Humoral and Memory B Cell Responses Following SARS-CoV-2 Infection and mRNA Vaccination

Abstract

Background: Understanding the duration and quality of immune memory following SARS-CoV-2 infection and vaccination is critical for informing public health strategies and vaccine development. While waning antibody levels have raised concerns about long-term protection, the persistence of memory B cells (MBCs) and T cells plays a vital role in sustaining immunity.

Materials and methods: We conducted a longitudinal prospective study over 12 months, enrolling 285 participants in total, either after natural infection or vaccination with BNT162b2 or mRNA-1273. Peripheral blood samples were collected at four defined time points (baseline, 1-2 months, 6-7 months, and 12-13 months after vaccination or disease onset). Immune responses were assessed through serological assays quantifying anti-RBD IgG and neutralizing antibodies, B-ELISPOT, and multiparameter flow cytometry for S1-specific memory B cells.

Results: Both mRNA vaccines induced robust B cell and antibody responses, exceeding those observed after natural infection. Memory B cell frequencies peaked at 6 months and declined by 12 months, but remained above the baseline. The mRNA-1273 vaccine elicited stronger and more durable humoral and memory B-cell-mediated immunity compared to BNT162b2, likely influenced by its higher mRNA dose and longer prime-boost interval. Class-switched memory B cells and S1-specific B cells were significantly expanded in vaccine recipients. Natural infection induced more heterogeneous immune memory.

Conclusions: Both mRNA vaccination and natural SARS-CoV-2 infection induce a comparable expansion of memory B cell subsets, reflecting a consistent pattern of humoral immune responses across all studied groups. These findings highlight the importance of vaccination in generating sustained immunological memory and suggest that the vaccine platform and dosage influence the magnitude and durability of immune responses against SARS-CoV-2.

Keywords: B cell memory; BNT162b2; COVID-19; SARS-CoV-2; antigen-specific B cells; humoral immune response; mRNA-1273.

Figure 1

Gating strategy for the identification and quantification of all conventional, general, and antigen-specific B cell subpopulations. Initial gating included time-based quality control, selection of CD45⁺ lymphocytes (SSC-A vs. CD45), singlet discrimination (FSC-H vs. FSC-A), and the exclusion of dead cells. CD19⁺ B cells were identified after removing CD14⁺ monocytes, CD3⁺ T cells, and CD56⁺ NK cells. Plasma cells were defined as CD19⁻CD138⁺CD38^hi. CD19⁺ B cells were characterized based on the expression of CD27 and CD38 to distinguish plasma cells and memory B cell subsets. Memory B cell subsets (CSMB and NCSMB cells) were further characterized based on CD27 and IgD expression, with additional refinement of transitional B cells using CD21 and CD24. S1-specific memory B cells were identified as double-positive for SA-S1-BV421 and SA-S1-BUV395, but negative for the SA-BV711 decoy SA. IgG⁺ S1-specific memory B cells were detected within the switched memory compartment.

Figure 2

SARS-CoV-2-specific antibody responses in the studied subjects. (A) Comparison of the levels of anti-RBD IgG antibody at different time points within each cohort. (B) Comparison of the levels of NAB at different time points within each cohort. All p-values are from two-sided unpaired t-tests. p-values below 0.05 were considered statistically significant. Abbreviations: BAUs/mL—binding antibody units per ml. Black dots indicate individual data points plotted as outliers.

Figure 3

Comparison of B-ELISpot results at different time points within each cohort. All p-values are from two-sided unpaired t-tests. p-values below 0.05 were considered statistically significant. Black dots indicate individual data points plotted as outliers.

Figure 4

Dynamics in the frequency of peripheral B cells as determined by multiparametric flow cytometry analysis. (A) Comparison of the frequency of CD19⁺ cells at different time points within each cohort. (B) Comparison of the frequency of memory B cells at different time points within each cohort. (C) Comparison of the frequency of class-switched memory B cells at different time points within each cohort. (D) Comparison of the frequency of non-class-switched memory B cells at different time points within each cohort. All p-values are from two-sided unpaired t-tests. p-values below 0.05 were considered statistically significant. Black dots indicate individual data points plotted as outliers.

Figure 5

Dynamics in the frequency of S1-specific peripheral B cells as determined by multiparametric flow cytometry analysis. (A) Comparison of the frequency of S1-specfic class-switched memory B cells between different cohorts after stratification to three different time points. (B) Comparison of the frequency of S1-specfic non-class-switched memory B cells between different cohorts after stratification to different time points. (C) Comparison of the frequency of S1-specfic class-switched memory B cells at different time points within each cohort. (D) Comparison of the frequency of S1-specfic non-class-switched memory B cells at different time points within each cohort. All data in (A–D) are from the same dataset. All p-values are from two-sided unpaired t-tests. p-values below 0.05 were considered statistically significant. Black dots indicate individual data points plotted as outliers.