key: cord-0794677-j8fhidqv
authors: Hu, Chao; Shen, Meiying; Han, Xiaojian; Chen, Qian; Li, Luo; Chen, Siyin; Zhang, Jing; Gao, Fengxia; Wang, Wang; Wang, Yingming; Li, Tingting; Li, Shenglong; Huang, Jingjing; Wang, Jianwei; Zhu, Ju; Chen, Dan; Wu, Qingchen; Tao, Kun; Pang, Da; Jin, Aishun
title: Identification of Cross-Reactive CD8+ T Cell Receptors with High Functional Avidity to a SARS-CoV-2 Immunodominant Epitope and Its Natural Mutant Variants
date: 2020-11-03
journal: bioRxiv
DOI: 10.1101/2020.11.02.364729
sha: 5c71441c53497c0bbcff0ebad4c4b5153c0f6cb4
doc_id: 794677
cord_uid: j8fhidqv

Despite the growing knowledge of T cell responses and their epitopes in COVID-19 patients, there is a lack of detailed characterizations for T cell-antigen interactions and T cell functions. Using a peptide library predicted with HLA class I-restriction, specific CD8+ T cell responses were identified in over 75% of COVID-19 convalescent patients. Among the 15 SARS-CoV-2 epitopes identified from the S and N proteins, N361-369 (KTFPPTEPK) was the most dominant epitope. Importantly, we discovered 2 N361-369-specific T cell receptors (TCRs) with high functional avidity, and they exhibited complementary cross-reactivity to reported N361-369 mutant variants. In dendritic cells (DCs) and the lung organoid model, we found that the N361-369 epitope could be processed and endogenously presented to elicit the activation and cytotoxicity of CD8+ T cells ex vivo. Our study evidenced potential mechanisms of cellular immunity to SARS-CoV-2, illuminating natural ways of viral clearance with high relevancy in the vaccine development.

predominant epitopes together with diverse TCRs carrying broader cross-reactivity to 117 SARS-CoV-2 and its mutant variants, which can provide detailed information to 118 understand the mechanisms of T cell-mediated viral clearance in COVID-19 patients. 119 In this study, we evaluated SARS-CoV-2 specific CD8 + T cell responses in 135 It has been reported that SARS-CoV-2 S and N proteins were the predominant T cell 136 targets in COVID-19 patients (Grifoni et al., 2020b) . In order to predict SARS-CoV-2 137 specific CD8 + T cell epitopes, the NetMHCpan 4.0 algorithm was used to analyse the 138 binding of 9-mer peptides derived from the full lengths of SARS-CoV-2 S and N 139 proteins, with HLA-A*02:01, HLA-A*24:02 and HLA-A*11:01, the 3 most frequent 140 HLA class I A alleles in the Asian population. We synthesized a peptide library of 72 141 predicted SARS-CoV-2 epitopes, which overlapped 29 SARS-CoV epitopes, together 142 with additional 6 confirmed SARS-CoV epitopes (Table S1 ) (Ahmed et al., 2020; 143 Grifoni et al., 2020a). Among these 78 peptides, 62 were from the S protein and 16 144 were from the N protein, with 3 different HLA-A restrictions (Table S2) . 145 In parallel, we collected a total of 37 COVID-19 convalescent PBMC samples and 146 subjected them for the HLA typing analysis. Among the 20 HLA-A2 + samples and the 147 12 HLA-A24 + samples revealed by flow cytometric assay, we selected 8 samples from 148 each group. A matching 8 samples with HLA-A*11:01 + genotype was identified by 149 RT-PCR, which included 2 overlapping with the HLA-A2 + group and 2 with the 150 HLA-A24 + group (Table S3) (Table 1) . Four dominant peptides were identified, with at least one 162 peptide from each HLA group (Table 1) . Two of these peptides, P63 (N 338-346 ) and P77 163 (N 219-227 ), were both HLA-A*02:01-restricted, and they could induce IFN-γ release in 164 4/8 and 2/8 samples, respectively (Figures 1A and S2; Table 1 ). The 165 HLA-A*24:02-restricted P45 (S 448-456 ) was located at the receptor binding domain of 166 the S protein and could be recognized by 2/8 samples tested ( Figure 1B ; Table 1 ). 167 Importantly, 5 out of 8 samples showed positive responses to the 168 HLA-A*11:01-restricted P64 (N 361-369 ). Hence we termed P64, or N 361-369 , as the most 169 dominant epitope in this study ( Figure 1C ; Table 1 ). Of note, IFN-γ induction was not 170 observed when these 4 dominant peptides were used to stimulate PBMCs of healthy 171 individuals ( Figure S4 ; Table S4 ). Overall, these data demonstrated that 15 predicted Furthermore, the specificities of the P64 reactive TCRs were verified in CD8 +

Jurkat cells, which were transduced with the top 4 dominant TCR clones (TCR 1-4).

Using the P64-HLA-A*11:01 tetramer, we confirmed that only TCR 1 and TCR 4 193 were P64-specific ( Figure 2C ; Table S6 ). Also, with co-culture of the transfection efficiency by N 361-369 -HLA-A*11:01 tetramer. We found that 9.86% 207 CD8 + T cells expressed TCR 1 and 6.34% expressed TCR 4 ( Figure 3A ).

Unexpectedly, we observed higher percentages of TCR-transduced CD4 + T cells, with 209 38.4% for TCR 1 and 11.7% for TCR 4 ( Figure 3A ). These results indicated that 210 TCR-transduced CD4 + T cells were able to recognize the N 361-369 -HLA complex with 211 the absence of CD8 co-receptor. Also, the similar fluorescence intensities detected in 212 the CD4 + T cells and in the CD8 + T cells implied that TCR 1 and TCR 4 might both 213 have high avidity for the N 361-369 -HLA-A*11:01 complex ( Figure 3A ). To confirm the 214 binding of these TCRs to N 361-369 -HLA-A*11:01 was independent of CD8 co-receptor, 215 TCR 1 and TCR 4 were separately transduced to CD8and CD8 + Jurkat cells. Both 216 TCR 1-and TCR 4-transduced CD8 -Jurkat cells showed positive tetramer staining, 217 and comparable fluorescent signal levels were found between CD8and CD8 + Jurkat 218 cells ( Figure 3B ). These results were in line with those of the CD4 + and CD8 + T cells 219 ( Figures 3A and 3B ). 220 Next, we evaluated the functionality of TCR 1 and TCR 4 in CD8 + T cells and 221 CD4 + T cells from expanded CD3 + T cells expressing TCR 1 or TCR 4. After 222 co-cultured with HLA-A*11:01 + K-562 cells pulsed with N 361-369 , remarkable levels 223 of T cell activation were detected in TCR 1-and TCR 4-CD8 + T cells, indicated by 224 the upregulation of T cell activation marker CD137 expression, compared to the 225 DMSO controls ( Figure 3C ). Similar results were obtained in TCR-transduced CD4 + 226 T cells stimulated with N 361-369 ( Figure 3C ). Moreover, we evaluated the cytotoxicity 227 of TCR 1-and TCR 4-transduced CD8 + T cells against HLA-A*11:01 + K-562 cells, 228 using Calcein-AM release assay. Significant levels of target cell lysis were detected in 229 the presence of N 361-369 , relative to the DMSO controls ( Figure 3D ). Increasing 230 cytotoxic efficiency was observed at the effector to target (E:T) cell ratios of 20:1 or 231 higher, and approximately 70% target cell lysis was achieved at E:T cell ratio of 40:1 232 ( Figure 3D ). The functional activity of TCR 1-and TCR 4-transduced CD4 + T cells 233 was analysed by the production of IFN-γ, IL-2 and TNF-α via cytokine secretion 234 assay. We found that N 361-369 could elicit significant induction of all three cytokines in 235 these CD4 + T cells, comparing to an equimolar amount of DMSO ( Figure 3E ). These 236 results proved that the N 361-369 -specific TCR 1 and TCR 4 could mediate the 237 cytotoxicity of CD8 + T cells and the effector cytokine secretion of CD4 + T cells. EC50s of 118 nM and 380.9 nM respectively, in a co-receptor independent manner 248 ( Figure 4A ). These data confirmed that TCR 1 and TCR 4 both had high affinity to 249 the N 361-369 -HLA complex in CD8 + and CD4 + T cells. Figure 4B ). Interestingly, we found that N T366I preferably led to an 268 activation of the TCR 1-T cells similar to that of the wild type N 361-369 , and the TCR 269 4-T cells were mainly activated by N T362I or N F363L ( Figure 4B ). These findings Also, we found significantly higher levels of IFN-γ secretion in the N protein loaded 286 groups, relative to the S protein controls ( Figure 5B ). To evaluate the N 361-369 activated CD8 + T cells could kill lung organoids, we 302 co-cultured HLA-A*11:01 + LO loaded with the N 361-369 peptide with TCR 1-CD8 + T 303 cells. After 24 h, we observed apparent organoid lysis, together with the recruitment 304 of TCR 1-T cells, only with the presence of N 361-369 ( Figure 5E ). We found a 305 significantly higher level of target cell lysis in the N 361-369 loaded group, when these 306 effects were quantified by Calcein-AM release assay ( Figure 5E ). To determine 307 whether the cytotoxic CD8 + T cells could eliminate infected lung organoids, we 308 co-cultured HLA-A*11:01 + N-LO or S-LO with TCR 1-CD8 + T cells. We found that 309 lung organoids with endogenous expression of the N protein became more susceptible 310 to the cytotoxicity of specific T cells, comparing to the S-LO group ( Figure 5F ). Such 311 difference of the cytotoxic effect was statistically significant, analysed by LDH 312 release assay ( Figure 5F ). Taken together, we identified a dominant T cell epitope, 313 N 361-369 , that could activate specific CD8 + T cells, and the cytotoxicity of these T cells Table S5 ). Strikingly, we found that 3 out of 4 dominant peptides were 336 derived from the N protein, and accounted for approximately 50% of tested CD8 + T 337 cell reactivity. These data suggested that the N protein might be a dominant target for (Table 1) , offering more 347 accurate information for vaccine designs. Of these dominant epitopes, the 348 HLA-A*11:01-restricted N 361-369 (KTFPPTEPK) was found in 5/8 COVID-19 349 PBMCs ( Figure 1C ; Table 1 ). Independently, this epitope has also been identified in (Table S3) (Table S4) . Samples were collected and processed as described (Table S1 ). Individual peptide's HLA-A specificity 512 was listed in Table S2 . All 78 epitopes were synthesized (GenScript) and dissolved in 513 the corresponding solvents for the subsequent experiments. Detailed information for 514 peptide grouping was summarized in Table S2 . The supernatants were used for measuring IFN-γ concentrations by ELISA. RFUs -Spontaneous death RFUs).

The ability of T cells to lyse organoids was also measured using LDH-Glo™ 675 Cytotoxicity Assay Kit (Promega, USA) according to the manufacturer's instructions.

The organoids were dissociated to single cells and counted, which was used to infer 677 the number of organoids to allow co-cultured with T cells at a 4:1 effector:target ratio. histograms and line charts were presented as mean ± standard deviation (mean ± SD).

Statistical significance was determined using ANOVA for multiple comparisons.

Student's t-tests were applied to compare two groups. A value of P < 0.05 was 709 considered statistically significant. 

Evidence for a TCR affinity threshold delimiting maximal CD8 T cell function. 

Preliminary Identification of 713

Potential Vaccine Targets for the COVID-19 Coronavirus (SARS-CoV-2) Based on 714 SARS-CoV Immunological Studies

IMGT/V-QUEST: the highly 716 customized and integrated system for IG and TR standardized V-J and V-D-J 717 sequence analysis

The Quest for the Best: How 720 TCR Affinity

Potent Neutralizing Antibodies against SARS-CoV-2

Throughput Single-Cell Sequencing of Convalescent Patients' B 725 Cells

SARS-CoV-2 728 infection protects against rechallenge in rhesus macaques

Virus-specific memory CD8 T cells provide substantial protection from lethal severe 732 acute respiratory syndrome coronavirus infection

Human monoclonal antibodies block the binding of SARS-CoV-2 spike 736 protein to angiotensin converting enzyme 2 receptor. Cellular & molecular 737 immunology

Shared Antigen-specific CD8+ T cell Responses 740

Against the SARS-COV-2 Spike Protein in HLAA*02:01 COVID-19 Participants. 741 medRxiv

Enhanced antitumor activity of murine-human hybrid T-cell receptor (TCR) in human 744 lymphocytes is associated with improved pairing and TCR/CD3 stability

SARS-CoV-2 mRNA vaccine design enabled by prototype pathogen preparedness

Evaluation of 752 the mRNA-1273 Vaccine against SARS-CoV-2 in Nonhuman Primates

England journal of medicine

An interactive web-based dashboard to 755 track COVID-19 in real time. The Lancet. Infectious diseases

Why are RNA virus mutation rates so damn high? PLoS biology

Form Memory CD8+ T Cells that Recognize a Small Set of Shared Immunodominant 762 Epitopes in SARS-CoV-2. medRxiv

T cell receptor binding kinetics required 766 for T cell activation depend on the density of cognate ligand on the antigen-presenting 767 cell

A Sequence Homology and Bioinformatic Approach Can Predict Candidate 771 Targets for Immune Responses to SARS-CoV-2

Cell Responses to SARS-CoV-2 Coronavirus in Humans with COVID-19 Disease 776 and Unexposed Individuals

A novel, 779 rapid and efficient method of cloning functional antigen-specific T-cell receptors from 780 single human and mouse T-cells. Biochemical and biophysical research 781 communications

HLA common and well-documented alleles in China

A Multibasic Cleavage 786

Site in the Spike Protein of SARS-CoV-2 Is Essential for Infection of Human Lung 787

Engineered T cells 790 targeting E7 mediate regression of human papillomavirus cancers in a murine model

Simplified method of the growth of human tumor infiltrating 794 lymphocytes in gas-permeable flasks to numbers needed for patient treatment

Human neutralizing antibodies elicited by SARS-CoV-2 infection

0: Improved Peptide-MHC Class I Interaction Predictions Integrating 801

Eluted Ligand and Peptide Binding Affinity Data

Phase 1-2 Trial of a SARS

Recombinant Spike Protein Nanoparticle Vaccine. The New England journal of 806 medicine

SARS-CoV-2-specific T cell 809 immunity in cases of COVID-19 and SARS, and uninfected controls

DMSO was used as negative control. T cells were stained with blue dye and lung 1004 organoids were stained in green. Target cell lysis was quantified using Calcein-AM 1005 release assay (right). (F) Cytotoxicity assay of TCR 1-CD8 + T cells against 1006 HLA-A*11:01 + lung organoids endogenously expressing the N protein (left). 1007 Organoids expressing the S protein was applied as negative control. T cells were 1008 stained in red and lung organoids were stained in green.Target cell lysis was 

Usage for the epitopes screeningSupplementary