A preprint version of the study is available on the bioRxiv* server while the article undergoes peer review.
The researchers used a QM model, an approach that uses large-scale calculations to extract systems' electronic density matrix, allowing investigation of intermolecular interactions. Using a model of the molecules, they calculated the electronic structure and drew a contact map to identify relevant chemical interactions between the RBD and other molecules, such as ACE2.
The strength of inter-residue interaction is quantified by Fragment Bond Order (FBO). This value is calculated using the structure of the system in the proximity of a residue. The FBO can identify residues that have a chemical interaction, such as the amino acids of a counter ligand that share a bond above a certain threshold with the ligand. The FBO metric is a non-empirical identification of steric hotspot interactions. The distribution of these hotspots is useful in describing the distance between the binding phenotypes of the different variants and characterizing the binding of the RBDs of these variants to nAbs.
Once these chemical connections have been identified, the scientists assigned each residue a value representing the contribution to the binding interaction. These can be split into two parts – the electrostatic attraction/repulsion and the chemical interaction, an attractive term representing chemical binding between the two fragments. The second part is only relevant if the electronic clouds of the fragments are within a very close range.
The scientists focused on the S1 subunit of the spike protein, specifically investigating the mutations K417N, N440K, G446S, S477N, T478K, E484A, Q493K, G496S, Q498R, N501Y, and Y505H. Mutations that have no strong electrostatic character or are not near the interface were not included. The aim was to identify the role of each mutation in the interactions between the RBD and ACE2 and categorize the mutations according to their stabilizing power both within and outside the interface, the hot spot region they belong in, and whether the effects of the mutations were additive or not.
The researchers found that Omicron showed much stronger binding enthalpy than the wild-type strain and the Delta variant. The interaction patterns suggest that the Omicron variant has a specific binding arrangement with ACE2 and that the interface residues of the spike differ significantly compared to the other variants.
There are two main hotspot regions on the ACE2 side when interacting with the wild-type RBD – the first is region A, involving residues 37-38, 41-42, and 353. The second is region B, involving residues D30, K31, and H34-E35.
In Omicron, region A differs, with E36 removed and K353 and D355 included. Region B shows an increased relevance of E35 and the removal of D30. This new arrangement decreases the short-term contribution of the interaction but increases the long-range contribution enough to favor Omicron binding. The Omicron E484A mutation pulls position 484 off the interface, while the opposite happens with N01Y. K417N compromises one of the strongest interactors in other strains, but the Q493K mutation increases its interface strength significantly enough to counteract this, especially considering the added electrostatic attraction to the E35 residue. The interface mutation Q489R also improves the binding, alongside Y505H and N501Y, which helps the D355 residue play the role the K353 residue occupies in other variants. N440K further increases electrostatic binding.
The researcher's model suggests that the increase in Omicron RBD binding to ACE2 is largely electrostatic. While the interaction with nAbs has not been examined, the interaction pattern will likely be affected.
The scientists hypothesize that the Omicron variant acquired some mutations in a specific order. The K417N mutation seems destabilizing, suggesting that it should be accompanied by other mutations stabilizing the same hotspot, such as Q493K. However, a destabilizing effect is still present even with these two mutations, suggesting another mutation was involved. Y505H follows a similar pattern and probably requires another mutation to stabilize it. While the approach taken is unusual, the authors highlight that their research could act as a proof-of-concept for using this modeling to explore new mutations.
bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information