High-resolution mapping of T-cell immunity to the entire sarscov-2 proteome

ABSTRACT

Background: Since the discovery of SARS-CoV-2, researchers have put major efforts towards the understanding of virus-specific cellular immunity. However, the identification of epitope-and protein specific T-cell responses is limited to bioinformatic approaches, use of total viral proteins or peptide mega pools. To overcome these current limitations, we performed a high-resolution mapping using IFN-y ELISpot and peptide sets covering the entire CoV-2 proteome. Methods: We synthetized a 15-mer peptide library of 2790 peptides (11 amino acid overlap) covering a CoV-2cons proteome sequence based on 1700 sequences. We designed a mega matrix of consecutive and non-consecutive peptide pools with 20 to 35 peptides per pool. We assessed T-cell responses in cryopreserved PBMCs from IgG+ SARS-CoV-2 infected individuals (N=13), who recovered from mild/moderate infection, 90-190 Days from off-set symptoms. Also, we expanded PBMCs in the presence of anti-CD3 and IL-2 during 3 weeks and performed a comparative ELISpot using total and expanded PBMCs. Results: Frequencies of T-cell responses from positive peptide pools revealed 40% of responses targeting S2, 20% against S1, 10% against M, and 6% against nsp3 and NP, respectively. The strongest responses were targeting S2 and S1 (median values of 540 and 240 IFN-y SFC/106, respectively), followed by nsp3, NP and M. We observed a median of 13 deconvoluted reactive peptides across the entire proteome per tested individual. The breadth of responses ranged from 1-8 targeted proteins with a median of 2. In addition, we mapped responses in subproteins 3C-LP, nsp6, nsp10 (Orf1ab), and alternative reading frames. We also identified responses to peptide sequences conserved across pan-coronavirus strains Orf1b (n=2), S (n=1) and M (n=1). Following expansion, we observed a loss of CD4+ T-cells in cultured cells and altered peptide-recognition profiles characterized by a loss of S2 and an increase of nsp3 responses. Conclusion: We characterize protein hierarchy in terms of breadth and magnitude by high-resolution mapping of T-cell responses against the entire CoV-2 proteome. The most frequently targeted and immunogenic regions were S2 and S1. We identify responses to small proteins, alternative reading frames and conserved regions across coronaviruses. This data brings new insight into the complexity of CoV-2 T-cell responses and crucial information for vaccine design.