Inhibiting the Evolution of Antibiotic Resistance

Abstract

Efforts to battle antimicrobial resistance (AMR) are generally focused on developing novel antibiotics. However, history shows that resistance arises regardless of the nature or potency of new drugs. Here, we propose and provide evidence for an alternate strategy to resolve this problem: inhibiting evolution. We determined that the DNA translocase Mfd is an "evolvability factor" that promotes mutagenesis and is required for rapid resistance development to all antibiotics tested across highly divergent bacterial species. Importantly, hypermutator alleles that accelerate AMR development did not arise without Mfd, at least during evolution of trimethoprim resistance. We also show that Mfd's role in AMR development depends on its interactions with the RNA polymerase subunit RpoB and the nucleotide excision repair protein UvrA. Our findings suggest that AMR development can be inhibited through inactivation of evolvability factors (potentially with "anti-evolution" drugs)-in particular, Mfd-providing an unexplored route toward battling the AMR crisis.

Keywords: Mfd; Mycobacteria; anti-evolution; antibiotic resistance; antimicrobial resistance; evolution; hypermutator; transcription-coupled repair.

Graphical abstract

Figure 1

Mfd Promotes Mutagenesis in Diverse Bacterial Species, Related to Figures S1 and S5 (A) Mutation rates of WT (black) and Δmfd (gray) strains to rifampicin for three indicated species (Bs, B. subtilis HM1; Pa, P. aeruginosa CF127; St, S. typhimurium ST19). Number of replicates for Bs = 75, Pa = 42, St = 36. Error bars are 95% confidence intervals. (B) Mutation rates of Mtb (H37Rv) to three different antibiotics for WT (black) and Δmfd (gray). Number of replicates for Mtb = 33–48. Error bars are 95% confidence intervals. ^∗Ciprofloxacin y-axis is mutations per 10⁸ cells per generation. (C) Mutation frequency of S. typhimurium in culture tubes and during infection of CACO-2 cells. Frequency was measured by plating on M9 glycerol with 5-flourocytosine for CFU enumeration. Error bars are standard error of the mean. Two-tailed Student’s t test determined statistical significance (^∗∗p value < 0.01, ^∗∗∗p value < 0.001). (D) CFU enumeration of WT and Δmfd S. typhimurium strains upon infection of CACO-2 cells.

Figure 2

Mfd Promotes Evolution to Various Classes of Antibiotics, Related to Figure S2, S3, and Table S1 Evolution of S. typhimurium ST19 to (A) rifampicin, (B) phosphomycin, (C) trimethoprim, (D) kanamycin, and (E) vancomycin; evolution of B. subtilis HM1 to (F) rifampicin. Heatmaps and line plots show median antibiotic concentration for WT and Δmfd strains at each sampled time point. Black bars represent median growth greater than highest concentration shown on the scale. Concentrations for all antibiotics are in μg/mL. Statistical significance was determined using a two-tailed Mann-Whitney U test (^∗p value < 0.05, ^∗∗p value < 0.01, ^∗∗∗p value < 0.001). Number of replicates for each strain and antibiotic of St and Bs are 12–30.

Figure 3

Mfd Promotes Evolution to Antibiotics in Mtb (A) Evolution of Mtb H37Rv to rifampicin. Heatmaps and line plots showing median rifampicin concentration for WT and Δmfd strains at each sampled passage from a representative experiment are shown. Black bars represent median growth greater than highest concentration shown on the scale. Concentrations are in ng/mL. Statistical significance was determined using a two-tailed Mann-Whitney U test (^∗p value < 0.10). Number of replicates for each strain of Mtb is 6. (B) Mtb Mfd and ST19 RpoB interact. Mtb Mfd RNAP interacting domain (RID) and S. typhimurium ST19 RpoB N-terminal domain were cloned into a luciferase-based bacterial 2-hybrid system. Interactions between these respective protein domains were measured by luminescence and normalized to OD₆₀₀. Results are from three independent experiments, and error bars indicate standard error of the mean. Statistical significance was determined using two-tailed Student’s t test (^∗∗p value < 0.01). (C) Mutation rate analyses were performed with indicated strains of S. typhimurium to rifampicin as in Figure 1. Number of replicates is 36–96. (D) Evolution of indicated strains of S. typhimurium to rifampicin. Plots and statistical testing for evolution assays were performed as described in Figure 2. Number of replicates per strain is 12–24. ^∗p value < 0.05 between WT and Δmfd strains and ^∗∗p value < 0.01 between Δmfd::Mtb-mfd and Δmfd strains.

Figure 4

Mfd-RpoB and Mfd-UvrA Interactions Are Essential for Mfd-Driven Mutagenesis and Evolution to Antibiotics, Related to Figure S4 (A) Mutation of Mfd L499R and R165A residues abrogates RNAP and UvrA interactions, respectively. Relevant domains of the RpoB, Mfd, and UvrA proteins of S. typhimurium ST19 were cloned into a luciferase-based bacterial 2-hybrid system. Interactions between the respective protein domains were measured as in Figure 3. Results are from three independent experiments, and error bars indicate standard error of the mean. Statistical significance was determined using two-tailed Student’s t test (^∗∗∗p value < 0.001). (B) Mutation rate analysis of indicated strains of S. typhimurium to rifampicin. Complement and point mutant (L499R and R165A) strains were expressed episomally. WT and Δmfd strains contain pUC19 empty vector control (see Tables S2 and S3). Number of replicates per strain is 36–112. Errors bars are 95% CI. (C) Evolution of indicated S. typhimurium strains to rifampicin. Complement and point mutant strains (L499R and R165A) were expressed episomally. WT and Δmfd strains contain pMMB67EH empty vector controls. Strains were grown in 50 μg/mL carbenicillin to maintain selection of episomes. Plots and statistical testing for evolution assays were performed as described in Figure 2. Number of replicates per strain is 12–24. ^∗∗p value < 0.01 between WT and Δmfd strains and ^∗p value < 0.05 between WT and Δmfd::mfd(L499R) and WT and Δmfd::mfd(R165A) strains.