Sofija Marković1*, Magdalena Đorđević2, Hong-Yu Ou3 and Marko Đorđević1
1 Faculty of Biology, University of Belgrade, Belgrade, Serbia
2 Institute of Physics Belgrade, University of Belgrade, Belgrade, Serbia
3 School of Life Sciences & Biotechnology, Shanghai Jiaotong University, Shanghai, China
sofija.markovic [at] bio.bg.ac.rs
Abstract
Antibiotic persistence refers to a phenomenon where a subset of genetically identical bacteria enters a dormant state, becoming highly resistant to environmental stresses. This phenomenon is crucial in understanding why biofilms, communities of bacteria attached to surfaces, often resist antibiotic treatments, leading to persistent and recurrent infections. Despite being recognized for almost a century, the precise processes triggering persister formation remain elusive.
Among the various biological systems implicated in persister formation, toxin-antitoxin systems within bacteria stand out. These systems consist of a toxic protein and its corresponding antitoxin, which neutralizes the toxin’s effects. In this study, we propose a biophysical model focusing on a type I toxin-antitoxin system where the antitoxin is a small RNA molecule. Our analysis involves both theoretical calculations and computer simulations to explore the stability of the model and its behavior under deterministic and stochastic conditions.
Our model successfully reproduces two distinct states within bacterial populations: a low-toxin state associated with normal growth and a high-toxin state leading to persister formation. We analytically derive a system stability diagram, allowing us to map under which conditions the low and high toxin states coexist in an isogenic bacterial population. Furthermore, we observe a stochastic transition from low to high-toxin. This bistability in our model arises from feedback loops governing toxin production. Specifically, a positive feedback loop controls toxin dilution rate, while a negative feedback loop slows down antitoxin degradation.
Our findings have significant implications for understanding bacterial persistence mechanisms. We have shown that type I toxin-antitoxin systems may play a role in stress-induced persister formation. However, they are unlikely to account for “spontaneous” persister formation, as toxin expression is markedly reduced during normal growth phases. These insights could lead to developing new therapeutic approaches that target the specific mechanisms of stress-induced persister formation, thereby improving the effectiveness of antibiotic treatments.
Keywords: Antibiotic persistence, Toxin-antitoxin systems, Bistability, Bifurcations, Stochastic simulations.
Acknowledgment: This work is supported by The Science Fund of the Republic of Serbia (Grant no. 7750294, q-bioBDS).