Resistance to colistin, one of the antibiotics of last resort against multidrug-resistant Gram-negative bacteria, is increasingly reported. Notably, MCR plasmids discovered in 2015 have now been reported worldwide in humans. To keep this antibiotic of last resort efficient, a way to tackle this mechanism seems essential. Terpene alcohols such as farnesol have been shown to improve the efficacy of some antibiotics. However, their high lipophilicity makes them difficult to use. This problem can be solved by encapsulating them in water-dispersible lipid nanoparticles (LNPs). The aim of this study was to discover, using checkerboard tests and time-kill curve experiments, an association between colistin and farnesol or geraniol loaded in LNPs, which would improve the efficacy of colistin against E. coli and, in particular, MCR-1 transconjugants. Then, the effect of the combination on E. coli inner membrane permeabilisation was evaluated using propidium iodide (PI) uptake and compared to human red blood cells plasma membrane permeabilisation. Both terpene alcohols were able to restore the susceptibility of E. coli J53 MCR-1 to colistin with the same efficacy (Emax = 16, i.e., colistin MIC was decreased from 8 to 0.5 mg/L). However, with an EC50 of 2.69 mg/L, farnesol was more potent than geraniol (EC50 = 39.49 mg/L). Time-kill studies showed a bactericidal effect on MCR-1 transconjugant 6 h after incubation, with no regrowth up to 30 h in the presence of 1 mg/L colistin (1/8 MIC) and 60 mg/L or 200 mg/L farnesol or geraniol, respectively. Colistin alone was more potent in increasing PI uptake rate in the susceptible strain (EC50 = 0.86 ± 0.08 mg/L) than in the MCR-1 one (EC50 = 7.38 ± 0.85 mg/L). Against the MCR-1 strain, farnesol-loaded LNP at 60 mg/L enhanced the colistin-induced inner membrane permeabilization effect up to 5-fold and also increased its potency as shown by the decrease in its EC50 from 7.38 ± 0.85 mg/L to 2.69 ± 0.25 mg/L. Importantly, no hemolysis was observed for LNPs loaded with farnesol or geraniol, alone or in combination with colistin, at the concentrations showing the maximum decrease in colistin MICs. The results presented here indicate that farnesol-loaded LNPs should be studied as combination therapy with colistin to prevent the development of resistance to this antibiotic of last resort.
Keywords: E. coli; Emax model; antibiotic-non-antibiotic combination; farnesol; lipid nanoparticle; mcr-1; time-kill curve.