Persistence of Antibody and Cellular Immune Responses in Coronavirus Disease 2019 Patients Over Nine Months After Infection

Abstract

Background

The duration of humoral and T and B cell response after the infection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) remains unclear.

Methods

We performed a cross-sectional study to assess the virus-specific antibody and memory T and B cell responses in coronavirus disease 2019 (COVID-19) patients up to 343 days after infection. Neutralizing antibodies and antibodies against the receptor-binding domain, spike, and nucleoprotein of SARS-CoV-2 were measured. Virus-specific memory T and B cell responses were analyzed.

Results

We enrolled 59 patients with COVID-19, including 38 moderate, 16 mild, and 5 asymptomatic patients; 31 (52.5%) were men and 28 (47.5%) were women. The median age was 41 years (interquartile range, 30–55). The median day from symptom onset to enrollment was 317 days (range 257 to 343 days). We found that approximately 90% of patients still have detectable immunoglobulin (Ig)G antibodies against spike and nucleocapsid proteins and neutralizing antibodies against pseudovirus, whereas ~60% of patients had detectable IgG antibodies against receptor-binding domain and surrogate virus-neutralizing antibodies. The SARS-CoV-2-specific IgG⁺ memory B cell and interferon-γ-secreting T cell responses were detectable in more than 70% of patients.

Conclusions

Severe acute respiratory syndrome coronavirus 2-specific immune memory response persists in most patients approximately 1 year after infection, which provides a promising sign for prevention from reinfection and vaccination strategy.

memory B cell, memory T cell, neutralizing antibody, persistence, SARS-CoV-2

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the coronavirus disease 2019 (COVID-19) agent, has caused the pandemic worldwide [1–3]. As of February 2, 2021, more than 100 million confirmed cases of SARS-CoV-2 had been reported, with more than 2 million deaths [4]. Effective vaccines are vital for controlling the pandemic. Therefore, understanding the long-term immunological memory response to SARS-CoV-2 after nature infection is critical for developing and implementing a SARS-CoV-2 vaccine. Recent studies showed that most patients persist in virus-specific antibody response 6–8 months after infection but displayed a waning trend of humoral immunity over time [5–12]. Severe acute respiratory syndrome coronavirus 2-specific memory CD8⁺ and CD4⁺ T cells were detected in most patients but declined with a half-life of 3–5 months [13], whereas virus-specific memory B cells (MBCs) continued to increase or were unchanged 5–6 months after the infection [7, 8, 13, 14]. However, the antibody and memory T and B cell response beyond 8 months or approximately 1 year after the infection is unclear. Therefore, in this study, we conducted a cross-sectional study of 59 patients up to 343 days after infection and assessed the virus-specific antibody and memory T cells and B cells.

MATERIALS AND METHODS

Study Design and Participants

Between December 7 and 30, 2020, patients who recovered from the COVID-19 with SARS-CoV-2 infection in Jiangsu, Shandong and Zhejiang Provinces, China were invited to participate in this study. All of the patients were laboratory-confirmed positive for SARS-CoV-2 by real-time reverse-transcription polymerase chain reaction (RT-PCR) results. Each enrolled patient provided a 3-mL blood sample for serum isolation and an additional 5-mL blood sample for peripheral blood mononuclear cell (PBMC) isolation. Demographics and clinical characteristics of patients were collected upon enrollment. Thirty age- and sex-matched healthy controls (HCs) enrolled before the pandemic were used as control. Each patient signed informed consent. The study protocol was approved by the Institutional Review Board (IRB) of the Academy of Military Medical Sciences (IRB number AF/SC-08/02.60).

According to the diagnostic and treatment guidelines for SARS-CoV-2 issued by the Chinese National Health Committee (Trail Version 8), the disease severity was defined as asymptomatic, mild, and moderate. Asymptomatic infection was defined as an individual who had a positive SARS-CoV-2 by RT-PCR without any associated clinical symptoms in the preceding 14 days and during hospitalization for observation as part of the control measures. Mild infection was defined as an individual who had mild clinical symptoms without radiological signs of pneumonia. Moderate was defined according to the following criteria: (1) fever and respiratory symptoms and (2) radiological signs of pneumonia.

Serum and Peripheral Blood Mononuclear Cell Isolation

Sera were separated by centrifugation at 2000 rpm for 10 minutes, aliquoted into 3 cryovials, and preserved at −80°C until testing. PBMCs were isolated by density gradient centrifugation with Lymphoprep in SepMate tubes (Stemcell Technologies) according to the manufacturer’s instruction. In brief, the blood was placed on top of Lymphoprep in SepMate tubes and centrifuged at 1200 ×g for 10 minutes. The PBMCs from the top layer were harvested and washed twice with phosphate-buffered saline (PBS) at 400 ×g for 10 minutes. Isolated PBMCs were frozen in cell recovery Media containing 10% dimethyl sulfoxide ([DMSO] GIBCO), supplemented with 90% heat-inactivated fetal bovine serum, and stored in liquid nitrogen before assay analyses.

Enzyme-Linked Immunosorbent Assay Analysis of Immunoglobulin G Antibody to Receptor-Binding Domain and Spike Trimer of Severe Acute Respiratory Syndrome Coronavirus 2

The recombinant receptor-binding domain (RBD) and spike (S) trimer derived from SARS-CoV-2 (Sino Biological, Beijing, China) were coated onto flat-bottom, 96-well, enzyme-linked immunosorbent assay (ELISA) plates overnight at 4°C with 0.1 μg/well. Plates were washed with PBS with 0.05% Tween 20 (PBS-T) and blocked with blocking buffer (5% skim milk and 2% bovine serum albumin in PBS) for 1 hour at room temperature. Duplicate 3-fold 8-point serial dilutions (starting at 1:100) of heat-inactivated serum samples diluted in 1% milk in PBS-T were added to the wells and incubated at 37°C for 1 hour. Wells were then incubated with the horseradish peroxidase (HRP)-labeled antihuman immunoglobulin (Ig)G antibody (1:5000; no. W4031, Promega) and TMB substrate (Kinghawk, Beijing, China). The optical density (OD) was measured by a spectrophotometer at 450 nm and 630 nm. Endpoint antibody titers were calculated by a fitted curve (4 parameter log regression), and 3 times the average value of HCs was used as the detection threshold.

Enzyme-Linked Immunosorbent Assay Analysis of Immunoglobulin G Antibody to the Nucleocapsid Protein of Severe Acute Respiratory Syndrome Coronavirus 2

Serum IgG to nucleocapsid (N) protein of SARS-CoV-2 was semiquantitatively measured by ELISA using a well validated commercial diagnostic ELISA kit (Beijing Wantai Biological Pharmacy Enterprise Co., Ltd) [15] according to manufacturer’s instruction. The anti-N IgG antibody was detected using an indirect ELISA kit based on a recombinant N protein of SARS-CoV-2. The cutoff value for IgG is the mean OD value of 3 negative controls (if the mean absorbance value for 3 negative calibrators is <0.03, take it as 0.03 following manufacturer’s instructions.) + 0.16. A serum sample with an OD value ≥cutoff OD value was considered anti-N IgG antibody positive.

Surrogate Virus Neutralization Test

Surrogate virus neutralization test (sVNT) assays were performed by using a Surrogate Virus Neutralization Test Kit (no. L00847, GenScript) according to the manufacturer’s instructions. In brief, we mixed the positive control, negative control, or diluted samples (1:10) with diluted HRP-RBD with a volume ratio of 1:1 in tubes and incubated at 37°C for 30 minutes. Then, 100 μL mixture of the sample was added to the 96-well microplates precoated with recombinant angiotensin-converting enzyme 2 (ACE2) protein and incubated at 37°C for 15 minutes. After washing the plate 4 times, 100 μL TMB solution was added to each well and incubated at room temperature for 15 minutes. Then we add 50 μL stop solution to the plate. The OD was measured by a spectrophotometer at 450 nm and 570 nm. The HRP-RBD alone and plasma with no HRP-RBD incubation were used as controls. The percentage inhibition was calculated as (1–sample od value/average negative control OD value) × 100. The inhibition rate above 30% was considered positive for the SARS-CoV-2 neutralizing antibody (NAb).

Pseudovirus Neutralization Test

The SARS-CoV-2 pseudovirus neutralization test (pVNT) was performed as described previously [4]. In brief, Huh7 cells were seeded in 96-well plates (200 000 cells/well) and incubated for approximately 24 hours until 90%–100% confluent. Serial 3-fold diluted serum, starting at 1:30, were incubated with 650 TCID₅₀ (50% tissue culture infective dose) of the pseudovirus for 1 hour at 37°C. Dulbecco’s modified Eagle’s medium was used as the negative control. The supernatant was then removed, and luciferase substrate was added to each well, followed by incubation for 2 minutes in darkness at room temperature. Luciferase activity was then measured using GloMax 96 Microplate Luminometer (Promega). Half-maximal inhibitory concentrations of the serum samples were determined by luciferase activity 48 hours after exposure to the virus-serum mixture with a 3-parameter nonlinear regression inhibitor curve in GraphPad Prism 8.4.1 (GraphPad Software). Titers were determined as the serum dilution that inhibited 50% virus infection (ID₅₀).

Enzyme-Linked Immunospot Assays

To assess B cells secreting IgG antibodies specific for SARS-CoV-2 RBD and cells secreting IgG (total IgG), we performed an enzyme-linked immunospot (ELISpot) assay using the Human IgG SARS-CoV-2 RBD ELISpot (HRP) kit (3850-4HPW-R1-1, Mabtech AB) according to the manufacturer’s protocol. In brief, PBMCs were incubated for 4 days in Roswell Park Memorial Institute (RPMI) 1640 medium with 10% fetal calf serum (FCS), supplemented with R848 (1 μg/mL; Mabtech AB) and recombinant human interleukin-2 (10 ng/mL) for stimulation of memory B cells. The ELISpot plates precoated with capturing monoclonal anti-human IgG antibodies were incubated with a total of 200 000 or 40 000 prestimulated cells per well for detection of RBD-specific IgG and total IgG secreting cells, respectively.

T cell responses were measured using Human IFN-γ SARS-CoV-2 ELISpot kit (ALP, 3420-4AST-P1-1, Mabtech AB) according to the manufacture’s protocol. In brief, plates were washed with filtered PBS (Sigma-Aldrich, St. Louis, MO) and blocked with RPMI 1640 culture media containing 10% batch tested fetal bovine serum (Gibco, Thermo Fisher Scientific, Waltham, MA, USA). The plates precoated with capturing monoclonal anti-IFN-γ were incubated for 18 hours in RPMI 1640 medium containing 10% FCS supplemented with a mixture containing the SARS-CoV-2-defined peptide pool. The peptide pool contained 47 synthetic peptides binding to human human leukocyte antigen, derived from the S, N, membrane (M), and the open reading frame (ORF)-3a and ORF-7a proteins (no. 3622-1, Mabtech AB) at a concentration of 2 μg/mL of each peptide, anti-CD28 (0.1 μg/mL), and 250 000 cells per well in a humidified incubator (5% CO₂, 37°C). Negative controls comprising DMSO and positive controls containing anti-CD3 were also included.

Spot numbers were analyzed by the CTL ImmunoSpot S6 universal analyzer (Cellular Technology Ltd). To determine the SARS-CoV-2-specific spots mean spots of the control wells were subtracted from the positive wells, and the results were expressed as spot-forming cells (SFCs) per 10⁶ PBMCs. We defined 3-fold higher SARS-CoV-2-specific spots versus background together with at least 3 spots above background as a positive response. This cutoff was set based on negative control values as described previously. If negative control wells had >30 SFC per 10⁶ PBMCs or positive control wells (anti-CD3 and CD28 stimulation) were negative, the results were excluded from further analysis.

Statistical Analysis

One-way analysis of variance with Least Significant Difference (LSD) post hoc testing (normal distribution) or Kruskal-Wallis test with false discovery rate method (nonnormal distribution) was used for multiple group comparisons. The Mann-Whitney U test was used to compare the difference between the 2 groups. Spearman correlations analyses were used to determine associations between analyzed parameters. All statistical analyses were performed using GraphPad Prism (version 8.4.2; GraphPad Software, La Jolla, CA), and all statistical tests were 2-sided with a significance level of 0.05.

RESULTS

Study Subjects

We enrolled 59 patients who recovered from COVID-19 with a time interval of 317 days (range 257 to 343 days, defined 9–11 months after infection) from symptom onset to sampling. Of the 59 patients, 38 were moderate, 16 were mild, and 5 were asymptomatic (Table 1). The median age was 41 (interquartile range [IQR], 30–55), and 31 (52.5%) were male. Fever (66.7%), cough (44.4%), and expectoration (16.7%) are the most reported common symptoms (Table 1). No significant differences in sex and age were observed between HCs and patients.

Table 1.

Demographic and Clinical Characteristics of the COVID-19 Patients Enrolled in This Study

Characteristics	Healthy Controls	Asymptomatic	Mild	Moderate	P Value
No. of participants	30	5	16	38
Onset to inclusion (days; median, IQR)	NA	303.0 (264.0–308.0)	315.0 (309.8–320.5)	319.5 (313.0–330.8)
Age (median, IQR)	44.5 (38.5–52.3)	43.0 (20.0–49.5)	34.0 (18.5–47.8)	43.0 (34.8–59.8)	.135
Age Group (years)
≤60	25 (83.3)	5 (100.0)	13 (81.3)	29 (76.3)	.764
>60	5 (16.7)	0	3 (18.8)	9 (23.7)
Sex
Male	15 (50.0)	0	8 (50.0)	23 (60.5)	.085
Female	15 (50.0)	5 (100.0)	8 (50.0)	15 (39.5)
Underlying Medical Condition
Yes	0	1 (20.0)	4 (25.0)	15 (39.5)	.566a
No	30 (100.0)	4 (80.0)	12 (75.0)	23 (60.5)
Signs and Symptoms
Fever	NA	No	10 (62.5)	26 (68.4)
Dry cough	NA	No	10 (62.5)	14 (36.8)
Expectoration	NA	No	3 (18.75)	6 (15.8)
Pharyngalgia	NA	No	2 (12.5)	3 (7.9)
Itchy throat	NA	No	0	2 (5.3)
Sneeze	NA	No	1 (6.25)	1 (2.6)
Stuffy nose	NA	No	1 (6.25)	2 (5.3)
Rhinorrhea	NA	No	0	4 (10.5)
Dyspnea	NA	No	0	1 (2.6)
Chest stuffiness	NA	No	0	1 (2.6)
Fatigue	NA	No	0	4 (10.5)
Headache	NA	No	1 (6.25)	4 (10.5)
Inappetence	NA	No	0	1 (2.6)
Nausea	NA	No	0	1 (2.6)
Diarrhea	NA	No	0	4 (10.5)
Chills	NA	No	0	3 (7.9)
Myalgia	NA	No	0	3 (7.9)

Abbreviations: COVID-19, coronavirus disease 2019; IQR, interquartile range; NA, not applicable.

^aP value was calculated between patient groups.

Persistent of Severe Acute Respiratory Syndrome Coronavirus 2-Specific Immunoglobulin G and Neutralizing Antibodies

We first assessed the SARS-CoV-2 RBD binding IgG antibody using ELISA and found that 55.9% (31 of 59) of patients had detectable anti-RBD IgG antibodies approximately 1 year after infection. Regarding the disease severity, 66.7% (14 of 21) of asymptomatic/mild and 50.0% (19 of 38) of moderate patients tested positive for the anti-IgG antibody, with a geometric mean endpoint titer of 53.2 (95% confidence interval [CI], 19.0–148.9) and 17.3 (95% CI, 7.9–33.7) (Figure 1A). We observed that 91.5% of patients had low binding anti-RBD antibody titer (<500) (Figure 1A). In contrast, 93.2% (55 of 59) of patients (100% for asymptomatic/mild and 89.5% for moderate patients) showed a positive detection of anti-S IgG (Figure 1B), with a geometric mean endpoint titer of 879 (95% CI, 545.8–1415.8) and 543.3 (95% CI, 319.9–922.6) (Figure 1B). Most (67.8%) patients showed a moderate and strong binding anti-IgG titer (>500). Although our focus was on the S protein, we also semiquantitatively measured the antibody response to the N protein of SARS-CoV-2 because this is the antigen targeted by multiple commercial assays. We observed that anti-N IgG antibodies are persistently high in recovered patients, and 98.3% (58 of 59) of patients (100% for asymptomatic/mild and 97.4% for moderate patients) were still positive (Figure 1C). As expected, in the serum of HCs, we observed a minimal reactivity of anti-S or anti-RBD IgG antibodies (Figure 1A and B), and one serum was positive for anti-IgG antibodies (Figure 1C).

Figure 1.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-specific antibody response in recovered patients 9–11 months after infection. (A–E, left) Immunoglobulin (Ig)G antibodies against the receptor-binding domain (RBD), spike, and nucleocapsid of SARS-CoV-2 and neutralizing antibodies (NAbs) of RBD-angiotensin-converting enzyme 2 (ACE2) binding inhibition and pseudovirus in serum samples collected from patients and healthy controls. (Right) Ranked anti-RBD, anti-spike, and anti-nucleocapsid IgG titers as well as RBD-ACE2 binding inhibition rate and pseudovirus NAb titer of each coronavirus disease 2019 (COVID-19) patients. Each dot represents a titer, optical density (OD) value, or inhibition rate for each serum sample. The black dashed line indicates the threshold for positivity (anti-RBD IgG = 29.9, anti-S IgG = 66.7, anti-N IgG = 0.19, inhibition rate = 30%, and NAb = 30). Boxplots indicate median and interquartile range (IQR); and the whiskers represent 1.5 times the IQR. Kruskal-Wallis test and Dunn’s multiple comparison test using false discovery rate was used for the comparisons. ***, P < .001; ****, P < .0001.

To determine whether patients maintain neutralizing antibodies, we detected SARS-CoV-2 neutralization indirectly using a cell-free assay of RBD-ACE2-binding inhibition by sVNT and directly pVNT. The sVNT assays showed that 66.1% (39 of 59) of patients (71.4% for asymptomatic/mild and 63.2% for moderate patients) had antibody to inhibit RBD binding to ACE2 (Figure 1D), and 44% (26 of 59) patients displayed ≥50% inhibit rate to ACE2 (Figure 1D). Further pVNT assay revealed that 89.8% (53 of 59) of patients (21 of 21 asymptomatic/mild and 32 of 38 moderate patients) had detectable NAb (Figure 1E). The geometric mean titer of NAb was significantly higher in asymptomatic/mild (88.0; 95% CI, 66.0–117.3) and moderate patients (56.4; 95% CI, 42.7–74.6) than HCs (Figure 1E). We found that IgG antibodies were strongly correlated with each other and neutralizing inhibition rate, whereas they were moderately correlated with pesudovirus-based NAb titer (Supplementary Figure 1).

Although no significant differences were observed for antibody response between asymptomatic/mild and moderate patients, moderate patients tended to have a slightly lower positive antibody response rate. Most moderate patients (median day of 320; IQR, 313–331) had a longer time interval from symptom onset to sampling than asymptomatic/mild patients (median day of 312; IQR, 306–318; P = .0025). Further analysis showed that the age, sex, and underlying medical conditions were not associated with antibody responses except for a higher anti-S IgG titer (Supplementary Figure 2).

Maintenance of Severe Acute Respiratory Syndrome Coronavirus 2-Specific B and T Cells

We further assessed virus-specific memory B and T cell responses among these patients using ELISpot assay (Figure 2A and D). We found that 74.6% (44 of 59) of patients (76.2% and 73.2% for asymptomatic/mild and moderate patients, respectively) had detectable RBD-specific IgG⁺ memory B cells, with a mean number of 132 (95% CI, 71–193) (Figure 2C). Similar to memory B cell response, 73.7% (28 of 38) of moderate patients and 66.7% (14 of 21) of asymptomatic/mild patients had detectable IFN-γ-secreting T cells, with a mean number of 368 (95% CI, 269–467) IFN-γ-secreting T cells and significantly higher than in HCs (Figure 2D and E). In addition, 2 of 20 (6.7%) (2 of 30) HCs had detectable IFN-γ-secreting T cells (Figure 2E). There was correlation between SARS-CoV-2-specific MBCs response and anti-RBD IgG (r = 0.350, P = .007), anti-S IgG (r = 0.330, P = .010), and NAb-inhibition rate (r = 0.456, P < .001), but not anti-N IgG (Figure 2F). However, no significant correlations were observed between the SFC of SARS-CoV-2 specific T cells secreting IFN-γ and antibody response (Supplementary Figure 3). Further analysis showed that patients older than 60 years old had fewer memory B cells secreting total IgG and RBD-specific IgG than patients under 60 years old (Supplementary Figure 4).

Figure 2.

B and T cell responses to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in coronavirus disease 2019 (COVID-19) convalescent patients. (A) A representative enzyme-linked immunospot (ELISpot) of total IgG⁺ and SARS-CoV receptor-binding domain (RBD)-specific IgG⁺ memory B cells from each type of subjects. (B) Total IgG⁺ memory B cell counting in patients and healthy controls. (C) RBD-specific IgG⁺ memory B cell counting in patients and healthy controls. (D) A representative ELISpot of IFN-γ producing T cells against a defined peptide pool of spike (S)/nucleoprotein (N)/membrane protein (M) and open reading frame (ORF)3a/7a from each type of subject, with blank as negative control and anti-CD3 as positive controls. (E) Summary data of patients studied according to S/N/M and ORF3a/7a in patients and healthy controls. (F) Correlations between spot-forming cells (SFCs) of memory B cells and antibody responses. The dotted line indicates the cutoff for positive responses (RBD-specific IgG secreting memory B cells = 3, and IFN-γ-secreting T cells = 75). Boxplots indicate median and interquartile range (IQR); and the whiskers represent 1.5 times the IQR. Data in the graph (B, C, and E) are represented as SFCs per 1 × 10⁶ peripheral blood mononuclear cells (PBMCs), and each point on the dot plot represents an individual subject. Kruskal-Wallis test and Dunn’s multiple comparison test using false discovery rate was used for the comparisons in B, C, and E. ***, P < .001; ****P < .0001. Spearman correlations were used in (F).

DISCUSSION

Recent studies have shown that most patients had detectable SARS-CoV-2 antibody responses 6–8 months after infection [6, 12, 16–19]. This study evaluated SARS-CoV-2-specific antibody and cellular immune responses in patients up to 11 months after infection. Our data reveal that approximately 70% of COVID-19 patients maintain anti-RBD and anti-S IgG, ACE2-RBD inhibition rate, and NAb at least 9–11 months after infection. Moreover, we observed that pseudovirus neutralizing antibodies and antibodies against the N and S proteins are longer-lived than those against RBD. Although antibodies to RBD of SARS-CoV-2 S protein accounted for the majority of IgG responses, antibodies to other epitopes such as the N-terminal domain and S2 subunit of S protein also contribute to the IgG response [20]. A previous study also showed that the combination of antibodies to other epitopes of S protein and antibodies against RBD could more effectively block the virus from invading host cells [21]. Therefore, the antibody titer and positivity rates of anti-RBD IgG and ACE2-RBD inhibition might be lower than the NAb and anti-S IgG response. We also observed relatively high (~98%) positive detection of anti-N IgG among patients compared with other antibodies. In human SARS-CoV-1 infection, antibodies against the SARS-CoV-1 N protein are abundant and longer-lived than other viral components such as the S, M, and envelope proteins [22]. In another study, antibodies against S1 or RBD persisted longer than antibodies against N protein in the sera of SARS survivors 17 years after infection [23]. Although this phenomenon’s biological significance and mechanistic characterization are beyond this study’s scope, further investigation is warranted.

Whether primary infection with SARS-CoV-2 protects individuals from reinfection and how long protection lasts has yet to be established. Patients with SARS-CoV-2 reinfection are still rare but on the rise [24]. Reinfections may imply that immunity against SARS-CoV-2 may be weak and decay relatively quickly, with implications not just for the risks facing recovered patients but also for how long future vaccines might protect people [25]. Notably, several cases with SARS-CoV-2 reinfection displayed a low or without producing antibody response after prior infection [26–30], which may hamper a more effective response to the second time around. In contrast, recent studies have shown that the presence of antibodies to SARS-CoV-2 was associated with a significantly reduced risk of SARS-CoV-2 reinfection for 6 to 7 months after prior infection [31, 32]. Moreover, prior infection also reduces the risk of asymptomatic infection and likely to be protective against severe disease or symptomatic infection [31, 32]. These data provide clues on the protection of humoral immune response against reinfection. Collectively, sustained humoral immunity after infection might help apply vaccines for reliable prevention of SARS-CoV-2 transmission, and patients who became seronegative for antibody response may be at higher risk of reinfection when they meet the next exposure compared with seropositive individuals.

Although most patients had detectable virus-specific antibodies 9–11 months after infection, the longevity of virus-specific memory T cells and MBCs is still unresolved. Previous studies have shown that memory CD4⁺ and CD8⁺ T cells can be detected in 70%–100% early convalescent COVID-19 patients [33–35]. In addition, broad and strong memory CD4⁺ and CD8⁺ T cells can be found in >90% of convalescent COVID-19 patients [36, 37]. Recent studies also showed that memory CD4⁺ and CD8⁺ T cell response could persist in 50%–100% of patients at least 3–6 months postsymptom onset [13, 14, 38, 39]. Regarding the virus-specific MBCs, previous studies have shown that virus-specific MBCs were detected in >90% of patients and increased over time even 5 months postsymptom onset [8, 14]. Moreover, more abundant S-specific MBCs were detected in approximately 80% of patients at 6 months compared with 1 month after infection [13], and S-specific and RBD-specific MBCs can even bedetected in 78% and 100% of patients, respectively, for up to 8 months postsymptom onset [40, 41]. Consistent with the above evidence, we found that more than 70% of patients had detectable virus-specific T cells producing IFN-γ as well as RBD-specific IgG⁺ MBCs response 9–11 months after infection. Our data indicate that infection of SARS-CoV-2 could produce long-lasting T and B cell memory in most patients up to approximately 1 year after infection, which is potentially beneficial for protecting against systemic disease upon reinfection.

There are several limitations of this study. First, it is a cross-sectional study and limited the observation of the dynamic changes of antibody or immune cells over time, which limits generalization. Second, it has a relatively small number of patients and a paucity of severe patients. However, previous studies have shown that severe patients had a higher antibody titer than other disease conditions [42, 43].

CONCLUSIONS

In summary, detectable humoral and cellular immunity in most of the patients 9–11 months after infection with SARS-CoV-2 offers insights into the long-term immune response to SARS-CoV-2 infection. Our findings will provide direct implications for COVID-19 vaccine development and implementation and other public health responses to the COVID-19 pandemic.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. We thank all of the patients for their participation in this study.

Disclaimer. The views expressed in this article are those of the authors and do not necessarily represent the official position of Wuxi, Dezhou, and Quzhou Municipal Centers for Diseases Control and Prevention. All the authors have declared no relationships or activities that could appear to have influenced this work.

Author contributions. M.-J. M. conceived the study. Y. S., Z.-Y. W., B.-D. Z., B. L., C. S., H.-H. P., Y.-M. G., G.-Q. W., D.-M. W., M.-D. J., and G.-P. C. collected patients’ samples and clinical data. G.-L. W., L. Y., L.-J. D., and H.-H. P. performed peripheral blood mononuclear cell isolation. L. Y. and L.-J. D. performed immunoglobulin G experiments; G.-L. W., L. Y., and L.-J. D. performed enzyme-linked immunospot assays. G.-L. W., L. Y., and L.-J. D. performed surrogate virus neutralization and pseudovirus neutralization tests. M.-J. M., G.-L. W., L. Y., and L.-J. D. analyzed the data. M.-J. M., L. Y., and G.-L. W. drafted the manuscript. All authors reviewed and approved the final manuscript.

Financial support. This work was funded by grants from the Beijing Natural Science Foundation (L202038), Natural Science Foundation of China (81773494), and the National Key Research and Development Program of China (2019YFC1200502.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.

Author notes

L. Y., G.-L. W., Y. S., Z.-Y.W., and B.-D. Z. contributed equally to this study.