Neutralization of BQ.1, BQ.1.1, and XBB with RBD-Dimer Vaccines

To the Editor: Several B.1.1.529 (omicron) subvariants of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) are emerging and have become the dominant strains, such as BF.7, BQ.1, BQ.1.1, and XBB. These variants contain more mutations in the spike protein receptor-binding domain (RBD) than the BA.2 and BA.5 strains (Figure 1A). Therefore, the potential of these omicron subvariants for immune evasion is a concern.

Figure 1. Vaccine-Elicited Neutralizing Antibodies against SARS-CoV-2 Variants in Human and Murine Serum Samples.

Panel A shows the mutations in the spike protein of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants analyzed in this study. BA.4 and BA.5 have the same spike protein sequence. The amino acid differences — including substitutions, deletions (blank squares), and insertions — in variant sequences from those in the SARS-CoV-2 prototype (PT) isolate are highlighted in gold. The third and fourth types of amino acid on the same locus are highlighted in red and blue, respectively. The amino acid in variant sequences identical to those in the PT are highlighted in the background color (light green). NTD denotes N-terminal domain, RBD receptor-binding domain, S2 S2-protein subunit, and SP signal peptide. The 50% pseudovirus neutralizing titers (pVNT₅₀) are shown for the ZF2001 group (16 vaccinees who received three injections of the ZF2001) in Panel B, the inactivated-vaccine group (16 vaccinees who received three doses of inactivated vaccine) in Panel C, and the heterologous-boost group (16 vaccinees who received two doses of inactivated vaccine plus a booster dose of ZF2001) in Panel D. Panel E shows the RBD-dimer immunogens that were evaluated in mice. The details of the experiments in mice are described in the Supplementary Appendix. The pVNT₅₀ in murine serum samples elicited by RBD-dimer vaccines are shown in Panel F for prototype RBD homodimer, in Panel G for the BA.2 RBD homodimer, in Panel H for the prototype–B.1.351 (beta) RBD heterodimer, in Panel I for the B.1.617.2 (delta)–BA.1 RBD heterodimer, and in Panel J for the delta–BA.2 RBD heterodimer. The pVNT₅₀ is shown as the geometric mean titer at the top in each panel. The 𝙸 bars indicate 95% confidence intervals. The dashed horizontal line indicates the limit of detection (LOD). A pVNT₅₀ value below the LOD was determined as half the LOD.

The coronavirus disease 2019 (Covid-19) vaccines with inactivated virus (CoronaVac and BBIBP-CorV) and the ZF2001 protein subunit vaccine¹ have been widely used in China and many other countries. In this study, we obtained serum samples from vaccinees (age range, 18 to 59 years) who had received three homologous doses of ZF2001 (the ZF2001 group, 16 participants) or inactivated vaccine (the inactivated-vaccine group, 16 participants) or two doses of inactivated vaccine plus a booster dose of ZF2001 (the heterologous-boost group, 16 participants) (Fig. S1 in the Supplementary Appendix, available with the full text of this letter at NEJM.org). The neutralizing activities in the serum samples from these participants were analyzed with a panel of pseudoviruses. Details of the study design are provided in the Supplementary Appendix.

In the ZF2001 group, three doses of ZF2001 induced high titers of neutralizing antibodies against the prototype and BA.2 strains (geometric mean titer [GMT], 1302 and 252, respectively), a finding that was consistent with the results of our previous study.² As compared with the neutralizing GMT of 181 against BA.4 and BA.5 (which have the same spike protein sequence), the neutralizing GMTs against BA.4.6 and BF.7 (which have an additional R346T mutation in the RBD) decreased to 67 and 63, respectively. Seropositivity (defined as a neutralization titer >10 [the limit of detection]) was 44% against BQ.1, 13% against BQ.1.1, and 13% against XBB, and the GMTs were below the limit of detection, indicating immune evasion of these more recent strains (Figure 1B, Figs. S2 and S3, and Table S1).

In the inactivated-vaccine group, the neutralizing GMT was 131 against the prototype isolate, 29 against BA.2, and 26 against BA.4 and BA.5. However, neutralization activities against BA.2.3.20, BA.4.6, BF.7, BQ.1, and BQ.1.1 were at background levels, and no serum sample showed neutralization activity against XBB (Figure 1C).

In the heterologous-boost group, the neutralization profile of the serum samples was similar to that in the ZF2001 group. The neutralizing GMT was 812 against the prototype isolate, 60 against BA.2, 15 against BA.2.3.20, 27 against BA.4 and BA.5, 20 against BA.4.6, and 22 against BF.7. In addition, seropositivity for neutralizing antibodies was 13% against XBB, 13% against BQ.1, and 13% against BQ.1.1 (Figure 1D). These results indicated that BQ.1, BQ.1.1, and XBB showed strong resistance to the vaccine-elicited humoral immunity, a finding that was consistent with the results of another recent analysis.³

Recently, we engineered the homotypic RBD-dimer immunogen⁴ (which is used in the ZF2001 vaccine) and designed heterotypic chimeric RBD dimers, including prototype–B.1.351 (beta) and B.1.617.2 (delta)–omicron vaccine candidates. They induced relatively broad immune responses against SARS-CoV-2 variants.⁵ In the current study, we tested their neutralizing activities in murine serum samples. BA.2 RBD homodimer and delta–BA.2 RBD heterodimer were also included as comparators (Figure 1E).

The results showed that XBB and BQ.1.1 strongly escape the antibody responses induced by the prototype RBD homodimer (Figure 1F and Fig. S4). For the BA.2 RBD homodimer, the neutralizing GMTs were more than 10⁵ against BA.1, BA.2, and BA.4 and BA.5; between 10⁴ and 10⁵ against the prototype isolate, beta, delta, BA.2.75, BA.2.3.20, BA.4.6, and BF.7; approximately 10³ against BQ.1 and BQ.1.1; and 393 against XBB (Figure 1G). The prototype–beta RBD heterodimer could induce high neutralizing GMTs against the currently circulating BF.7, which probably resulted from the K417N, E484K, and N501Y substitutions in beta RBD (Figure 1H). In comparison, delta–BA.1 and delta–BA.2 RBD heterodimers induced balanced neutralization profiles against the early circulating strains (such as the prototype, beta, and delta) and omicron subvariants (Figure 1I and 1J).

The currently circulating omicron subvariants, especially BQ.1, BQ.1.1, and XBB, showed immune escape to the humoral immunity elicited by prototype strain sequence-based vaccines, such as inactivated vaccine and ZF2001. Our study showed that next-generation and updated Covid-19 vaccines are needed for better protection and pandemic control. Our newly updated delta–omicron BA.1 and BA.2 RBD heterodimers had high neutralizing activities against the emerging omicron subvariants. An analysis of clinical data regarding the delta–omicron BA.1 RBD-dimer vaccine ZF2202 is currently under way (ClinicalTrials.gov number, NCT05616754).

This letter was published on March 1, 2023, at NEJM.org.