Omicron: A Heavily Mutated SARS-CoV-2 Variant Exhibits Stronger Binding to ACE2 and Potently Escapes Approved COVID-19 Therapeutic Antibodies

Abstract

The new SARS-CoV-2 variant of concern "Omicron" was recently spotted in South Africa and spread quickly around the world due to its enhanced transmissibility. The variant became conspicuous as it harbors more than 30 mutations in the Spike protein with 15 mutations in the receptor-binding domain (RBD) alone, potentially dampening the potency of therapeutic antibodies and enhancing the ACE2 binding. More worrying, Omicron infections have been reported in vaccinees in South Africa and Hong Kong, and that post-vaccination sera poorly neutralize the new variant. Here, we investigated the binding strength of Omicron with ACE2 and monoclonal antibodies that are either approved by the FDA for COVID-19 therapy or undergoing phase III clinical trials. Computational mutagenesis and free energy perturbation could confirm that Omicron RBD binds ACE2 ~2.5 times stronger than prototype SARS-CoV-2. Notably, three substitutions, i.e., T478K, Q493K, and Q498R, significantly contribute to the binding energies and almost doubled the electrostatic potential (ELE) of the RBD^Omic-ACE2 complex. Omicron also harbors E484A substitution instead of the E484K that helped neutralization escape of Beta, Gamma, and Mu variants. Together, T478K, Q493K, Q498R, and E484A substitutions contribute to a significant drop in the ELE between RBD^Omic-mAbs, particularly in etesevimab, bamlanivimab, and CT-p59. AZD1061 showed a slight drop in ELE and sotrovimab that binds a conserved epitope on the RBD; therefore, it could be used as a cocktail therapy in Omicron-driven COVID-19. In conclusion, we suggest that the Spike mutations prudently devised by the virus facilitate the receptor binding, weakening the mAbs binding to escape the immune response.

Keywords: ACE2; Omicron; SARS-CoV-2; antibodies; immune escape; therapeutic.

Figure 1

(A) Phylogeny of the Omicron and annotation of the mutation in Spike protein. The unrooted phylogenetic tree was constructed from the Nextstrain servers. Wuhan-Hu-1/2019 strains were taken as a reference sequence. (B) The full-length Delta and Omicron Spikes were built to annotate the relative (not exact) positions of the mutations on the surface map of Spike. (C) The amino acids mutated in the RBD of Omicron are shown concerning the ACE2 interface. Residues are colored according to the electrostatic map of the WT strain. The respective Omicron mutations are depicted in the panel below the RBD surface map.

Figure 2

Relative effect of mutations in Omicron RBD on the ACE2 binding. (A) Effect of 15 individual mutations on the binding and stability of RBD^Omic–ACE2 was monitored relative to that of RBD^WT–ACE2. (B) The binding free energies (measured through MMPBSA) as consequences of all 15 mutations at once were monitored for both RBD^Omic–ACE2 and RBD^WT–ACE2. (C) Per-residue energy contribution was monitored, and the hotspots of RBD were labeled. The change in the hydrogen bond network of the selected hotspots is shown at the right. (D, E) Root mean square deviation and hydrogen bonds at the RBD–ACE2 interface as a function of time are displayed for both RBD^Omic–ACE2 and RBD^WT–ACE2 complexes.

Figure 3

Mutations in the Omicron RBD distort the epitopes of therapeutic mAbs. (A–D) Crude epitopes of seven selected mAbs are shown on the RBD. Antibodies used as cocktails are labeled with their sponsors. All variable light chains are colored yellow or orange and variable heavy chains are colored red. (E) Changes in the binding affinity of the RBD^Omic–mAbs relative to RBD^WT–mAbs are shown. The binding energies were calculated through endpoint MMGBSA. (F) The binding free energies (measured through MMPBSA) as consequences of all 15 mutations at once were monitored for RBD^Omic–etesevimab and RBD^WT–etesevimab. (G) The binding free energies (measured through MMPBSA) as consequences of all 15 mutations at once were monitored for RBD^Omic–CT-p59 and RBD^WT–CT-p59.

Figure 4

Per-residue changes in the binding affinity of RBD–mAbs were monitored and the hotspots on CDRs of (A) bamlanivimab and (B) CT-p59 are labeled. The change in the hydrogen bond network of the selected hotspots is shown at the right.