SARS-CoV-2 booster vaccination rescues attenuated IgG1 memory B cell response in primary antibody deficiency patients

Abstract

Background: Although SARS-CoV-2 vaccines have proven effective in eliciting a protective immune response in healthy individuals, their ability to induce a durable immune response in immunocompromised individuals remains poorly understood. Primary antibody deficiency (PAD) syndromes are among the most common primary immunodeficiency disorders in adults and are characterized by hypogammaglobulinemia and impaired ability to mount robust antibody responses following infection or vaccination.

Methods: Here, we present an analysis of both the B and T cell response in a prospective cohort of 30 individuals with PAD up to 150 days following initial COVID-19 vaccination and 150 days post mRNA booster vaccination.

Results: After the primary vaccination series, many of the individuals with PAD syndromes mounted SARS-CoV-2 specific memory B and CD4⁺ T cell responses that overall were comparable to healthy individuals. Nonetheless, individuals with PAD syndromes had reduced IgG1⁺ and CD11c⁺ memory B cell responses following the primary vaccination series, with the defect in IgG1 class-switching rescued following mRNA booster doses. Boosting also elicited an increase in the SARS-CoV-2-specific B and T cell response and the development of Omicron-specific memory B cells in COVID-19-naïve PAD patients. Individuals that lacked detectable B cell responses following primary vaccination did not benefit from booster vaccination.

Conclusion: Together, these data indicate that SARS-CoV-2 vaccines elicit memory B and T cells in most PAD patients and highlights the importance of booster vaccination in immunodeficient individuals.

Keywords: B cells; SARS-CoV-2; common variable immunodeficiency; hypogammaglobulinemia; immune memory; primary antibody deficiency; specific antibody deficiency; vaccination.

Figure 1

COVID-19-experienced PAD patients have elevated spike-specific B cell response following primary vaccination series. (A) Schematic of study design including time points in which PBMCs were obtained and number of samples per time point for each group. (B) Representative flow cytometry plots of the gating strategy used to identify IgD^lo Spike⁺ B cells. Percentage of (C) IgD^lo, (D) memory (IgD^lo CD20⁺ CD38^int-lo CD27⁺), double negative (IgD^lo CD20⁺ CD38^int-lo CD27^-), and (E) IgD⁺ CD20⁺ CD38^int-lo CD27⁺ Spike⁺ cells amongst the B (Live CD19⁺ CD3^-) cell population in the healthy donor (left, white), COVID-19-naïve PAD (middle left, red), and COVID-19-experienced PAD (middle right, green) groups. Median percentage of B cells that comprise each population in all groups is shown on right. Statistical analyses in (B–E) were performed using a mixed-effects model (for trends found between time points) or two-way ANOVA (for trends found between groups shown on the median graphs) with Fisher’s least significant difference testing. Significance testing between time points was limited to comparisons relative to T1. Above the median graphs, an orange asterisk indicates a comparison between the COVID-19-naïve and -experienced groups, and a purple asterisk indicates a comparison between the COVID-19-experienced and healthy donor groups (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.001). See also Figure S2 .

Figure 2

COVID-19-naïve PAD patients display elevated spike-specific B cell response following booster vaccination. Percentage of (A) IgD^lo, (B) memory (IgD^lo CD20⁺ CD38^int-lo CD27⁺), (C) double negative (IgD^lo CD20⁺ CD38^int-lo CD27^-), and (D) IgD⁺ CD20⁺ CD38^int-lo CD27⁺ Spike⁺ cells among the B (Live CD19⁺ CD3^-) cell population in the COVID-19-naïve PAD (left, red) and COVID-19-experienced PAD (middle, green) cohorts. Median percentage of B cells that comprise each population in all groups is shown on right. Correlation between percentage of B cells that are Spike⁺ (left) or RBD⁺ (right) memory cells prior to boosting and the serum neutralizing activity against (E) WA1/2020, (F) B.1.617.2, and (G) BA.1. The pre-boost group consists of the last sample obtained from each patient prior to booster vaccination. Statistical analyses were performed using a mixed effects model (for trends found between time points) with Fisher’s least significant difference testing in (A–D), or a Pearson rank correlation (with Pearson trend lines for visualization) in (E–G). (*p < 0.05; **p < 0.01). Significance testing between time points was limited to comparisons relative to pre-boost. See also Figure S3 .

Figure 3

Spike-specific memory B cells from COVID-19-naïve PAD patients display reduced IgG1 class switching following primary vaccination series. (A) Representative flow cytometry plots of the gating strategy used to identify the immunoglobulin subclass of Spike⁺ memory (IgD^lo CD20⁺ CD38^int-lo CD27⁺) B cells. Percentage of Spike⁺ memory B cells that are (B) IgG1⁺, (C) IgM⁺, (D) IgA⁺, (E) IgG2⁺, and (F) IgG3⁺ in healthy donors (left, white), and COVID-19-naïve (middle left, red) and -experienced (middle right, green) PAD cohorts. Median percentage of B cells that comprise each population in all groups is shown on right. Statistical analyses in B-D were performed using a mixed effects model (for trends found between time points) or two-way ANOVA (for trends found between groups shown on the median graphs) with Fisher’s least significant difference testing. Significance testing between time points was limited to comparisons relative to T1. Above the median graphs, a green asterisk indicates a comparison between the COVID-19-naïve and healthy donor groups, an orange asterisk indicates a comparison between the COVID-19-naïve and -experienced groups, and a purple asterisk indicates a comparison between the COVID-19-experienced and healthy donor groups (*p < 0.05; **p < 0.01; ***p < 0.001). See also Figure S4 .

Figure 4

Spike-specific memory B cells from PAD patients display reduced CD11c expression. (A) Representative flow cytometry plots of the expression of CD11c, CD71, and CXCR5 on Spike⁺ memory (IgD^lo CD20⁺ CD38^int-lo CD27⁺) B cells. (B) Representative flow cytometry plots of the expression of CD11c and CXCR5 on Spike⁺ double negative (IgD^lo CD20⁺ CD38^int-lo CD27^-) B cells. (C) Percentage of Spike⁺ memory B cells that are CD11c⁺ in the healthy donor (left, white), COVID-19-naïve PAD (middle left, red), and COVID-19-experienced PAD (middle right, green) cohorts. Median percentage of CD11c⁺ cells in all groups is shown on right. (D) Correlation between percentage of Spike⁺ memory B cells that are IgG1⁺ and CD11c⁺ at T1 (left) and B1 (middle) in the COVID-19-naïve PAD patients, and at T1 (right) in the. COVID-19-experienced PAD patients. Associations for D are calculated using Pearson rank correlation and shown with Pearson trend lines for visualization. Percentage of Spike⁺ memory B cells that are (E) CD71⁺ or (F) CXCR5⁺ in healthy donors (left, white), COVID-19-naïve PAD (middle left, red), and COVID-19-experienced PAD (middle right, green) cohorts. Median percentage of B cells that comprise each population in all groups is shown on right. Percentage of Spike⁺ double negative B cells that are (G) CD11c⁺ CXCR5^- or (H) CD11c^- CXCR5⁺ in the healthy donor (left, white), COVID-19-naïve PAD (middle left, red), and COVID-19-experienced PAD (middle right, green) cohorts. Median percentage of double negative B cells that comprise each population in all groups is shown on right. Statistical analyses in (C, E–H) were performed using a mixed effects model (for trends found between time points) or two-way ANOVA (for trends found between groups shown on the median graphs) with Fisher’s least significant difference testing. Significance testing between time points was limited to comparisons relative to T1. Above the median graphs, a green asterisk indicates a comparison between the COVID-19-naïve and healthy donor groups, an orange asterisk indicates a comparison between the COVID-19-naïve and COVID-19-experienced groups, and a purple asterisk indicates a comparison between the COVID-19-experienced and healthy donor groups (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). See also Figure S5 .

Figure 5

PAD patients have elevated Omicron-specific B cell responses following booster vaccination. (A) Representative flow cytometry plots of the percentage of Omicron-specific B cells among the memory (IgD^lo CD20⁺ CD38^int-lo CD27⁺) B cell population prior to and post booster vaccination in COVID-19-naïve (left) and -experienced (right) individuals with PAD syndromes. Percentage of (B) IgD^lo, (C) memory, (D) double negative (IgD^lo CD20⁺ CD38^int-lo CD27^-), and (E) IgD⁺ CD20⁺ CD38^int-lo CD27⁺ Omicron⁺ cells among the B (Live CD19⁺ CD3^-) cell population in healthy donors (left, white), COVID-19-naïve PAD (middle left, red), and COVID-19-experienced PAD (middle right, green) cohorts. Median percentage of B cells that comprise each population in all groups is shown on right. Statistical analyses in (B–E) were performed using a mixed effects model (for trends found between time points) or two-way ANOVA (for trends found between groups shown on the median graphs) with Fisher’s least significant difference testing. Significance testing between time points was limited to comparisons relative to pre-boost. Above the median graphs, an orange asterisk indicates a comparison between the COVID-19-naïve and -experienced groups (*p < 0.05). See also Figure S6 .

Figure 6

PAD patients display unimpaired SARS-CoV-2-specific CD4⁺ T cell responses following vaccination. (A) Representative flow cytometry plots of the gating strategy used to identify S_167-180 ⁺ and S_816-830 ⁺ CD4⁺ T cells. Percentage of CD4⁺ (Live CD3⁺ CD19^- CD4⁺ CD8^-) T cells that are (B) S_167-180 ⁺, (C) S_816-830 ⁺, or (D) Tetramer⁺ (S_167-180 ⁺ or S_816-830 ⁺) in healthy donors (left, white), COVID-19-naïve PAD (middle left, red), and COVID-19-experienced PAD (middle right, green) cohorts. Median percentage of CD4⁺ T cells that comprise each population in all groups is shown on right. Correlation between percentage of B cells that are Spike⁺ (left) or RBD⁺ (right) memory (IgD^lo CD20⁺ CD38^int-lo CD27⁺) cells and the percentage of CD4⁺ T cells that are Tetramer⁺ at (E) T1 or (F) B1. Associations for (E, F) are calculated using Pearson rank correlation and shown with Pearson trend lines for visualization. Statistical analyses in (B–D) were performed using a mixed effects model (for trends found between time points) or two-way ANOVA (for trends found between groups shown on the median graphs) with Fisher’s least significant difference testing. Significance testing between time points was limited to comparisons relative to T1. Above the median graphs, an orange asterisk indicates a comparison between the COVID-19-naïve and COVID-19-experienced groups, and a purple asterisk indicates a comparison between the COVID-19-experienced and healthy donor groups (*p < 0.05; **p < 0.01). .

Figure 7

Phenotype of SARS-CoV-2-specific CD4⁺ T cell response following vaccination in PAD patients. (A) Representative flow cytometry plots of the expression of PD1, ICOS, CD38, HLA-DR, and CD45RO on Tetramer⁺ CD4⁺ (Live CD3⁺ CD19^- CD4⁺ CD8^- S_167-180 ⁺ or S_816-830 ⁺) T cells. (B) Representative flow cytometry plots of the expression of CD27 and CCR7 on CD45RO⁺ Tetramer⁺ CD4⁺ T cells. Percentage of Tetramer⁺ CD4⁺ T cells that are (C) PD1⁺, (D) ICOS⁺, (E) CD38⁺, (F) HLA-DR+, (G) central memory (CD45RO⁺ CD27⁺ CCR7⁺), or (H) effector memory (CD45RO⁺ CD27⁺ CCR7^-) in healthy donors (left, white), COVID-19-naïve PAD (middle, red), and COVID-19-experienced PAD (middle, green) cohorts. Median percentage of Tetramer⁺ T cells that that comprise each population in all groups is shown on right. See also Figure S7 .