MD simulation investigation of noncovalent interaction hotspots in the bacterial ribosome upon translocation as potential antibiotic targets

The steady increase in antimicrobial resistance to existent antibiotics is a growing threat. The ribosome remains one of the most convenient targets for antibiotics due to its conservation, multiple copies of the rRNA genes, and universal central biochemical role. Multiple classes of ribosomal antibiotics bind to several conservative sites at or close to the crutial functional centers in this huge target. Antimicrobial resistance is provided by genes, inactivating or pumping these antibiotics out of the cell or modifying the binding sites, which are ancient products of evolution. A promising strategy to overcome evolutionarily acquired resistance is the identification of compounds with novel binding sites on the ribosome.

There are several reliable datasets about antibiotics affecting the ribosome translocation, a multistep stochastic movement that is catalysed by elongation factor G (EF-G), which transfers the ribosome from the A/A,P/P- to the P/P,E/E-state. For example, octapeptides argyrins trap the EF-G in the conformation corresponding to the pre-translocational state [1]. Even for aminoglycosides, trapping the ribosome in one of its intermediate states was recently considered to be the main mechanism of action [2]. Thus, there can be more than two sites in which drug binding can arrest the translocation process.

In order to identify another potential site for trapping ribosome translocation, we performed all-atom MD simulations of the pre- and post-translocational states of the 70S ribosome. In the obtained trajectories, our comparative and cross-correlation analyses revealed reproducible differences in non-covalent interactions (hydrogen bonds and stacking interactions) in elongation factor G and intersubunit bridges. A time series analysis of occurencies of these interactions detected "hotspots" of multiple conserted interactions that define one of the intermediate states of translocation.

All the MD simulations were performed using the Lomonosov-II supercomputer of Lomonosov Moscow State University's supercomputer center.

1. Proc Natl Acad Sci USA. 2022 119(19):e2114214119.

2. Nucleic Acids Res. 2021 49(12):6880–6892