Immune repertoire sequencing reveals an abnormal adaptive immune system in COVID‐19 survivors

Abstract

Accumulating evidence suggests that severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) impairs the adaptive immune system during acute infection. Still, it remains largely unclear whether the frequency and functions of T and B cells return to normal after the recovery of Coronavirus Disease 2019 (COVID‐19). Here, we analyzed immune repertoires and SARS‐CoV‐2‐specific neutralization antibodies in a prospective cohort of 40 COVID‐19 survivors with a 6‐month follow‐up after hospital discharge. Immune repertoire sequencing revealed abnormal T‐ and B‐cell expression and function with large T cell receptor/B cell receptor clones, decreased diversity, abnormal class‐switch recombination, and somatic hypermutation. A decreased number of B cells but an increased proportion of CD19⁺CD138⁺ B cells were found in COVID‐19 survivors. The proportion of CD4⁺ T cells, especially circulating follicular helper T (cTfh) cells, was increased, whereas the frequency of CD3⁺CD4⁻ T cells was decreased. SARS‐CoV‐2‐specific neutralization IgG and IgM antibodies were identified in all survivors, especially those recorded with severe COVID‐19 who showed a higher inhibition rate of neutralization antibodies. All severe cases complained of more than one COVID‐19 sequelae after 6 months of recovery. Overall, our findings indicate that SARS‐CoV‐2‐specific antibodies remain detectable even after 6 months of recovery. Because of their abnormal adaptive immune system with a low number of CD3⁺CD4⁻ T cells and high susceptibility to infections, COVID‐19 patients might need more time and medical care to fully recover from immune abnormalities and tissue damage.

1. INTRODUCTION

Since the broke out of severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) in early 2020, the ongoing pandemic of Coronavirus Disease 2019 (COVID‐19) has caused excess morbidity and mortality globally. According to the daily update of the World Health Organization (WHO), the number of confirmed COVID‐19 cases reached >621 million (including >6 million deaths) by October 18, 2022, and the mortality rate is approximately 1.05% in a global situation. Most patients survive, and the disease outcome and long‐term quality of life now receive the most attention. ¹ Many discharged patients still experience long‐term complications, such as fatigue, muscle weakness, and depression, ² , ³ because host immune abnormalities might contribute to disease severity and disease outcome. ⁴ Due to the emerging spread of SARS‐CoV‐2 variants (e.g., Omicron ⁵ ), it remains essential to investigate whether the adaptive immune system of recovered patients would provide long‐term immunity against SARS‐CoV‐2 reinfections. Therefore, it is critical to investigate whether the immune system's function and status impact the quality of life of COVID‐19 survivors.

It is known that the human adaptive immune system plays an important role in the defense against viral infections. The adaptive immune system consists of three major components: (i) CD4⁺ T cells that express surface molecules and produce cytokines to help other immune cells, (ii) CD8⁺ T cells that kill virus‐infected host cells, and (iii) B cells that produce antibodies against viral infections. ⁶ Virus‐specific CD4⁺ T cells usually differentiate into some non‐Tfh and Tfh cells that subsequently instruct B cells to mediate the production of neutralizing antibodies and the generation of B‐cell memory. ⁷ Accumulating evidence suggests a strong association between the disease severity of COVID‐19 and serum levels of Tfh cells in the peripheral blood. ⁸ , ⁹ Unlike CD4⁺ T cells, SARS‐CoV‐2‐specific CD8⁺ T cells are less frequently observed. ¹⁰ However, better clinical outcomes might be associated with the presence of SARS‐CoV‐2‐specific CD8⁺ T cells. ⁸ With the help of CD4⁺ T cells, B cells produce anti‐SARS‐CoV‐2 neutralizing antibodies with a wide range of heavy chain and light chain V genes. ¹¹ , ¹² Although many studies have reported the adaptive immune system during the acute infection of SARS‐CoV‐2, the function and frequency of adaptive immune cells after recovery remain rarely reported.

Previously, we characterized the clinical features of COVID‐19 patients during their hospitalization in Changsha, China ¹³ , ¹⁴ and reported SARS‐CoV‐2 genome diversity, ¹⁵ anti‐SARS‐CoV‐2 IgG/IgM antibodies, ¹⁶ and inflammatory cytokine dynamics during patient hospitalization. ¹⁷ After these patients had undetectable SARS‐CoV‐2 and fulfilled discharge criteria, they were discharged and closely monitored in a follow‐up of 6 months. Here, we report immune repertoires and SARS‐CoV‐2‐specific neutralization antibodies in this follow‐up cohort of 40 COVID‐19 survivors. Based on immune repertoire, flow cytometry, and ELISA analyses, our findings suggest that, compared with healthy adults, COVID‐19 survivors experience an impaired and disturbed immune system, with large T cell receptor/B cell receptor (TCR/BCR) clones, decreased diversity, abnormal class‐switch recombination (CSR) and somatic hypermutation (SHM), a decreased number of B cells but an increased proportion of CD19⁺CD138⁺ B cells, and an increased proportion of CD4⁺ T cells but a decreased frequency of CD3⁺CD4⁻ T cells, even 6 months after their hospital discharge. In addition, all survivors retained SARS‐CoV‐2‐specific neutralization IgG and IgM antibodies, while high titers of neutralization antibodies were found in patients with a history of severe COVID‐19.

2. MATERIALS AND METHODS

2.1. Patients and study approval

A total of 40 adults (median age: 51 years, male proportion: 52.5%) with a history of COVID‐19 from January to March 2020 were monitored for 6 months in our follow‐up study. These patients were recorded as having mild (57.5%, N = 23) or severe (32.5%, N = 13) COVID‐19 during their hospitalization. After their hospital discharge, neither death nor SARS‐CoV‐2 reinfection was reported in these COVID‐19 survivors. This study was conducted under the Helsinki Declaration and was approved by the Ethics Committee of The First Hospital of Changsha (approval ID: KL‐2020005). Written informed consent was obtained from all patients.

2.2. Cell separation and RNA extraction

Peripheral blood mononuclear cells (PBMCs) were isolated from 1 ml of whole peripheral venous blood using density gradient centrifugation over Ficoll‐Paque (GE Healthcare). In brief, superimpose 1 ml of whole blood diluted in phosphate‐buffered saline (PBS) on Ficoll‐Paque. After centrifugation, the middle layer was aspirated, and PBMCs were obtained after washing with PBS.

Total RNA of PBMCs was extracted by RNeasy mini kit (Qiagen) and followed the instructions. In brief, add 350 μl Buffer RLT to harvest cells and homogenize, add one volume of 70% ethanol to the lysate and mix well. Transfer sample to an RNeasy MinElute spin column, centrifuge, and discard the flow‐through. Add 350 μl Buffer RW1, 500 μl Buffer RPE, and 500 μl of 80% ethanol successively, centrifuge each time when a reagent is added. Finally, add 14 μl RNase‐free water to elute the RNA. The concentration of total RNA was measured by NanoDrop ND‐2000 spectrophotometer, and the RNA samples were reserved at −80°C.

2.3. Flow cytometry

Flow cytometry detection of T/B lymphocyte subsets was performed using Cytek™ Aurora. PBMCs were stained with PE/Cyanine7 anti‐human CD3 (Biolegend, No. 300316), Brilliant Violet 421 anti‐human CD4 (Biolegend, No. 317434), PerCP/Cyanine5.5 anti‐human CD19 (Biolegend, No. 982412), PE/Dazzle 594 anti‐human CD27 (Biolegend, No. 124228), Brilliant Violet 605 anti‐human CD45RA (BD, No. 562886), APC anti‐human CD45RO (eBioscience No. 17‐0457‐42), FITC anti‐human CD138 (Biolegend, No. 352303), APC/Cyanine7 anti‐human CXCR5 (Biolegend, No. 356926), and PE anti‐human PD‐1 (Biolegend No. 329906).

2.4. IgG and IgM antibody levels

Serum levels of IgG and IgM antibodies were measured using magnetic chemiluminescence enzyme immunoassay kits (Bioscience), which were approved by the Chinese National Medical Products Administration (approved ID: 20203400498, 20203400499).

2.5. Neutralization assay of serum antibodies

As described previously, ¹⁸ a SARS‐CoV‐2 pseudovirus‐based neutralization assay was used to measure the neutralization activity of blood samples against SARS‐CoV‐2 (Sino Biological Inc.). Briefly, the neutralization assay was performed on HEK293T cells expressing ACE2 (Cat# OEC001). Diluted serum samples (1:30) were incubated with HIV‐1‐based luciferase‐expressing lentiviruses (Cat# PSV001) pseudotyped with wildtype SARS‐CoV‐2 spike (YP_009724390.1, Cat# VG40589⁻UT) in a 96‐well plate. After 3 days of incubation, the relative luminescence units of luciferase activity were measured in the cells lysed with 30 μl of lysis buffer (Promega) using the Luciferase Assay Kit (Promega).

2.6. RNA amplification

Extracted total RNA was amplified using a commercially available iR‐RepSeq‐plus 7⁻Chain Cassette (iRepertoire Inc.), which covers all 7 chains of human TCR and BCR. A strategy allowing the incorporation of unique molecular identifiers (UMIs) was performed during the reverse transcription (RT) step. The library preparation for each sample was performed with a disposable cassette, in which all amplification and purification reagents were preloaded into the cassette.

2.7. Construction of human TCR/BCR libraries

For the construction of human TCR and BCR libraries, Qiagen OneStep RT PCR mix (Qiagen) was used. The first cDNA strand was selected, and SPRIselect bead selection (Beckman Coulter) removed unused primers. Then, the second round of PCR bound and extended sequences with the V‐gene primer mix. Afterward, both the first‐ and second‐strand products were purified using SPRIselect beads. A pair of primers that are specific for communal sites engineered onto the 5′ end of V‐ and C‐ primers employed in both the first‐ and second‐strand synthesis were used for library amplification. The final library included adapters of dual index sequencing from Illumina, which was a 10‐nucleotide UMI region and an 8‐nucleotide C‐gene internal barcode primer.

2.8. Immune repertoire sequencing

Libraries that underwent amplification were multiplexed and pooled for sequencing on the Illumina NovaSeq platform. The use of 300 base‐pair paired‐end sequencing allowed complete coverage of the full variable region, including CDR1, CDR2, and CDR3. Analysis of the raw data from sequencing was performed using the iRmap program, which has been described previously.

In brief, sequence reads were demultiplexed according to both Illumina dual indices incorporated during the amplification process and barcode sequences from the constant region at the 5′ end of reads. Then, reads were trimmed by base qualities individually with a 2‐basesliding window. If any value of quality in the window was <20, the sequence that stretched from the 3′ end window was trimmed out from the original read. A modified Needleman‐Wunsch algorithm was used to join trimmed pair‐end reads together by overlapping alignment. If a perfect match did not exist in the pair of reverse and forward reads from overlapping regions, both reads were discarded from further investigation. A Smith‐Waterman algorithm was performed to merge reads in order of mapping V, D, J, and C germline sequences according to an IMGT reference library. To define the complementary determining regions 3 (CDR3s) region, the reads from the mapping results were associated with the CDR3 boundary positions of the IMGT reference sequences. Then, the extracted CDR3 regions were translated, and the amino acid sequences were recorded.

2.9. Statistical analysis

The Mann–Whitney U‐test or Student's t‐test was performed by SPSS to calculate the statistical significance of mean values. Statistical significance was considered as a p‐value less than 0.05. The sample size was not predetermined by any statistical method. Figures and tables were generated using GraphPad Prism 9, Python, R, www.chiplot.online, and smart.servier.com.

3. RESULTS

3.1. Large TCR/BCR clones continuously existed in COVID‐19 survivors

T/B lymphocytes are the foundation of the human adaptive immune response and immune memory. To investigate the long‐term effects of SARS‐CoV‐2 on the survivors' immune system, specifically on TCR/BCR at the clonal and chain levels, we performed immune repertoire sequencing of PBMCs from 40 COVID‐19 survivors at 6 months after acute infection (SARS‐CoV‐2 nucleic acid detection shifted to negative), and samples from 12 age‐ and sex‐matched healthy individuals were used as controls (Figure 1). All seven TCR/BCR chains, including TCR alpha (TRA), TCR beta (TRB), TCR delta (TRD), TCR gamma (TRG), BCR heavy (IGH), BCR kappa (IGK), and BCR lambda (IGL), were detected simultaneously. Overall, as many as 6.2 million reads of receptors were detected, and 1.5 million unique complementarity‐determining regions (uCDR3s) were recognized from each survivor individually (Supporting Information: Figure 1).

Figure 1.

Overview of the immune repertoires, anti‐SARS‐CoV‐2 antibody detection, and lymphocyte subsets from the COVID‐19 survivor analysis strategy. Schematic diagram of the analysis strategy for immune repertoires, anti‐SARS‐CoV‐2 antibody detection, and lymphocyte subsets from COVID‐19 survivors 6 months after acute infection. COVID‐19, Coronavirus Disease 2019; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2.

At the clonal level, compared with healthy controls who had not come across with SARS‐CoV‐2, COVID‐19 survivors still reserved numerous large clones of TCR and BCR at 6 months after acute infection (Figure 2A and Supporting Information: Figure 2). These large clones mainly existed in the TRA, TRB, IGH, IGK, and IGL chains. Due to that current sequencing data on immune repertoires of COVID‐19 survivors are still focused on the TRB CDR3 sequences, we identified clones with the top 1% expression in each sample as the large clones and compared these sequences with the TRB CDR3 sequences which have been considered to be highly specific to SARS‐CoV‐2 in literature. ¹⁹ The results showed that 77.5% (31/40) of the survivors had at least one large clone highly specific to SARS‐CoV‐2. Among them, a certain survivor had as many as 11 large clones specific to SARS‐CoV‐2 (Figure 2B). Therefore, we believe that the majority of survivors reserve large TCR/BCR clones specific to SARS‐CoV‐2, which maintained an efficient immune response and immune memory in peripheral blood at 6 months after acute viral infection.

Figure 2.

Tree maps of COVID‐19 survivors’ immune repertoire and TCR diversity. (A) Tree maps of the immune repertoire of COVID‐19 survivors. Each unique CDR3 clone is treated as a pane with different colors, and the size of the panes is based on expression. Panes were gathered from top to bottom and from left to right as TRA, TRB, TRD, TRG, IGH, IGK, and IGL chains. (B) The count number of SARS‐CoV‐2 specific clones in the top 1% clones and part of their sequences. (C) Expression percentage calculated by reads and uCDR3s in each chain from COVID‐19 survivors and healthy controls. The outside circle represents COVID‐19 survivors, whereas the inside circle represents healthy controls. (D) The Pielou evenness, Shannon index, Simpson index, Gini index, D50‐full index, and D50‐head index of TRA, TRB, TRD, and TRG chains. For (C), *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, and ns is no significance. For all column plots, the bars show the mean ± SEM. COVID‐19, Coronavirus Disease 2019; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2; uCDR3, unique complementarity‐determining region.

When we calculated the read numbers of TCR/BCR chains, the immune repertoire from both COVID‐19 survivors and healthy controls was dominated by the expression of TRA, TRB, and IGK chains. In our measurement of the uCDR3s number, COVID‐19 survivors and healthy individuals were both dominated by the expressions of TRA, TRB, and IGH chains (Figure 2C). However, analyses of read numbers and uCDR3 numbers showed no significant difference in the expression of each of the 7 chains between COVID‐19 survivors and healthy controls (Supporting Information: Figure 3). In contrast to the extremely compressed expression of TRB during the SARS‐CoV‐2 acute infection period, ²⁰ the compositions of the immune repertoire from COVID‐19 survivors, especially their TRB expression, returned to normal levels at 6 months after acute infection.

3.2. Decreased TCR/BCR diversity in COVID‐19 survivors

We utilized multiple models to characterize four TCR chains (TRA, TRB, TRD, TRG). The Pielou Evenness and Gini index places particular emphasis on the evenness of a repertoire, and the Shannon index and Simpson index focus on the diversity of a repertoire. The D50‐full index and D50‐head index were used to describe the effect of large clones on a repertoire. In the TCR repertoires, the Pielou evenness, Shannon index, Simpson index, D50‐full index, and D50‐head index were all significantly reduced in the TRA and TRB chains of COVID‐19 survivors at 6 months after acute infection compared to healthy controls. Furthermore, the Gini index of the TRA, TRB, and TRG repertoires from COVID‐19 survivors was significantly increased compared to that of healthy individuals (Figure 2D), where a higher Gini index indicated a more uneven repertoire. COVID‐19 survivors had lower diversity and higher unevenness of the TCR repertoire than healthy controls at 6 months after acute infection, particularly in the TRA and TRB chains. Moreover, there was a significant increase in the proportion of large clones in the TRA and TRB repertoires of COVID‐19 survivors. The appearance of large clones from TRA and TRB significantly influenced the diversification and evenness of the TCR repertoire.

Similar to the TCR repertoire, a significant decrease in diversification and evenness was also observed in the BCR repertoire from COVID‐19 survivors. The Pielou evenness, Shannon index, and D50‐full index of IGH, IGK, and IGL showed significant decreases in COVID‐19 survivors compared with those in healthy controls, while the Simpson index significantly decreased in IGH, and the D50‐head index significantly decreased in the IGH and IGL. Additionally, the Gini index of all three BCR chains increased compared to healthy controls with statistical significance. Meanwhile, the proportion of large clones in the IGH repertoire from COVID‐19 survivors increased significantly versus controls (Figure 3A).

Figure 3.

BCR repertoire from COVID‐19 survivors. (A) The Pielou evenness, Shannon index, Simpson index, Gini index, D50‐full index, and D50‐head index of IGH, IGK, and IGL chains. (B) Expression percentage calculated by reads and uCDR3s in each IGH chain isotype from COVID‐19 survivors and healthy controls. The outside circle represents COVID‐19 survivors, whereas the inside circle represents healthy controls. (C) Networks of IGHs from COVID‐19 survivors. Each node represents a unique CDR3 sequence based on expression, and the sequences that underwent CSR or SHM are linked by sticks. Purple represents IgA, light blue represents IgM, dark blue represents IgG, orange represents IgD and green represents IgE. (D) Schematic diagram of B cells and the structure of BCRs. (E) Mean CDR3 length of IGL, IGK, and IGH. For (A) and (E), *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, and ns is no significance. For all column plots, the bars show the mean ± SEM. BCR, B cell receptor; COVID‐19, Coronavirus Disease 2019; CSR, class‐switch recombination; SHM, somatic hypermutation; uCDR3, unique complementarity‐determining region.

3.3. Abnormal CSR and SHM in the BCR repertoire of COVID‐19 survivors

Our further analyses focused on the IGH repertoire with different immunoglobulin subtypes, including IgM, IgD, IgA, IgG, and IgE. In the analyses of read numbers and uCDR3s numbers, both IgM and IgA expression dominated the IGH repertoire of COVID‐19 survivors at 6 months after acute infection (Figure 3B). The proportion of IgE expression was significantly higher in COVID‐19 survivors than in healthy controls (Supporting Information: Figure 4A). Consistently, a significant increase in IgM/D‐IgE and IgG‐IgE CSR was observed in COVID‐19 survivors (Supporting Information: Figure 4B).

Figure 4.

Lymphocyte subsets from COVID‐19 survivors. (A) The B‐cell subsets of COVID‐19 survivors at 6 months after acute infection, including CD19⁺CD138⁺ B cells, CD19⁺CD27⁻ naïve B cells, and CD19⁺CD27⁺ memory B cells. (B) The T cell subsets of COVID‐19 survivors at 6 months after acute infection, including CD4⁺ T cells, CD3⁺CD4⁻ T cells, CD45RO⁺ memory T cells, CD45RA⁺ naïve T cells, and CD4⁺PD‐1⁺CXCR5⁺ Tfh‐like cells. For (A) and (B), *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, and ns is no significance. For all column plots, the bars show the mean ± SEM. COVID‐19, Coronavirus Disease 2019; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2.

To visualize the sophisticated SHM and CSR occurring among IGH clones of both COVID‐19 survivors and healthy controls, we illustrated all IGH clones from the same individual on a network diagram. Each colored circle represents a specific IGH clone of a certain subtype, the size of the circle is related to the expression level, and clones linked by short lines indicate only one different amino acid between their primary structure (sequences from CDR2 to framework 4). We observed the intricate SHM and CSR occurring in IGH repertoires from COVID‐19 survivors at 6 months after acute infection, which was not present in healthy individuals who had not suffered from SARS‐COV‐2 infection (Figure 3C and Supporting Information: Figure 5). At the same time, the SHM frequency of BCR light chains, including IGK and IGL, was significantly higher in COVID‐19 survivors than in healthy controls (Supporting Information: Figure 6). At 6 months after acute infection with SARS‐CoV‐2, not only the constant region but also the variable region of BCR showed a difference from healthy controls (Figure 3D).

3.4. Shortened CDR3 length of BCR light chains in COVID‐19 survivors

With a degree of peptide sequence variability, CDR3 is the major domain that can specifically bind to antigens. The average CDR3 peptide chain length of IGK and IGL was significantly shorter in COVID‐19 survivors than in healthy controls (Figure 3E), which might be due to immune responses against SARS‐CoV‐2. A large number of high‐affinity SARS‐CoV‐2‐specific antibodies with short CDR3 sequences were produced, resulting in a decreased average length of CDR3 peptide chains at 6 months after acute infection.

3.5. Decreased B‐cell proportion but increased CD19⁺CD138⁺ cell proportion in COVID‐19 survivors

To verify the findings from the immune repertoire sequence in T/B lymphocyte subsets, flow cytometry was performed in PBMCs from 36 COVID‐19 survivors at 6 months after acute infection, as well as 23 age‐ and sex‐matched healthy individuals as controls. The proportion of CD19⁺ B lymphocytes was significantly lower in COVID‐19 survivors than in healthy controls, indicating that the damage to lymphatic tissue caused by SARS‐CoV‐2 had not recovered completely. Notably, the proportion of CD19⁺CD138⁺ cells in B lymphocytes was significantly higher, and more B cells proliferated and differentiated into an antibody‐secreting phenotype, suggesting that humoral immunity still played an important role after 6 months of SARS‐CoV‐2 infection. However, there was no significant difference in the proportion of CD19⁻CD138⁺ B cells, CD19⁺CD27⁻ naïve B cells, or CD19⁺CD27⁺ memory B cells between COVID‐19 survivors and healthy controls (Figure 4A and Supporting Information: Figure 7).

The proportion of CD3⁺ T cells in PBMCs showed no significant difference between COVID‐19 survivors and healthy controls. In subsets of T cells, the proportion of CD4⁺ T cells was higher in COVID‐19 survivors, but the CD3⁺CD4⁻ T cell proportion was significantly lower than that in healthy controls. Due to the limitation of fluorescence dye in our flow machine, we had no CD8 staining in this experiments. CD3⁺CD4⁻T cells are also mostly CD8⁺ T cells except tiny proportion of CD4⁻CD8⁻ T cells and double‐positive T cells, we think the proportion of CD3⁺CD4⁻T cells can partially reflect the frequency of CD8⁺T cells. This implies the long‐term effects of SARS‐CoV‐2 in survivors. In both subsets of CD4⁺ T cells and CD3⁺CD4⁻ T cells, COVID‐19 survivors showed a significant increase in the proportion of CD45RO⁺ memory T cells and a significant decrease in CD45RA⁺ naïve T cells (Figure 4B), which indicated that increased T cells were activated to participate in immune responses against SARS‐CoV‐2.

CD4⁺PD‐1⁺CXCR5⁺ Tfh‐like cells are a group of T helper cells that trigger B‐cell differentiation into plasma cells, which secrete specific antibodies and memory B cells. The significantly increased proportion of Tfh‐like cells was one of the notable characteristics that distinguished COVID‐19 survivors from healthy controls after 6 months (Figure 4B). Active humoral immune responses were maintained at least half a year after the SARS‐CoV‐2 infection.

3.6. Anti‐SARS‐CoV‐2 neutralizing antibodies in COVID‐19 survivors

To explore the existence of anti‐SARS‐CoV‐2 antibodies in COVID‐19 survivors, we examined the serum levels of SARS‐CoV‐2‐specific IgM and IgG, as well as SARS‐CoV‐2 pseudovirus‐based neutralizing antibodies, in 40 COVID‐19 survivors and 23 age‐sex‐matched healthy controls. The results showed that both anti‐SARS‐CoV‐2 IgM and IgG serum levels were significantly higher in COVID‐19 survivors than in healthy controls (Supporting Information: Figure 8). The rate of anti‐SARS‐CoV‐2 neutralizing antibodies was significantly higher in COVID‐19 survivors than in healthy individuals (Supporting Information: Figure 8). All survivors recorded with a history of severe COVID‐19 achieved a high viral inhibition rate. These severe COVID‐19 survivors also showed a significantly lower T‐cell proportion than other survivors. The Pielou evenness and D50‐full index of the TRA repertoire were significantly lower in severe COVID‐19 survivors than in other survivors. In addition, the Gini index of TRA increased significantly in severe COVID‐19 survivors. Moreover, the frequency of SHM in IGHG3/4 was significantly lower in severe COVID‐19 survivors than in other survivors (Supporting Information: Figure 9A). These results indicate that severe COVID‐19 leads to intense immune responses of anti‐SARS‐CoV‐2 neutralizing antibodies.

When COVID‐19 survivors were classified into two subgroups based on a high (≥40%) or low (<40%) inhibition rate, we found that all survivors recorded with a history of severe COVID‐19 had a high virus inhibition rate of ≥40% (Supporting Information: Figure 8). The serological antibody detection showed that the concentration of anti‐SARS‐CoV‐2 IgG in the high‐inhibition group was significantly higher than that in the low‐inhibition group. However, the globulin concentration and the proportion of plasma cells were significantly lower in the high‐inhibition group than in the low‐inhibition group. In the analysis of immune repertoires, the diversity of the TRA repertoire (Shannon index) and the frequency of SHM in IGHD were significantly lower in the high‐inhibition group than in the low‐inhibition group (Supporting Information: Figure 9B).

3.7. Sequelae occurred more often in severe COVID‐19 patients

To correlate COVID‐19 sequelae and immune responses, we analyzed the association of 21 common sequelae with immune functions in 36 COVID‐19 survivors. Twenty‐one sequelae included chronic cough, expectoration, wheezing, chest pain, myalgia, fatigue, smell disorder, taste disorder, diarrhea or vomiting, alopecia, joint pain, oral ulcer, skin rash, palpitation, dizziness, headache, decreased appetite, sore throat or difficulty swallowing, sleep difficulties, anxiety or depression, and hypomnesia. In total, 50% (18/36) of COVID‐19 survivors had at least one sequela after 6 months of recovery. Compared with other regular survivors, survivors who had suffered from severe COVID‐19 had higher incidences of myalgia, fatigue, dizziness, headache, and sleep difficulties. However, there was no significant difference between the high and low viral inhibition rate groups (Table 1).

Table 1.

Sequela of COVID‐19 survivors at 6 months after acute infection

Sequelae	All survivors (n = 36)		Regular (n = 25)		Severe (n = 10)		Low inhibition (n = 13)		High inhibition (n = 23)
Any one of the following symptoms	50.00%	(18/36)	44%	(11/25)	70%	(7/10)	38.46%	(5/13)	56.52%	(13/23)
Chronic cough	5.56%	(2/36)	8%	(2/25)	0%	(0/10)	15.38%	(2/13)	0%	(0/23)
Expectoration	2.78%	(1/36)	4%	(1/25)	0%	(0/10)	7.69%	(1/13)	0%	(0/23)
Wheeze	11.11%	(4/36)	8%	(2/25)	20%	(2/10)	15.38%	(2/13)	8.70%	(2/23)
Cheat pain	11.11%	(4/36)	8%	(2/25)	20%	(2/10)	7.69%	(1/13)	13.04%	(3/23)
Myalgia*	11.11%	(4/36)	4%	(1/25)	30%	(3/10)	7.69%	(1/13)	13.04%	(3/23)
Fatigue*	16.67%	(6/36)	8%	(2/25)	40%	(4/10)	7.69%	(1/13)	21.74%	(5/23)
Smell disorder	8.33%	(3/36)	4%	(1/25)	20%	(2/10)	0%	(0/13)	13.04%	(3/23)
Taste disorder	0%	(0/36)	0%	(0/25)	0%	(0/10)	0%	(0/13)	0%	(0/23)
Diarrhea or vomiting	2.78%	(1/36)	0%	(0/25)	10%	(1/10)	0%	(0/13)	4.35%	(1/23)
Alopecia	19.44%	(7/36)	20%	(5/25)	20%	(2/10)	15.38%	(2/13)	21.74%	(5/23)
Joint pain	8.33%	(3/36)	4%	(1/25)	20%	(2/10)	7.69%	(1/13)	8.70%	(2/23)
Oral ulcer	0%	(0/36)	0%	(0/25)	0%	(0/10)	0%	(0/13)	0%	(0/23)
Skin rash	0%	(0/36)	0%	(0/25)	0%	(0/10)	0%	(0/13)	0%	(0/23)
Palpitation	2.78	(1/36)	0%	(0/25)	10%	(1/10)	0%	(0/13)	4.35%	(1/23)
Dizziness*	11.11%	(4/36)	4%	(1/25)	30%	(3/10)	7.69%	(1/13)	13.04%	(3/23)
Headache**	8.33%	(3/36)	0%	(0/25)	30%	(3/10)	0%	(0/13)	13.04%	(3/23)
Decreased appetite	2.78%	(1/36)	0%	(0/25)	10%	(1/10)	0%	(0/13)	4.35%	(1/23)
Sore throat or difficult to swallow	2.78%	(1/36)	0%	(0/25)	10%	(1/10)	0%	(0/13)	4.35%	(1/23)
Sleep difficulties*	11.11%	(4/36)	4%	(1/25)	30%	(3/10)	7.69%	(1/13)	13.04%	(3/23)
Anxiety/depression	0%	(0/36)	0%	(0/25)	0%	(0/10)	0%	(0/13)	0%	(0/23)
Hypomnesia	8.33%	(3/36)	4%	(1/25)	20%	(2/10)	0%	(0/13)	13.04%	(3/23)

Note: Comparison between regular and severe group, *p < 0.05; **p < 0.01.

Abbreviation: COVID‐19, Coronavirus Disease 2019.

Compared with survivors with no sequelae, COVID‐19 survivors with sequelae showed a significantly lower CSR frequency of IgD/M to IgG and a lower SHM frequency of IGHG3/4 (Supporting Information: Figure 9C). This suggests that the BCR repertoire, especially the IgG repertoire, is likely associated with COVID‐19 sequelae.

4. DISCUSSION

In the past 2 years, immune responses in COVID‐19 patients have been intensively investigated, not only for protection but also for immunopathology. ²¹ , ²² Several immunomodulatory drugs, such as dexamethasone and cytokine blockers, have become the main therapy for immunopathology after SARS‐CoV‐2 infection. ²³ Some studies have already revealed persisting antibody responses and B and T cells in COVID‐19 patients several months after recovery. ²⁴ , ²⁵ , ²⁶ Different immune statuses have been described by flow cytometry in patients with mild and severe COVID‐19. ²⁷ However, to our knowledge, no study has comprehensively described the adaptive immune system with regard to cell surface markers, immune repertoires in 7 chains, and antibody titers. Therefore, this study attempted to fill this research gap.

Viral infections can affect the host immune system greatly beyond our current knowledge. From this study, even half a year after recovery, an obvious difference in the adaptive immune system was still observed in the majority of SARS‐CoV‐2 survivors. Large TCR/BCR clones, decreased TCR/BCR diversity, decreased B‐cell proportion but increased plasma cell proportion, and SARS‐CoV‐2 neutralization antibodies continuously existed in COVID‐19 survivors. These changes can affect the host immune system as a whole, which may dampen the ability of the immune system to fight other infections. These findings indicate that the immune system is still active in recovered patients, even though some of them have no COVID‐19 symptoms.

Our findings suggest that PBMCs samples of COVID‐19 harbored CD4⁺ and CD8⁺ T cells and B cells with large‐scale clones of TRA, TRB, IGH, IGK, and IGL chains. The majority of γδT cells (encoded by TRD and TRG) lack CD4 and CD8 expression and do not recognize MHC molecules. Most αβT cells (encoded by TRA and TRB) express CD4 and CD8 molecules and recognize antigens presented by MHC‐I and ‐II. ²⁸ , ²⁹ This finding is consistent with previous reports that a high frequency of virus‐specific T cells was observed in convalescent patients. ²⁴ Decreased TCR and BCR diversity is also observed in recovered patients, which is consistent with the previous finding that TCR repertoires decreased in COVID‐19 patients and increased in the top clonotypes in the repertoire space for COVID‐19 samples, suggesting an expansion of a small number of functional clones targeting SARS‐CoV‐2. ³⁰ However, this decreased diversity and large clones persist in convalescent patients, highlighting the reduced immune responses to fight other pathogens.

We observed dominant expression of IgM, IgG, and IgA in the IGH repertoire of COVID‐19 survivors, which is consistent with other studies showing increased serum levels of IgM, IgG, and IgA in COVID‐19 patients. In addition, abnormal CSR and SHM in BCR repertoires have been found in convalescent patients. CSR and SHM improve the quality of the B‐cell response after viral infection. ³¹ Defects or abnormalities in CSR and SHM lead to disordered immune responses, and CSR and SHM have not been described in other COVID‐19 studies. Although few differences in IgE expression were found in convalescent patients, dramatically increased IgD/M/G‐IgE CSR was observed in convalescent patients in this study. Very few studies have focused on the serum levels of IgE in COVID‐19 patients. Only one recent report pointed out that more severe COVID‐19 patients had higher levels of IgE. ³² In addition, IgE is a key player in airway disease, ³³ which is unlikely to be associated with SARS‐CoV‐2‐specific reactions but a consequence of airway inflammation.

The proportion of antibody‐producing plasma cells, B cell helper Tfh cells, and the serum levels of SARS‐CoV‐2 pseudovirus‐based neutralizing antibodies remained higher in convalescent patients, indicating that the impact of SARS‐CoV‐2 persists for more than half a year. Combined with immune repertoire sequencing, all these parameters can be utilized as a panel to evaluate the immune status of recovered patients. In addition, sequelae, such as chronic cough, expectoration, wheezing, chest pain, myalgia, fatigue, smell disorder, and taste disorder, more often occur in patients with a history of severe COVID‐19, indicating that SARS‐CoV‐2 may dampen our health not only in the immune system but also in other organs. ³⁴

This study has limitations. First, the sample size is a limitation when we study an important and global pandemic. Future studies need to recruit a larger cohort with detailed information on SARS‐CoV‐2 variants. Second, due to the very strict regulation of COVID‐19 samples, we were not able to obtain COVID‐19 samples before hospital discharge, and the ethics committees only permitted our study to collect PBMCs samples from convalescent patients. Future studies need to collect samples multiple times, especially before COVID‐19, during COVID‐19, and after COVID‐19. Third, a 6‐month follow‐up study may be insufficient to fully understand the long‐term impact of COVID‐19. Despite our wishes to obtain 1‐year samples, a majority of survivors already had COVID‐19 vaccinations which prevented us from extending our study to 1 year or even longer. Nevertheless, we believe that a 6‐month follow‐up already addressed an important impact of COVID‐19 on those survivors. Future studies need to address the immune responses of COVID‐19 patients after 1 year and even longer. Finally, no patient in our study was reported to have any SARS‐COV‐2 reinfection. Further analyses also need to address the interplay between vaccines and immune memory in COVID‐19 survivors.

In conclusion, our data suggest that SARS‐CoV‐2‐specific antibodies remain detectable even after 6 months of recovery. Due to abnormal adaptive immune system with a low number of CD3⁺CD4⁻ T cells and high susceptibility to infections, COVID‐19 patients might need a long time and medical care to fully recover from immune abnormalities and tissue damage.

AUTHOR CONTRIBUTIONS

Chen Jia, Pan Zhang, and Song Li performed statistical analyses and drafted the manuscript. Zhiguo Zhou, Wenjing Pan, and Ming Yang performed data acquisition. Bo Li, Qianqian Zhang, Mingming Zhao, and Xianglong Kong conducted clinical practice and data interpretation. Keyu Li, Tingting Yue, Ting Cai, Zijun Wang, and Erik De Clercq edited the writing of the manuscript. Guangdi Li, Jiyang Liu, Haijing Wu, and Qianjin Lu supervised the study, obtained funding, and revised the manuscript. All authors contributed to the final article.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

Supplementary information.

Jia C, Zhou Z, Pan W, et al. Immune repertoire sequencing reveals an abnormal adaptive immune system in COVID‐19 survivors. J Med Virol. 2022;95:e28340. 10.1002/jmv.28340

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.