Excellent treatment activity of biscoumarins and dihydropyrans against P. aeruginosa pneumonia and reinforcement learning for designing novel inhibitors

Abstract

Pseudomonas aeruginosa (P. aeruginosa) is a common clinical pathogen, which can easily cause cystic fibrosis and even bacteremia. In recent years, the antibiotic resistance of P. aeruginosa has been increasing. In an attempt to develop novel antibacterial agents, a series of biscoumarins ( 1 – 5 ) and dihydropyrans ( 6 – 10 ) were successfully prepared. The molecular structures of two representative compounds, that is, 1 and 6 were confirmed by single crystal X-ray diffraction study. The anti-bacterial activity of these synthesized compounds in vitro was evaluated by measuring the MIC values, as well as the P. aeruginosa growth curves. Next, the in vivo treatment activity of these compounds against the P. aeruginosa pneumonia infection was assessed by observing the survival rate of the infected mice and counting the bacterial load with colony plate counting assay. Additionally, the ELISA detection was conducted to evaluate the inflammatory response levels by measuring the IL-1β and TNF-α content released into the plasma, nasal lavage fluid and alveolar lavage fluid. The HE staining was also carried out to detect the protective effect of the compounds on the lung tissue damage. Further, novel anti-bacterial structures that are based on biscoumarin 5 are predicted and evaluated using reinforcement learning technic, and two outstanding pharmaceutical structures with low binding energy and high SA and QED scores are analyzed in detail using molecular docking simulation.