Fig. 5A) (28–31); for example, the S494P mutation was found within <1% of the Alpha variants. In addition, we assessed whether previously characterized T cell escape mutations could be identified within the S protein region of the VOCs (Fig. 5B). Currently, a total of 12 mutations within the S protein are known to reduce T cell responses (47, 48, 50, 63, 64). Only three T cell escape mutations were identified within ≥1% of the sequences derived from the VOCs (i.e., Y144Δ/X for Alpha, K417N for Beta, and L452R for Delta VOCs).

FIGURE 5.

Comparison of mutations within SARS-CoV-2 S protein associated with VOCs and T cell escape to the peptides selected for ex vivo assessment.

(A and B) The proportion of S protein sequences containing mutations that define Alpha, Beta, and Delta VOCs (A) and the proportion of Alpha, Beta, and Delta VOC S protein sequences containing known T cell escape mutations (B) are shown. The proportion of S protein sequences with ambiguous amino acid residues (X) at the mutation sites is marked by blue bars (A and B). The threshold for determining the most significant mutations for Alpha, Beta, and Delta VOCs is denoted by the dotted line (≥1%; A and B). (C and D) All of the mutations were compared with the peptides selected from S subunit 1 (C) and subunit 2 (D). The 25 peptides are indicated by dark (≥12-mer) and light (9-mer) maroon bars. Detailed legends are provided within the figure panels (C and D).

FIGURE 5.

Comparison of mutations within SARS-CoV-2 S protein associated with VOCs and T cell escape to the peptides selected for ex vivo assessment.

(A and B) The proportion of S protein sequences containing mutations that define Alpha, Beta, and Delta VOCs (A) and the proportion of Alpha, Beta, and Delta VOC S protein sequences containing known T cell escape mutations (B) are shown. The proportion of S protein sequences with ambiguous amino acid residues (X) at the mutation sites is marked by blue bars (A and B). The threshold for determining the most significant mutations for Alpha, Beta, and Delta VOCs is denoted by the dotted line (≥1%; A and B). (C and D) All of the mutations were compared with the peptides selected from S subunit 1 (C) and subunit 2 (D). The 25 peptides are indicated by dark (≥12-mer) and light (9-mer) maroon bars. Detailed legends are provided within the figure panels (C and D).

Close modal

When assessing whether the 23 selected S peptides contain the mutations or deletions associated with VOCs, we found that five peptides contain changes that appear within ≥1% of the S protein sequences derived from the Alpha and Beta VOCs (Fig. 5C, 5D). However, all of our peptides avoided T cell escape mutations experimentally verified by other studies (Fig. 5C, 5D) (47, 48, 50, 63, 64).

We also compared the peptides to the mutations associated with the recently identified SARS-CoV-2 variants, Omega and its sublineages (Fig. 6). A total of 41 mutations within S had been reported for these variants by August 2022, including K417N and L452R mutations previously characterized for T cell immune escape (29, 32, 33, 63, 70, 71). We found a total of three peptides containing the mutations associated with the Omicron variants. However, none of the peptides contained the T cell escape mutations. Overall, six peptides identified from the S protein had a total of seven sites that contained mutations and/or deletions that appear within ≥1% of sequences derived from Alpha, Beta, and Omicron VOCs. Importantly, our S-derived peptides avoided 55 of 60 sites (92%) containing the mutations and/or deletions associated with all four VOCs.

FIGURE 6.

Comparison of mutations within SARS-CoV-2 S protein associated with Omicron VOC and its sublineages to the peptides selected for ex vivo assessment.

(A and B) The mutations that define the Omicron SARS-CoV-2 variants were compared with the selected peptides for S subunit 1 (A) and subunit 2 (B) protein regions.

FIGURE 6.

Comparison of mutations within SARS-CoV-2 S protein associated with Omicron VOC and its sublineages to the peptides selected for ex vivo assessment.

(A and B) The mutations that define the Omicron SARS-CoV-2 variants were compared with the selected peptides for S subunit 1 (A) and subunit 2 (B) protein regions.

Close modal

The core of our immunoinformatics analysis pipeline is the selection of peptides from highly networked regions of a viral protein where amino acid residues within these regions form many molecular interactions with other neighboring residues (54). These molecular interactions include hydrophobic and hydrophilic bonding. Therefore, we assessed whether the hydrophobicity and isoelectric point are affected by the VOC-associated substitution mutations and/or deletions found within the six peptides selected from the S protein (Fig. 7A, 7B). In addition, we assessed whether the VOC-associated mutations and/or deletions can affect the weight and size of the peptides by converting the amino acid residues within the peptides to their atomic chemical structure (Fig. 7C, 7D). We found that most substitution mutations and/or deletions do not substantially change the hydrophobicity and isoelectric point of the S-derived peptides (Fig. 7A, 7B). The only exception was the N969K mutation associated with the Omicron variant, which was found within QAL as underlined (Figs. 6, 7B). When this peptide contains the N969K mutation, the isoelectric point of the peptide increases from 9.2 to 10.8. However, this peptide is positively charged with or without the mutation at physiological pH. The deletions associated with Alpha and Omicron VOCs decreased the weight and changed the length of the two peptides (LPFFSNVTWFHAIHV and FSNVTWFHAIHVSGT) as expected (Figs. 5, 6, 7C, 7D). Overall, our result shows that the physicochemical properties of the peptides are mostly preserved even though the mutations/deletions associated with the VOCs can be found within these peptides.

FIGURE 7.

Physicochemical properties of the six S-derived peptides with and without major mutations and deletions found within Alpha, Beta, and Omicron VOCs.

(AD) The mean hydrophobicity (A), isoelectric point (B), molecular mass (C), and geometric distance (D) of each S-derived peptide with or without major mutations and deletions associated with Alpha, Beta, and Omicron VOCs are shown. The units for molecular mass and geometric distance are Daltons and angstroms, respectively. The six S-derived peptides include those that contained mutations/deletions that were found within ≥1% of Alpha and Beta variant sequences, or those that contained at least 1 of 32 mutations associated with the Omicron variant. The peptide sequences that do not contain the mutations/deletions are underlined. The mutations/deletions are indicated by bolded red texts in the peptide sequences.

FIGURE 7.

Physicochemical properties of the six S-derived peptides with and without major mutations and deletions found within Alpha, Beta, and Omicron VOCs.

(AD) The mean hydrophobicity (A), isoelectric point (B), molecular mass (C), and geometric distance (D) of each S-derived peptide with or without major mutations and deletions associated with Alpha, Beta, and Omicron VOCs are shown. The units for molecular mass and geometric distance are Daltons and angstroms, respectively. The six S-derived peptides include those that contained mutations/deletions that were found within ≥1% of Alpha and Beta variant sequences, or those that contained at least 1 of 32 mutations associated with the Omicron variant. The peptide sequences that do not contain the mutations/deletions are underlined. The mutations/deletions are indicated by bolded red texts in the peptide sequences.

Close modal

In-depth mutational analysis has not been performed for the SARS-CoV-2 NC protein, an essential coronavirus protein for viral replication (86). Therefore, we determined the frequency of mutations found within the NC protein sequences derived from the Alpha, Beta, and Delta VOCs (Supplemental Figs. 3, 4). We extracted the NC protein region from 1.7 million whole genome SARS-CoV-2 sequences. Similar to the S protein mutation analysis, genetically identical NC protein sequences were counted only once, allowing our sequence alignment to represent all circulating SARS-CoV-2 variants. The genetic diversity of the NC protein region was 1.2% for Alpha, 1.5% for Beta, and 1.6% for Delta VOCs. Moreover, the frequency of mutations was calculated at each amino acid residue position within the N- and C-terminal domains and the α-helix loop of the NC protein.

A total of 27 aa mutations and one region containing a deletion were found within ≥1% of the NC protein sequences derived from all the VOCs combined (Fig. 8A, Supplemental Figs. 3, 4). Of the three regions of the NC protein, the α-helix loop, which is 72 aa residues in length, contained 10 mutations. Therefore, the frequency of mutations within the α-helix loop is 0.14. This frequency was 2- to 4-fold greater than the N- and C-terminal domains of the NC protein (0.057 and 0.040 for the N- and C-terminal, respectively). Of note, all of the selected HN peptides are located outside of the α-helix loop. Next, we compared all mutations and deletions associated with the VOCs to the peptides selected for T cell response studies (Fig. 8B). We found that only two peptides contained the L139F mutation associated with the Alpha and Delta SARS-CoV-2 variants. However, this mutation was only found within <2% of the NC protein sequences derived from these two VOCs (Fig. 8B). When comparing the HN peptides to the NC protein sequences derived from the Omicron VOC and its sublineages, we found that none of the peptides contained the mutations associated with the Omicron variants within the NC region (Fig. 8C).

FIGURE 8.

Comparison of mutations within SARS-CoV-2 NC protein associated with VOCs to the peptides selected for ex vivo assessment.

(A) The mutations that appear within at least 1% of the NC protein sequences derived from Alpha, Beta, and Delta VOCs were indicated and compared with the seven peptides selected for ex vivo assay. (B) The proportion of NC protein sequences derived from Alpha and Delta VOCs that contained L139, the L139F mutation, ambiguous amino acid residues, and other mutations is shown. (C) The mutations associated with NC protein sequences of the Omega and its sublineages were compared with the peptides selected for ex vivo assay.

FIGURE 8.

Comparison of mutations within SARS-CoV-2 NC protein associated with VOCs to the peptides selected for ex vivo assessment.

(A) The mutations that appear within at least 1% of the NC protein sequences derived from Alpha, Beta, and Delta VOCs were indicated and compared with the seven peptides selected for ex vivo assay. (B) The proportion of NC protein sequences derived from Alpha and Delta VOCs that contained L139, the L139F mutation, ambiguous amino acid residues, and other mutations is shown. (C) The mutations associated with NC protein sequences of the Omega and its sublineages were compared with the peptides selected for ex vivo assay.

Close modal

It is becoming increasingly clear that T cells play an important role in the resolution of COVID-19 and also in limiting the severity of disease (1, 5, 6, 87). Boosting vaccine-induced T cell responses would support breadth and longevity of immunity both in supporting Ab production through follicular helper T cells and other T cell effector mechanisms. It could be especially beneficial for the aging (88, 89), particularly in the face of viral variants that evade neutralizing Abs. To further understand and define the SARS-CoV-2–specific T cell immunity, diagnostic tools are needed that can distinguish cross-reactive T cells specific for seasonal coronaviruses and retain sensitivity. This is important, as it has been debated whether such pre-existing T cell cross-reactivity will allow and enhance specific SARS-CoV-2–specific T cell immunity after exposure or focus this response on the previously infecting coronavirus (“original antigenic sin” [90]) In this regard, we identified a pool of selected peptides derived from HN regions of the S and NC protein of SARS-CoV-2 that had two advantages over an overlapping peptide pool: 1) being able to detect immunity to viral variants by avoiding mutations, and 2) allowing us to exclude epitopes that are cross-reactive with seasonal coronaviruses, thus avoiding false negatives and false positives, respectively. This HN peptide pool may be ideal for a T cell–based immune test and could inform development of a broad T cell boosting vaccine. In our cohort of 37 recovered COVID-19 participants, all of the HN peptides tested were immunogenic for IFN-γ and/or IL-2 and we observed a broad T cell response (reflecting in vivo T cell responses) in 78% of subjects with a varied peptide repertoire recognized between individuals. This is consistent with previous reports of most people recognizing ∼20 CD4+ and 17 CD8+ SARS-CoV-2 T cell epitopes out of more than a thousand identified (45, 91). The peptide YLQPRTFLL (S 269–277 epitope) has been identified as immunodominant in multiple studies (reviewed in Ref. 80) and was part of our HN peptide pool. However, in our small cohort of HLA-A*02+ donors we found a higher response rate with peptides VLYENQKLI and SVTTEILPV. Several peptides, including these two, also had a higher magnitude of IFN-γ and IL-2 response than YLQPRTFLL. As far as we are aware, this is the first demonstration of the immunogenicity of these two individual peptides. Nathan et al. (52) also identified highly networked HLA-A2–restricted peptide epitopes in the S protein for stimulating CD8+ T cells, although their approach was different from our network calculation method. In our approach, we primarily focused on the physicochemical interactions at individual amino acid residues and the position of these residues within the SARS-CoV-2 protein structures. In addition, we extended our analysis to NC protein, which is more challenging to define SARS-CoV-2–specific epitopes due to a high degree of genetic homology with other coronaviruses (92–94). Moreover, our approach was also applied to identify HLA-II–restricted epitopes, which are more difficult to define due to greater HLA promiscuity and multiple overlapping epitopes that are often found within them.

Examining T cell responses by intracellular cytokine staining complemented the ex vivo FluoroSpot assays by separating CD4+ and CD8+ T cell responses and examining a broader range of responses, including markers for CD8 T cell degranulation and TNF-α. To obtain sufficient cells, in vitro expansion was necessary, so this revealed a memory rather than an effector response. Although the in vivo effector responses shown by FluoroSpot are most relevant to recovery from acute COVID, we also considered memory responses and their relationship to effector responses to be important for durability of immunity and protection against subsequent reinfection, and as a comparison with vaccine-induced memory responses. CD4+ and CD8+ T cells from most of the subjects tested (100% for CD4+; 56% for CD8+) showed polyfunctional cytokine responses of two or three cytokines, dominated by IFN-γ and TNF-α. IL-2–producing cells could be expected to increase at later times postinfection as central memory T cells accumulate (95). CD8+ T cells additionally showed evidence of degranulation with surface expression of CD107a/b. High levels of TNF-α production are associated with activation and proliferation of memory T cells and are consistent with other reports for SARS-CoV-2–specific CD4+ T cells (91, 96–99). This may be a feature of SARS-CoV-2, with TNF-α being associated with disease severity (82). Furthermore, one study noted that the classic Th1 hierarchy of IFN-γ > TNF-α in memory CD4+ T cells was observed for influenza but reversed for SARS-CoV-2 (99). It will be important to study the dynamics of this response over time to know whether the strong TNF-α response persists after COVID-19 recovery. Our observation that CD4+ T cells responded more frequently than CD8+ T cells in ICS is most likely due to the in vitro expansion of the PBMCs prior to testing, as SARS-CoV-2–specific CD4+ T cells have been reported to expand more readily than CD8+ T cells (100, 101). All of the class II peptides that elicited CD8+ T cell responses by ICS have embedded predicted class I epitopes restricted to HLA molecules other than A*02:01.

In agreement with previous reports, we observed cross-reactivity to the commercial S peptide pool in 75% of unexposed pre-COVID samples, although compared with convalescent samples, this was weaker both in terms of the number of T cells responding and the amount of cytokine each cell produced, probably reflective of more distant immunological memory. Although this rate is higher than in some studies (reviewed in Ref. 1), it has been observed by others (85). There was little cross-reactivity directed toward the commercial NC pool. The rate of cross-reactivity to the HN peptide pool was much lower, 30%, and the great majority was due to QAL in our cohort, although this pool contained a smaller number of peptides than the commercial one, which contains >300. There has been much discussion about the potential advantages or disadvantages of this T cell cross-reactivity in enhancing natural or vaccine-induced immunity or diminishing it through original antigenic sin. However, in line with the conclusion drawn by Mateus et al. (101), that COVID-19 convalescent subjects could increase their response to SARS-CoV-2 peptide pools beyond that of unexposed individuals and generate novel T cells argues against the original antigenic sin phenomenon in most subjects, at least in the case of natural infection. As there is evidence that cross-reactive T cells could protect from symptomatic disease (24, 102), as for influenza (16, 17), QAL could be excluded from the HN peptide pool to probe specifically for SARS-CoV-2 T cell responses but may be of value in a T cell vaccine to boost existing cross-reactive responses.

As predicted, our immunoinformatics pipeline identified peptides with physicochemical properties that are essential to the structure of the viral proteins. The VOC mutations that overlap with our peptides are very conservative in terms of their effect on the physicochemical properties of the peptides and are unlikely to influence the peptide-binding affinity to MHC or TCR receptors, thereby avoiding T cell escape. In addition, our peptides avoided all VOC-associated mutations reported by Tarke et al. (27) that result in a decrease in the predicted MHC binding affinity within the S protein region. Although a negligible decrease in T cell responses has been observed to Alpha, Beta, and Gamma VOCs and a modest decrease to Omicron to date (26, 27, 97), this may not always be the case as the virus continues to evolve. Indeed, a recent study showed that 20% of COVID-19 recovered or vaccinated individuals had significant reductions of 50% or more in their T cell responses against the Omicron variant (103). The HN peptide pool identified in the current study will likely continue to avoid radical mutations given the structural importance of these regions to the virus, which provides insurance against T cell escape for future VOCs. Of note, our mutational analysis included the 66,518 S protein sequences derived from Alpha, Beta, and Delta VOCs, representing genetic diversity of 1.9–2.3%. The genetic diversity within our sequence alignment was ∼2-fold higher than the diversity calculated from a total of 15 SARS-CoV-2 sequences previously analyzed to compare T cell epitopes to mutations/deletions associated with Alpha, Beta, Gamma, and CAL20C variants (27). Even when we included a large number of S protein sequences from all VOCs, our immunoinformatics analysis pipeline allowed us to select the peptides that avoided 92% of the mutations associated with all VOCs. This indicates that our immunoinformatics analysis pipeline selected genetically conserved T cell epitopes from the original 2019 SARS-CoV-2 sequences that remain conserved in new SARS-CoV-2 variants (54, 104).

Our in-depth mutational analysis revealed that 23 out of 30 peptides selected for ex vivo assessment avoided the major mutations within the S or NC protein regions derived from the Alpha, Beta, and Delta SARS-CoV-2 variants. This finding suggests that the selected peptides combined would be effective in generating T cell responses against SARS-CoV-2 viral variants. For the ≥12-mer peptides that contained VOC-associated mutations, the impact of the mutations can be circumvented by the allelic diversity of heterodimeric HLA-II molecules and their open-ended peptide-binding groove, which can allow a broader range of peptide epitopes to be presented (105–107). For the 9-mer peptides that contained the VOC-associated mutations, only one NC-derived peptide had a genetic change (i.e., L139F) at an anchor position, which is important for binding to HLA-I molecules, implying that this mutation can reduce the HLA-I Ag presentation of this particular peptide (108). However, this L139F mutation was marginally classed as a major mutation associated with the Delta variant because the frequency of NC sequences containing this mutation was 1.3%, just above our cutoff of 1%. Nevertheless, all of the peptides selected for ex vivo assessment avoided previously characterized T cell escape mutations within the S protein region, indicating that the combination of these 30 peptides would be effective against global viral variants (47, 48, 50, 63, 64).

When performing the sequence analysis of the NC protein region, we noted that 10 mutations were densely clustered within the α-helix loop compared with the other protein domains. Of these mutations, the genetic changes at positions 203 and 204 were of particular interest, as they were associated with ≥80% of the NC protein sequences derived from the Alpha and Delta VOCs. In addition, ∼98% of sequences derived from the Omicron variants contained arginine-to-lysine (R203K) and glycine-to-arginine (G204R) mutations at these positions (70, 71, 109). Similar to the G614D mutation within the S protein, this R023K/G204R-linked mutation within the NC protein has been shown to be positively selected, contributing to increased viral transmission and replication (34–39, 110–112). Also, we found that the frequency of mutations within this α-helix loop that were associated with the three VOCs was 2- to 4-fold greater than the other domains of the NC protein. Our finding supports recent studies that suggest the α-helix loop contains “mutational hotspots” that are associated with the selective advantage of SARS-CoV-2 variants (111–113). Of note, our immunoinformatics analysis pipeline excluded the α-helix loop when identifying highly networked T cell epitope–derived peptides (54). This indicates that our in silico analysis pipeline identifies T cell epitope–derived peptides within evolutionarily constrained regions of the NC protein.

This study has several limitations, including the relatively small cohort used to test the peptides. These participants were primarily infected with the 2019 strain of SARS-CoV-2, and testing of VOCs was limited to the Delta strain. This was because the Alpha and Beta VOCs did not circulate widely in Australia and 95% of the Australian population was vaccinated prior to the emergence of the Omicron VOCs, making it impossible to determine an Omicron-specific T cell response. Longitudinal testing of the cohort is underway. Another limitation of our study is that the ex vivo assessment of MHC class I peptides was performed only in HLA-A*02(+) participants. However, our class I peptides are predicted to bind to multiple HLA class I alleles resulting in a global population coverage of 90 to 96% (54). Inclusion of HN peptides from the SARS-CoV-2 membrane and ORF1ab proteins will also be beneficial, as these Ags also contain a high number of T cell epitopes (80), and specificity for these could explain the lack of T cell responses to both the commercial and HN pools in two of our subjects. In some studies activation markers have been used for detecting cells producing low levels of cytokines. However, this was countered in the current study by assessing the HN peptide pool–expanded PBMCs by ICS. Further breakdown of CD4+ Th cells into the functional subsets of Th1, Th2, and Tfh, among others, in future could provide a deeper insight into the nature of the SARS-CoV-2–specific and cross-reactive responses to the HN peptide pool.

In summary, our immunoinformatics pathway focusing on highly networked regions of the SARS-CoV-2 virus has identified several CD4+ and CD8+ T cell epitopes from S and NC that should remain resistant to viral mutations that will arise in the future, and which can be pooled to include or omit cross-reactivity to seasonal coronaviruses. These peptides have application as specific diagnostic tools and as a guide to constructing vaccines with enhanced T cell boosting vaccine activity with broad specificity against all VOCs. Our results also strengthen the argument for including NC Ags in next-generation vaccines.

The authors have no financial conflicts of interest.

We acknowledge with gratitude the participants who donated samples for this study. The Centre for Immunology and Allergy at The Westmead Institute for Medical Research and the COSIN (Collection of COVID-19 Outbreak Samples in NSW) cohort, biobanked at The Kirby Institute (Kensington, NSW, Australia), are acknowledged for processing PBMCs of pre–COVID-19 blood donors and COSIN for Delta donors for this study. Flow cytometry was performed at the Westmead Scientific Platforms, which are supported by the Westmead Research Hub, the Cancer Institute New South Wales, the National Health and Medical Research Council, and the Ian Potter Foundation.

This work is supported by funding from the National Health and Medical Research Council, NSW Health, NSW Health Pathology, The Snow Foundation, The Peter Weiss Foundation, Sandra and David Ansley, and The University of Sydney Institute for Infectious Diseases.

The online version of this article contains supplemental material.

GISAID

Global Initiative on Sharing All Influenza Data

HLA-I

HLA class I

HLA-II

HLA class II

HN

highly networked

ICS

intracellular cytokine staining

IQR

interquartile range

NC

nucleocapsid

QAL

QALNTLVKQLSSNFGAI

SFU

spot-forming units

S

spike

VOC

variant of concern

1
Sette
,
A.
,
S.
Crotty
.
2021
.
Adaptive immunity to SARS-CoV-2 and COVID-19
.
Cell
184
:
861
880
.
2
Bastard
,
P.
,
L. B.
Rosen
,
Q.
Zhang
,
E.
Michailidis
,
H. H.
Hoffmann
,
Y.
Zhang
,
K.
Dorgham
,
Q.
Philippot
,
J.
Rosain
,
V.
Béziat
, et al
COVID Human Genetic Effort
.
2020
.
Autoantibodies against type I IFNs in patients with life-threatening COVID-19
.
Science
370
:
eabd4585
.
3
Galani
,
I. E.
,
N.
Rovina
,
V.
Lampropoulou
,
V.
Triantafyllia
,
M.
Manioudaki
,
E.
Pavlos
,
E.
Koukaki
,
P. C.
Fragkou
,
V.
Panou
,
V.
Rapti
, et al
.
2021
.
Untuned antiviral immunity in COVID-19 revealed by temporal type I/III interferon patterns and flu comparison
.
Nat. Immunol.
22
:
32
40
.
4
Hadjadj
,
J.
,
N.
Yatim
,
L.
Barnabei
,
A.
Corneau
,
J.
Boussier
,
N.
Smith
,
H.
Péré
,
B.
Charbit
,
V.
Bondet
,
C.
Chenevier-Gobeaux
, et al
.
2020
.
Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients
.
Science
369
:
718
724
.
5
Tan
,
A. T.
,
M.
Linster
,
C. W.
Tan
,
N.
Le Bert
,
W. N.
Chia
,
K.
Kunasegaran
,
Y.
Zhuang
,
C. Y. L.
Tham
,
A.
Chia
,
G. J. D.
Smith
, et al
.
2021
.
Early induction of functional SARS-CoV-2-specific T cells associates with rapid viral clearance and mild disease in COVID-19 patients
.
Cell Rep.
34
:
108728
.
6
Rydyznski Moderbacher
,
C.
,
S. I.
Ramirez
,
J. M.
Dan
,
A.
Grifoni
,
K. M.
Hastie
,
D.
Weiskopf
,
S.
Belanger
,
R. K.
Abbott
,
C.
Kim
,
J.
Choi
, et al
.
2020
.
Antigen-specific adaptive immunity to SARS-CoV-2 in acute COVID-19 and associations with age and disease severity
.
Cell
183
:
996
1012.e19
.
7
Brouwer
,
P. J. M.
,
T. G.
Caniels
,
K.
van der Straten
,
J. L.
Snitselaar
,
Y.
Aldon
,
S.
Bangaru
,
J. L.
Torres
,
N. M. A.
Okba
,
M.
Claireaux
,
G.
Kerster
, et al
.
2020
.
Potent neutralizing antibodies from COVID-19 patients define multiple targets of vulnerability
.
Science
369
:
643
650
.
8
Piccoli
,
L.
,
Y. J.
Park
,
M. A.
Tortorici
,
N.
Czudnochowski
,
A. C.
Walls
,
M.
Beltramello
,
C.
Silacci-Fregni
,
D.
Pinto
,
L. E.
Rosen
,
J. E.
Bowen
, et al
.
2020
.
Mapping neutralizing and immunodominant sites on the SARS-CoV-2 spike receptor-binding domain by structure-guided high-resolution serology
.
Cell
183
:
1024
1042.e21
.
9
Robbiani
,
D. F.
,
C.
Gaebler
,
F.
Muecksch
,
J. C. C.
Lorenzi
,
Z.
Wang
,
A.
Cho
,
M.
Agudelo
,
C. O.
Barnes
,
A.
Gazumyan
,
S.
Finkin
, et al
.
2020
.
Convergent antibody responses to SARS-CoV-2 in convalescent individuals
.
Nature
584
:
437
442
.
10
Khoury
,
D. S.
,
D.
Cromer
,
A.
Reynaldi
,
T. E.
Schlub
,
A. K.
Wheatley
,
J. A.
Juno
,
K.
Subbarao
,
S. J.
Kent
,
J. A.
Triccas
,
M. P.
Davenport
.
2021
.
Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection
.
Nat. Med.
27
:
1205
1211
.
11
Peng
,
Y.
,
A. J.
Mentzer
,
G.
Liu
,
X.
Yao
,
Z.
Yin
,
D.
Dong
,
W.
Dejnirattisai
,
T.
Rostron
,
P.
Supasa
,
C.
Liu
, et al
ISARIC4C Investigators
.
2020
.
Broad and strong memory CD4+ and CD8+ T cells induced by SARS-CoV-2 in UK convalescent individuals following COVID-19
.
Nat. Immunol.
21
:
1336
1345
.
12
Wagner
,
K. I.
,
L. M.
Mateyka
,
S.
Jarosch
,
V.
Grass
,
S.
Weber
,
K.
Schober
,
M.
Hammel
,
T.
Burrell
,
B.
Kalali
,
H.
Poppert
, et al
.
2022
.
Recruitment of highly cytotoxic CD8+ T cell receptors in mild SARS-CoV-2 infection
.
Cell Rep.
38
:
110214
.
13
McMahan
,
K.
,
J.
Yu
,
N. B.
Mercado
,
C.
Loos
,
L. H.
Tostanoski
,
A.
Chandrashekar
,
J.
Liu
,
L.
Peter
,
C.
Atyeo
,
A.
Zhu
, et al
.
2021
.
Correlates of protection against SARS-CoV-2 in rhesus macaques
.
Nature
590
:
630
634
.
14
Crotty
,
S.
2015
.
A brief history of T cell help to B cells
.
Nat. Rev. Immunol.
15
:
185
189
.
15
Poon
,
M. M. L.
,
K.
Rybkina
,
Y.
Kato
,
M.
Kubota
,
R.
Matsumoto
,
N. I.
Bloom
,
Z.
Zhang
,
K. M.
Hastie
,
A.
Grifoni
,
D.
Weiskopf
, et al
.
2021
.
SARS-CoV-2 infection generates tissue-localized immunological memory in humans
.
Sci. Immunol.
6
:
eabl9105
.
16
Sridhar
,
S.
,
S.
Begom
,
A.
Bermingham
,
K.
Hoschler
,
W.
Adamson
,
W.
Carman
,
T.
Bean
,
W.
Barclay
,
J. J.
Deeks
,
A.
Lalvani
.
2013
.
Cellular immune correlates of protection against symptomatic pandemic influenza
.
Nat. Med.
19
:
1305
1312
.
17
Wilkinson
,
T. M.
,
C. K.
Li
,
C. S.
Chui
,
A. K.
Huang
,
M.
Perkins
,
J. C.
Liebner
,
R.
Lambkin-Williams
,
A.
Gilbert
,
J.
Oxford
,
B.
Nicholas
, et al
.
2012
.
Preexisting influenza-specific CD4+ T cells correlate with disease protection against influenza challenge in humans
.
Nat. Med.
18
:
274
280
.
18
Schmidt
,
M. E.
,
S. M.
Varga
.
2018
.
The CD8 T cell response to respiratory virus infections
.
Front. Immunol.
9
:
678
.
19
Breton
,
G.
,
P.
Mendoza
,
T.
Hägglöf
,
T. Y.
Oliveira
,
D.
Schaefer-Babajew
,
C.
Gaebler
,
M.
Turroja
,
A.
Hurley
,
M.
Caskey
,
M. C.
Nussenzweig
.
2021
.
Persistent cellular immunity to SARS-CoV-2 infection
.
J. Exp. Med.
218
:
e20202515
.
20
Zuo
,
J.
,
A. C.
Dowell
,
H.
Pearce
,
K.
Verma
,
H. M.
Long
,
J.
Begum
,
F.
Aiano
,
Z.
Amin-Chowdhury
,
K.
Hoschler
,
T.
Brooks
, et al
.
2021
.
Robust SARS-CoV-2-specific T cell immunity is maintained at 6 months following primary infection. [Published erratum appears in 2021 Nat. Immunol. 22: 928.]
Nat. Immunol.
22
:
620
626
.
21
Dan
,
J. M.
,
J.
Mateus
,
Y.
Kato
,
K. M.
Hastie
,
E. D.
Yu
,
C. E.
Faliti
,
A.
Grifoni
,
S. I.
Ramirez
,
S.
Haupt
,
A.
Frazier
, et al
.
2021
.
Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection
.
Science
371
:
eabf4063
.
22
Sherina
,
N.
,
A.
Piralla
,
L.
Du
,
H.
Wan
,
M.
Kumagai-Braesch
,
J.
Andréll
,
S.
Braesch-Andersen
,
I.
Cassaniti
,
E.
Percivalle
,
A.
Sarasini
, et al
.
2021
.
Persistence of SARS-CoV-2-specific B and T cell responses in convalescent COVID-19 patients 6–8 months after the infection
.
Med
2
:
281
295.e4
.
23
Jung
,
J. H.
,
M. S.
Rha
,
M.
Sa
,
H. K.
Choi
,
J. H.
Jeon
,
H.
Seok
,
D. W.
Park
,
S. H.
Park
,
H. W.
Jeong
,
W. S.
Choi
,
E. C.
Shin
.
2021
.
SARS-CoV-2-specific T cell memory is sustained in COVID-19 convalescent patients for 10 months with successful development of stem cell-like memory T cells
.
Nat. Commun.
12
:
4043
.
24
Kundu
,
R.
,
J. S.
Narean
,
L.
Wang
,
J.
Fenn
,
T.
Pillay
,
N. D.
Fernandez
,
E.
Conibear
,
A.
Koycheva
,
M.
Davies
,
M.
Tolosa-Wright
, et al
.
2022
.
Cross-reactive memory T cells associate with protection against SARS-CoV-2 infection in COVID-19 contacts
.
Nat. Commun.
13
:
80
.
25
Swadling
,
L.
,
M. O.
Diniz
,
N. M.
Schmidt
,
O. E.
Amin
,
A.
Chandran
,
E.
Shaw
,
C.
Pade
,
J. M.
Gibbons
,
N.
Le Bert
,
A. T.
Tan
, et al
COVIDsortium Investigators
.
2022
.
Pre-existing polymerase-specific T cells expand in abortive seronegative SARS-CoV-2
.
Nature
601
:
110
117
.
26
Geers
,
D.
,
M. C.
Shamier
,
S.
Bogers
,
G.
den Hartog
,
L.
Gommers
,
N. N.
Nieuwkoop
,
K. S.
Schmitz
,
L. C.
Rijsbergen
,
J. A. T.
van Osch
,
E.
Dijkhuizen
, et al
.
2021
.
SARS-CoV-2 variants of concern partially escape humoral but not T-cell responses in COVID-19 convalescent donors and vaccinees
.
Sci. Immunol.
6
:
eabj1750
.
27
Tarke
,
A.
,
J.
Sidney
,
N.
Methot
,
E. D.
Yu
,
Y.
Zhang
,
J. M.
Dan
,
B.
Goodwin
,
P.
Rubiro
,
A.
Sutherland
,
E.
Wang
, et al
.
2021
.
Impact of SARS-CoV-2 variants on the total CD4+ and CD8+ T cell reactivity in infected or vaccinated individuals
.
Cell Rep. Med.
2
:
100355
.
28
Khan
,
A.
,
T.
Khan
,
S.
Ali
,
S.
Aftab
,
Y.
Wang
,
W.
Qiankun
,
M.
Khan
,
M.
Suleman
,
S.
Ali
,
W.
Heng
, et al
.
2021
.
SARS-CoV-2 new variants: characteristic features and impact on the efficacy of different vaccines
.
Biomed. Pharmacother.
143
:
112176
.
29
World Health Organization
.
2022
.
Tracking SARS-CoV-2 variants
. .
30
Rambaut
,
A.
,
E. C.
Holmes
,
Á.
O’Toole
,
V.
Hill
,
J. T.
McCrone
,
C.
Ruis
,
L.
du Plessis
,
O. G.
Pybus
.
2020
.
A dynamic nomenclature proposal for SARS-CoV-2 lineages to assist genomic epidemiology
.
Nat. Microbiol.
5
:
1403
1407
.
31
O’Toole
,
Á.
,
E.
Scher
,
A.
Underwood
,
B.
Jackson
,
V.
Hill
,
J. T.
McCrone
,
R.
Colquhoun
,
C.
Ruis
,
K.
Abu-Dahab
,
B.
Taylor
, et al
.
2021
.
Assignment of epidemiological lineages in an emerging pandemic using the pangolin tool
.
Virus Evol.
7
:
veab064
.
32
Karim
,
S. S. A.
,
Q. A.
Karim
.
2021
.
Omicron SARS-CoV-2 variant: a new chapter in the COVID-19 pandemic. [Published erratum appears in 2022 Lancet 399: 142.]
Lancet
398
:
2126
2128
.
33
Sun
,
Y.
,
W.
Lin
,
W.
Dong
,
J.
Xu
.
2022
.
Origin and evolutionary analysis of the SARS-CoV-2 Omicron variant
.
J. Biosaf. Biosecur.
4
:
33
37
.
34
Yurkovetskiy
,
L.
,
X.
Wang
,
K. E.
Pascal
,
C.
Tomkins-Tinch
,
T. P.
Nyalile
,
Y.
Wang
,
A.
Baum
,
W. E.
Diehl
,
A.
Dauphin
,
C.
Carbone
, et al
.
2020
.
Structural and functional analysis of the D614G SARS-CoV-2 spike protein variant
.
Cell
183
:
739
751.e8
.
35
Korber
,
B.
,
W. M.
Fischer
,
S.
Gnanakaran
,
H.
Yoon
,
J.
Theiler
,
W.
Abfalterer
,
B.
Foley
,
E. E.
Giorgi
,
T.
Bhattacharya
,
M. D.
Parker
, et al
.
2020
.
Tracking Changes in SARS-CoV-2 Spike: Evidence that D614G Increases Infectivity of the COVID-19 Virus
.
Cell
36
Zhou
,
B.
,
T. T. N.
Thao
,
D.
Hoffmann
,
A.
Taddeo
,
N.
Ebert
,
F.
Labroussaa
,
A.
Pohlmann
,
J.
King
,
S.
Steiner
,
J. N.
Kelly
, et al
.
2021
.
SARS-CoV-2 spike D614G change enhances replication and transmission
.
Nature
592
:
122
127
.
37
Deng
,
X.
,
M. A.
Garcia-Knight
,
M. M.
Khalid
,
V.
Servellita
,
C.
Wang
,
M. K.
Morris
,
A.
Sotomayor-González
,
D. R.
Glasner
,
K. R.
Reyes
,
A. S.
Gliwa
, et al
.
2021
.
Transmission, infectivity, and antibody neutralization of a spike L452R SARS-CoV-2 variant
.
Cell
184
:
3426
3437.e8
.
38
Shen
,
X.
,
H.
Tang
,
C.
McDanal
,
K.
Wagh
,
W.
Fischer
,
J.
Theiler
,
H.
Yoon
,
D.
Li
,
B. F.
Haynes
,
K. O.
Sanders
, et al
.
2021
.
SARS-CoV-2 variant B.1.1.7 is susceptible to neutralizing antibodies elicited by ancestral spike vaccines
.
Cell Host Microbe
29
:
529
539.e3
.
39
Volz
,
E.
,
V.
Hill
,
J. T.
McCrone
,
A.
Price
,
D.
Jorgensen
,
Á.
O’Toole
,
J.
Southgate
,
R.
Johnson
,
B.
Jackson
,
F. F.
Nascimento
, et al
COG-UK Consortium
.
2021
.
Evaluating the effects of SARS-CoV-2 spike mutation D614G on Transmissibility and pathogenicity
.
Cell
184
:
64
75.e11
.
40
Collier
,
D. A.
,
A.
De Marco
,
I. A. T. M.
Ferreira
,
B.
Meng
,
R. P.
Datir
,
A. C.
Walls
,
S. A.
Kemp
,
J.
Bassi
,
D.
Pinto
,
C.
Silacci-Fregni
, et al
.
2021
.
Sensitivity of SARS-CoV-2 B.1.1.7 to mRNA vaccine-elicited antibodies
.
Nature
593
:
136
141
.
41
Yadav
,
P. D.
,
G. N.
Sapkal
,
P.
Abraham
,
R.
Ella
,
G.
Deshpande
,
D. Y.
Patil
,
D. A.
Nyayanit
,
N.
Gupta
,
R. R.
Sahay
,
A. M.
Shete
, et al
.
2022
.
Neutralization of variant under investigation B.1.617.1 with sera of BBV152 vaccinees
.
Clin. Infect. Dis.
74
:
366
368
.
42
Garcia-Beltran
,
W. F.
,
E. C.
Lam
,
K.
St Denis
,
A. D.
Nitido
,
Z. H.
Garcia
,
B. M.
Hauser
,
J.
Feldman
,
M. N.
Pavlovic
,
D. J.
Gregory
,
M. C.
Poznansky
, et al
.
2021
.
Multiple SARS-CoV-2 variants escape neutralization by vaccine-induced humoral immunity. [Published erratum appears in 2021 Cell 184: 2523.]
Cell
184
:
2372
2383.e9
.
43
Annavajhala
,
M. K.
,
H.
Mohri
,
P.
Wang
,
M.
Nair
,
J. E.
Zucker
,
Z.
Sheng
,
A.
Gomez-Simmonds
,
A. L.
Kelley
,
M.
Tagliavia
,
Y.
Huang
, et al
.
2021
.
Emergence and expansion of SARS-CoV-2 B.1.526 after identification in New York
.
Nature
597
:
703
708
.
44
Altmann
,
D. M.
,
C. J.
Reynolds
,
R. J.
Boyton
.
2021
.
SARS-CoV-2 variants: subversion of antibody response and predicted impact on T cell recognition
.
Cell Rep. Med.
2
:
100286
.
45
Tarke
,
A.
,
J.
Sidney
,
C. K.
Kidd
,
J. M.
Dan
,
S. I.
Ramirez
,
E. D.
Yu
,
J.
Mateus
,
R.
da Silva Antunes
,
E.
Moore
,
P.
Rubiro
, et al
.
2021
.
Comprehensive analysis of T cell immunodominance and immunoprevalence of SARS-CoV-2 epitopes in COVID-19 cases
.
Cell Rep. Med.
2
:
100204
.
46
Moss
,
P.
2022
.
The T cell immune response against SARS-CoV-2
.
Nat. Immunol.
23
:
186
193
.
47
Agerer
,
B.
,
M.
Koblischke
,
V.
Gudipati
,
L. F.
Montaño-Gutierrez
,
M.
Smyth
,
A.
Popa
,
J.-W.
Genger
,
L.
Endler
,
D. M.
Florian
,
V.
Mühlgrabner
, et al
.
2021
.
SARS-CoV-2 mutations in MHC-I-restricted epitopes evade CD8+ T cell responses
.
Sci. Immunol.
6
:
eabg6461
.
48
de Silva
,
T. I.
,
G.
Liu
,
B. B.
Lindsey
,
D.
Dong
,
S. C.
Moore
,
N. S.
Hsu
,
D.
Shah
,
D.
Wellington
,
A. J.
Mentzer
,
A.
Angyal
, et al
.
2021
.
The impact of viral mutations on recognition by SARS-CoV-2 specific T cells
.
iScience
24
:
103353
.
49
Motozono
,
C.
,
M.
Toyoda
,
J.
Zahradnik
,
A.
Saito
,
H.
Nasser
,
T. S.
Tan
,
I.
Ngare
,
I.
Kimura
,
K.
Uriu
,
Y.
Kosugi
, et al
Genotype to Phenotype Japan (G2P-Japan) Consortium
.
2021
.
SARS-CoV-2 spike L452R variant evades cellular immunity and increases infectivity
.
Cell Host Microbe
29
:
1124
1136.e11
.
50
Dolton
,
G.
,
C.
Rius
,
M. S.
Hasan
,
A.
Wall
,
B.
Szomolay
,
E.
Behiry
,
T.
Whalley
,
J.
Southgate
,
A.
Fuller
,
T.
Morin
, et al
.
2022
.
Emergence of immune escape at dominant SARS-CoV-2 killer T-cell epitope
.
Cell
185
:
2936
2951.e19
.
51
Nersisyan
,
S. A.
,
M. Y.
Shkurnikov
,
A. P.
Zhiyanov
,
V. O.
Novosad
,
A. G.
Tonevitsky
.
2022
.
Differences in presentation of SARS-CoV-2 Omicron strain variant BA.1-BA.5 peptides by HLA molecules
.
Dokl. Biochem. Biophys.
507
:
298
301
.
52
Nathan
,
A.
,
E. J.
Rossin
,
C.
Kaseke
,
R. J.
Park
,
A.
Khatri
,
D.
Koundakjian
,
J. M.
Urbach
,
N. K.
Singh
,
A.
Bashirova
,
R.
Tano-Menka
, et al
.
2021
.
Structure-guided T cell vaccine design for SARS-CoV-2 variants and sarbecoviruses
.
Cell
184
:
4401
4413.e10
.
53
Gaiha
,
G. D.
,
E. J.
Rossin
,
J.
Urbach
,
C.
Landeros
,
D. R.
Collins
,
C.
Nwonu
,
I.
Muzhingi
,
M. N.
Anahtar
,
O. M.
Waring
,
A.
Piechocka-Trocha
, et al
.
2019
.
Structural topology defines protective CD8+ T cell epitopes in the HIV proteome
.
Science
364
:
480
484
.
54
Lee
,
E.
,
K.
Sandgren
,
G.
Duette
,
V. V.
Stylianou
,
R.
Khanna
,
J. S.
Eden
,
E.
Blyth
,
D.
Gottlieb
,
A. L.
Cunningham
,
S.
Palmer
.
2021
.
Identification of SARS-CoV-2 nucleocapsid and spike T-cell epitopes for assessing T-cell immunity
.
J. Virol.
95
:
e02002-20
.
55
Zhang
,
H.
,
H.
Hong
,
D.
Li
,
S.
Ma
,
Y.
Di
,
A.
Stoten
,
N.
Haig
,
K.
Di Gleria
,
Z.
Yu
,
X. N.
Xu
, et al
.
2009
.
Comparing pooled peptides with intact protein for accessing cross-presentation pathways for protective CD8+ and CD4+ T cells
.
J. Biol. Chem.
284
:
9184
9191
.
56
Roederer
,
M.
,
J. L.
Nozzi
,
M. C.
Nason
.
2011
.
SPICE: exploration and analysis of post-cytometric complex multivariate datasets
.
Cytometry A
79
:
167
174
.
57
Elbe
,
S.
,
G.
Buckland-Merrett
.
2017
.
Data, disease and diplomacy: GISAID’s innovative contribution to global health
.
Glob. Chall.
1
:
33
46
.
58
Shu
,
Y.
,
J.
McCauley
.
2017
.
GISAID: global initiative on sharing all influenza data—from vision to reality
.
Euro Surveill.
22
:
30494
.
59
Katoh
,
K.
,
D. M.
Standley
.
2013
.
MAFFT multiple sequence alignment software version 7: improvements in performance and usability
.
Mol. Biol. Evol.
30
:
772
780
.
60
Pagès
,
H.
,
P.
Aboyoun
,
R.
Gentleman
,
S.
DebRoy
.
2021
.
Biostrings: efficient manipulation of biological strings
.
R package version 2.62.0. Available at: https://bioconductor.org/packages/Biostrings
.
61
Wickham
,
H.
2019
.
stringr: simple, consistent wrappers for common string operations
. .
62
Charif
,
D.
,
J. R.
Lobry
.
2007
.
SeqinR 1.0-2: a contributed package to the R project for statistical computing devoted to biological sequences retrieval and analysis
. In
Structural Approaches to Sequence Evolution: Molecules, Networks, Populations.
U.
Bastolla
,
M.
Porto
,
H. E.
Roman
,
M.
Vendruscolo
, eds.
Springer
,
Berlin
, p.
207
232
.
63
Zhang
,
H.
,
S.
Deng
,
L.
Ren
,
P.
Zheng
,
X.
Hu
,
T.
Jin
,
X.
Tan
.
2021
.
Profiling CD8+ T cell epitopes of COVID-19 convalescents reveals reduced cellular immune responses to SARS-CoV-2 variants
.
Cell Rep.
36
:
109708
.
64
Wu
,
D.
,
A.
Kolesnikov
,
R.
Yin
,
J. D.
Guest
,
R.
Gowthaman
,
A.
Shmelev
,
Y.
Serdyuk
,
G. A.
Efimov
,
B. G.
Pierce
,
R. A.
Mariuzza
.
2022
.
Structural assessment of HLA-A2-restricted SARS-CoV-2 spike epitopes recognized by public and private T-cell receptors
.
Nat. Commun.
65
Chongsuvivatwong
,
V.
2018
.
epiDisplay: epidemiological data display package
. .
66
Stewart
,
J. J.
,
C. Y.
Lee
,
S.
Ibrahim
,
P.
Watts
,
M.
Shlomchik
,
M.
Weigert
,
S.
Litwin
.
1997
.
A Shannon entropy analysis of immunoglobulin and T cell receptor
.
Mol. Immunol.
34
:
1067
1082
.
67
Boni
,
C.
,
D.
Cavazzini
,
A.
Bolchi
,
M.
Rossi
,
A.
Vecchi
,
C.
Tiezzi
,
V.
Barili
,
P.
Fisicaro
,
C.
Ferrari
,
S.
Ottonello
.
2021
.
Degenerate CD8 epitopes mapping to structurally constrained regions of the spike protein: a T cell-based way-out from the SARS-CoV-2 variants storm
.
Front. Immunol.
12
:
730051
.
68
Kearse
,
M.
,
R.
Moir
,
A.
Wilson
,
S.
Stones-Havas
,
M.
Cheung
,
S.
Sturrock
,
S.
Buxton
,
A.
Cooper
,
S.
Markowitz
,
C.
Duran
, et al
.
2012
.
Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data
.
Bioinformatics
28
:
1647
1649
.
69
Gangavarapu
,
K.
,
A. A.
Latif
,
J. L.
Mullen
,
M.
Alkuzweny
,
E.
Hufbauer
,
G.
Tsueng
,
E.
Haag
,
M.
Zeller
,
C. M.
Aceves
,
K.
Zaiets
, et al
GISAID Core and Curation Team
.
2023
.
Outbreak.info genomic reports: scalable and dynamic surveillance of SARS-CoV-2 variants and mutations
.
Nat. Methods
20
:
512
522
.
70
Gangavarapu
,
K.
,
A. A.
Latif
,
J. L.
Mullen
,
M.
Alkuzweny
,
E.
Hufbauer
,
G.
Tsueng
,
E.
Haag
,
M.
Zeller
,
C. M.
Aceves
,
K.
Zaiets
, et al
.
2023
.
Outbreak.info genomic reports: scalable and dynamic surveillance of SARS-CoV-2 variants and mutations
.
Nat. Methods
20
:
512
522
.
71
Khare
,
S.
,
C.
Gurry
,
L.
Freitas
,
M. B.
Schultz
,
G.
Bach
,
A.
Diallo
,
N.
Akite
,
J.
Ho
,
R. T.
Lee
,
W.
Yeo
,
S.
Maurer-Stroh
.
2021
.
GISAID’s role in pandemic response
.
China CDC Wkly
3
:
1049
1051
.
72
Madeira
,
F.
,
Y. M.
Park
,
J.
Lee
,
N.
Buso
,
T.
Gur
,
N.
Madhusoodanan
,
P.
Basutkar
,
A. R. N.
Tivey
,
S. C.
Potter
,
R. D.
Finn
,
R.
Lopez
.
2019
.
The EMBL-EBI search and sequence analysis tools APIs in 2019
.
Nucleic Acids Res.
47
(
W1
):
W636
W641
.
73
Backman
,
T. W.
,
Y.
Cao
,
T.
Girke
.
2011
.
ChemMine tools: an online service for analyzing and clustering small molecules
.
Nucleic Acids Res.
39
(
Suppl.
):
W486
W91
.
74
Todeschini
,
R.
,
V.
Consonni
.
2003
.
Descriptors from molecular geometry
. In
Handbook of Chemoinformatics.
J.
Gasteiger
, ed.
Wiley-VCH
, p.
1004
1033
.
75
Kim
,
M.
,
J.
Taylor
,
J.
Sidney
,
Z.
Mikloska
,
N.
Bodsworth
,
K.
Lagios
,
H.
Dunckley
,
K.
Byth-Wilson
,
M.
Denis
,
R.
Finlayson
, et al
.
2008
.
Immunodominant epitopes in herpes simplex virus type 2 glycoprotein D are recognized by CD4 lymphocytes from both HSV-1 and HSV-2 seropositive subjects
.
J. Immunol.
181
:
6604
6615
.
76
Fleri
,
W.
,
S.
Paul
,
S. K.
Dhanda
,
S.
Mahajan
,
X.
Xu
,
B.
Peters
,
A.
Sette
.
2017
.
The Immune Epitope Database and Analysis Resource in epitope discovery and synthetic vaccine design
.
Front. Immunol.
8
:
278
.
77
Yang
,
J.
,
E.
James
,
M.
Roti
,
L.
Huston
,
J. A.
Gebe
,
W. W.
Kwok
.
2009
.
Searching immunodominant epitopes prior to epidemic: HLA class II-restricted SARS-CoV spike protein epitopes in unexposed individuals
.
Int. Immunol.
21
:
63
71
.
78
Cheung
,
Y. K.
,
S. C.
Cheng
,
F. W.
Sin
,
K. T.
Chan
,
Y.
Xie
.
2007
.
Induction of T-cell response by a DNA vaccine encoding a novel HLA-A*0201 severe acute respiratory syndrome coronavirus epitope
.
Vaccine
25
:
6070
6077
.
79
Tsao
,
Y. P.
,
J. Y.
Lin
,
J. T.
Jan
,
C. H.
Leng
,
C. C.
Chu
,
Y. C.
Yang
,
S. L.
Chen
.
2006
.
HLA-A*0201 T-cell epitopes in severe acute respiratory syndrome (SARS) coronavirus nucleocapsid and spike proteins
.
Biochem. Biophys. Res. Commun.
344
:
63
71
.
80
Grifoni
,
A.
,
J.
Sidney
,
R.
Vita
,
B.
Peters
,
S.
Crotty
,
D.
Weiskopf
,
A.
Sette
.
2021
.
SARS-CoV-2 human T cell epitopes: adaptive immune response against COVID-19. [Published erratum appears in 2022 Cell Host Microbe 30: 1788.]
Cell Host Microbe
29
:
1076
1092
.
81
Vita
,
R.
,
S.
Mahajan
,
J. A.
Overton
,
S. K.
Dhanda
,
S.
Martini
,
J. R.
Cantrell
,
D. K.
Wheeler
,
A.
Sette
,
B.
Peters
.
2019
.
The Immune Epitope Database (IEDB): 2018 update
.
Nucleic Acids Res.
47
(
D1
):
D339
D343
.
82
Del Valle
,
D. M.
,
S.
Kim-Schulze
,
H. H.
Huang
,
N. D.
Beckmann
,
S.
Nirenberg
,
B.
Wang
,
Y.
Lavin
,
T. H.
Swartz
,
D.
Madduri
,
A.
Stock
, et al
.
2020
.
An inflammatory cytokine signature predicts COVID-19 severity and survival
.
Nat. Med.
26
:
1636
1643
.
83
Le Bert
,
N.
,
A. T.
Tan
,
K.
Kunasegaran
,
C. Y. L.
Tham
,
M.
Hafezi
,
A.
Chia
,
M. H. Y.
Chng
,
M.
Lin
,
N.
Tan
,
M.
Linster
, et al
.
2020
.
SARS-CoV-2-specific T cell immunity in cases of COVID-19 and SARS, and uninfected controls
.
Nature
584
:
457
462
.
84
Grifoni
,
A.
,
D.
Weiskopf
,
S. I.
Ramirez
,
J.
Mateus
,
J. M.
Dan
,
C. R.
Moderbacher
,
S. A.
Rawlings
,
A.
Sutherland
,
L.
Premkumar
,
R. S.
Jadi
, et al
.
2020
.
Targets of T cell responses to SARS-CoV-2 coronavirus in humans with COVID-19 disease and unexposed individuals
.
Cell
181
:
1489
1501.e15
.
85
Steiner
,
S.
,
F.
Sotzny
,
S.
Bauer
,
I. K.
Na
,
M.
Schmueck-Henneresse
,
V. M.
Corman
,
T.
Schwarz
,
C.
Drosten
,
D. J.
Wendering
,
U.
Behrends
, et al
.
2020
.
HCoV- and SARS-CoV-2 cross-reactive T cells in CVID patients
.
Front. Immunol.
11
:
607918
.
86
McBride
,
R.
,
M.
van Zyl
,
B. C.
Fielding
.
2014
.
The coronavirus nucleocapsid is a multifunctional protein
.
Viruses
6
:
2991
3018
.
87
Rébillard
,
R. M.
,
M.
Charabati
,
C.
Grasmuck
,
A.
Filali-Mouhim
,
O.
Tastet
,
N.
Brassard
,
A.
Daigneault
,
L.
Bourbonnière
,
S. P.
Anand
,
R.
Balthazard
, et al
.
2021
.
Identification of SARS-CoV-2-specific immune alterations in acutely ill patients
.
J. Clin. Invest.
131
:
e145853
.
88
Bartleson
,
J. M.
,
D.
Radenkovic
,
A. J.
Covarrubias
,
D.
Furman
,
D. A.
Winer
,
E.
Verdin
.
2021
.
SARS-CoV-2, COVID-19 and the ageing immune system
.
Nat. Aging
1
:
769
782
.
89
Boutry
,
C.
,
A.
Hastie
,
J.
Diez-Domingo
,
J. C.
Tinoco
,
C. J.
Yu
,
C.
Andrews
,
J.
Beytout
,
C.
Caso
,
H. S.
Cheng
,
H. J.
Cheong
, et al
.
2022
.
The adjuvanted recombinant zoster vaccine confers long-term protection against herpes zoster: interim results of an extension study of the pivotal phase 3 clinical trials ZOE-50 and ZOE-70
.
Clin. Infect. Dis.
74
:
1459
1467
.
90
Rijkers
,
G. T.
,
F. J.
van Overveld
.
2021
.
The “original antigenic sin” and its relevance for SARS-CoV-2 (COVID-19) vaccination
.
Clin. Immunol. Commun.
1
:
13
16
.
91
Heide
,
J.
,
S.
Schulte
,
M.
Kohsar
,
T. T.
Brehm
,
M.
Herrmann
,
H.
Karsten
,
M.
Marget
,
S.
Peine
,
A. M.
Johansson
,
A.
Sette
, et al
.
2021
.
Broadly directed SARS-CoV-2-specific CD4+ T cell response includes frequently detected peptide specificities within the membrane and nucleoprotein in patients with acute and resolved COVID-19. [Published erratum appears in 2022 PLoS Pathog. 18: e1010220.]
PLoS Pathog.
17
:
e1009842
.
92
Dutta
,
N. K.
,
K.
Mazumdar
,
J. T.
Gordy
.
2020
.
The nucleocapsid protein of SARS-CoV-2: a target for vaccine development
.
J. Virol.
94
:
e00647-20
.
93
Grifoni
,
A.
,
J.
Sidney
,
Y.
Zhang
,
R. H.
Scheuermann
,
B.
Peters
,
A.
Sette
.
2020
.
A sequence homology and bioinformatic approach can predict candidate targets for immune responses to SARS-CoV-2
.
Cell Host Microbe
27
:
671
680.e2
.
94
Rota
,
P. A.
,
M. S.
Oberste
,
S. S.
Monroe
,
W. A.
Nix
,
R.
Campagnoli
,
J. P.
Icenogle
,
S.
Peñaranda
,
B.
Bankamp
,
K.
Maher
,
M. H.
Chen
, et al
.
2003
.
Characterization of a novel coronavirus associated with severe acute respiratory syndrome
.
Science
300
:
1394
1399
.
95
Wherry
,
E. J.
,
V.
Teichgräber
,
T. C.
Becker
,
D.
Masopust
,
S. M.
Kaech
,
R.
Antia
,
U. H.
von Andrian
,
R.
Ahmed
.
2003
.
Lineage relationship and protective immunity of memory CD8 T cell subsets
.
Nat. Immunol.
4
:
225
234
.
96
Sandberg
,
J. T.
,
R.
Varnaitė
,
W.
Christ
,
P.
Chen
,
J. R.
Muvva
,
K. T.
Maleki
,
M.
García
,
M.
Dzidic
,
E.
Folkesson
,
M.
Skagerberg
, et al
Karolinska COVID‐19 Study Group
.
2021
.
SARS-CoV-2-specific humoral and cellular immunity persists through 9 months irrespective of COVID-19 severity at hospitalisation
.
Clin. Transl. Immunology
10
:
e1306
.
97
Gao
,
Y.
,
C.
Cai
,
A.
Grifoni
,
T. R.
Muller
,
J.
Niessl
,
A.
Olofsson
,
M.
Humbert
,
L.
Hansson
,
A.
Osterborg
,
P.
Bergman
, et al
.
2022
.
Ancestral SARS-CoV-2-specific T cells cross-recognize the Omicron variant
.
Nat. Med.
28
:
472
476
.
98
Skittrall
,
J. P.
,
M.
Wilson
,
A. A.
Smielewska
,
S.
Parmar
,
M. D.
Fortune
,
D.
Sparkes
,
M. D.
Curran
,
H.
Zhang
,
H.
Jalal
.
2021
.
Specificity and positive predictive value of SARS-CoV-2 nucleic acid amplification testing in a low-prevalence setting
.
Clin. Microbiol. Infect.
27
:
469.e9
469.e15
.
99
Law
,
J. C.
,
W. H.
Koh
,
P.
Budylowski
,
J.
Lin
,
F.
Yue
,
K. T.
Abe
,
B.
Rathod
,
M.
Girard
,
Z.
Li
,
J. M.
Rini
, et al
.
2021
.
Systematic examination of antigen-specific recall T cell responses to SARS-CoV-2 versus influenza virus reveals a distinct inflammatory profile
.
J. Immunol.
206
:
37
50
.
100
Habel
,
J. R.
,
T. H. O.
Nguyen
,
C. E.
van de Sandt
,
J. A.
Juno
,
P.
Chaurasia
,
K.
Wragg
,
M.
Koutsakos
,
L.
Hensen
,
X.
Jia
,
B.
Chua
, et al
.
2020
.
Suboptimal SARS-CoV-2-specific CD8+ T cell response associated with the prominent HLA-A*02:01 phenotype
.
Proc. Natl. Acad. Sci. USA
117
:
24384
24391
.
101
Mateus
,
J.
,
A.
Grifoni
,
A.
Tarke
,
J.
Sidney
,
S. I.
Ramirez
,
J. M.
Dan
,
Z. C.
Burger
,
S. A.
Rawlings
,
D. M.
Smith
,
E.
Phillips
, et al
.
2020
.
Selective and cross-reactive SARS-CoV-2 T cell epitopes in unexposed humans
.
Science
370
:
89
94
.
102
Sagar
,
M.
,
K.
Reifler
,
M.
Rossi
,
N. S.
Miller
,
P.
Sinha
,
L. F.
White
,
J. P.
Mizgerd
.
2021
.
Recent endemic coronavirus infection is associated with less-severe COVID-19
.
J. Clin. Invest.
131
:
e143380
.
103
Naranbhai
,
V.
,
A.
Nathan
,
C.
Kaseke
,
C.
Berrios
,
A.
Khatri
,
S.
Choi
,
M. A.
Getz
,
R.
Tano-Menka
,
O.
Ofoman
,
A.
Gayton
, et al
.
2022
.
T cell reactivity to the SARS-CoV-2 Omicron variant is preserved in most but not all individuals
.
Cell
185
:
1041
1051.e6
.
104
Wu
,
F.
,
S.
Zhao
,
B.
Yu
,
Y. M.
Chen
,
W.
Wang
,
Z. G.
Song
,
Y.
Hu
,
Z. W.
Tao
,
J. H.
Tian
,
Y. Y.
Pei
, et al
.
2020
.
A new coronavirus associated with human respiratory disease in China. [Published erratum appears in 2020 Nature 580: E7.]
Nature
579
:
265
269
.
105
Gfeller
,
D.
,
M.
Bassani-Sternberg
.
2018
.
Predicting antigen presentation—what could we learn from a million peptides?
Front. Immunol.
9
:
1716
.
106
Marty Pyke
,
R.
,
W. K.
Thompson
,
R. M.
Salem
,
J.
Font-Burgada
,
M.
Zanetti
,
H.
Carter
.
2018
.
Evolutionary pressure against MHC class II binding cancer mutations. [Published erratum appears in 2018 Cell 175: 1991.]
Cell
175
:
416
428.e13
.
107
Chen
,
B.
,
M. S.
Khodadoust
,
N.
Olsson
,
L. E.
Wagar
,
E.
Fast
,
C. L.
Liu
,
Y.
Muftuoglu
,
B. J.
Sworder
,
M.
Diehn
,
R.
Levy
, et al
.
2019
.
Predicting HLA class II antigen presentation through integrated deep learning
.
Nat. Biotechnol.
37
:
1332
1343
.
108
Bassani-Sternberg
,
M.
,
C.
Chong
,
P.
Guillaume
,
M.
Solleder
,
H.
Pak
,
P. O.
Gannon
,
L. E.
Kandalaft
,
G.
Coukos
,
D.
Gfeller
.
2017
.
Deciphering HLA-I motifs across HLA peptidomes improves neo-antigen predictions and identifies allostery regulating HLA specificity
.
PLOS Comput. Biol.
13
:
e1005725
.
109
Gu
,
H.
,
P.
Krishnan
,
D. Y. M.
Ng
,
L. D. J.
Chang
,
G. Y. Z.
Liu
,
S. S. M.
Cheng
,
M. M. Y.
Hui
,
M. C. Y.
Fan
,
J. H. L.
Wan
,
L. H. K.
Lau
, et al
.
2022
.
Probable transmission of SARS-CoV-2 Omicron variant in quarantine hotel, Hong Kong, China, November 2021
.
Emerg. Infect. Dis.
28
:
460
462
.
110
Rochman
,
N. D.
,
Y. I.
Wolf
,
G.
Faure
,
P.
Mutz
,
F.
Zhang
,
E. V.
Koonin
.
2021
.
Ongoing global and regional adaptive evolution of SARS-CoV-2
.
Proc. Natl. Acad. Sci. USA
118
:
e2104241118
.
111
Wu
,
H.
,
N.
Xing
,
K.
Meng
,
B.
Fu
,
W.
Xue
,
P.
Dong
,
W.
Tang
,
Y.
Xiao
,
G.
Liu
,
H.
Luo
, et al
.
2021
.
Nucleocapsid mutations R203K/G204R increase the infectivity, fitness, and virulence of SARS-CoV-2
.
Cell Host Microbe
29
:
1788
1801.e6
.
112
Johnson
,
B. A.
,
Y.
Zhou
,
K. G.
Lokugamage
,
M. N.
Vu
,
N.
Bopp
,
P. A.
Crocquet-Valdes
,
B.
Kalveram
,
C.
Schindewolf
,
Y.
Liu
,
D.
Scharton
, et al
.
2021
.
Nucleocapsid mutations in SARS-CoV-2 augment replication and pathogenesis
.
PLoS Pathog.
18
:
e1010627
.
113
Plante
,
J. A.
,
B. M.
Mitchell
,
K. S.
Plante
,
K.
Debbink
,
S. C.
Weaver
,
V. D.
Menachery
.
2021
.
The variant gambit: COVID-19’s next move
.
Cell Host Microbe
29
:
508
515
.
This article is distributed under the terms of the CC BY-NC-ND 4.0 Unported license.

Supplementary data