De Novo Resistance to Arg10-Teixobactin Occurs Slowly and Is Costly.

Bacterial pathogens are rapidly evolving resistance to all clinically available antibiotics. One part of the solution to this complex issue is to better understand the resistance mechanisms to new and existing antibiotics. Here, we focus on two antibiotics. Teixobactin is a recently discovered promising antibiotic that is claimed to "kill pathogens without detectable resistance" (L. L. Ling, T. Schneider, A. J. Peoples, A. L. Spoering, et al., Nature 517:455-459, 2015, https://doi.org/10.1038/nature14098). Moenomycin A has been extensively used in animal husbandry for over 50 years with no meaningful antibiotic resistance arising. However, the nature, mechanisms, and consequences of the evolution of resistance to these "resistance-proof" compounds have not been investigated. Through a fusion of experimental evolution, whole-genome sequencing, and structural biology, we show that Staphylococcus aureus can develop significant resistance to both antibiotics in clinically meaningful timescales. The magnitude of evolved resistance to Arg10-teixobactin is 300-fold less than to moenomycin A over 45 days, and these are 2,500-fold and 8-fold less than evolved resistance to rifampicin (control), respectively. We have identified a core suite of key mutations, which correlate with the evolution of resistance, that are in genes involved in cell wall modulation, lipid synthesis, and energy metabolism. We show the evolution of resistance to these antimicrobials translates into significant cross-resistance against other clinically relevant antibiotics for moenomycin A but not Arg10-teixobactin. Lastly, we show that resistance is rapidly lost in the absence of antibiotic selection, especially for Arg10-teixobactin. These findings indicate that teixobactin is worth pursuing for clinical applications and provide evidence to inform strategies for future compound development and clinical management.