The SARS-CoV-2 Omicron (B.1.1.529) variant of concern is now prevalent in large parts of the world. It has been divided into three main sub-lineages BA.1, BA.2 and BA.3, which are characterized by a heavily mutated spike protein (Supplementary Fig. 1), leading to substantial escape from antibodies induced by previous infections and/or vaccinations 1,2. Patients with primary BA.1 infections were shown to mount strongly reduced neutralizing antibody responses to strains circulating before Omicron 3–6, although vaccine-boosting regimens led to efficient cross-neutralization that may protect from disease 7–9. The rapid increase in BA.2 infections, which is currently replacing BA.1 as the dominant variant 1, has raised the possibility that it may be more transmissible than BA.1 10 and/or may escape antibody-mediated immunity, potentially including the protection gained from BA.1 infections 3.
Here, we report antibody neutralization data of Omicron BA.1, BA.2, and Delta variants in serum samples collected from individuals who had recovered from primary Omicron BA.1 or BA.2 infection or who had been vaccinated with or without previous infection. We analyzed the neutralizing capacity of serum samples obtained after primary wildtype (WT), Omicron BA.1 and BA.2 infections against a WT strain (isolated early in the pandemic with the D614G mutation) and the three variants of concern: Delta; Omicron BA.1; and Omicron BA.2. In addition, we tested samples from individuals who were infected with an Omicron variant following prior COVID-19 vaccination (breakthrough infection), and from individuals who had received three doses of a vaccine with and without prior infection. The characteristics of these cohorts are summarized in Table 1 and the Supplementary Tables 1 and 2.
Table 1
Cohorts and samples analyzed in this study.
| | | Median [IQR]a NT titer |
| | n | WT | Delta | Omicron BA.1 | Omicron BA.2 |
| Vaccinated (3 doses, mRNA vaccine) |
| 3–4 weeks | 15 | 640 (480–1120) | 320 (120–400) | 160 (80–280) | 160 (100–240) |
| 3 months | 15 | 160 (120–560) | 120 (70–280) | 40 (25–90) | 60 (30–100) |
| Infected (WT + 3 doses mRNA vaccine) |
| 3–4 weeks | 9 | 640 (480–1280) | 480 (320–640) | 120 (80–320) | 240 (160–320) |
| Infected (WT) |
| 3–4 weeks | 11 | 640 (400–640) | 160 (80–280) | 30 (25–40) | 40 (35–70) |
| 6 months | 11 | 160 (100–280) | 60 (40–100) | 20 (15–25) | 30 (20–35) |
| Infected (Omicron BA.1) |
| 3–4 weeks | 18 | ≤ 10 | ≤ 10 | 60 (20–60) | 20 (10–40) |
| Infected (Omicron BA.2) |
| 3–4 weeks | 7 | ≤ 10 | ≤ 10 | ≤ 10 | 30 (12.5–40) |
| Omicron breakthroughs |
| 3–4 weeks | 11 | 640 (480–960) | 480 (320–640) | 320 (280–800) | 480 (240–800) |
| aIQR = interquartile range |
Serum samples obtained from vaccinees three weeks and three months after the third dose of an mRNA vaccine (Table 1 and Supplementary Table 1) efficiently cross-neutralized Omicron variants. Neutralization titers were significantly lower than for WT (Fig. 1a; Table 1), but there was no significant difference between BA.1 and BA.2 (Mann-Whitney test, p > 0.05). We also detected Omicron cross-neutralization in individuals who had a WT infection before being vaccinated three times (Fig. 1b).
Samples obtained from primary infections with WT, Omicron BA.1, and Omicron BA.2 neutralized the heterologous strain to a much lesser extent than the homologous virus (Fig. 1c-e). Furthermore, Omicron NT titers were significantly reduced in samples obtained 3 weeks and 6 months after infection with the ancestral WT strain (Fig. 1c). Sera from Omicron BA.1 convalescents neutralized WT and Delta to a lesser extent, and even BA.2 neutralization was significantly lower (Fig. 1d). In contrast, samples from Omicron BA.2 convalescents did not cross-neutralize any other virus strain tested (Fig. 1e). However, all samples obtained from individuals after an Omicron infection who had been previously vaccinated (breakthrough infection) were able to neutralize the two Omicron variants as efficiently as the WT and Delta viruses (Fig. 1f).
Consistent with the antigenic equidistance of BA.1 and BA.2 from the original SARS-CoV-2 strain 3, we show that neutralizing antibodies present in serum samples from primary Omicron patients are highly variant-specific. While samples from individuals after primary BA.1 infection exhibited some cross-neutralization of BA.2 (Fig. 1d), samples from primary BA.2-infected individuals showed no cross-neutralizing capabilities against any other variant (Fig. 1e). The two strains have 22 amino acid mutations in the spike protein in common, however there are considerable differences in the N-terminal domain (NTD), which is recognized by neutralizing antibodies 11 (Supplementary Fig. 1). Changes in the NTD might also affect its packing contacts with the receptor-binding domain (RBD)12, the major target of neutralizing antibodies 13,14.
Multiple exposures to pre-Omicron-SARS-CoV-2, via three vaccinations (Fig. 1a) or mixed infection-vaccination scenarios (Fig. 1b), led to high titers with efficient neutralization of both Omicron variants. These data are in agreement with other studies that suggested an increase in the magnitude as well as the breath of neutralizing antibody responses by repeated exposure to the original antigen 3,4,7,8,15,16. The Omicron NT titers, however, were lower in these cohorts than against pre-Omicron variants and a fast waning of neutralizing antibodies was observed (Fig. 1a).
Our data clearly show that primary infections with the currently circulating Omicron variants BA.1 and BA.2 result in sub-lineage-specific neutralization. With respect to the co-circulation as well as emergence of variants, our study additionally highlights the importance of booster vaccinations in immune protection.