Efflux pump activity potentiates the evolution of antibiotic resistance across S. aureus isolates

Abstract

The rise of antibiotic resistance in many bacterial pathogens has been driven by the spread of a few successful strains, suggesting that some bacteria are genetically pre-disposed to evolving resistance. Here, we test this hypothesis by challenging a diverse set of 222 isolates of Staphylococcus aureus with the antibiotic ciprofloxacin in a large-scale evolution experiment. We find that a single efflux pump, norA, causes widespread variation in evolvability across isolates. Elevated norA expression potentiates evolution by increasing the fitness benefit provided by DNA topoisomerase mutations under ciprofloxacin treatment. Amplification of norA provides a further mechanism of rapid evolution in isolates from the CC398 lineage. Crucially, chemical inhibition of NorA effectively prevents the evolution of resistance in all isolates. Our study shows that pre-existing genetic diversity plays a key role in shaping resistance evolution, and it may be possible to predict which strains are likely to evolve resistance and to optimize inhibitor use to prevent this outcome.

Fig. 1. Experimental evolution of ciprofloxacin resistance.

a Maximum-likelihood phylogeny of S. aureus strains included in this study. The tree was constructed using a whole-genome alignment of 222 strains mapped to the S. aureus MRSA252 reference genome and corrected for recombination. b Population growth during experimental evolution. Five heatmaps show optical density (λ = 595) of bacterial populations at the end of each transfer. The optical density varies from low (no growth, OD₅₉₅ < 0.08, black) to high density (high growth, OD₅₉₅ > 1, bright yellow). In each heatmap, the columns correspond to 12 replicate populations, and the rows correspond to 222 strains. c The correlation of population survival and intrinsic resistance in 14 phylogenetic clusters. Each data point represents one strain where population survival is the proportion of replicate populations that evolved resistance, and intrinsic resistance is IC₅₀ (half growth inhibition dose) of a parental strain. Colour and shape for each strain indicate MLST (multilocus sequence type). An open circle is shown when MLST could not be determined. 222 strains were assigned to 14 phylogenetic clusters based on a genetic distance matrix obtained from the phylogenetic tree in (a). The black curves show a model fit for the logistic regression between survival and intrinsic resistance across all 222 strains (the same for all clusters, GLM: χ² = 393.93, d.f. = 1, residual d.f. = 220, p < 2.2e−16), and the red curves show cluster-specific effects. Post hoc tests for cluster-specific effects are shown in Supplementary Table 6.

Fig. 2. Genomic basis of evolved ciprofloxacin resistance.

a Resistance mutations in the evolved S. aureus populations identified by whole-genome sequencing (N = 121). The panels shows the identified mutations in gyrA, grlB and grlA (red) or the amplification of norA gene (blue). The populations are ranked by the evolvability of their parental strains shown as barplots on top of the panel. b Mutation rate in high and low evolvability strains. We measured the mutation rate to rifampicin resistance using a Luria-Delbruck fluctuation test. Lines in the figure connect pairs of high and low evolvability strains (N = 11 pairs). Two-sided Wilcoxon signed-rank test: W = 0.8311, P = 0.8311, N = 12. c Copy number across sites spanning the 7239 bp region that is amplified in ST3535 and ST291 evolved strains. Copy number was calculated by summing the number of reads per site, normalized by the mean sequencing depth across all sites mapping to the ST291 reference genome, and smoothed using a generalized additive model. Gene annotations are shown below: yellow = ISSau1 transposases; blue = norA.

Fig. 3. Gene expression analysis.

a Plotted points show the average difference (high evolvability−low evolvability) in gene expression between 14 pairs of high and low evolvability strains. Gene expression was measured after 1.5 h of exposure to ciprofloxacin (1 mg l⁻¹) and each point represents a gene in the MRSA252 transcriptome (the number of genes is n = 2047). Significantly differentially expressed genes are coloured red (p < 0.05, two-sided Wald tests) and norA is coloured blue. P-values were adjusted using Benjamini-Hochberg method. b The expression of norA in low and high evolvability strains. The read counts for the norA gene were normalised by sequencing depth. Pairs of strains are connected with a line (the number of pairs was N = 14). Two-sided Wald test: log₂ fold-change = 0.979, standard error = 0.246, t = 3.979, adjusted p = 0.006802703).

Fig. 4. The role of norA in resistance evolution.

a The effect of norA overexpression on ciprofloxacin resistance. Resistance was measured in the RN4220 cells overexpressing norA from vector pRMC2 under a native promoter (red), compared to an pRMC2 empty vector control (grey) and a vector-free control (blue). N = 3 independent cultures were used per treatment per concentration. Dose-response curves show model fit from the analysis presented in Supplementary Table 8. b The effect of norA overexpression on the evolution of ciprofloxacin resistance. Optical density (λ = 595) was measured for five daily transfers with 1 mg l⁻¹ of ciprofloxacin in RN4220 cells overexpressing norA from the pRMC2-norA vector (red), cells with the empty pRMC2 vector (blue) and the cells without the vector (grey). N = 40 independent cultures were used for each type of cells. Statistical analysis in shown in Supplementary Table 10. c The effect of reserpine on intrinsic resistance. Intrinsic resistance to ciprofloxacin (IC₅₀, mg l⁻¹) was determined for a representative set of 27 strains in the presence (y-axis) or absence of 33 µM reserpine (x-axis) (N = 5 per dose/reserpine combination). Two-sided Wilcoxon signed-rank test: W = 308, d.f. = 26, p = 0.003. d Evolvability was determined for the same set of 27 strains with 33 µM reserpine (y-axis) or without reserpine (x-axis). Evolvability was measured as the probability of population survival after 5 serial transfers at 1 mg l⁻¹ of ciprofloxacin (N = 16 replicate populations for each strain/reserpine combination). Two-sided Wilcoxon signed-rank test: W = 226, N = 27, p-value = 0.000128.

Fig. 5. The mechanism of norA potentiation.

a Exponentially growing cells were exposed to 1 mg l⁻¹ of ciprofloxacin and viable cells counts were estimated by plating on TSB agar. Four treatments were compared: (i) norA overexpression = red solid line, (ii) norA overexpression and reserpine (33 μM) inhibition = orange solid line, (iii) empty vector control = grey line, and (iv) no vector control = blue line. In addition, cell density was measured in treatments (i) and (ii) without ciprofloxacin (red dash-dotted line and orange dash-dotted line, correspondingly). For each time-point, six independent replicates per treatment were measured (with the exception of the pRCM-norA cells in ciprofloxacin at 0 h which had N = 5). The lines show model fit for polynomial regression (F_{23, 191} = 622.6, p-value < 2.2e−16). Post-hoc comparisons are shown in Supplementary Table 14. b Growth rate of grlA mutants with or without norA overexpression. We estimated the growth rate of the RN4220 wild type (WT) and three independently obtained mutants carrying each grlA substitution in the presence of 1 mg l⁻¹ of ciprofloxacin. Growth rate in cells carrying pRMC2-norA (red) or cells without the vector (blue) was determined using growth curves shown in Supplementary Fig. 10b. Black horizontal lines show the mean growth rate of independent cultures (N = 6 per mutant/treatment combination for A116E, E84K, S80F and S80Y, N = 11 for WT pRMC2-norA and N = 12 for WT no vector). Paired two-sided t-test comparing the means in the mutants with and without pRMC2-norA: t = 11.958, d.f. = 11, p = 1.206e−07. Two-sided t-test comparing pRMC2-norA and no vector control in WT: t = 0.96281, d.f. = 12.136, p = 0.3544. c Representative dose-response experiments for grlA mutants. Red lines show mean optical density with norA overexpression, and blue lines show mean density without overexpression. Black arrow shows the concentration used during experimental evolution. Five independent cultures were used per mutant/treatment/concentration. The results for all mutants are shown in Supplementary Fig. 10a. The difference of means between pRMC2-norA and no vector cells at 1 mg l⁻¹ for all mutants was tested using two-sided paired t- test: t = −18.712, P = 1.09e−09, d.f. = 11.