Recall of pre-existing cross-reactive B cell memory following Omicron BA.1 breakthrough infection

Understanding immune responses following severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) breakthrough infection will facilitate the development of next-generation vaccines. Here, we profiled spike (S)-specific B cell responses following Omicron/BA.1 infection in mRNA-vaccinated donors. The acute antibody response was characterized by high levels of somatic hypermutation (SHM) and a bias toward recognition of ancestral SARS-CoV-2 strains, suggesting the early activation of vaccine-induced memory B cells (MBCs). BA.1 breakthrough infection induced a shift in B cell immunodominance hierarchy from the S2 subunit, which is highly conserved across SARS-CoV-2 variants of concern (VOCs), and toward the antigenically variable receptor binding domain (RBD). A large proportion of RBD-directed neutralizing antibodies isolated from BA.1 breakthrough infection donors displayed convergent sequence features and broadly recognized SARS-CoV-2 VOCs. Together, these findings provide insights into the role of pre-existing immunity in shaping the B cell response to heterologous SARS-CoV-2 variant exposure.

Other Supplementary Material for this manuscript includes the following: Table S3. Sequence, binding, and neutralization properties of isolated antibodies. (Excel datasheet) Table S4. Raw data. (Excel datasheet)

Ethics permits and sample collection.
Blood samples among breakthrough infection donors and uninfected/two-dose vaccinated donors were collected at Dartmouth-Hitchcock Hospital. Venous blood was collected using BD Vacutainer® tubes with acid citrate dextrose (ACD), and plasma and PBMCs were isolated using a Ficoll 1077 (Sigma) gradient, washed, and counted with an anti-human CD45 stain on a volumetric flow cytometer. PBMC were frozen in 12.5% human serum and 10% DMSO diluted in RPMI-1040 and stored in liquid nitrogen until use. Plasma was isolated and frozen at -80 ˚C.
Venous blood was collected from uninfected/three-dose vaccinated donors at Umeå University. PBMCs and plasma isolated in BD EDTA Vacutainer® CPT™ tubes. PBMCs were frozen in 90% fetal calf serum supplemented with 10% DMSO and stored in liquid nitrogen until use. Plasma and serum were stored at -80 ˚C.

Pseudovirus neutralization assay.
HeLa-hACE2 reporter cells (BPS Bioscience Cat #79958) were seeded overnight at 10,000 cells per well in 96-well tissue culture plates (Corning). Human plasma and serum samples were heat-inactivated at 56 ˚C for 30 min. Next, monoclonal antibodies or heatinactivated sera were serially diluted in MEM/EBSS media supplemented with 10% FBS with 50 µl of MLV viral stock and incubated for 1 h at 37 ˚C with 5% carbon dioxide. Cell culture media was removed, and cells were washed two times with PBS. The virus-antibody mixture was subsequently added to HeLa-hACE2 cells and incubated for 48 h at 37 ˚C with 5% carbon dioxide. Cells were then lysed with Luciferase Cell Culture Lysis 5× reagent (Promega), and luciferase activity was measured using the Luciferase Assay System (Promega) following manufacturer's protocols. Infectivity was measured as relative luminescence units (RLUs) using a luminometer (Perkin Elmer). The percentage neutralization was calculated as 100*(1-[RLUsample-RLUbackground]/[ RLUisotype control mAb-RLUbackground]), and the 50% neutralization concentration was interpolated from four-parameter non-linear regression fitted curves in GraphPad Prism (version 9.3.1).
The proportion of class-switched RBD-specific B cells that reacted with WT and/or BA.1 RBD was calculated by dividing the number of BA.1/WT cross-reactive or WT-specific IgG + and IgA + (swIg + ) B cells by the total number of RBD + S + swIg + B cells. The proportion of Sreactive B cells that recognized each subdomain (NTD, RBD, or S2) was calculated by dividing the number of IgG + and IgA + (swIg + ) B cells that recognize both S and the subdomain by the total number of S + swIg + cells.

Binding affinity measurements by biolayer interferometry.
All steps were performed at 25 o C and at an orbital shaking speed of 1000 rpm. All reagents were formulated in PBSF buffer (PBS with 0.1% w/v BSA). Recombinant biotinylated antigens were diluted (100 nM) in PBSF and loaded onto streptavidin biosensors (Sartorius) to a sensor response of 0.6-1.0 nm and then allowed to equilibrate in PBSF for a minimum of 30 min. After a 60 s baseline step in PBSF, antigen-loaded sensors were exposed (180 s) to Fab or IgG fragments (100 nM) and then dipped (180 s) into PBSF to measure any dissociation of the antigen from the biosensor surface. Fab binding data with detectable binding responses (>0.1 nm) were aligned, inter-step corrected (to the association step) and fit to a 1:1 binding model using the FortéBio Data Analysis Software, version 11.1.

Epitope binning by biolayer interferometry.
Antibody competition with recombinant human ACE2 and comparator antibodies for binding to SARS-CoV-2 RBD was determined by BLI using a ForteBio Octet HTX (Sartorius). All binding steps were performed at 25 o C and at an orbital shaking speed of 1000 rpm. All reagents were formulated in PBSF (1X PBS with 0.1% w/v BSA). For ACE2 competition experiments, test antibodies (100 nM) were captured onto anti-human IgG capture (AHC) biosensors (Molecular Devices) to a sensor response of 1.0 nm-1.4 nm. IgG-loaded sensors were then soaked (20 min) in an irrelevant IgG1 solution (0.5 mg/ml) to block remaining Fc binding sites, followed by a 30 min incubation in PBSF. To assess any potential cross interactions between sensor-loaded IgG and ACE2, the IgG-loaded and blocked sensors were exposed (90 s) to a 300 nM ACE2 (Sino Biological, Cat# 10108-H08H). Sensors were next allowed to baseline (60 s) before exposing (180 s) to recombinant SARS-CoV-2 RBD (100 nM; Acro Biosystems, Cat # SPD-C52H3) and then exposed (180 s) to ACE2 (300 nM). Increased sensor responses following ACE2 exposure represented a non-ACE2-competitive binding profile, whereas antibodies showing unchanged sensor responses were designated as ACE2-competitive. Antibody competition with comparator antibodies (REGN10933, ADI-62113, COV2-2130, REGN10987, and S309) was performed using the same method as described above but with a different assay orientation: comparator antibodies were captured to anti-human IgG capture biosensors (Molecular Devices) and then exposed to antibodies of interest (300 nM) in solution.

Fig. S3. FACS gating strategy for SARS-CoV-2 antigen-specific B cell staining. (A)
Representative FACS gating strategy to determine frequencies of WT-and BA.1-RBD-reactive B cells among IgG + and IgA + B cells. (B) Representative FACS gating strategy to determine the proportion of CD71 + S + RBD + cells that are WT-specific or BA.1/WT cross-reactive. (C) Representative FACS gating strategy used to calculate the proportion S-specific B cells directed to the NTD, RBD, and S2 subdomains. FSC-A, forward scatter area; FSC-H, forward scatter height; SSC-A, side scatter area. Statistical comparisons were determined by Kruskal-Wallis test with subsequent Dunn's multiple comparisons and two-sided Mann Whitney U test for two-dose and three-dose vaccinated individuals, respectively. 1M, one month; 6M, six months; WT, wild type. *P < 0.05, **P < 0.01.       (22). Statistical comparisons were made by Fisher's exact test compared to the baseline repertoire. IGHV, immunoglobulin heavy variable domain. **P < 0.01, ***P < 0.001, ****P < 0.0001.