RR:C19 Evidence Scale rating by reviewer:
SARS-CoV-2 variants escape to varying degrees from neutralizing antibodies induced by current vaccines against the spike (S) protein of the original Wuhan-1 strain. This is especially the case for the recently emerged Omicron variant and its sublineages, resulting in substantial numbers of breakthrough infections in vaccinated individuals. Nevertheless, most Omicron breakthrough infections have been mild or asymptomatic, suggesting that vaccination generates strong protection beyond neutralizing antibody induction. Recent reports, including that of Kaku and colleagues, indicate that cross-reactive memory B cells (MBCs) generated by SARS-CoV-2 vaccination respond to variant viruses and play a key role in the protective response to breakthrough infection.
Kaku et al. provide a comprehensive profile of the S-specific B cell response after PCR-confirmed breakthrough infection in mRNA-vaccinated subjects (n = 7). The infecting virus was not sequenced, but was likely to be Omicron/BA.1. Sera and peripheral blood mononuclear cells were collected at a single time point 2-4 weeks after infection. Comparison was with mRNA-vaccinated, but uninfected subjects sampled 1 or 6 months after the last vaccine dose.
Serum antibody analysis (ELISA, neutralization assay) demonstrated that breakthrough infection generated IgG and IgA against the S protein receptor-binding domain (RBD) of BA.1 and neutralizing activity against BA.1 and other SARS-CoV-2 variants. Although not stated in the paper, an important point is that mRNA vaccination generated BA.1 RBD-binding antibodies and BA.1 neutralizing antibodies, suggesting that BA.1 RBD-reactive MBCs would also be generated. The flow cytometric data is important and utilizes probes to identify B cells reactive with regions of the S protein. However, this component of the work could be better described. The authors refer to RBD-reactive IgG+ or IgA+ B cells. Are these to be considered MBCs? If not, what are they? The MBC marker CD27 is listed in the flow panel, but it does not appear in the gating strategy to identify CD19+ IgG+ S+ RBD+ B cells (for example). The gating strategy is hard to follow and some dot plots have peculiar features. Otherwise, the results nicely demonstrate (i) generation of class-switched MBCs by mRNA vaccination that are cross-reactive with WT and BA.1 RBD, and (ii) expansion of the proportion of cross-reactive versus WT RBD-specific MBCs by breakthrough infection, consistent with activation of cross-reactive MBCs. The analysis of the proportions of S-reactive MBCs that recognize different regions of S (RBD, NTD, or S2) after breakthrough infection versus mRNA vaccination is interesting. Breakthrough infections increase the proportion of RBD-reactive cells and decrease the proportion of S2-reactive cells. The authors suggest that infection redirected the B cell immunodominance hierarchy from S2 to RBD. This statement could be misleading. It was the presence of RBD-reactive MBCs generated by mRNA vaccination that mediated an RBD-focused rather than an S2-focused B cell response to breakthrough infection.
Kaku et al. also analyzed a large number of antibodies cloned from class-switched RBD+ B cells collected after breakthrough infection. Interestingly, the vast majority of antibodies were BA.1/WT RBD cross-reactive, indicating selection by BA.1 infection. Most of the antibodies also contained somatic mutations, consistent with selection in germinal centers and an MBC origin. Sequence analysis of the cloned antibodies identified both known and novel germline gene families strongly represented in the response to BA.1 Generally, the cloned antibodies had broad binding and neutralizing activity against SARS-CoV-2 variants. Competition ELISAs were used to investigate binding to RBD antigenic sites. Kaku et al. have generated important data that add to our picture of the S-reactive B cell response after breakthrough infection in mRNA vaccinated subjects. The work is well-performed and key conclusions well-supported by the data. Overall, findings indicate that mRNA vaccines generate RBD-reactive MBCs that respond to variant viruses, including Omicron, and rapidly generate protective antibodies. Importantly, anti-RBD antibodies generated in the response to breakthrough infection were broadly-reactive with a range of SARS-CoV-2 variants, suggesting that boosting with heterologous S protein could enhance broad and potent B cell immunity to coronaviruses. The long-term persistence of antigen-reactive MBC populations gives confidence that these cells provide durable protection.
Since our solicitation of reviews, this preprint has been published in Science Immunology journal and the link to the published manuscript can be found here.