Back

Multigenerational Proteolytic Inactivation of Restriction Upon Subtle Genomic Hypomethylation

Shmidov, E.; Villani, A.; Mendoza, S. D.; Avihu, E.; Lebenthal-Loinger, I.; Karako-Lampert, S.; Shoshani, S.; Ye, C.; Wang, Y.; Yan, H.; Tang, W.; Bondy-Denomy, J.; Banin, E.

2024-10-23 microbiology
10.1101/2024.10.21.619520 bioRxiv
Show abstract

Restriction-modification (R-M) systems, present in most bacterial genomes, protect against phage infection by detecting and degrading invading foreign DNA. However, like many prokaryotic anti-phage systems, R-M systems pose a significant risk of auto-immunity, exacerbated by the presence of hundreds to thousands of potential cleavage sites in the bacterial genome. In Pseudomonas aeruginosa, restriction inactivation upon growth at high temperatures was previously described, however, which system is being inactivated, the underlying mechanism, as well as the timing of recovery, remain unknown. Here, we report that P. aeruginosa Type I methyltransferase (HsdMS) and restriction endonuclease (HsdR) components are degraded by two Lon-like proteases when replicating above 41 {degrees}C, which induces partial genome hypomethylation and simultaneously prevents self-targeting, respectively. Interestingly, upon return to 37 {degrees}C, methyltransferase activity returns gradually, with restriction activity not fully recovering for over 60 bacterial generations, representing the longest bacterial memory to our knowledge. Forced expression of HsdR over the first 45 generations is toxic, demonstrating the fitness benefit of HsdR inactivation. Our findings demonstrate that type I R-M is tightly regulated post-translationally with a remarkable memory effect to ensure genomic stability and emphasize the importance of mitigating auto-toxicity for bacterial defense systems.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.