High activation levels maintained in receptor-binding domain-specific memory B cells in people with severe coronavirus disease 2019

Abstract

The long-term health consequences of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection are still being understood. The molecular and phenotypic properties of SARS-CoV-2 antigen-specific T cells suggest a dysfunctional profile that persists in convalescence in those who were severely ill. By contrast, the antigen-specific memory B-cell (MBC) population has not yet been analyzed to the same degree, but phenotypic analysis suggests differences following recovery from mild or severe coronavirus disease 2019 (COVID-19). Here, we performed single-cell molecular analysis of the SARS-CoV-2 receptor-binding domain (RBD)-specific MBC population in three patients after severe COVID-19 and four patients after mild/moderate COVID-19. We analyzed the transcriptomic and B-cell receptor repertoire profiles at ~2 months and ~4 months after symptom onset. Transcriptomic analysis revealed a higher level of tumor necrosis factor-alpha (TNF-α) signaling via nuclear factor-kappa B in the severe group, involving CD80, FOS, CD83 and TNFAIP3 genes that was maintained over time. We demonstrated the presence of two distinct activated MBCs subsets based on expression of CD80^hi TNFAIP3^hi and CD11c^hi CD95^hi at the transcriptome level. Both groups revealed an increase in somatic hypermutation over time, indicating progressive evolution of humoral memory. This study revealed distinct molecular signatures of long-term RBD-specific MBCs in convalescence, indicating that the longevity of these cells may differ depending on acute COVID-19 severity.

Keywords: CD11c; CD95; RBD; SARS-CoV-2; memory B cells.

Figure 1

Transcriptomic analysis reveals activated receptor‐binding domain (RBD)–specific memory B cells. (a) Experimental design showing the number of patients (n = 9) and workflow used in the B cell sorting strategy and single‐cell RNA sequencing (RNA‐seq) pipeline. (b) Uniform manifold approximation projection (UMAP) generated using single‐cell RNA sequencing (RNA‐seq) of RBD‐specific memory B cells (MBCs) across five clusters (MBC1, MBC2, MBC2, actBC1 and actBC2). (c) UMAP showing distribution of nine patients across two uninfected controls (UCs; 2850955 and 2854766) and seven individuals infected with severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2; severe, S group: 289036, 247004 and 250002, mild/moderate, M group: 250011, 213007, 250017 and 213021). (d) Stack plot showing distribution of disease severity subtypes in UC, the S group (S_t1 and S_t2) and the M group (M_t1 and M_t2). (e) Dot plot showing log‐normalized average expression (color scale) and percentage of expressing cells (size scale) of selected genes across five distinct clusters identified and named as MBC1, MBC2, MBC3, actBC1 and actBC2. (f) Average gene expression of isotype transcripts (IgG1, IgG2, IgG3, IgG4, IgA1, IgA2, IgD, IgM and IgE) across the five memory B cell clusters. Ig, immunoglobulin; actBC1, activated MBC cluster 1; actBC2, activated MBC cluster 2; MBC1, MBC cluster 1; MBC2, MBC cluster 2; MBC3, MBC cluster 3.

Figure 2

Higher and maintained gene expression profile associated with B cell activation. (a) Gene Set Enrichment Analysis (GSEA) showing normalized enrichment score (NES) across the severe (S; n = 3) and mild/moderate (M; n = 4) groups with NES shown as gradient from blue to red. (b) Dot plot showing log‐normalized average expression of genes in tumor necrosis factor‐alpha (TNF‐α) signaling via the nuclear factor‐kappa B (NF‐κB) pathway and the genes involved in (c) protein secretion and glycolysis. (d) Violin plots of the distribution of activated genes (CD69, CD80, CD83, CXCR3, DUSP2, FOS, PHB, TNFAIP3) across uninfected controls (UCs; n = 2), the severe (S) group (n = 3) and the mild/moderate (M) group (n = 4). (e) Violin plots showing change in the expression level over time of CD83, CD80, TNFAIP3, NFKBIA, CD69 and CXCR3 in the S and (f) M groups. Statistical differences across disease severity and timepoints were calculated using the unpaired two‐tailed Wilcoxon rank sum test with P‐values as *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. ns, nonsignificant.

Figure 3

Maturation of receptor‐binding domain (RBD)–specific memory B cells over time. (a) Frequency of gene usage, as a percentage, of heavy chains across seven patients at t1 and t2. (b) Percentage mutated from germline in IgD⁻ heavy chain across t1 and t2 in both severe (S; n = 3: 289036, 247004, 250002) and moderate (M; n = 4: 250011, 213007, 250017 and 213021) groups. (c) Percentage mutated from germline in IgD⁻ heavy chain in the S and M group in both time points combined and (d) between t1 and t2 in the S and (e) M groups. (f) Percentage of clonal B cell receptor (BCR) population (singlets, n = 1; double, n = 2; more than two BCRs in a clone, n > 2) at t1 and t2 across seven individuals infected with severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2). (g) Increase in the percentage mutated from germline in heavy chain over time clones of the S and (h) M groups of individuals infected with SARS‐CoV‐2. (i) Decrease in aromaticity of CDR3H sequences from t1 to t2 over time clones of the S group. Statistical differences across disease severity were calculated using the two‐tailed unpaired t‐test (Mann–Whitney U‐test), with adjusted P‐values as *P < 0.05, **P < 0.01, ****P < 0.0001, ns, nonsignificant. The red horizontal line depicts median at that stage of disease. Nonparametric paired t‐test (Wilcoxon) was performed for statistical significance analysis for testing change in percentage mutation from germline and aromaticity in paired over time clones. Ig, immunoglobulin.