Researchers computationally design immunogens that elicit antibodies against different SARS-CoV-2 variants

A recent study published in PLOS Computational Biology We designed immunogens to induce antibodies against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants.

Background

SARS-CoV-2 Vaccine Is Highly Effective, But Several Influencing Mutants Are Emerging efficacy of available vaccine. Current vaccines are effective in preventing severe disease if infected with mutant variants, but their emergence suggests that more immune-evading variants may emerge in the future. In addition, other pathogenic CoVs may evolve for zoonotic diseases. Therefore, vaccines against potential SARS-CoV-2 variants and other zoonotic CoVs are needed to combat future outbreaks.

About research

In the current study, researchers computationally designed immunogens to elicit antibodies against different SARS-CoV-2 variants. Structural and sequence data were leveraged to identify CoV variable and conserved residues. spike proteinEach spike residue was assigned a conserved fraction that was the average of the biochemically and structurally conserved fractions.

Structural conservation percentages were calculated using structure-based sequence alignments with a specific CoV as a reference. It was some of the other CoVs that were structurally aligned at specific residues. Biochemical conservation rates were calculated from multiple sequence alignments using specific CoVs as references. This was the proportion of sequences with amino acid residues of the same class (hydrophobicity, polarity, etc.) as the consensus residue of the reference CoV.

It was hypothesized that a cocktail of variant spike sequences could induce polyclonal responses to protect against SARS-CoV-2 variants. The idea was to elicit strain-specific antibodies that could protect against multiple variants rather than broad spectrum. Neutralizing antibody (bnAbs), which target conserved epitopes. For this purpose, a list of 37 mutations containing 10 residues was generated. Most were spatially close to the receptor binding domain (RBD) of the spike protein.

Researchers trained a neural network using RBD with spatially separated mutations. A neural network was trained using a dataset of approximately 169,000 her RBD variants with measured formulas. In addition, he used a dataset of approximately 135,000 RBD mutants to train another neural network to predict ACE2 binding affinities. A neural network predicted the expression of the antigen of interest and its ACE2 binding properties.

A coarse-grained stochastic model was used to simulate the processes occurring during affinity maturation (AM). A computational AM model evaluated how (immunizing) antibodies are generated in response to wild-type (WT) spikes and engineered antigens. The main goal was to identify the most effective vaccination scheme against the variant, using designed antigens, than vaccination with the WT spike alone.

findings

Six antigens (sequences 1 to 6) were designed using conservation analysis. Neural networks showed high correlation on the test set. Therefore, the predicted binding affinities and RBD expressions of the engineered antigens were compared with the circulating variants. The engineered antigens did not significantly reduce binding affinity or expression over the corresponding circulating variants.

Although the antigen was designed before the emergence of omicron variants, mutations at residues 417, 478, 484, 493, and 501 are common to engineered antigens and omicron variants. AM was simulated following one or two WT spike immunizations. Anti-WT spike titers were high after one or two WT spike immunizations. Still, titers against variant spike proteins decreased according to the mutational distance between variant and WT spikes.

Simulations were then performed to find the optimal vaccination scheme using the designed antigens. The authors performed 1) 1 immunization with all 6 designed antigens, 2) 2 immunizations with 6 antigens, 3) 1 immunization with 3 antigens (sequences 1–3), and 4) simulated immunizations with three antigens (sequences 1-3). , followed by sequences 4-6.

Antigen cocktail immunizations had lower anti-WT titers than WT immunizations because the cocktail antigens were mutated from the WT spike. Yet, two immunizations with cocktail antigens still retained moderately high anti-WT titers. Anti-beta, gamma, and delta titers were higher with cocktail antigens than with WT. The authors noted that immunization with cocktail antigens after exposure to WT spike increased titers more than boosting with WT.

Anti-alpha, delta, and WT antibody titers were higher with two WT immunizations than with hybrid immunizations that included one exposure to WT and the other to cocktail antigens. This is due to the low mutation of these variants. However, anti-beta and anti-gamma titers were comparable between the two WT immunizations and the hybrid immunization scheme (WT | cocktail).

Finally, researchers calculated spike conservation across 12 CoVs from multiple genera to identify targets for a pan-CoV vaccine. None of the selected CoVs had conserved sequences in the RBD, as different CoVs have different host cell receptors. However, the S2 domain of the spike has various residues that are structurally conserved in many CoVs.

Residues are conserved across multiple CoVs as they overlap the fusion peptide and the S2′ cleavage site. These sites are functionally important and should be conserved. Developing antibodies against the stem of the CoV spike is therefore feasible and may serve as a pan-CoV vaccine strategy.

Conclusion

The researchers designed six antigens based on analysis of sequence and structural data and found through neural networks that the antigens were viable and capable of binding to ACE2. They showed that an antigen cocktail provided stronger immunity against pre-existing SARS-CoV-2 variants than WT vaccination, and predicted that a cocktail of six antigens would be optimal from several possibilities. Did.