key: cord-0881856-9h7tdotx
authors: Kang, Chang Kyung; Kim, Minji; Lee, Soojin; Kim, Gwanghun; Choe, Pyoeng Gyun; Park, Wan Beom; Kim, Nam Joong; Lee, Chang-Han; Kim, Ik Soo; Jung, Keehoon; Lee, Dong-Sup; Shin, Hyun Mu; Kim, Hang-Rae; Oh, Myoung-don
title: Longitudinal Analysis of Human Memory T-Cell Response according to the Severity of Illness up to 8 Months after SARS-CoV-2 Infection
date: 2021-03-23
journal: J Infect Dis
DOI: 10.1093/infdis/jiab159
sha: 6d336317677f73bda537c22e43d95b2706f6fd8d
doc_id: 881856
cord_uid: 9h7tdotx

BACKGROUND: Understanding the memory T-cell response to severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) is crucial for assessing the longevity of protective immunity after SARS-CoV-2 infection or coronavirus disease-2019 (COVID-19) vaccination. However, the longitudinal memory T-cell response up to 8 months post-symptom onset (PSO) according to the severity of illness is unknown. METHODS: We analyzed peripheral blood mononuclear cells (PBMCs) from healthy volunteers or patients with COVID-19 who experienced asymptomatic, mild, or severe illness at 2, 5, and 8 months PSO. SARS-CoV-2 spike, nucleocapsid, and membrane protein-stimulated PBMCs were subjected to flow cytometry analysis RESULTS: A total of 24 patients—seven asymptomatic and nine with mild and eight with severe disease—as well as six healthy volunteers were analyzed. SARS-CoV-2-specific OX40 (+)CD137 (+) CD4 (+) T cells and CD69 (+)CD137 (+) CD8 (+) T cells persisted at 8 months PSO. Also, antigen-specific cytokine-producing or polyfunctional CD4 (+) T cells were maintained for up to 8 months PSO. Memory CD4 (+) T-cell responses tended to be greater in patients who had severe illness than in those with mild or asymptomatic disease. CONCLUSIONS: Memory response to SARS-CoV-2, based on the frequency and functionality, persists for 8 months PSO. Further investigations involving its longevity and protective effect from reinfection are warranted.

M a n u s c r i p t 4

Although 1 year has elapsed the first report of coronavirus disease-19 , caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the pandemic is ongoing [1, 2] . Recent reports on COVID-19 vaccines with high efficacy raise hope for pandemic control [3, 4] . However, evaluations of the duration of SARS-CoV-2-specific immune responses after SARS-CoV-2 infection or vaccination are needed [5, 6] .

The cell-mediated immune response is important in both the acute and convalescent phases of COVID-19. Delayed clinical deterioration in severe COVID-19 after an early peak in the viral load suggests the importance of the immune response in the progression of COVID-19 [7] , and hyperactivated and uncontracted T cells and their infiltration of vital organs have been implicated. [8, 9] In addition, the memory T-cell response after the acute phase of COVID-19 is crucial for protection against recurrence or progression of the disease [6, 10] .

Understanding the memory T-cell response to SARS-CoV-2 is critical to further examine the longevity of protective immunity after SARS-CoV-2 infection or vaccination [11] . The T-cell response at 2 months post-symptom onset (PSO) of COVID-19 has been reported [12, 13] , and was stronger in patients who experienced severe illness than in those who had mild disease. However, there are few studies of the memory T-cell response after the first 2 months PSO. Although Dan et al. reported that SARS-CoV-2-specific CD4 + or CD8 + T cells declined with a half-life of 3-5 months after analyzing samples within and beyond 6 months PSO [14] , most of the patients had mild COVID-19 and the evaluation was performed at a single time point. Therefore, the longitudinal memory T-cell response to SARS-CoV-2 up to 8 months PSO is unclear, especially depending on the severity of illness.

We examined the longitudinal memory T-cell response to SARS-CoV-2 at 2, 5, and 8 months PSO, stimulated by various SARS-CoV-2 antigens, in patients who experienced asymptomatic, mild, or severe illness. M a n u s c r i p t 5

We analyzed peripheral blood mononuclear cells (PBMCs) of patients with COVID-19, which had been collected at 2, 5, and 8 months (± 2, 4, and 4 weeks, respectively) after diagnosis (asymptomatic patients) or disease onset. All patients were laboratory-confirmed with reversetranscription polymerase chain reaction and were treated or monitored at Seoul National University Hospital or at a community treatment center in Daegu, Republic of Korea [15] . All asymptomatic patients were diagnosed during contact tracing in the midst of Daegu metropolitan city outbreak, March 2020 [16] . The Institutional Review Board of Seoul National University Hospital approved the study (No. H-2004 (No. H- -158-1118 , and all participants provided written informed consent in accordance with the Declaration of Helsinki.

The asymptomatic patients were defined as those with a body temperature of < 37.5℃ and without symptoms during their stay in the community treatment center despite undergoing a comprehensive medical interview twice daily [17] . Severe cases were defined as radiological pneumonia and an oxygen saturation of ≤ 93% in ambient air during their illness [18] . Others were classified as mild cases. Information on clinical characteristics-age, gender, the day of onset or diagnosis of COVID-19, maximal oxygen demand, and antiviral or anti-inflammatory drugs-was collected from the electronic medical records.

We analyzed PBMCs of SARS-CoV-2-seronegative healthy volunteers (healthy controls, HCs) during the pandemic (HC [2020]) who had neither been diagnosed of COVID-19 nor had received COVID-19 exposure notification [17] . Additionally, we analyzed PBMCs of those who had been 

PBMCs were purified from heparinized peripheral whole blood using a Ficoll-Histopaque gradient (1.077 g/mL; GE Healthcare Life Sciences, Piscataway, NJ). They were stored in liquid nitrogen until analysis in freezing medium comprising 50% fetal bovine serum, 10% dimethyl sulfoxide (DMSO), and 40% RPMI-1640 (all reagents from Thermo Fisher Scientific, Waltham, MA) [19] .

After thawing, the PBMCs (1 × 10 6 cells/mL) were stimulated with 2 μg/mL SARS-CoV-2 spike glycoprotein peptide pool, 2 μg/mL SARS-CoV-2 NCAP (nucleocapsid) peptide pool, or 2 μg/mL SARS- A c c e p t e d M a n u s c r i p t 7 Stained PBMCs were analyzed using an LSR II flow cytometer (BD Biosciences) and FACSDiva software with a minimum target event count of 500,000 cells. Data were analyzed using FlowJo software version 9.9.6 (TreeStar, Ashland, OR).

The frequencies of SARS-CoV-2-specific T cells (activation-induced markers, AIM + T cells; OX40 + CD137 + CD4 + T cells or CD69 + CD137 + CD8 + T cells) [20] or SARS-CoV-2-specific cytokine-producing cells (IFN-, TNF-, and IL-2) among CD137 + T cells were evaluated. T cells expressing two or more of IFN-, TNF-, and IL-2 were determined by sequential gating and were regarded as polyfunctional cells (Supplementary Figure S1 ) [21] .

The percentages of target populations in the unstimulated specimens (DMSO control) were subtracted from that in the antigen stimulated specimens to account for a nonspecific response [21] .

If there was no available unstimulated specimen at the same time point, the mean percentages of samples at other time points from the same patient were used. The responses to the three SARS-CoV-2 antigens were calculated by summing the final value of the response to each antigen [20] .

To compare the clinical characteristics of the asymptomatic, mild, and severe patients, the Kruskal-Wallis rank sum test or linear-by-linear association was performed. Data are expressed as means ± standard errors of the mean (SEMs) and as dot plots. When comparing the proportions of activated or cytokine-producing T cells between COVID-19 patients and HCs (2020), the Mann-Whitney U test with the Benjamini-Hochberg method for multiple comparisons was used. The Kruskal-Wallis ranksum test with Dunn's post hoc test for multiple comparisons was used to compare frequencies according to disease severity. P < 0.05 was considered indicative of statistical significance. All statistical analyses were twotailed and performed using PASW for Windows (version 25.0; IBM Corp., Armonk, NY) and GraphPad Prism 8 (GraphPad Software, La Jolla, CA). Graphs were generated using Prism 8. M a n u s c r i p t 8

A total of 24 patients-seven asymptomatic and nine with mild and eight with severe disease-were analyzed ( Table 1) . No patient had evidence of immunodeficiency or a history of re-exposure to COVID-19 or confirmed patients during the follow-up period. The median (range) ages of the asymptomatic, mild, and severe patients were 25 (20) (21) (22) (23) (24) (25) (26) (27) (28) , 48 (24-69), and 63 (39-76) years, respectively (P = 0.001). Regarding anti-inflammatory treatment, one mild and three severe patients received baricitinib [22] , and two severe patients received a therapeutic dose of steroid. The demographics, disease severity, treatment, details on availability of samples from each patient, and timing of sample collections are shown in Supplementary Table S1 .

The HC (2020) group comprised six healthy volunteers who donated their blood in September 2020. Their median (range) age was 35 (28-47) years, and five (83.3%) were male ( Table 1 ). The HC (MERS) group comprised seven blood samples from MERS survivors obtained in October 2019. Their median (range) age at the time of donation was 60 (38-64) years, and five (85.7%) were male.

The frequency of SARS-CoV-2-specific (OX40 + CD137 + ) CD4 + T cells ( Figure 1A ) in the patients with COVID-19, especially in those with severe disease, was higher at 8 months PSO than those in the HCs 

To assess the functional competence of SARS-CoV-2-specific (CD137 + ) memory CD4 + T cells, we measured cytokine production by CD4 + T cells responding to spike, nucleocapsid, and membrane proteins ( Figure 3A -C) from patients who were asymptomatic and from those with mild and severe disease. The levels of IFN-, TNF-, and IL-2 in memory CD4 + T cells at 2 and 5 months PSO in patients with COVID-19 tended to be higher than those in HCs (2020) ( Figure 3D -F and Supplementary Figure S1 ). The proportion of IL-2-producing memory CD4 + T cells responding to spike protein from patients with mild and severe disease was higher than that from HCs (2020) even IFN-, TNF-, and IL-2-production by Ag-specific memory CD4 + T cells in patients with severe disease was significantly increased compared to that of asymptomatic patients at 2 months PSO ( Figure 3D ). Similar to the proportion of Ag-specific memory CD4 + T cells (Figure 1 ), the functionality of Ag-specific memory CD4 + T cells declined over time, and the significance of the differences among patients with asymptomatic, mild, and severe disease decreased. However, the proportions of cytokine-producing Ag-specific memory CD8 + T cells were not significantly different according to disease severity ( Figure 4A-F) . Therefore, the functionality of memory CD4 + T cells responding to SARS-CoV-2 antigens was greatest in symptomatic patients. The similar trends are also observed in paired dot-plots or heatmaps of cytokine productions stimulated by spike protein (Supplementary Figure S4B and S5 ).

To evaluate further the functionality of memory CD4 + T cells responding to SARS-CoV-2 antigens, we examined the frequencies of polyfunctional T cells. [20] IFN- + TNF- + , IFN- + IL-2 + , TNF- + IL-2 + , or triple-positive cells among CD4 + T cells responding to spike protein were more dominant in patients with mild or severe disease up to 5 months PSO compared to the HCs (2020) ( Figure 5A-D) . The proportion of polyfunctional CD4 + T cells also tended to be higher in patients with severe disease than in those with mild disease or in asymptomatic patients ( Figure 5 ).

We analyzed longitudinal memory T-cell responses up to 8 months PSO, in terms of frequency and functionality, to SARS-CoV-2 antigens in patients with COVID-19 according to disease severity. SARS-CoV-2-specific memory CD4 + or CD8 + T cells slowly decline up to 8 months PSO. Memory T-cell responses tended to be stronger in symptomatic than in asymptomatic patients, especially in those A c c e p t e d M a n u s c r i p t 11 with severe disease. The spike, nucleocapsid, and membrane proteins stimulated similar memory Tcell response patterns.

After the early reports on efficacy of COVID-19 vaccines [3, 4] , several countries have initiated national vaccination programs. However, the correlation between protection against COVID-19 and the longevity of the immunity induced by vaccination is unclear [5] . Our findings on the memory Tcell response of patients with COVID-19 of differing severities could be used as reference data for studies of the cellular immunogenicity of COVID-19 vaccines. [23, 24] . However, the degree and longevity of the memory response to SARS-CoV-2 according to disease severity was unknown. Our results show that a memory T-cell response persists up to 8 months PSO, particularly in patients with severe COVID-19. Further long-term and larger studies are warranted to characterize the magnitude and duration of the protective effect.

Although we could not explore the relationship between the numbers of AIM + T cells and the magnitude of antibody response [25] , decreased, but persistent cellular response to COVID-19 up to 8 months PSO were similar to the humoral response [14, 26] . Since Dan et al reported that circulating follicular helper (Tfh) memory CD4 + T cells which enhance B-cell function were maintained until 8 months PSO [14] , similar kinetics might be mediated by circulating Tfh memory CD4 + T cells. In addition, increased memory T-cell response with disease severity was also similar to the humoral response [27] . Such a severity-dependent response may be attributed to the delayed but strong type I IFN response in the acute phase of severe COVID-19 [28] , because the type I IFN response contributes to the memory formation in response to viral infection [29] . Similar responses were observed when PBMCs were stimulated with SARS-CoV-2 nucleocapsid and membrane proteins, in agreement with previous reports [12, 20] . Pre-existing SARS-CoV-2-reactive T cells might have been induced by seasonal coronaviruses [13, 20] . One could accurately examine the degree of COVID-19-specific T-cell responses if they had pre-COVID-19 PBMCs, which is impractical. To minimize this concern, we subtracted the frequencies in unstimulated samples to determine the SARS-CoV-2-specific response. In addition, the responses in terms of all three cytokines in CD4 + T cells were robust at 2 months but decreased over time.

These kinetics imply that the responses measured in this study were COVID-19-specific.

We analyzed two control groups to compensate for confounding by seasonal coronavirusand/or MERS-CoV-reactive T cells. Interestingly, the frequency of SARS-CoV-2 spike-specific CD4 + T cells in HCs (2020), which represents HC during the COVID-19 pandemic, was considerably higher than that of HCs (MERS) ( Figure 1B) . Cytokine production by Ag-specific memory CD4 + T cells from HCs (2020) was lower than that of HCs (MERS) and similar to that of asymptomatic patients ( Figure   5 ). The mechanism underlying this frequency-functionality discordance is unclear. The composition of a truly HC group in the COVID-19 era necessitates further research.

The memory response was less prominent in CD8 + T cells than in CD4 + T cells, as reported previously [30] . However, the possibility of suboptimal stimulation of CD8 + T cells by the 15-mer peptides pool could not be excluded because major histocompatibility complex (MHC) class Ⅰ has a shorter binding groove (typically 8-10 residues) than MHC class II [31] . Moreover, Tarke et al.

recently reported that epitope pools could be helpful to optimize detection of T cell responses because of HLA binding-related immunodominance. Therefore, further evaluation using either a shorter peptide or epitope megapool is warranted [32] .

This study had several limitations. First, since we could analyze a small number of samples, the results of statistical analyses should be interpreted with caution. Similarly, we could not draw a meaningful severity-specific decay rate of SARS-CoV-2-specific T cells in this study. Second, the age distribution differed among the severity groups. Further validation using a larger, age-matched A c c e p t e d M a n u s c r i p t 13 cohort is therefore needed. Third, baricitinib or steroid treatment in severe group could have affected memory response [33] . Lastly, we could not account initial viral load in the present study.

Although inoculum size of SARS-CoV-2 might affect severity of COVID-19 [34] , viral load itself could influence establishment or longevity of memory T-cell response.

In conclusion, a memory T-cell response, in terms of frequency and functionality, persisted up A c c e p t e d M a n u s c r i p t 26 Figure 5 

A Novel Coronavirus from Patients with Pneumonia in China

Novel Coronavirus (2019-nCoV) situation reports

Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

A New Vaccine to Battle Covid-19

Correlates of protection against SARS-CoV-2 in rhesus macaques

COVID-19: consider cytokine storm syndromes and immunosuppression

Aberrant hyperactivation of cytotoxic T-cell as a potential determinant of COVID-19 severity

Pathological findings of COVID-19 associated with acute respiratory distress syndrome

T cells in COVID-19 -united in diversity

Immunological considerations for COVID-19 vaccine strategies

Broad and strong memory CD4(+) and CD8(+) T cells induced by SARS-CoV-2 in UK convalescent individuals following COVID-19

Robust T Cell Immunity in Convalescent Individuals with Asymptomatic or Mild COVID-19

Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection

Operating Protocols of a Community Treatment Center for Isolation of Patients with Coronavirus Disease

Cohort Patients in Daegu Metropolitan City Outbreak in 2020

Antibody Responses to SARS-CoV-2 at 8 Weeks Postinfection in Asymptomatic Patients

Characteristics of and Important Lessons From the

COVID-19) Outbreak in China: Summary of a Report of 72314 Cases From the Chinese Center for Disease Control and Prevention

Altered IL-7Ralpha expression with aging and the potential implications of IL-7 therapy on CD8+ T-cell immune responses

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

Cell-Mediated Immunogenicity of Influenza Vaccination in Patients With Cancer Receiving Immune Checkpoint Inhibitors

Baricitinib plus Remdesivir for Hospitalized Adults with Covid-19

Memory T cell responses targeting the SARS coronavirus persist up to 11 years post-infection

Recovery from the Middle East respiratory syndrome is associated with antibody and T-cell responses

Comprehensive analysis of T cell immunodominance and immunoprevalence of SARS-CoV-2 epitopes in COVID-19 cases

Antibody Responses 8 Months after Asymptomatic or Mild SARS-CoV-2 Infection

Waning Antibody Responses in Asymptomatic and Symptomatic SARS-CoV-2 Infection

Understanding SARS-CoV-2-related multisystem inflammatory syndrome in children

Type I interferons act directly on CD8 T cells to allow clonal expansion and memory formation in response to viral infection

Suboptimal SARS-CoV-2-specific CD8(+) T cell response associated with the prominent HLA-A*02:01 phenotype

Major Histocompatibility Complex (MHC) Class I and MHC Class II Proteins: Conformational Plasticity in Antigen Presentation

PD-1-Expressing SARS-CoV-2-Specific

Cells Are Not Exhausted, but Functional in Patients with COVID-19

Selective inhibition of low-affinity memory CD8(+) T cells by corticosteroids

Inoculum at the time of SARS-CoV-2 exposure and risk of disease severity

A c c e p t e d M a n u s c r i p t 16