Antibody and Memory B-Cell Immunity in a Heterogeneously SARS-CoV-2-Infected and -Vaccinated Population

Abstract

Global population immunity to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is accumulating through heterogeneous combinations of infection and vaccination. Vaccine distribution in low- and middle-income countries has been variable and reliant on diverse vaccine platforms. We studied B-cell immunity in Mexico, a middle-income country where five different vaccines have been deployed to populations with high SARS-CoV-2 incidences. Levels of antibodies that bound a stabilized prefusion spike trimer, neutralizing antibody titers, and memory B-cell expansion correlated with each other across vaccine platforms. Nevertheless, the vaccines elicited variable levels of B-cell immunity, and the majority of recipients had undetectable neutralizing activity against the recently emergent omicron variant. SARS-CoV-2 infection, experienced before or after vaccination, potentiated B-cell immune responses and enabled the generation of neutralizing activity against omicron and SARS-CoV for all vaccines in nearly all individuals. These findings suggest that broad population immunity to SARS-CoV-2 will eventually be achieved but by heterogeneous paths. IMPORTANCE The majority of studies on SARS-CoV-2 vaccine-elicited immunity and immune evasion have focused on single vaccines corresponding to those distributed in high-income countries. However, in low- and middle-income countries, vaccine deployment has been far less uniform. It is therefore important to determine the levels of immunity elicited by vaccines that have been deployed globally. Such data should help inform policy. Thus, this paper is very much a "real-world" study that focuses on a middle-income country, Mexico, in which five different vaccines based on mRNA, adenovirus, and inactivated-virus platforms have been extensively deployed, while (as documented in our study) SARS-CoV-2 variants with increasing degrees of immune evasiveness have propagated in the Mexican population, culminating in the recent emergence of B.1.1.529 (omicron).

Keywords: SARS-CoV-2; infection; memory B cells; neutralizing antibodies; vaccination.

FIG 1

(A) Frequency of SARS-CoV-2 variants in Mexico between February 2020 and January 2022. A total of 48,221 viral genome sequences obtained from samples collected in Mexico and downloaded from the GISAID on 28 January 2022 were analyzed. The frequency of variants was estimated over periods of a few months or individual months based on the numbers of complete genomes sequenced. The most common lineages include variants circulating at frequencies above 10% nationally in at least one period. In addition, less common lineages classified as variants of concern (VOC) and variants of interest (VOI) by the World Health Organization (WHO) were also included. Additional non-VOC/VOI lineages that circulated at frequencies below 10% nationally during all periods were aggregated into the “Other” category. (B) Geographical distribution of viral genomes obtained in Mexico between February 2020 and January 2022. The number of genome sequences per 100,000 persons in the 32 states of Mexico is represented by a color gradient.

FIG 2

Plasma neutralization activity against SARS-CoV-2 variants in vaccine recipients. NT₅₀ values of plasma samples from recipients of one of five SARS-CoV-2 vaccines against B.1 or other SARS-CoV-2 variants were determined. (A) Recipients with no documented prior infection with SARS-CoV-2. (B) Recipients who were infected with SARS-CoV-2 prior to the study as documented by a PCR positive test (closed circles) or at an unknown time prior to sample collection as indicated by the presence of anti-N antibodies (closed triangles). Individuals with prior positive PCR tests but who were seronegative for anti-N are indicated by open circles. The median values from 2 to 4 independent experiments for each plasma sample are plotted. Dashed lines indicate the lowest plasma dilution tested (1:50). Lines indicate group median NT₅₀ values.

FIG 3

Plasma neutralization activity against SARS-CoV in vaccine recipients. NT₅₀ values were measured in recipients who were infected with SARS-CoV-2 either prior to the study as documented by a positive PCR test (closed circles) or at an unknown time prior to sample collection as indicated by the presence of anti-N antibodies (closed triangles). Individuals with prior positive PCR tests but who were seronegative for anti-N are indicated by open circles. The medians from 2 independent experiments for each plasma sample are plotted. Dashed lines indicate the lowest plasma dilution tested (1:50). Lines indicate group median NT₅₀ values.

FIG 4

Quantification of SARS-CoV-2 spike-specific memory B cells in vaccine recipients. Memory B cells (Bm) in recipient PBMCs were enumerated using fluorescence-activated cell sorter (FACS) analysis and a trimeric recombinant SARS-CoV-2 (B.1) spike protein. (A) Percentage of spike-binding (S⁺) memory B cells in recipients of one of the five vaccines who were uninfected (left) (black circles) or infected (right) (red circles) (open red circles, anti-N-negative individuals, as described in the legend of Fig. 2B). Horizontal lines indicate the median percentages of S⁺ cells (of memory B cells). (B) Correlation between neutralizing antibody titers (NT₅₀) and percentages of S-binding memory B cells across all vaccine recipients. (C) Correlation of neutralizing antibody titers (NT₅₀) with percentages of S-binding memory B cells in each separate uninfected (black symbols) or infected (red symbols) vaccine recipient group. The r values indicate Spearman correlation coefficients.

FIG 5

Quantification of antibodies that bind a prefusion SARS-CoV-2 spike protein. A conformationally stabilized trimer of a fusion protein between spike (B.1) and NanoLuc (S-6P-NanoLuc) was used to measure spike-binding antibodies. Antibodies in serially diluted participant plasma samples were captured using protein G magnetic beads and then incubated with S-6P-NanoLuc, and bound NanoLuc activity was quantified. (A) Captured NanoLuc activity expressed as relative light units (RLU) per microliter of plasma from uninfected (left) (black circles) or infected (right) (red circles) vaccine recipients (open red circles indicate anti-N-negative samples, as described in the legend of Fig. 2B). Means from two independent experiments are shown. Lines indicate group median RLU per microliter. (B and C) Correlation between neutralizing antibody (Ab) titers (NT₅₀) (B) or spike-specific memory B-cell expansion (C) and spike-binding antibodies (RLU per microliter) across all vaccine recipients. (D and E) Correlation between neutralizing antibody titers (NT₅₀) (D) or spike-specific memory B-cell expansion (E) and spike-binding antibodies (RLU per microliter) for each uninfected (black) or infected (red) vaccine recipient group. The r values indicate Spearman correlation coefficients.

FIG 6

Longitudinal analysis of neutralizing antibodies after vaccination. (A and B) Comparison of neutralizing antibody titers between the first and second plasma samples for each participant. (A) Uninfected participants; (B) SARS-CoV-2-infected participants (open circles, individuals who were seronegative for anti-N, as described in the legend of Fig. 2B). (C) Change in NT₅₀ values for participants who were infected between the times of acquisition of the 2 samples as indicated by the acquisition of, or a large increase in, anti-N antibodies. (D and E) Neutralizing antibody titers against SARS-CoV-2 variants (D) and SARS-CoV (E) in the second samples from participants who were infected between the times of collection of the two samples. One sample from the Ad5-nCoV group with a prior positive PCR test and anti-N antibodies in the second but not the first sample was included in this group. The medians from 2 independent experiments for each plasma sample are plotted. Dashed lines indicate the lowest plasma dilution tested (1:50). Horizontal lines indicate group median NT₅₀ values.