SARS-CoV-2 BA.1 variant is neutralized by vaccine booster-elicited serum but evades most convalescent serum and therapeutic antibodies

Abstract

The rapid spread of the highly contagious Omicron variant of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) along with its high number of mutations in the spike gene has raised alarms about the effectiveness of current medical countermeasures. To address this concern, we measured the neutralization of the Omicron BA.1 variant pseudovirus by postvaccination serum samples after two and three immunizations with the Pfizer/BioNTech162b2 SARS-CoV-2 mRNA (Pfizer/BNT162b2) vaccine, convalescent serum samples from unvaccinated individuals infected by different variants, and clinical-stage therapeutic antibodies. We found that titers against the Omicron variant were low or undetectable after two immunizations and in many convalescent serum samples, regardless of the infecting variant. A booster vaccination increased titers more than 30-fold against Omicron to values comparable to those seen against the D614G variant after two immunizations. Neither age nor sex was associated with the differences in postvaccination antibody responses. We also evaluated 18 clinical-stage therapeutic antibody products and an antibody mimetic protein product obtained directly from the manufacturers. Five monoclonal antibodies, the antibody mimetic protein, three antibody cocktails, and two polyclonal antibody preparations retained measurable neutralization activity against Omicron with a varying degree of potency. Of these, only three retained potencies comparable to the D614G variant. Two therapeutic antibody cocktails in the tested panel that are authorized for emergency use in the United States did not neutralize Omicron. These findings underscore the potential benefit of mRNA vaccine boosters for protection against Omicron and the need for rapid development of antibody therapeutics that maintain potency against emerging variants.

Fig. 1.

Sensitivity of the Omicron variant to neutralization by Pfizer/BNT162b2 vaccinee serum samples. (A) Neutralization assays used lentiviral pseudoviruses bearing SARS-CoV-2 spike proteins from D614G, Delta, or Omicron. Serum samples from 39 healthcare workers were obtained at a mean of 30 ± 11 or 43 ± 17 days after the 2^nd or 3^rd immunization, respectively. The geometric means titers (GMT), the number of serum samples above threshold (1:40, the lowest dilution tested, dotted lines), and the fold change are indicated. Titers below 1:40 were set at 20 to calculate GMTs. Black arrows indicate fold decrease relative to D614G after the 2^nd or 3^rd immunization. Purple arrows indicate fold increase for each variant after the 3^rd vaccination compared to the 2^nd vaccination. Connecting lines indicate serum from the same individual. Demographic information is provided in Table 1. (B) NT₅₀ values are shown by sex after 2^nd or 3^rd vaccinations. (C) NT₅₀ values are shown by age after 2^nd and 3^rd vaccinations. (D) NT₅₀ values are shown after 2^nd or 3^rd vaccination according to anti-N (nucleocapsid protein) seroconversion between 2^nd and 3^rd vaccinations. Presence of anti-N antibodies suggests prior SARS-CoV-2 infection. (E) Ratio of NT₅₀ values after 3^rd versus 2^nd immunizations are shown for D614G, Delta, and Omicron. The mean of ratios is shown below each variant. Black squares correspond to D614G, blue triangles correspond to Delta, and red circles correspond to Omicron. Data shown represent two independent experiments each with an intra-assay duplicate. Significance for (A) and (E) was assessed by one-way ANOVA with the Geisser-Greenhouse correction, followed by Tukey’s multiple comparison test. Significance for (B) and (D) was assessed by Kruskal-Wallis test, followed by Dunn’s multiple comparison test. Significance values are indicated as *P < 0.05; **P < 0.01; ***P < 0.001; ns, nonsignificant.

Fig. 2.

Sensitivity of the Omicron variant to neutralization by convalescent serum samples. Neutralization assays were performed using convalescent serum samples from individuals infected with genotyped variants from B.1, B.1.2, B.1.1.7, B.1.351, B.1.617.2, AY.14, AY.25, AY.44, AY.47, AY.62, AY.74 or AY.119 lineages (Table 2, tables S1 and S2). B.1 and B.1.2 have no mutations in the receptor binding domain and were therefore considered D614G, whereas some of the AY mutants have additional non-RBD spike mutations relative to B.1.617.2. Titers above threshold (1:40) and fold changes are indicated. Titers below threshold were set as 20 for GMT calculations. Arrows indicate decrease relative to D614G (for D614G, Alpha and Beta serum samples) or Delta (for Delta serum samples). Connecting lines indicate serum from the same individual. Data shown represent two independent experiments, each with an intra-assay duplicate. Squares correspond to D614G, triangles correspond to Delta, and circles correspond to Omicron. Significance was assessed using a Kruskal-Wallis test followed by a Dunn’s post-test. Significance values are indicated as *P < 0.05; **P < 0.01; ***P < 0.001; ns, nonsignificant.

Fig. 3.. Antigenic cartography of convalescent and vaccinee serum samples against D614G, Delta, and Omicron variants.

Antigenic maps were generated from (A) convalescent, (B) 2^nd vaccination, or (C) 3^rd vaccination serum samples. Convalescent serum samples are shown in open diamonds as follows: B.1 (black), B.1.2 (gray), B.1.1.7 (magenta), B.1.351 (green), AY variants (light blue), and B.1.617.2 (dark blue). Solid gray diamonds correspond to post-vaccination serum. Each grid square (1 antigenic unit) corresponds to a 2-fold dilution in the neutralization assay. Antigenic distance is interpretable in any direction. Black squares correspond to D614G variant. Blue triangles correspond to the Delta variant. Red circles correspond to the Omicron variant. The confidence area of the position of individual viruses or serum sample displayed in the bottom panels was estimated with stress parameter 0.5 and is shown as a rounded shape in the maps.

Fig. 4.

Neutralization of the Omicron variant pseudovirus by therapeutic antibodies. (A to C) Neutralization curves against D614G (black) and Omicron (red) are shown for eleven monoclonal neutralizing antibodies (nAb) and one tri-specific antibody mimetic protein (DARPin) (A), five cocktail neutralizing antibody products (B), and two polyclonal antibody preparations (C). (D) IC₅₀ values against D614G and Omicron are shown for all therapeutic antibodies. IC₅₀ values were calculated from two independent experiments, each with an intra-assay duplicate.