Yersinia pseudotuberculosis doxycycline tolerance strategies include modulating expression of genes involved in cell permeability and tRNA modifications

Author summaryThe initial goal of this study was to determine how Yersinia pseudotuberculosis changes transcriptionally in response to doxycycline exposure, to then determine if these transcriptional changes are required to survive prolonged drug exposure. Our data suggests that porin expression is not just involved in antibiotic entry into bacterial cells, and may also promote passive diffusion of doxycycline out of the cell. Unexpectedly, we also found that overexpression of tusB, a gene product involved in tRNA modifications, resulted in potent doxycycline bactericidal activity at the typical minimum inhibitory concentration for Y. pseudotuberculosis. Doxycycline is classically viewed as a bacteriostatic antibiotic, which reversibly binds ribosomes, resulting in bacterial growth inhibition. However, growing evidence suggests that bacteriostatic antibiotics can also have bactericidal activity. We believe our results may be the first demonstration of the bactericidal activity of doxycycline under physiological conditions; with a bacterial pathogen, at a drug concentration that can be obtained within mouse and human tissues. Antibiotic tolerance is typically associated with a phenotypic change within a bacterial population, resulting in a transient decrease in antibiotic susceptibility that can contribute to treatment failure and recurrent infections. Although tolerant cells may emerge prior to treatment, the stress of prolonged antibiotic exposure can also promote tolerance. Here, we sought to determine how Yersinia pseudotuberculosis responds to doxycycline exposure, to then verify if these gene expression changes could promote doxycycline tolerance in culture and in our mouse model of infection. Only four genes were differentially regulated in response to a physiologically-relevant dose of doxycycline: osmB and ompF were upregulated, tusB and cnfy were downregulated; differential expression also occurred during doxycycline treatment in the mouse. ompF, tusB and cnfy were also differentially regulated in response to chloramphenicol, indicating these could be general responses to ribosomal inhibition. cnfy has previously been associated with persistence and was not a major focus here. We found deletion of the OmpF porin resulted in increased antibiotic accumulation, suggesting expression may promote diffusion of doxycycline out of the cell, while OsmB lipoprotein had a minor impact on antibiotic permeability. Overexpression of tusB significantly impaired bacterial survival in culture and in the mouse, suggesting that tRNA modification by tusB, and the resulting impacts on translational machinery, promotes survival during treatment with an antibiotic classically viewed as bacteriostatic. We believe this may be the first observation of bactericidal activity of doxycycline under physiological conditions, which was revealed by reversing tusB downregulation.