Dysfunctional Sars-CoV-2-M protein-specific cytotoxic T lymphocytes in patients recovering from severe COVID-19

Ogura, Hideki; Gohda, Jin; Lu, Xiuyuan; Yamamoto, Mizuki; Takesue, Yoshio; Son, Aoi; Doi, Sadayuki; Matsushita, Kazuyuki; Isobe, Fumitaka; Fukuda, Yoshihiro; Huang, Tai-Ping; Ueno, Takamasa; Mambo, Naomi; Murakami, Hiromoto; Kawaguchi, Yasushi; Inoue, Jun-ichiro; Shirai, Kunihiro; Yamasaki, Sho; Hirata, Jun-Ichi; Ishido, Satoshi

doi:10.1038/s41467-022-34655-1

Download PDF

Article
Open access
Published: 16 December 2022

Dysfunctional Sars-CoV-2-M protein-specific cytotoxic T lymphocytes in patients recovering from severe COVID-19

Nature Communications volume 13, Article number: 7063 (2022) Cite this article

4257 Accesses
4 Citations
13 Altmetric
Metrics details

Subjects

Abstract

Although the importance of virus-specific cytotoxic T lymphocytes (CTL) in virus clearance is evident in COVID-19, the characteristics of virus-specific CTLs related to disease severity have not been fully explored. Here we show that the phenotype of virus-specific CTLs against immunoprevalent epitopes in COVID-19 convalescents might differ according to the course of the disease. We establish a cellular screening method that uses artificial antigen presenting cells, expressing HLA-A^*24:02, the costimulatory molecule 4-1BBL, SARS-CoV-2 structural proteins S, M, and N and non-structural proteins ORF3a and nsp6/ORF1a. The screen implicates SARS-CoV-2 M protein as a frequent target of IFNγ secreting CD8⁺ T cells, and identifies M_198–206 as an immunoprevalent epitope in our cohort of HLA-A^*24:02 positive convalescent COVID-19 patients recovering from mild, moderate and severe disease. Further exploration of M_198–206-specific CD8⁺ T cells with single cell RNA sequencing reveals public TCRs in virus-specific CD8⁺ T cells, and shows an exhausted phenotype with less differentiated status in cells from the severe group compared to cells from the moderate group. In summary, this study describes a method to identify T cell epitopes, indicate that dysfunction of virus-specific CTLs might be an important determinant of clinical outcomes.

An immunodominant NP105–113-B*07:02 cytotoxic T cell response controls viral replication and is associated with less severe COVID-19 disease

Article Open access 01 December 2021

Activation or exhaustion of CD8+ T cells in patients with COVID-19

Article Open access 19 August 2021

Characterization of pre-existing and induced SARS-CoV-2-specific CD8+ T cells

Article 12 November 2020

Introduction

Even at present, SARS-CoV-2 is spreading across the world and causing COVID-19 with diverse clinical features; severe/critical respiratory infections such as pneumonia and acute respiratory distress syndrome is fatal outcomes especially in elder patients¹. Therefore, the mechanisms behind severe outcomes have been intensively investigated^2,3. Among these, an inability to engage timely anti-viral immune responses is proposed^4,5. In general, type I interferon (IFN-I) suppresses viral propagation, supports tissue repair and stimulates adaptive immunity, so IFN-I plays a critical role in innate immunity⁶. This is the case with SARS-CoV-2 infection^7,8; in some critically ill COVID-19 patients, insufficient IFN-I activation was reported⁹. In this scenario, impaired IFN-I activation leads to incomplete elimination of virus followed by subsequent leukocyte-related inflammation, the so-called two-step model of pathogenesis⁵.

As mechanisms other than impaired IFN-I signaling, maladapted acquired immunity is also proposed^10,11. Among essential cells in acquired immunity against COVID-19, cytotoxic T cell (CTL) and B cells were demonstrated to provide sterilizing immunity and humoral immunity, respectively^12,13. Autoreactive B cells have been reported to produce autoantibody against IFN-I, which suppress IFN-I signaling^14,15,16. Profound depletion of T cells in peripheral bloods was present presumably due to overproduction of inflammatory cytokines¹⁷. Thus, dysregulation of both innate and acquired immunity could be associated with COVID-19 severity. However, it is not fully understood how differentiation status of antigen/virus-specific CTLs is related to disease severity. Since antigen/virus-specific CTLs are essential sentinels against invading pathogens, thorough characterization of these in COVID-19 patients is thought to be necessary to unveil the mechanisms of COVID-19 progression, as well as to inform therapeutic strategies.

In this work, we explore immunoprevalent epitopes from COVID-19 convalescents for examining SARS-CoV-2-specific CTLs. We perform deep and thorough analysis to characterize the SARS-CoV-2 immunoprevalent epitope-specific CTLs for searching the features related to disease severity. The identified traits of dysfunction of the cells of severe COVID-19 convalescents highlight the impaired virus-specific CTL development as a possible determinant of clinical outcomes.

Results

Generation of artificial antigen-presenting cells

To explore and better define SARS-CoV-2-specific CD8⁺ T cell responses in convalescents, we sought to prepare artificial antigen-presenting cells (aAPCs) stably expressing SARS-CoV-2 proteins. Human myeloid leukemia K562 cells were employed as parental cells as we previously observed minimum antigenicity of K562 to human CD45RO⁺CD8⁺ T cells¹⁸. Since 70.6% (36/51) of COVID-19 convalescents (mild, 60.0% (12/20); moderate, 84.6% (11/13); and severe, 72.2% (13/18)) and 75.0% (9/12) of healthy controls enrolled in this study carried an HLA-A^*24:02 allele, K562 cells were firstly engineered to express HLA-A^*24:02, and the costimulatory molecule 4-1BBL (Supplementary Fig. 1a, b)^18,19. The function of antigen presentation was confirmed using influenza virus and cytomegalovirus (Flu/CMV)-derived peptides and peptide-dependent IFNγ secretion from a specific T cell line (Supplementary Fig. 1c). Then, three SARS-CoV-2 structural proteins (S, M, and N) and two non-structural proteins (ORF3a and nsp6/ORF1a), which were reported as viral antigens of SARS-CoV-2-specific CD8⁺ T cells, were selected to be expressed in aAPCs²⁰. Stable expression of each SARS-CoV-2 protein was confirmed by quantitative RT-PCR and western blot (Fig. 1d, e). Because these aAPCs endogenously express viral proteins, the proteins are processed and presented as viral peptides associated with MHC class I via the intracellular antigen processing pathway²¹. With these aAPCs, a new screening system was established (Fig. 1a).

**Fig. 1: Cellular screening of SARS-CoV-2-specific response.**

M protein is an immunocompetent viral protein

In order to achieve an in-depth analysis of SARS-CoV-2-specific CD8⁺ T cells, we generated CD8⁺ T cell libraries from peripheral blood cells of COVID-19 convalescents. These libraries are useful for the definitive characterization of low frequency, but potentially crucial, CD8⁺ T cell populations in peripheral blood because rare, virus-specific CD8⁺ T cells are polyclonally expanded followed by stimulation with relevant antigen-expressing aAPCs (Fig. 1a). COVID-19 convalescents enrolled in this study were admitted to, or consulted with a physician, at hospitals affiliated with Hyogo College of Medicine or Kyowa-kai Medical Corporation from late 2020. Convalescents had experienced mild, moderate, or severe COVID-19 (Supplementary Table 1). Disease severity of COVID-19 was classified according to a recent report²² (see Methods). Since HLA-A^*24:02 was the dominant allele in our study group, we focused on HLA-A^*24:02⁺ subjects and a total of 36 COVID-19 convalescents and 9 healthy volunteers were enrolled in the study (Supplementary Table 1), and 20 convalescents and 8 healthy volunteers among them were subjected to the library assay. A small number (two thousand) CD45RO⁺ CD8⁺ T cells isolated from each participant’s PBMC were seeded into each well of 96-well-plates and polyclonally expanded with PHA in the presence of allogenic PBMC and cytokines to establish a library (Fig. 1a). All the established libraries from each participant were evenly divided into 7 groups (i.e., 5 groups for a series of SARS-CoV-2 proteins, 1 group for mixed peptides derived from influenza virus and cytomegalovirus, and 1 group for a non-antigen negative control) and incubated with relevant aAPCs individually. IFNγ in the culture supernatant was measured as a functional read-out for antigen-specific CD8⁺ T cells, because it is a major effector cytokine of various virus-specific human CD8⁺ T cells^{23,24,25,26,27} and because IFNγ⁺ CD8⁺ T cells are strongly associated with low disease severity among acute cases of COVID-19²⁸. The response to the antigens was defined as positive when the IFNγ level was above mean +3 SD of that in the negative control (i.e., incubation with parental aAPCs). In total, 5403 libraries were examined in COVID-19 convalescents and healthy volunteers (3808 and 1595, respectively). Frequencies of IFNγ−positive libraries upon each antigen stimulation were compared among participant groups (i.e., COVID-19 convalescents and healthy volunteers). There were minimum responses observed in the non-antigen negative control groups of healthy and COVID-19 subjects; representative data are shown in Fig. 1b. Consistent with the previous reports, the response was observed against various antigens including structural and non-structural proteins (Supplementary Figs. 2, 3). Of note, those included reported immunodominant epitopes: S_1208-1216²⁹ and ORF3a_112-120³⁰ (Supplementary Fig. 3b). Importantly, SARS-CoV-2-M protein induced high responses in the libraries of convalescents compared with those of healthy volunteers (p = 0.0019, Mann–Whitney U test) (Fig. 1c, d) (Fig. 2) in our cohort. Interestingly, the frequency is the highest in the convalescents recovered from moderate severity COVID-19 (Fig. 1c, e). Although pre-existing cross-reactive immune memory to SARS-CoV-2 has been suggested³¹, no significant responses observed against the proteins we tested were detectable in samples from healthy subjects of the cohort.

**Fig. 2: Identification of M198–206 as a major epitope.**

Identification of an immunoprevalent CTL epitope M_198–206

To further examine the importance of M protein-specific CD8⁺ T cells in the context of SARS-CoV-2 infection, we sought to identify immunodominant epitopes of the M protein. The candidate M protein epitopes were obtained in silico as follows: candidate binders to HLA-A^*24:02 were screened using full-length M protein amino-acid sequence (YP_009724393.1) and the Immune Epitope Database (https://www.iedb.org), and the top 5 peptides with the highest score (0.5 <Score, Percentile rank <0.2) and having a typical amino acid length were chosen (Table 1). Once M protein-responding libraries were obtained, they were further expanded and divided into multiple wells to examine the response to each candidate peptide by coculturing with aAPCs pulsed with the individual peptide. Next, the supernatant was collected to measure IFNγ levels again. For each candidate peptide, the response was defined as positive when IFNγ levels were statistically higher than those of controls (i.e., stimulation with aAPCs alone) and the mean of IFNγ levels was more than twice that of controls. The majority of the libraries responded to only a single peptide, but some responded to two or more peptides, indicating the presence of two or more SARS-CoV-2-M-specific T cell clones among the 2000 original CD45RO⁺CD8⁺ T cells. In total, 37 libraries responded to M protein-expressing aAPCs (aAPC-M) from the 7 convalescents were tested for their antigen specificity. Surprisingly, 81.1% (30/37) of the libraries were found to respond to the M_198–206 peptide (Fig. 2a, b). Moreover, the response to M_198–206 was detected in the libraries from 6 out of 7 convalescents (Fig. 2b, c). In addition, in 5 out of 6 convalescents whose libraries responded to M_198–206, the frequency in the M_198–206 responding library was more than half (Fig. 2c). Lastly, the presence of M_198–206-specific CD8⁺ T cells in the libraries was confirmed by staining with MHC class I tetramer generated with HLA-A^*24:02 and synthetic M_198–206 peptide (Fig. 2d). By confirming its direct association to HLA-A^*24:02 by competition assay (Supplementary Fig. 4), we concluded that M_198–206 is one of the SARS-CoV-2 major epitopes in M protein. In order to examine the reliability of the library assay, we performed activation induced marker (AIM) assay with the same set of M peptides. As expected, M_198–206 response was observed in the convalescents having M-responding libraries (CV-004, and CV-007, Supplementary Fig. 5), but not in the one having no responding library (CV-006, Supplementary Fig. 5). There was minor difference in the peptide response between these assays. This might be due to the different populations to be identified by assays; the library assay identified IFNγ⁺CD8⁺CD45RO⁺ cells whereas AIM assay identified the cells which became positive for selected activation markers after peptide stimulation.

Table 1 The candidates of SARS-CoV-2-M epitopes for HLA-A*24:02

Full size table

Next, we further expanded the libraries obtained from moderate convalescents and performed the following downstream experiments because the libraries from severe convalescents tended not to expand well and did not reach enough cells. The effector features of M_198–206-specific CD8⁺ T cells, cytokine profiles were examined by intracellular cytokine staining (ICS). As shown in Fig. 3a, M_198–206-specific CD8⁺ T cells produced the typical inflammatory cytokines: IFNγ and TNFα. Further, we examined cytotoxic activity by employing M_198–206 peptide-pulsed Calu-3 cells as a model of SARS-CoV-2-infected lung epithelial cells. M_198–206-specific CD8⁺ T cell lines (Supplementary Fig. 6) from the libraries were subjected to the assay. As shown in Fig. 3b, antigen-specific cytotoxic activities were proportional to the effector/target (E/T) ratio. Furthermore, a TCR αβ pair cloned from an M_198–206-specific CD8⁺ T cell line was functional (Supplementary Table 2) (Supplementary Fig. 7). To examine direct effector functions of M_198–206-specific CD8⁺ T cells to SARS-CoV-2-infected lung epithelial cells, Calu-3 cells were infected with SARS-CoV-2-Wuhan strain for the coculture with M_198–206-specific CD8⁺ T cells. Remarkably, as shown in Fig. 3c, M_198–206-specific CD8⁺ T cells suppressed not only intracellular viral RNA replication, but also suppressed propagation of infectious virus in SARS-CoV-2-infected Calu-3 cells, identifying M_198–206 as a SARS-CoV-2 epitope of CTLs. To examine immunological relevance of M_198–206-specific CD8⁺ T cells, we studied the M_198–206-specific CD8⁺ T cells in the peripheral blood of COVID-19 convalescents with different clinical severities without in vitro expansion. As expected, tetramer-positive CD8⁺ T cells were detected (Fig. 4a, b) with significantly higher frequency in peripheral blood of moderate or severe COVID-19 convalescents, which is consistent with the results from the libraries (Fig. 2d). In addition, we examined the frequency of virus-specific CD8⁺ T cells which were reported as prevalent virus-specific CD8⁺ T cells in previous reports^30,32,33. As shown in Supplementary Fig. 8d, in our cohort, the frequency of M_198–206-specific CD8⁺ T was higher than that of ORF1b or ORF3a-specific CD8⁺ T cells. Further, additional convalescents who suffered moderate/severe COVID-19 in early 2022 (CV-052, 057, 062, 065, 071, and 073) (Supplementary Table 1), when Omicron strain spread in Japan, were examined. In fact, among the six convalescents from moderate/severe COVID-19, three were tested for Omicron and all of them were found to be positive for Omicron. Five out of six moderate-severe COVID-19 convalescents harbored M_198–206-specific CD8⁺ T cells with similar frequency to late 2020 (Fig. 4g). Importantly, M_198–206 specific CD8⁺ T cells were detected in the peripheral blood of COVID-19 convalescents for more than a year (Fig. 4f). We also found that an M_198–206-specific CTL line suppressed propagation of Omicron strain (Supplementary Fig. 8c). Taken together, we concluded that M_198–206 is an immunoprevalent CTL epitope in our study cohort.

**Fig. 3: Cytotoxic and antiviral activity of M198–206-specific CD8+ T cells.**

**Fig. 4: Phenotypic analysis of M198–206-specific CD8+ T cells in PBMC.**

Phenotypes and signatures of M_198–206-specific CTLs

By using M_198–206 MHC tetramer, we addressed the question as to whether the status of the virus-specific CD8⁺ T cells, such as differentiation, exhaustion, or senescence, reflects disease severity. With flow cytometry we found that the tetramer-positive CD8⁺ T cells in peripheral blood were significantly skewed toward an effector-memory (CCR7^- CD45RA^-) phenotype in moderate group, but not in severe group (Fig. 4c). The frequency of inhibitory receptor PD-1⁺ cells in tetramer-positive CD8⁺ T cells was higher than that in total CD8⁺ T cells in the severe group, suggesting increased exhaustion of M_198–206-specific CTLs; this observation supports several recent reports (Fig. 4d)^34,35,36. Interestingly, the senescence marker CD57⁺ was significantly higher in tetramer-positive CD8⁺ T cells of patients in the moderate group, but not those of patients in the severe group (Fig. 4e). These observations were not due to the difference in the time point after the infection (Supplementary Fig. 8a). The gating strategy for these analyses is demonstrated in Supplementary Fig. 9. Functionality of M_198–206-specific CD8⁺ T cells in the moderate group was confirmed by detection of response to peptide stimulation; they secreted IFNγ and TNFα upon the peptide stimulation as observed in the libraries (Supplementary Fig. 8b).

Next, we performed single-cell RNA-sequencing (scRNA-seq, 10X Genomics Platform) analysis on M_198–206-specific CD8⁺ T cells. A total of 18,222 tetramer-positive CD8⁺ T cells were isolated via florescence-activated cell sorting (FACS) from 10 PBMC samples derived from 6 convalescents (three moderate and three severe convalescents; in two moderate convalescents, samples from different time points after disease onset were included (Supplementary Table 3). Each sample was individually stained with Hashtag antibodies, followed by FACS-based isolation, then mixed and subjected to sequencing. As a result, single-cell transcriptomic data were obtained from 4,452 single cells. Uniform manifold approximation and projection (UMAP), a bioinformatic dimension reduction algorithm, identified 11 clusters (cluster 0 to 10) (Fig. 5a). Since cluster 9 (C9) and C10 did not include enough number of cells (<1%), these clusters were removed from the further analysis. Differentially expressed genes (DEGs, one cluster vs. rest of the cells) of each cluster and selected featured genes are shown in Fig. 5b. Clusters could be roughly divided into two groups (group 1 and 2); group 1 includes C1, C2, C3, C4 and C7 and group 2 includes C0, C5, C6, and C8. Of note, group 1 highly expressed cytotoxic-effector genes including GZMB, GZMA, and TBX21 with some preferences in the expression of FCGR3A, PRF1, GNLY, CX3CR1, GZMH, or activation markers HLA-DR and CD38 etc. (Fig. 5c and Supplementary Fig. 10a). Additionally, Gene set enrichment analysis with a consensus list of cytotoxicity signature genes³⁷, also showed high score in group 1 (Fig. 5g), demonstrating that these are the cytotoxic-effector or memory cells (T_cyto-eff/mem). B3GAT1 (CD57) expression was also localized in group 1 clusters (Supplementary Fig. 10a). Furthermore, DEG analysis between group 1 and 2 showed that in group 1, several cytotoxicity markers were upregulated while some of the naïve markers (e.g., CCR7, TCF7, LEF1) were downregulated, demonstrating that the group 2 includes naïve/less differentiated cells (Fig. 5d–f, Supplementary Fig. 10a). In fact, naive markers including CCR7, CD62L, CD28, and CD27 were highly expressed in C8, suggesting that C8 is the cluster of naïve cells or memory stem cells (hereafter designated as ‘T_naive-like’) (Fig. 5c, d and Supplementary Fig. 10a). The expression of naive markers was gradually decreased from C8 towards C5 (Fig. 5d), indicating the early differentiated status of C5. To elucidate the features distinguishing C0 and C6 from the rest of clusters, we explored marker genes (see “Methods”). Curiously, GZMK was identified as a selectively expressed gene in both C0 and C6, whose expression was slightly decreased toward C6 (Fig. 5c, d). Recent papers have reported GZMK as a marker of predysfunctional cells or precursor of exhausted T(T_PEX) cells, which have a distinct fate commitment to exhausted cells (T_EX) in human memory T cell pool³⁸. In line with this, C0 highly expressed TCF7 and IL7R and intermediately expressed ZNF683, PDCD1 (Fig. 5d, Supplementary Fig. 10a) consistent with gene signature of T_PEX cells. As expected, C6 highly expressed T_EX markers including inhibitory receptors such as PDCD1 and TIGIT (Fig. 5d). Regarding other inhibitory receptors, CD244 was expressed in C6 (T_EX) as previously reported³⁹ (Supplementary Fig. 10a). Another inhibitory receptor LAG3 was expressed in C6 (T_EX)and C0 (T_PEX)^40,41 (Supplementary Fig. 10a). Cytotoxic genes PRF1 and GZMB were highly expressed in C6 and moderately in C0, that are similar to the characters of T_EX and T_PEX cells^39,41 (Fig. 5d and Supplementary Fig. 10a). The expression pattern of transcription factor genes such as EOMES, TOX, TCF7, TBX21, are also consistent with previously reported signatures of T_EX and T_PEX cells^{41,42,43,44,45,46,47} (Fig. 5d and Supplementary Fig. 10a). These findings could support that C6 and C0 are corresponding to T_EX and T_PEX, respectively. Moreover, C6 showed high score by analysis with exhaustion signature gene list employed in a recent work³⁷ (Fig. 5h). Trajectory analysis revealed the sequential distribution of the heterogenous cell-states along the C0-C6 axis towards C6 (Fig. 5i). Thus, we concluded that trajectory from C0 to C6 represents a unique trajectory towards the exhaustion of SARS-CoV-2-specific CD8⁺ T cells. Although the number of convalescent subjects was small, cells from subjects with moderate disease tended to group into T_cyto-eff/mem cells (group 1 clusters). In contrast, cells from subjects in the severe category tended to group into T_EX cells (C6) (Fig. 5j). This observation supported the results from the flow cytometry analysis where the severe group had higher PD-1 and lower CD57 expression (Fig. 4d, e). Furthermore, to confirm the increased exhaustion phenotype of M_198–206-specific CD8⁺ T cells in the severe group, we focused on TIGIT, an additional exhaustion marker for T cell, because C6 specifically co-expressed PDCD1 and TIGIT (Supplementary Fig. 10b). We compared the frequency of PD-1⁺TIGIT⁺ cells between moderate and severe group through flow cytometry analysis of SARS-CoV-2-M_198–206-specific CD8⁺ T cells. As expected by scRNA-seq results, the frequency of the cells was significantly higher in the cells from severe group compared with those from moderate group (Fig. 5k). Of note, time post viral clearance was not correlated with the frequency of PD-1⁺TIGIT⁺ cells (Supplementary Fig. 8e).

**Fig. 5: Single-cell RNA-sequencing of M198–206-specific CD8+ T cells.**

Identification of public TCRs of virus-specific CD8⁺ T cells

scRNA-seq analysis provided TCR sequences of the SARS-CoV-2-M_198–206 specific CD8⁺ T cells. TCR clonal expansion analysis revealed that in all the clusters except C8 (T_naive-like) and C5, a majority of the clones expanded well (more than three cells observed, Fig. 6a) as expected by the signature of those clusters (T_cyto-eff/mem, T_PEX, and T_EX). Among TCR sequences of the top 20 expanded clones, clone 14 from the CV-001 convalescent had the same sequences as TCRα_rank1β_rank1 clone isolated from CD8⁺ T cell libraries (Supplementary Table 2, Supplementary Table 5). Interestingly, clone 58 from the same convalescent had the same amino acid sequences of α and β chains as clone 14 while there is a difference of a single nucleotide of α chain, suggesting that this TCR was advantageous in forming significant proportions of CTLs in this convalescent. Longitudinal analysis of the top 20 expanded clones from two convalescents (CV-001 and CV-004) detected long-lived memory CD8⁺ T cells resided likely in C0, C2, and C4, suggesting that these clusters included memory cells. (Fig. 6c).

**Fig. 6: Single-cell TCR-sequencing of M_198–206-specific CD8+ T cells.**

Although obtained TCR sequences showed the heterogeneity of the TCR clones, there were a few TCRs shared among different convalescents; they are demonstrated as common TCRβ-1, -2, -3, and -4 in Fig. 6b and Table 2. Of note, amino acid sequence of CDR3 in common TCRβ-1 was shared among all members of the moderate group of convalescents (CV-001, CV-003, and CV-004) with difference in a single nucleotide in all the subjects. Among TCR clones with common TCRβ−1, clone 151 and clone 893 had identical amino acid sequence of CDR3 in TCRα while clone 91 had similar but different amino acid sequence of CDR3 in TCRα, in which an amino acid at position 4 of CDR3 was different (Table 2). Moreover, amino acid sequence of CDR3 in TCRβ-2 was different only in an amino acid at position 8 from TCRβ-1. Among TCR clones with TCRβ-1 or TCRβ-2, clone 29, clone 151 and clone 893 had identical amino acid sequence of CDR3 in TCRα while clone 209 had very closed, but different amino acid sequence of CDR3 in TCRα, in which an amino acid at position 4 of CDR3 was different from clone 151, 893, 91, and 29. Collectively, we found public TCRαβ motif: “CAVXYNQGGKLIF” for α motif and “CASSDSGXDGYTF” for β motif. The identification of the public TCRs also could highlight the importance of M_198–206 immunoprevalent epitope recognition in SARS-CoV-2 clearance.

Table 2 Identification of Public TCRs

Full size table

Discussion

Even at present, SARS-CoV-2 is spreading across the world and accumulating mutations and causing COVID-19 with diverse clinical features. Especially in elder populations, life-threating outcomes are manifested; age-specific infection fatality rate has been estimated to exponentially increase to 15% at age 85¹. Intensive efforts to tackle this issue are on-going from immunological point of view. Recent comprehensive analyses demonstrated innate immunity (e.g., type I interferon) is a critical contributing factor to the course of COVID-19. In parallel with innate immunity, contributions of the adaptive immune system (e.g., cellular immunity) have been demonstrated^3,5. In this study, we focused on CTLs, a critical effector population for virus clearance, and performed a thorough characterization of virus-specific CTLs from COVID-19 convalescents with different severities focused on M_198–206, an immunoprevalent CTL epitope to understand how differentiation status of virus-specific CTLs is related to disease severity.

We screened T cell libraries as previously reported^18,48,49,50 by constructing several artificial antigen-presenting cells which express SARS-CoV-2 proteins, individually. The system was optimized to screen for virus-specific IFNγ⁺ CD8⁺ T cells. M protein was identified as an immunocompetent viral protein, and M_198–206 was the dominant epitope of the M protein. Compared to the recent papers reporting comprehensive analyses of peptide epitopes in which many structural and non-structural protein epitopes were identified^22,51, our screening results were skewed toward the M protein. This discrepancy could come from the difference in the epitope preference of each HLA haplotype since our study focused on HLA-A^*24:02. Additionally, the other papers identified cross-reactive T cells present prior to the COVID-19 pandemic in their cohorts^52,53, which we could not detect in our cohort. It could also influence the memory T cell pool in the convalescents and could affect the screening results, accordingly. The other possibility was that the polyclonal expansion step in our system, driven by TCR stimulation and cytokines etc. could favor specific memory or effector T cell clones because theoretically there should be T cell subpopulations having different proliferative capacities within the population of CD45RO⁺ cells. Thus, we might have underestimated the antigen-specific responses. On the other hand, one of the strong points of the assay is the ease of downstream analysis such as cytotoxicity assays, cytokine profiling by ICS, and tetramer staining, etc. The recently developed activation-induced marker (AIM) assay and ELISpot assay have successfully identified antigen-specific cells and are employed by many researchers. In a simple way, which increases the accessibility and feasibility of the assays, the assays picked the antigen-specific cells primed by the epitope in vivo. But since these assays need to consume most of subjected cells for analysis, it is difficult to perform the downstream analysis, which requires sufficient numbers of cells. Above all, it should be noted that all these assays focus on different aspects of the cells; for example, the AIM assay focuses on activation marker-positive cells shortly after the peptide stimulation, whereas the library assay focuses on IFNγ-producing CD8⁺ T cells from CD45RO⁺ population in this study.

Importantly, M_198–206-specific CD8⁺ T cells were detected with significantly higher frequency at the periphery of convalescents suffering moderate/severe COVID-19 from late 2020 through early 2022, consistent with the reports demonstrating M protein as well conserved viral protein among various SARS-CoV-2 strains (https://nextstrain.org/ncov/gisaid/global). In addition, the M_198–206-specific CD8⁺ T cell lines showed antigen-dependent IFNγ and TNFα production as well as cytotoxic activity and suppressed propagation of SARS-CoV-2 strains: Wuhan and Omicron. Thus, we concluded that M_198–206 was one of the immunoprevalent CTL epitopes in our cohort. Previously M_198–206 was examined as an HLA-A*30:01 or HLA-A*24:02-restricted CD8⁺ T cell epitope, but has not been highlighted as a crucial viral epitope associated with COVID-19 severity^29,33,54.

Next, we performed in-depth analysis of the immunoprevalent epitope M_198–206-specific CD8⁺ T cells to examine their characteristics. M_198–206-specific CD8⁺ T cells were detected at substantially high frequency in the moderate and severe groups, which enabled us to study their phenotypic differences by conventional flow cytometry analysis. We found that the exhaustion marker PD-1 was significantly high in the severe group, compared with the moderate group. In contrast, senescence/terminal differentiation marker CD57 was significantly lower in the severe group than the in moderate group. Additionally, tetramer-positive cells of the moderate group were highly skewed towards effector-memory cells, which was disturbed in the cells of the severe group.

We further took advantage of the high frequency of M_198–206-specific cells in the moderate and severe groups of our cohort and performed scRNA-seq analysis by isolating a total of 18,222 tetramer-positive cells from moderate and severe convalescents by FACS. As a result, this revealed a highly heterogenous state of the virus-specific CD8⁺ T cells. In addition to cytotoxic-effector/memory populations, we found exhausted/pre-exhausted populations along a unique trajectory; the trajectory was from GZMK-expressing progenitors of exhaustion/pre-exhausted cells to related exhausted cells, which had the highest signature scored by gene set enrichment analysis with inhibitory receptor expression such as PDCD1 and TIGIT and limited cytotoxic gene expression. Surprisingly, T_PEX cluster accounted for the biggest population on UMAP. It is of note that the cells from severe group tended to be accumulated into T_EX cell cluster, compared with those from the moderate group, which supported the results from the PD-1 and/or TIGIT staining experiment by flow cytometry analysis (Figs. 4c–e, 5k). As reported in chronic viral infection or cancer⁵⁵, functional loss of exhausted SARS-CoV-2-specific CTLs could result in a failure of proper elimination of the virus, which results in severe outcomes. There has been a debate regarding the exhaustion state of T cells in COVID-19. Recently Dr. Shin’s group beautifully showed that PD-1 expressing CD8⁺ T cells in COVID-19 were functionally active in terms of IFNγ production^56,57. It would not be contradictory because the loss of IFNγ production occurs only in severely exhausted PD-1-expressing CD8⁺ T cells as reviewed in elsewhere^57,58.

There are several papers performing scRNA-seq on SARS-CoV-2-specific CD8⁺ T cells^37,59. Early studies successfully analyzed AIM⁺ cells as SARS-CoV-2-specific CD8⁺ T cells³⁷, but MHC tetramer technology would capture different status of the cells because this could isolate the cells without stimulation. Recently an excellent study was reported by Dr Dong’s group; NP_105-113-B*07:02-sepecific CD8⁺ T cells were extensively analyzed with MHC-NP_105-113 peptide tetramer^51,59. NP_105-113-specific CD8⁺ T cells were frequently detected in mild group, but less in the severe group, and clonotypes with different functional avidity were detected in their cohort. Consistent with our results, scRNA-seq analysis showed high expression of granzyme K in NP_105-113-specific CD8⁺ T cell from severe group⁵⁹. Thus, together with this report, our data strengthened the basic concept that CTLs are as crucial immune effector cells to determine disease severity.

Single-cell TCR-sequencing analysis identified numerous clonotypes among the COVID-19 convalescents. Of note, several public TCRs were identified. These included public TCRαβ-motifs shared in 3/3 of the moderate convalescents, which could further highlight the importance of the M_198–206 epitope recognition in COVID-19. As expected, the transcriptomic status of the cells (i.e., distribution of the cells on UMAP) within and across the clonotypes were heterogenous, suggesting that the fate of CTLs such as cell division, differentiation, survival or cell death should have been regulated by various factors including basal state of the cells, amount and frequency of the antigen stimulation or co-stimulation, cytokines etc., during T cell activation in the clinical course of disease. Recent advanced platforms are accumulating information on public TCRs in infectious diseases and cancers^60,61,62. In HIV-infected patients, there are rare subjects who could control viral propagation without therapy, called HIV controllers⁶¹. Public TCRs were reported to be crucial for control of HIV in those subjects; these TCRs showed high affinity to Gag293, which the most immunoprevalent CD4 epitope in HIV capsid⁶¹. Although further analyses are required, public TCRs identified in our experiment might contribute to recovery or lead to less severity. Since public clonotypes of SARS-CoV-2-specific T cells has been accumulating^63,64, our information on public clonotypes of M-specific CTL could be important to consider further strategies against unresolved disaster.

We analyzed CTL phenotype in the context of a single M_198–206 epitope without including other epitopes or bystander T cells as controls for scRNA-seq analysis. Also, these analyzes were performed with low numbers of moderate/severe COVID-19 convalescents. Additionally, one HLA haplotype: HLA-A^*24:02 was focused on in this study because of high frequency in our study cohort. These are limitations of this study, therefore, continuous accumulation of the data from different cohorts with different immunocompetent epitopes is required. Furthermore, deeper association of exhausted phenotype and signature with SARS-CoV-2-M_198–206-specific CTLs from convalescents of severe disease could raise the possibility; it is a consequence of heightened immune activation that is associated with severer disease. However, since we observed less coordinated differentiation status of SARS-CoV-2-M_198–206-specific CTLs in the severe group comparing to the moderate group (Fig. 4c), it could be interpreted as a cause of severe COVID-19. In order to corroborate such an interpretation, detailed investigation of virus-specific CTLs in immune compromised hosts with different severity are required.

In conclusion, we propose the trajectory towards exhaustion as a SARS-CoV-2-specific CTL fate to dysfunction in COVID-19. This could lead to poor outcomes presumably due to insufficient innate immune system in COVID-19 (e.g., type 1 IFN signaling etc.). Moreover, M_198–206 could be highlighted as a crucial CTL epitope to determine COVID-19 severity. These results could provide a platform for understanding severe COVID-19 pathogenesis in relation with dysfunction of cellular immunity.

Methods

Study participants and ethics

COVID-19 convalescents were recruited from hospitals affiliated with Hyogo Medical University or Kyowakai Medical Corporation. COVID-19 convalescents were classified into three groups (i.e., mild, moderate, and severe) based on the extent of oxygen supplementation and requirement of mechanical ventilation (mild: no oxygen supplementation, moderate: oxygen supplementation FiO₂ < 0.5, severe: heavy oxygen supplementation FiO₂ > 0.5 and/or mechanical ventilation) according to a recent report²². COVID-19 RNA was detected by Ampdirect 2019.nCov2 (SHIMADZU) at Hyogo College of Medicine, one Step RT PCR at SRL or Aptima SARS-CoV-2 at Hokenkagaku-Nishinihon. COVID-19 antigen was detected by Lumipulse G SARS-CoV-2 Ag (Fujirebio Inc.) at Kyowakai Medical Corporation, ESPURAIN at SRL. Diagnosis of COVID-19 was performed based on the presence of either mRNA or antigen. SARS-CoV-2 antigen. Individuals without known significant health problems including suspected COVID-19 and without the COVID-19 vaccine treatment were enrolled as healthy volunteers. Participant information is shown in Supplementary Table 1. Ethical approval was given by the ethics committee of Hyogo College of Medicine (reference: 202104-144). The consent to publish clinical characteristics of all the participants was obtained through written informed consent. Peripheral blood was drawn from convalescents recovered from COVID-19 with different severities or healthy volunteers after written informed consent was given. PBMCs were isolated by Ficoll-Hypaque gradient centrifugation and genomic DNA were purified using QIAamp DNA blood mini kit (51104, Qiagen). All the subjects were tested for HLA-A DNA typing (GenoDive Pharma Inc.) and the ones positive for HLA-A^*24:02 were subjected to the CD8⁺ T cell library assay and/or MHC tetramer staining etc. Our reporting of clinical data complies to the STROBE guidelines.

Cell lines

VeroE6/TMPRSS2⁶⁵ (JCRB 1819) cells were at 37 °C in 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM) (Thermo Fisher Scientific) supplemented with 10% heat-inactivated fetal bovine serum and 1 mg/ml G418. Calu-3 (ATCC HTB-55) cells, a human lung epithelial cell line, were maintained in Minimum Essential Medium (MEM) (Thermo Fisher Scientific) supplemented with 20% heat-inactivated fetal bovine serum. For the establishment of another VeroE6 cells expressing TMPRSS2, a vesicular stomatitis virus (VSV)-G pseudotyped lentivirus having human tmprss2 gene was produced using 293FT cells. VeroE6 (ATCC) cells infected with the pseudotyped virus were selected with 300 mg/ml hygromycin for at least 1 week. These bulk-selected cells were used for detecting SARS-CoV-2 viral RNA in supernatants from infected Calu-3 cells. TG40/CD8a cells were cultured at 37 °C in 5% CO₂ in RPMI medium (Wako) supplemented with 10% heat-inactivated fetal bovine serum.

Artificial antigen-presenting cell preparation

SARS-CoV-2 total RNA was provided by National Institute of Infectious Diseases. Viral genes including M, N, S, ORF3a, and ORF1ab nsp6 were cloned into pMSCVpuro (Clontech) plasmid DNA and transfected to retrovirus producing phoenix 293 T cells. For the artificial antigen-presenting cells, K562 cells were transfected with HLA-A^*24:02:01 cDNA⁶⁶ (provided by RIKEN BRC DNA Bank) and costimulatory molecule 4-1BBL cDNA. Next the cells were transduced with the viral genes via retroviral overexpression system (Clontech) and antibiotic selection.

Antibodies and reagents

The following monoclonal antibodies were used for flow cytometry analysis and cell sorting; anti-CD8a-Alexa Fluor 488 (300916, BioLegend, 1:200), anti-CD8a-FITC (FITC-65135, Proteintech, 1:50), anti-CD3-PE (317307, BioLegend, 1:200), anti-CD4-APC (300552, BioLegend, 1:200), anti-CD45RA-FITC (304148, BioLegend, 1:200), anti-CD45RO-PerCP/Cy5.5 (304222, BioLegend, 1:200), anti-PD-1-APC/Cy7 (329921, BioLegend, 1:50), anti-CD57-PerCP/Cy5.5 (359621, BioLegend, 1:200), anti-IFNγ-APC (502511, Biolegend, 1:200), anti-TNFα-BV421 (502931, BioLegend, 1:100), anti-CCR7-PE (353203, BioLegend, 1:50), anti-TIGIT-PerCP/Cy5.5 (372717, BioLegend, 1:50), anti-pan HLA (M0736, Dako, 1:200), control mouse IgG2a (401501, BioLegend, 1:1111), and anti-mouse IgG-FITC (406001, BioLegend, 1:200). LIVE/DEAD Fixable-Near IR Dead Cell Stain Kit (L10119, Thermo Fisher Scientific, 1:400), LIVE/DEAD Fixable-Aqua Dead Cell Stain Kit (L34957, Thermo Fisher Scientific, 1:400), and Human BD Fc Block (564220, BD Pharmingen, 1:50) was also used. MHC tetramer was prepared using QuickSwitch Quant HLA-A^*24:02 Tetramer Kit-PE (TB-7302-K1, MBL Life Science) and QuickSwitch Quant HLA-A^*24:02 Tetramer Kit-BV421 (TB-7302-K4, MBL Life Science) with appropriate peptides. CMV pp65_341-349 peptide (TS-0020-P, MBL) was purchased and all the other peptides were synthesized with a purity of >95% (Scrum Inc., or Toyobo).

CD8⁺ T cell library assay

CD8⁺ T cell library assay was performed as previously described¹⁸ with minor modifications. Briefly, CD8⁺CD45RO⁺ cells sorted from PBMC were cultured in 96-well round-bottom plates at 2 × 10³ cells per well in complete DMEM medium supplemented with 5 mM HEPES, pH 7.3 (345-06681, Dojindo), 0.1 mM nonessential amino acids (139-15651, Fujifilm), 1 mM sodium pyruvate (190-14881, Fujifilm), 5% human serum (H3667, Sigma-Aldrich), 50 U/ml penicillin and 50 U/ml streptomycin (168-23191, Fujifilm) in the presence of 1 µg/ml PHA-L (11249738001, Roche) with 20 U/ml IL-2 (589104, BioLegend), 20 ng/ml IL-7 (581906, BioLegend), and 20 ng/ml IL-15 (570306, BioLegend) with 2 × 10⁴ cells/well of irradiated allogeneic feeder cells. Fresh cytokines were added every 3 days. On day 9, the libraries were screened for antigen specificity by coculturing with irradiated-artificial antigen-presenting cells (aAPCs) expressing a series of viral proteins described above. CMV pp65_341-349 (QYDPVAALF) and Influenza PA_130-138 (YYLEKANKI) pulsed aAPCs (without expressing SARS-CoV-2 viral proteins) were also used. Then, the culture supernatant was harvested for IFNγ measurements by ELISA. In some experiments, CD45RO⁺CD8⁺ T cell libraries with positive responses were further expanded by adding cytokine cocktail every 3 days for approximately 14 days and then restimulated with control, SARS-CoV-2 viral protein-expressing aAPCs, or antigen peptide-pulsed aAPC.

Flow cytometry and cell sorting

Flow cytometry were performed as previously described¹⁸. For cell-surface labeling, CD8⁺ T cell libraries or PBMCs were stained with the antibodies for 30 min on ice. The cells were then analyzed by BD LSRFortessa (BD Biosciences) with BD FACSDiva (V8.0) or MACSQuant Analyzer (Miltenyi Biotech) with MACSQuantify (version 2.4) or sorted using BD FACSAria II (BD Biosciences). Data analyses were performed with FlowJo (v10.4.2) (TreeStar). For the intracellular cytokine staining, cells were stimulated with indicated peptide for 2 hours and then further incubated in the presence of brefeldin A for 4 hours. After cell-surface staining, cells were fixed and permeabilized. Intracellular cytokines were detected with specific monoclonal antibodies using FoxP3/transcription factor staining buffer set (00-5523-00, eBioscience) according to the manufacturer’s instructions.

Real-time quantitative PCR

A GeneAmp 5700 sequence detection system (Applied Biosystems) and KAPA SYBR FAST (KK4621, Kapa Biosystems) were used to quantify the levels of indicated mRNA expressions. Thermal cycles was; 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 sec and 60 °C for 1 min. The PCR primer sets used for real-time PCRs were as follows: S (5′-AACGCCACCAGATTTGCATC-3′,’-GTTTGCCCTGGAGCGATTTG-3′); M (5′-CGCGTTCCATGTGGTCATTC-3′, 5′-CCTTGATGTCACAGCGTCCT-3′); N (5′-CCTCGGCAAAAACGTACTGC-3′, 5′-TGGCACCTGTGTAGGTCAAC-3′); ORF3a (5′-CGGATGGCTTATTGTTGGCG-3′, 5′-GCAACGAGCAAAAGGTGTGA-3′); ORF1ab (5′-CTGCTAGTTGGGTGATGCGT-3′, 5′-AGCCCACATGGAAATGGCTT-3′); human Actb (5′-ACAGAGCCTCGCCTTTGC-3′ and 5′-CCACCATCACGCCCTGG-3′). In SARS-CoV-2 virus infection experiments, quantitative real-time RT-PCR was carried out using THUNDERBIRD^TM SYBR qPCR Mix (TOYOBO) at 95 °C for 3 min, followed by 50 cycles of 95 °C for 10 sec and 60 °C for 1 min. Detection of fluorescence during the thermal cycling process and quantification studies were performed using CFX Connect^TM Real-Time PCR detection system (BIO-RAD). The level of rpl13a expression in each sample was used to standardize the data.

The following primer sets were used: SARS-CoV-2 N (5′-AAATTTTGGGGACCAGGAAC-3′, 5′-TGGCAG CTGTGTAGGTCAAC-3′), human rpl13a (5′-TGTTTGACGGCATCCCAC-3′, 5′-CTGTCACTGCCTGGTACTTC-3′), and African green monkey rpl13a (5′-CTCAAGGTTGTGCGTCTGAA-3′, 5′-CTGTCACTGCCTGGTACTTCCA-3′).

Western blot

Cells were harvested, lysed with lysis buffer (20 mM Tris-HCl; pH 7.5, 100 mM NaCl, 2 mM EDTA, 0.5% Triton X-100) supplemented with cOmplete protease inhibitor cocktail (11697498001, Roche). The lysates were electrophoresed on 8-20% SDS-polyacrylamide gels. Then samples were transferred to polyvinylidene difluoride membranes (Millipore), and blocked with 5% milk in Tris-buffered saline with 0.05% Tween-20 (TBS-Tween). Primary antibodies were as follows: anti-SARS-CoV-2-S antibody (GTX632604, GeneTex, 1/1000 in 5% milk TBS-Tween), anti-SARS-CoV-1-M antibody (AP6008b, Abgent, 1/1000 in 5% milk TBS-Tween), anti-SARS-CoV-2-N antibody (GTX135357, GeneTex, 1/1000 in 5% milk TBS-Tween), anti-SARS-CoV-2-ORF3a antibody (A20234, ABclonal, 1/1000 in 5% milk TBS-Tween), anti-SARS-CoV-2-NSP6 antibody (9177, ProSci, 1/1000 in 5% milk TBS-Tween), and anti-tubulin (T5168, Sigma-Aldrich, 1/10,000 in 5% milk TBS-Tween). Secondary antibodies were as follows: anti-mouse IgG-HRP (330, MBL, 1/10,000 in 5% milk TBS-Tween), anti-rabbit IgG (458, MBL, 1/10,000 in 5% milk). The signals were detected through the enhanced chemiluminescent substrate (WBKLS0100, Millipore, or 34094, Thermo Fisher Scientific) by Intelligent Dark BoxII (Fujifilm) using IR LAS-1000 Pro (version 2.5) software. Uncropped scans of the most important western blots are provided in the Source Data file.

Cytotoxicity assay

M_198–206-specific CD8⁺ T cells were enriched as effector cells as follows. M_198–206-responding libraries from COVID-19 convalescents were further expanded and enriched with cytokines in the presence of irradiated (45 Grey) M expressing aAPCs for a couple of weeks. Then the cells were harvested and M_198–206 specific cells were purified with M_198–206 MHC tetramer-PE and anti-PE microbeads (Miltenyi). The resulted cells were confirmed to be M_198–206 tetramer⁺ with the purity of >95%. Next, in a 96-well plate, peptide-pulsed Calu-3 cells or unpulsed control cells were labeled with Calcein-AM (Dojin) as previously indicated⁶⁷. Then, the effector cells were added or not added to the target cells with different E/T ratios as indicated. 24 hours later, the cells were extensively washed and intracellular calcein levels were measured using a fluorescence microplate reader Infinite M200 Pro (TECAN) (λ_Em = 490 nm, λ_Ex = 520 nm). Wells were triplicated for each condition and the percentage of killing was calculated as ((OD of no-effector added wells)-(OD of effector added wells))/(OD of no-effector added wells)×100.

Preparation of SARS-CoV-2 virus stock

The SARS-CoV-2 isolate (UT-NCGM02/Human/2020/Tokyo)⁶⁸ and the Omicron isolate (BA.1 linage, TY38-873) from the National Institute of Infectious Diseases, Japan were propagated in VeroE6/TMPRSS2 (JCRB 1819) cells in DMEM containing 5% heat-inactivated fetal bovine serum at 37 °C in 5% CO₂. Briefly, SARS-CoV-2 was added at a multiplicity of infection (MOI) of 0.01 to VeroE6/TMPRSS2 (JCRB 1819) cells and incubated for 30 min at 37 °C. The culture medium was replaced with fresh medium. Cells were incubated for an additional 48 hours. The supernatant was centrifuged at 800×g for 5 minutes to remove cell debris. The supernatant was stored as virus stocks at −80 °C. The virus titer was determined by plaque assay using VeroE6/TMPRSS2 (JCRB 1819) cells.

SARS-CoV-2 infection assay

Calu-3 cells were seeded at 2 × 10⁴ cells per well in a 96-well cell culture plate. The following day, cells were infected with SARS-CoV-2 for 30 min at an MOI of 0.1 for Wuhan strain and MOI of 1 for Omicron strain. Cells were washed with PBS and incubated in fresh medium for 24 hours. The medium with or without 1 × 10⁵ or 2 × 10⁵ CTLs were added to the cells. Cells were incubated for an additional 24 hours. Supernatants were collected and stored at −80 °C after cell debris were removed by centrifugation at 800 g for 5 min. To measure the amount of viral RNA amplified in Calu-3 cells, the cells were washed three times with PBS and cell-lysis and cDNA synthesis were performed using SuperPrep II Cell Lysis & RT Kit for qPCR (TOYOBO) according to the manufacturer’s instructions. To measure the amount of infectious viral particles released from infected Calu-3 cells, 10 µl of the supernatants were incubated with VeroE6/TMPRSS2 (ATCC) cells seeded in a 96-well cell culture plate for 24 h. After washed three times with PBS, cells were lysed and cDNA was synthesized using SuperPrep II Cell Lysis & RT Kit for qPCR (TOYOBO) according to the manufacturer’s instructions.

Antigenic peptide screening

Immune Epitope Database (IEDB) was used for the antigenic peptide screening (https://www.iedb.org).

Peptide competition assay

Direct association of the M_198–206 peptide with HLA-A*24:02 was examined using the components of QuickSwitch Quant Tetramer Kit-PE (TB-7302-K1) according to the manufacturer’s instructions. Briefly, HLA-ABC Magnetic Capture Beads were mixed with or without QuickSwitch Tetramer or M_198–206 tetramer. Then, the beads were rinsed and stained with FITC-labeled Exiting Peptide antibody. The beads were rinsed and subjected to flowcytometry analysis using BD LSRFortessa (BD Biosciences). Mean fluorescent intensity (MFI) of FITC channel were calculated using FlowJo (v10.4.2) (TreeStar). The frequency of Exiting Peptide was calculated as follows; no-tetramer control and QuickSwitch Tetramer control were calculated as 0 and 100 percent, individually. Then, generate a linear curve by plotting the MFIs obtained with two controls against percent Exiting Peptide. Finally use the MFI of M_198–206 tetramer for calculating the percentage of peptide exchange.

TCR analysis

For the repertoire analysis, total RNA from SARS-CoV-2-M_198–206 specific cell line was purified and subjected to next-generation sequencing (Repertoire Genesys Inc., Osaka, Japan). cDNAs of TCR alpha and beta chains were linked by T2A sequence and subcloned into pMX-IRES-GFP vector. Ecotropic 293 T cells were used as packaging cells and resulted retroviral supernatant was collected. Viral transduction of the genes to TG40/CD8 cells were performed as previously described with minor modifications^67,69.

Single-cell transcriptome analysis

Single-cell libraries were prepared with reagents and instructions from 10x Genomics. cDNA was amplified for 14 cycles, and up to 50 ng of cDNA were used for gene expression libraries. Doublets were removed by using Scrublet⁷⁰. The top 4000 highly variable genes were selected, and used for clustering. Further data analysis was done with BBrowser platform (version 3.3.6, Bio Turing). Cytotoxicity signature and exhaustion signature scores were generated using published lists of genes³⁷ (Supplementary Table 4) with BBrowser⁷¹. Dimensionality reduction was done by UMAP (uwot package: https://github.com/jlmelville/uwot.), the number of neighbors is set at 30. Louvain clustering on the PCA results was run by igraph package⁷² with a flexible number of nearest neighbors. To detect marker genes, a non-parametric Venice method was used⁷³. Venice was also utilized for differential expression analysis between two groups. For trajectory analysis, information about cell embeddings on UMAP were fed to monocle3′s algorithm to obtain graph’s structure⁷⁴.

AIM assay

AIM assay was performed as previously described²⁰. PBMCs were cultured for 24 hours in the presence of peptide (10 µg/ml) or DMSO in 96-wells U bottom plates at 1 × 10⁶ cells/well. CD69⁺CD137⁺ cells in CD8⁺ T-cell population were detected as AIM⁺ by flow cytometry.

Statistical analysis

Comparisons were made using the indicated statistical tests using GraphPad software (version 7.02). Unless indicated, Mann-Whitney or Wilcoxon tests were applied for unpaired or paired comparisons, respectively. For library studies, wells greater than mean + 3 SD of the IFNγ levels for wells cultured with aAPCs (without expressing SARS-CoV-2 viral proteins) were calculated for each subject were considered positive. The percentage of positive library wells is presented as: (number of positive wells/total number of wells) × 100.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Immune Epitope Database was accessible online (https://www.iedb.org/). scRNA-seq data of SARS-CoV-2-M_198–206-specific CD8⁺ T cells generated in this study have been deposited in the Gene Expression Omnibus datasets under accession code GSE209676. The remaining data are available within the paper and Source Data file provided with this paper. Source data are provided with this paper.

References

Levin, A. T. et al. Assessing the age specificity of infection fatality rates for COVID-19: systematic review, meta-analysis, and public policy implications. Eur. J. Epidemiol. 35, 1123–1138 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kousathanas, A. et al. Whole genome sequencing reveals host factors underlying critical Covid-19. Nature 607, 97–103 (2022).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Q., Bastard, P., Cobat, A. & Casanova, J. L. Human genetic and immunological determinants of critical COVID-19 pneumonia. Nature 603, 587–598 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Zhang, Q. et al. Life-threatening COVID-19: defective interferons unleash excessive inflammation. Med. (N. Y) 1, 14–20 (2020).
Google Scholar
Merad, M., Blish, C. A., Sallusto, F. & Iwasaki, A. The immunology and immunopathology of COVID-19. Science 375, 1122–1127 (2022).
Article ADS CAS PubMed Google Scholar
Stark, G. R. et al. How cells respond to interferons. Annu. Rev. Biochem. 67, 227–264 (1998).
Article CAS PubMed Google Scholar
Kolumam, G. A. et al. Type I interferons act directly on CD8 T cells to allow clonal expansion and memory formation in response to viral infection. J. Exp. Med. 202, 637–650 (2005).
Article CAS PubMed PubMed Central Google Scholar
Vanderheiden, A. et al. Type I and Type III interferons restrict SARS-CoV-2 infection of human airway epithelial cultures. J. Virol. 94, e00985–00920 (2020).
Article CAS PubMed PubMed Central Google Scholar
Hadjadj, J. et al. Impaired type I interferon activity and inflammatory responses in severe COVID-19 patients. Science 369, 718–724 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Zhou, R. et al. Acute SARS-CoV-2 infection impairs dendritic cell and T Cell responses. Immunity 53, 864–877.e865 (2020).
Article CAS PubMed PubMed Central Google Scholar
Unterman, A. et al. Single-cell multi-omics reveals dyssynchrony of the innate and adaptive immune system in progressive COVID-19. Nat. Commun. 13, 440 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Ni, L. et al. Detection of SARS-CoV-2-specific humoral and cellular immunity in COVID-19 convalescent individuals. Immunity 52, 971–977.e973 (2020).
Article CAS PubMed PubMed Central Google Scholar
Moss, P. The T cell immune response against SARS-CoV-2. Nat. Immunol. 23, 186–193 (2022).
Article CAS PubMed Google Scholar
Bastard, P. et al. Autoantibodies against type I IFNs in patients with life-threatening COVID-19. Science 370, eabd4585 (2020).
Article CAS PubMed PubMed Central Google Scholar
Wang, E. Y. et al. Diverse functional autoantibodies in patients with COVID-19. Nature 595, 283–288 (2021).
Article ADS CAS PubMed Google Scholar
Chang, S. E. et al. New-onset IgG autoantibodies in hospitalized patients with COVID-19. Nat. Commun. 12, 5417 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Mathew, D. et al. Deep immune profiling of COVID-19 patients reveals distinct immunotypes with therapeutic implications. Science 369, eabc8511 (2020).
Article CAS PubMed PubMed Central Google Scholar
Ogura, H. et al. Identification and analysis of islet antigen-specific CD8(+) T cells with T cell libraries. J. Immunol. 201, 1662–1670 (2018).
Article CAS PubMed Google Scholar
Yan, X., Johnson, B. D. & Orentas, R. J. Murine CD8 lymphocyte expansion in vitro by artificial antigen-presenting cells expressing CD137L (4-1BBL) is superior to CD28, and CD137L expressed on neuroblastoma expands CD8 tumour-reactive effector cells in vivo. Immunology 112, 105–116 (2004).
Article CAS PubMed PubMed Central Google Scholar
Grifoni, A. et al. Targets of T cell responses to SARS-CoV-2 coronavirus in humans with COVID-19 disease and unexposed individuals. Cell 181, 1489–1501.e1415 (2020).
Article CAS PubMed PubMed Central Google Scholar
Neefjes, J., Jongsma, M. L., Paul, P. & Bakke, O. Towards a systems understanding of MHC class I and MHC class II antigen presentation. Nat. Rev. Immunol. 11, 823–836 (2011).
Article CAS PubMed Google Scholar
Tan, A. T. et al. 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 (2021).
Article CAS PubMed PubMed Central Google Scholar
Tzannou, I. et al. Immunologic profiling of human metapneumovirus for the development of targeted immunotherapy. J. Infect. Dis. 216, 678–687 (2017).
Article CAS PubMed Google Scholar
Sridhar, S. et al. Cellular immune correlates of protection against symptomatic pandemic influenza. Nat. Med. 19, 1305–1312 (2013).
Article CAS PubMed Google Scholar
Steinke, J. W. et al. Immune surveillance by rhinovirus-specific circulating CD4+ and CD8+ T lymphocytes. PLoS One 10, e0115271 (2015).
Article PubMed PubMed Central Google Scholar
Jozwik, A. et al. RSV-specific airway resident memory CD8+ T cells and differential disease severity after experimental human infection. Nat. Commun. 6, 10224 (2015).
Article ADS CAS PubMed Google Scholar
Savic, M. et al. Epitope specific T-cell responses against influenza A in a healthy population. Immunology 147, 165–177 (2016).
Article CAS PubMed Google Scholar
Rydyznski Moderbacher, C. et al. Antigen-specific adaptive immunity to SARS-CoV-2 in acute COVID-19 and associations with age and disease severity. Cell 183, 996–1012 (2020).
Article CAS PubMed PubMed Central Google Scholar
Rowntree L. C. et al. SARS-CoV-2-specific CD8(+) T-cell responses and TCR signatures in the context of a prominent HLA-A*24:02 allomorph. Immunol. Cell Biol. 99, 990-1000 (2021).
Ferretti, A. P. et al. Unbiased screens show CD8(+) T cells of COVID-19 patients recognize shared epitopes in SARS-CoV-2 that largely reside outside the spike protein. Immunity 53, 1095–1107.e1093 (2020).
Article CAS PubMed PubMed Central Google Scholar
Lipsitch, M., Grad, Y. H., Sette, A. & Crotty, S. Cross-reactive memory T cells and herd immunity to SARS-CoV-2. Nat. Rev. Immunol. 20, 709–713 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kared, H. et al. SARS-CoV-2-specific CD8+ T cell responses in convalescent COVID-19 individuals. J. Clin. Invest. 131, e145476 (2021).
Article CAS PubMed Central Google Scholar
Saini, S. K. et al. SARS-CoV-2 genome-wide T cell epitope mapping reveals immunodominance and substantial CD8(+) T cell activation in COVID-19 patients. Sci. Immunol. 6, eabf7550 (2021).
Article PubMed PubMed Central Google Scholar
Diao, B. et al. Reduction and functional exhaustion of T cells in patients with coronavirus disease 2019 (COVID-19). Front. Immunol. 11, 827 (2020).
Article CAS PubMed PubMed Central Google Scholar
Song, J. W. et al. Immunological and inflammatory profiles in mild and severe cases of COVID-19. Nat. Commun. 11, 3410 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
De Biasi, S. et al. Marked T cell activation, senescence, exhaustion and skewing towards TH17 in patients with COVID-19 pneumonia. Nat. Commun. 11, 3434 (2020).
Article ADS PubMed PubMed Central Google Scholar
Kusnadi, A. et al. Severely ill COVID-19 patients display impaired exhaustion features in SARS-CoV-2-reactive CD8(+) T cells. Sci. Immunol. 6, eabe4782 (2021).
Article PubMed PubMed Central Google Scholar
Galletti, G. et al. Two subsets of stem-like CD8(+) memory T cell progenitors with distinct fate commitments in humans. Nat. Immunol. 21, 1552–1562 (2020).
Article PubMed PubMed Central Google Scholar
Wieland, D. et al. TCF1(+) hepatitis C virus-specific CD8(+) T cells are maintained after cessation of chronic antigen stimulation. Nat. Commun. 8, 15050 (2017).
Article ADS PubMed PubMed Central Google Scholar
Kallies, A., Zehn, D. & Utzschneider, D. T. Precursor exhausted T cells: key to successful immunotherapy? Nat. Rev. Immunol. 20, 128–136 (2020).
Article CAS PubMed Google Scholar
Brummelman, J. et al. High-dimensional single cell analysis identifies stem-like cytotoxic CD8(+) T cells infiltrating human tumors. J. Exp. Med. 215, 2520–2535 (2018).
Article CAS PubMed PubMed Central Google Scholar
Utzschneider, D. T. et al. T cell factor 1-expressing memory-like CD8(+) T cells sustain the immune response to chronic viral infections. Immunity 45, 415–427 (2016).
Article CAS PubMed Google Scholar
Siddiqui, I. et al. Intratumoral Tcf1(+)PD-1(+)CD8(+) T cells with stem-like properties promote tumor control in response to vaccination and checkpoint blockade immunotherapy. Immunity 50, 195–211.e110 (2019).
Article CAS PubMed Google Scholar
Khan, O. et al. TOX transcriptionally and epigenetically programs CD8(+) T cell exhaustion. Nature 571, 211–218 (2019).
Article CAS PubMed PubMed Central Google Scholar
Wang, X. et al. TOX promotes the exhaustion of antitumor CD8(+) T cells by preventing PD1 degradation in hepatocellular carcinoma. J. Hepatol. 71, 731–741 (2019).
Article CAS PubMed Google Scholar
Alfei, F. et al. TOX reinforces the phenotype and longevity of exhausted T cells in chronic viral infection. Nature 571, 265–269 (2019).
Article CAS PubMed Google Scholar
Scott, A. C. et al. TOX is a critical regulator of tumour-specific T cell differentiation. Nature 571, 270–274 (2019).
Article CAS PubMed PubMed Central Google Scholar
Geiger, R., Duhen, T., Lanzavecchia, A. & Sallusto, F. Human naive and memory CD4+ T cell repertoires specific for naturally processed antigens analyzed using libraries of amplified T cells. J. Exp. Med. 206, 1525–1534 (2009).
Article CAS PubMed PubMed Central Google Scholar
Cao, Y. et al. Functional inflammatory profiles distinguish myelin-reactive T cells from patients with multiple sclerosis. Sci. Transl. Med. 7, 287ra274 (2015).
Article Google Scholar
Latorre, D. et al. T cells in patients with narcolepsy target self-antigens of hypocretin neurons. Nature 562, 63–68 (2018).
Article ADS CAS PubMed Google Scholar
Peng, Y. et al. 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 (2020).
Article CAS PubMed PubMed Central Google Scholar
Swadling, L. et al. Pre-existing polymerase-specific T cells expand in abortive seronegative SARS-CoV-2. Nature 601, 110–117 (2022).
Article ADS CAS PubMed Google Scholar
Buckley, P. R. et al. HLA-dependent variation in SARS-CoV-2 CD8 (+) T cell cross-reactivity with human coronaviruses. Immunology 166, 78–103 (2022).
Article CAS PubMed Google Scholar
Francis, J. M. et al. Allelic variation in class I HLA determines CD8(+) T cell repertoire shape and cross-reactive memory responses to SARS-CoV-2. Sci. Immunol. 7, eabk3070 (2022).
CAS PubMed Google Scholar
Hashimoto, M. et al. CD8 T cell exhaustion in chronic infection and cancer: opportunities for interventions. Annu. Rev. Med. 69, 301–318 (2018).
Article CAS PubMed Google Scholar
Rha, M. S. et al. PD-1-expressing SARS-CoV-2-specific CD8(+) T cells are not exhausted, but functional in patients with COVID-19. Immunity 54, 44–52.e43 (2021).
Article CAS PubMed Google Scholar
Rha, M. S. & Shin, E. C. Activation or exhaustion of CD8(+) T cells in patients with COVID-19. Cell Mol. Immunol. 18, 2325–2333 (2021).
Article CAS PubMed PubMed Central Google Scholar
Mackerness, K. J. et al. Pronounced virus-dependent activation drives exhaustion but sustains IFN-γ transcript levels. J. Immunol. 185, 3643–3651 (2010).
Article CAS PubMed Google Scholar
Peng, Y. et al. An immunodominant NP(105-113)-B*07:02 cytotoxic T cell response controls viral replication and is associated with less severe COVID-19 disease. Nat. Immunol. 23, 50–61 (2022).
Article CAS PubMed Google Scholar
Price, D. A. et al. Public clonotype usage identifies protective Gag-specific CD8+ T cell responses in SIV infection. J. Exp. Med. 206, 923–936 (2009).
Article CAS PubMed PubMed Central Google Scholar
Benati, D. et al. Public T cell receptors confer high-avidity CD4 responses to HIV controllers. J. Clin. Invest. 126, 2093–2108 (2016).
Article PubMed PubMed Central Google Scholar
Teng, Y. H. F. et al. Analysis of T cell receptor clonotypes in tumor microenvironment identifies shared cancer-type-specific signatures. Cancer Immunol. Immunother. 71, 989–998 (2022).
Article CAS PubMed Google Scholar
Shomuradova, A. S. et al. SARS-CoV-2 epitopes are recognized by a public and diverse repertoire of human T cell receptors. Immunity 53, 1245–1257.e1245 (2020).
Article CAS PubMed PubMed Central Google Scholar
Lu, X. et al. Identification of conserved SARS-CoV-2 spike epitopes that expand public cTfh clonotypes in mild COVID-19 patients. J. Exp. Med. 218, e20211327 (2021).
Article CAS PubMed PubMed Central Google Scholar
Matsuyama, S. et al. Enhanced isolation of SARS-CoV-2 by TMPRSS2-expressing cells. Proc. Natl Acad. Sci. USA 117, 7001–7003 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Akatsuka, Y. et al. Efficient cloning and expression of HLA class I cDNA in human B-lymphoblastoid cell lines. Tissue Antigens 59, 502–511 (2002).
Article CAS PubMed Google Scholar
Kobayashi, E. et al. A new cloning and expression system yields and validates TCRs from blood lymphocytes of patients with cancer within 10 days. Nat. Med. 19, 1542–1546 (2013).
Article CAS PubMed Google Scholar
Imai, M. et al. Syrian hamsters as a small animal model for SARS-CoV-2 infection and countermeasure development. Proc. Natl Acad. Sci. USA 117, 16587–16595 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Ueno, T. et al. Functionally impaired HIV-specific CD8 T cells show high affinity TCR-ligand interactions. J. Immunol. 173, 5451–5457 (2004).
Article CAS PubMed Google Scholar
Wolock, S. L., Lopez, R. & Klein, A. M. Scrublet: computational identification of cell doublets in single-cell transcriptomic data. Cell Syst. 8, 281–291.e289 (2019).
Article CAS PubMed PubMed Central Google Scholar
Le T. et al. BBrowser: making single-cell data easily accessible. bioRxiv, 2020.2012.2011.414136 (2020).
Cs·rdi G. b., & Nepusz T. s. The igraph software package for complex network research. (2006).
Wang, T., Li, B., Nelson, C. E. & Nabavi, S. Comparative analysis of differential gene expression analysis tools for single-cell RNA sequencing data. BMC Bioinformatics 20, 40 (2019).
Article PubMed PubMed Central Google Scholar
Cao, J. et al. The single-cell transcriptional landscape of mammalian organogenesis. Nature 566, 496–502 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

This work was funded by AMED under Grant Number JP20wm0125002, JP20fk0108454, JP20fk0108104, and JP21fk0108534 (J.G., M.Y., Y.K., J.I., and S.Y.), 2021 Hyogo College of Medicine Strategic Grant (H.O.) and partially by Osaka basic medical research promotion foundation (S.I.). We thank all participants for contributing the samples to this study; Dr. Masafumi Sakagami and Dr. Kazumasa Yoshinaga for introducing the collaborators working at Hyogo Medical University and Kyowa-kai Medical corporation to the study; Dr. Wataru Kamitani for sharing anti-SARS-CoV-1-M antibody; Dr. Masayuki Saijo for providing SARS-CoV-2 RNA; Dr. Yoshihiro Kawaoka for providing the SARS-CoV-2 isolate (UT-NCGM02/Human/2020/Tokyo); Dr. Daisuke Motooka, Dr. Daisuke Okuzaki, and Dr. Yu-Chen Liu for single-cell transcriptome analysis and data processing; Dr. Welkin Johnson for critical reading of the manuscript; Ms. Akie Teratani for the excellent technical supports.

Author information

Authors and Affiliations

Department of Microbiology, Hyogo Medical University, Hyogo, Japan
Hideki Ogura, Aoi Son & Satoshi Ishido
Research Center for Asian Infectious Diseases, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
Jin Gohda, Mizuki Yamamoto & Yasushi Kawaguchi
Laboratory of Molecular Immunology, Immunology Frontier Research Center, Osaka University, Suita, Japan
Xiuyuan Lu & Sho Yamasaki
Department of Infection Control and Prevention, Hyogo Medical University, Hyogo, Japan
Yoshio Takesue
Tokoname City Hospital, Aichi, Japan
Yoshio Takesue
Kawanishi City Hospital, Hyogo, Japan
Sadayuki Doi
Kyoritsu Hospital, Hyogo, Japan
Kazuyuki Matsushita
Kyowa Marina Hospital/Wellhouse Nishinomiya, Hyogo, Japan
Fumitaka Isobe
Dainikyoritsu Hospital, Hyogo, Japan
Yoshihiro Fukuda
Kyoritsu Onsen Hospital, Hyogo, Japan
Tai-Ping Huang
Joint Research Center for Human Retrovirus Infection, Kumamoto University, Kumamoto, Japan
Takamasa Ueno
Department of Emergency and Critical Care Medicine, Hyogo Medical University, Hyogo, Japan
Naomi Mambo, Hiromoto Murakami, Kunihiro Shirai & Jun-Ichi Hirata
Division of Molecular Virology, Department of Microbiology and Immunology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
Yasushi Kawaguchi
Research Platform Office, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan
Jun-ichiro Inoue
Department of Molecular Immunology, Research Institute for Microbial Diseases, Osaka University, Suita, Japan
Sho Yamasaki
Division of Molecular Design, Medical Institute of Bioregulation, Kyushu University, Fukuoka, Japan
Sho Yamasaki
Division of Molecular Immunology, Medical Mycology Research Center, Chiba University, Chiba, Japan
Sho Yamasaki

Authors

Hideki Ogura
View author publications
You can also search for this author in PubMed Google Scholar
Jin Gohda
View author publications
You can also search for this author in PubMed Google Scholar
Xiuyuan Lu
View author publications
You can also search for this author in PubMed Google Scholar
Mizuki Yamamoto
View author publications
You can also search for this author in PubMed Google Scholar
Yoshio Takesue
View author publications
You can also search for this author in PubMed Google Scholar
Aoi Son
View author publications
You can also search for this author in PubMed Google Scholar
Sadayuki Doi
View author publications
You can also search for this author in PubMed Google Scholar
Kazuyuki Matsushita
View author publications
You can also search for this author in PubMed Google Scholar
Fumitaka Isobe
View author publications
You can also search for this author in PubMed Google Scholar
Yoshihiro Fukuda
View author publications
You can also search for this author in PubMed Google Scholar
Tai-Ping Huang
View author publications
You can also search for this author in PubMed Google Scholar
Takamasa Ueno
View author publications
You can also search for this author in PubMed Google Scholar
Naomi Mambo
View author publications
You can also search for this author in PubMed Google Scholar
Hiromoto Murakami
View author publications
You can also search for this author in PubMed Google Scholar
Yasushi Kawaguchi
View author publications
You can also search for this author in PubMed Google Scholar
Jun-ichiro Inoue
View author publications
You can also search for this author in PubMed Google Scholar
Kunihiro Shirai
View author publications
You can also search for this author in PubMed Google Scholar
Sho Yamasaki
View author publications
You can also search for this author in PubMed Google Scholar
Jun-Ichi Hirata
View author publications
You can also search for this author in PubMed Google Scholar
Satoshi Ishido
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

S.I. and H.O. conceived research plans and wrote the manuscript. H.O. performed most parts of experiments including construction of library assay system with the help of A.S. J.G. and M.Y. performed experiments with SARS-CoV-2 and wrote the manuscript. Y.K. and J.I. organized and supervised the experiments with SARS-CoV-2. S.Y. organized and supervised single cell RNA-seq. H.O. and X.L. performed single cell RNA-seq. J.H., Y.T., N.M., H.M. and K.S. organized clinical study; patient enrollment, sample collection and summarizing clinical data at Hyogo Medical University. T.U. provided critical materials for experiments. J.H. and S.D. organized clinical study; patient enrollment and summarizing clinical data at Kyowa-kai Medical corporation. K.M., F.I., Y.F. and T-P.H. performed patient enrollment.

Corresponding authors

Correspondence to Hideki Ogura or Satoshi Ishido.

Ethics declarations

Competing interests

Hyogo Medical University has filed for patent protection for the identification of specific epitopes. The authors declare no other competing interests.

Peer review

Peer review information

Nature Communications thanks Antonio Bertoletti and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Ogura, H., Gohda, J., Lu, X. et al. Dysfunctional Sars-CoV-2-M protein-specific cytotoxic T lymphocytes in patients recovering from severe COVID-19. Nat Commun 13, 7063 (2022). https://doi.org/10.1038/s41467-022-34655-1

Download citation

Received: 04 July 2022
Accepted: 02 November 2022
Published: 16 December 2022
DOI: https://doi.org/10.1038/s41467-022-34655-1

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.