Back

Public T-cell epitopes shared among SARS-CoV-2 variants are presented on prevalent HLA class I alleles

Meyer, S.; Blaas, I.; Bollineni, R. C.; Delic-Sarac, M.; Tran, T. T.; Knetter, C.; Dai, K.-Z.; Stove Madssen, T.; Vaage, J. T.; Gustavsen, A.; Yang, W.; Haug Nissen-Meyer, L. S.; Douvlataniotis, K.; Laos, M.; Nielsen, M. M.; Thiede, B.; Soraas, A.; Lund-Johansen, F.; Rustad, E. H.; Olweus, J.

2021-10-14 immunology
10.1101/2021.10.13.463911 bioRxiv
Show abstract

T-cell epitopes with broad population coverage may form the basis for a new generation of SARS-CoV-2 vaccines. However, published studies on immunoprevalence are limited by small test cohorts, low frequencies of antigen-specific cells and lack of data correlating eluted HLA ligands with T-cell responsiveness. As the protective role of pre-existing cross-reactivity to homologous peptides is unclear, we aimed to identify SARS-CoV-2-specific minimal epitopes recognized by CD8 T-cells among 48 peptides eluted from prevalent HLA alleles, and an additional 84 predicted binders, in a large cohort of convalescents (n=83) and pre-pandemic control samples (n=19). We identified nine conserved SARS-CoV-2-specific epitopes restricted by four of the most prevalent HLA class I alleles in the Norwegian study cohort, to which responding CD8 T cells were detected in 70-100% of convalescents expressing the relevant HLA allele. Only two of these were derived from the Spike protein, included in current vaccines. We found a strong correlation between immunoprevalence and immunodominance. Thus, the CD8 T-cell response to SARS-CoV-2 is more focused than previously believed. Using a new algorithm, we predict that a vaccine including these epitopes could induce a T-cell response in 83% of Caucasians.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.