Surface electrostatic shift on spike protein decreased antibody activities against SARS-CoV-2 Omicron variant.

Dear Editor,

In this Journal, Pascarella and colleagues recently described the value of electrostatic potentials of the spike receptor binding and N-terminal domains in addressing transmissibility and infectivity of SARS-CoV-2 variants of concern. They interestingly found that the Omicron B.A1 variant has the highest net surface charge of SARS-CoV-2 receptor-binding domain (RBD), which may enhance affinity of Omicron RBD binding to the receptor angiotensin-converting enzyme 2 (ACE2). However, whether the protein surface electrostatic shift affect antibody activities is still unknown.

A vaccine's efficacy or effectiveness against SARS-CoV-2 infection usually ranges from 19% to 99%. However, some vaccines show notably lower efficacy, and the reason for this remains unknown. Marked reductions in neutralizing activity have been observed against Omicron relative to the ancestral pseudovirus in plasma from convalescent individuals and from individuals who had been vaccinated against SARS-CoV-2.3, 4, 5 The SARS-CoV-2 Omicron variant encodes 37 amino acid substitutions in the spike protein, 15 of which are in RBD, thereby raising concerns about the effectiveness of available vaccines and antibody-based therapeutics. In addition to these mutations, the other possible reasons for the considerable decline in the neutralizing activity against Omicron have yet to be documented.

In this study, we computed sequence-based antibody epitopes on spike proteins of SARS-CoV-2. Four epitopes with high surface accessible scores have been found and named #RBD, #CS (S1/S2 cleavage site), #S2–1 (S2 subunit-1) and #S2–2 (S2 subunit-2), respectively (Fig. 1 A and Supplementary Fig. 1, and Supplementary Table 1). Then these epitopes were synthesized chemically. The ascites production of monoclonal antibodies against these epitopes was generated by inoculation of mice. However, unlike the other three epitopes, the epitope #CS failed to generate any ascites antibodies, which may be because the cleavage impairs its antigenicity.

Fig. 1

Glycosylation affects antibody activities against SARS-CoV-2 original strain and Omicron variant. (A) Distribution of glycosylation sequons and antibody epitopes on SARS-CoV S. S1/S2 cleavage sites (CS) are marked with the dark purple color. The receptor-binding domain (RBD) is marked with the pale lavender color. Putative epitopes with different surface accessibilities (SA) are marked in yellow (SA 1.0–2.0), orange (SA 2.0–3.0), and red (SA > 3.0) respectively. CV30 binding residues conserved in all SARS-CoV-2 strains are marked with the sky-blue color. CV30 binding residues mutated in Omicron are marked with the dark blue color. Glycosylation sequons are marked with the green color. To present the sites more clearly, only one of the three monomers is labeled. (B) Endpoint (dilution) titers of mouse monoclonal antibodies against epitopes #RBD, #S2–1, and #S2–2 respectively. (C) Endpoint titers of human monoclonal antibodies against RBD, S1 subunit, and S2 subunit, respectively. (D) Endpoint titers of mouse anti-#RBD, anti-#S2–1, and anti-#S2–2 antibodies after PNGase F treatments. (E) Endpoint titers of human anti-RBD, anti-S1, and anti-S2 antibodies after PNGase F treatments. Bars represent standard deviations of three independent replicates. Values followed by different letters are significantly different at P < 0.05 according to Duncan's multiple range test.

Mouse monoclonal antibody against the epitope #RBD showed a relatively high endpoint (dilution) titer against the SARS-CoV-2 original strain (Fig. 1 B). A human neutralizing monoclonal antibody, CV30, , also in complex with the RBD, showed a much higher endpoint titer to the original strain than mouse anti-#RBD antibody (Fig. 1 C). This may be because CV30 binds more residues (32 residues) than the epitope #RBD (14 residues). However, endpoint titers of both antibodies against RBD were significantly lower when binding to the Omicron S protein, by a factor of 10 for CV30. This is consistent with a previous report stating that the neutralizing activity against Wuhan-Hu-1 and retained detectable neutralization against Omicron, with decreases about 21–39-fold. Five of 9 residues binding to CV30 at the C-terminal of RBD were found to be mutated in Omicron (Supplementary Fig. 1). This might be an important reason for its decreased binding activity.

The epitope #S2–1 with the highest SA score of 4.431 is located on the interface between subunits S1 and S2, which might be uncovered by transmembrane protease serine 2 (TMPRSS2) cleavage (Supplementary Fig. 2). However, neither TMPRSS2 nor its inhibitor Camostat affected antibody activity (Supplementary Fig. 2).

Unexpectedly, mouse monoclonal antibody against the epitope #S2–2 showed the highest endpoint (dilution) titer to the original strain (3 times higher than the antibody against #RBD; Fig. 1 B). Human monoclonal antibody against S2 subunit confirmed this finding (2 times higher than the antibody titer against S1 subunit; Fig. 1 C). The unexpectedly high activity of non-neutralizing antibodies against S2 subunit were consistent with the fact that the vaccine efficacies against severe disease are usually higher than 90%, no matter how low the vaccine efficacies against SARS-CoV-2 infection are. Nevertheless, activities of both anti-S2 antibodies declined 11–23-fold when binding with Omicron S protein (Fig. 1 BC).

Possible reasons for the dramatically reduced antibody activity against the Omicron variant were further investigated. Previous studies have suggested that N-linked glycosylation on the S protein may compromise its antibody activities. , Most SARS-CoV-2 epitopes are shielded by glycans, and only areas of the protein surface at the apex of the S1 domain (RBD region) are not surrounded by glycosylation sequons (Fig. 1 A). Consistent with the structure analysis, oligosaccharide-removing treatments involving PNGase F increased all antibody titers, especially for anti-S2 antibodies (Fig. 1 DE). However, after PNGase F treatments, antibody titers against Omicron were still much lower than those against the original strain (Fig. 1 DE).

Given that all the three epitopes #RBD, #S2–1, and #S2–2 are completely conserved in all SARS-CoV-2 strains (Supplementary Fig. 1), the large decline in the antibody activity against Omicron cannot be attributed to the mutations. We noticed that 11 residues had been substituted into the alkaline amino acid on Omicron S protein, which may change the electrostatic potential on the surface of the protein. We computed the electrostatic potential on the S protein. The surface of the Omicron S protein is uniformly positively charged. However, a large part of original SARS-CoV-2 S protein surface is electrically neutral or negatively charged, but its RBD is positively charged (Fig. 2 A). For the original strain, epitopes #RBD and #S2–2 distribute on electrically neutral areas, which are positively charged in the Omicron strain. When the pH value of the binding system was adjusted to 5.0 (the positive charge could be neutralized), antibody titers against Omicron increased exponentially. This effect was doubled for anti-RBD antibodies, 7–11-fold increases for anti-S2 antibodies (Fig. 2 BC). However, pH 9.0 decreased all antibody activity (Fig. 2 DE).

Fig. 2

Shift in electrostatic potential affect antibody activities against SARS-CoV-2 original strain and Omicron variant. (A) Electrostatic potential of SARS-CoV-2 S. The red-to-blue color on the molecular surface indicates the electrostatic potential (red: −1.8; blue: 1.8). Epitopes #RBD, #S2–1, and #S2–2 and CV30 binding residues are marked with the dark blue color. (B) Endpoint titers of mouse anti-#RBD, anti-#S2–1 and anti-#S2–2 antibodies at pH 5.0. (C) Endpoint titers of human anti-RBD, anti-S1, and anti-S2 antibodies at pH 5.0. (D) Endpoint titers of mouse anti-#RBD, anti-#S2–1, and anti-#S2–2 antibodies at pH 9.0. (E) Endpoint titers of human anti-RBD, anti-S1, and anti-S2 antibodies at pH 9.0. Bars represent standard deviations of three independent replicates. Values followed by different letters are significantly different at P < 0.05 according to Duncan's multiple range test.

The large decline in antibody activity against Omicron RBD (neutralization) may be attributed to both residue mutations and the shift in electrostatic potential. The large decline in antibody activity against the Omicron S2 subunit (non-neutralizing antibody activity) may be mainly attributable to a shift in electrostatic potential on the surface of the protein. Because the heterogeneity in antigen surface-charge distribution causes charge-related heterogeneity in monoclonal antibodies, new vaccines against the full-length Omicron S protein may be developed that have both negatively charged neutralizing antibodies and negatively charged non-neutralizing antibodies with high affinity to the Omicron variant.

Declaration of Competing Interest

The authors declare no competing interests.