Multidrug-resistant bacteria compensate for the epistasis between resistances

Abstract

Mutations conferring resistance to antibiotics are typically costly in the absence of the drug, but bacteria can reduce this cost by acquiring compensatory mutations. Thus, the rate of acquisition of compensatory mutations and their effects are key for the maintenance and dissemination of antibiotic resistances. While compensation for single resistances has been extensively studied, compensatory evolution of multiresistant bacteria remains unexplored. Importantly, since resistance mutations often interact epistatically, compensation of multiresistant bacteria may significantly differ from that of single-resistant strains. We used experimental evolution, next-generation sequencing, in silico simulations, and genome editing to compare the compensatory process of a streptomycin and rifampicin double-resistant Escherichia coli with those of single-resistant clones. We demonstrate that low-fitness double-resistant bacteria compensate faster than single-resistant strains due to the acquisition of compensatory mutations with larger effects. Strikingly, we identified mutations that only compensate for double resistance, being neutral or deleterious in sensitive or single-resistant backgrounds. Moreover, we show that their beneficial effects strongly decrease or disappear in conditions where the epistatic interaction between resistance alleles is absent, demonstrating that these mutations compensate for the epistasis. In summary, our data indicate that epistatic interactions between antibiotic resistances, leading to large fitness costs, possibly open alternative paths for rapid compensatory evolution, thereby potentially stabilizing costly multiple resistances in bacterial populations.

Fig 1. Faster compensatory evolution in double-antibiotic–resistant bacteria.

(A-C) Dynamics of a fluorescent neutral marker during compensation in rich media without antibiotics of ~1:1 mixtures of yellow fluorescent protein (YFP)/cyan fluorescent protein (CFP) cells in (A) 12 independent E. coli populations resistant to rifampicin (RpoB^H526Y); in (B) 12 populations resistant to streptomycin (RpsL^K43T), and in (C) 24 populations resistant to both rifampicin and streptomycin (RpoB^H526Y RpsL^K43T). (D-F) Competitive fitness of RpoB^H526Y (D), RpsL^K43T (E), and RpoB^H526Y RpsL^K43T (F) evolving populations, at different days during adaptation, measured against a reference nonfluorescent-sensitive strain. Each circle corresponds to a population with similar colour shown in panels A, B, and C, respectively. Red dashed lines correspond to the competitive fitness of the sensitive strain. The black dots at time 0 represent the competitive fitness of each of the founder resistant genotypes. Box plots represent the median and quartiles Q2 and Q3, and whiskers show the last quartiles of the data.

Fig 2. Compensatory mutations have stronger effects in double-resistant mutants.

Estimates of the rate of acquisition of beneficial mutations (A) and mean selective effects (B) of arising beneficial mutations are shown in the violin plots and correspond to the distribution of the top 20 simulations, from the inference algorithm in the One Bi-Allelic Marker Algorithm (OBAMAv2), that best explain the experimental data. Red lines indicate the median of the distributions. NS, non-significant, **p < 0.01 and ****p < 0.0001 in a two-sample Kolmogorov test for comparing the distributions, adjusted for multiple comparisons using the Benjamini and Hochberg method [38].

Fig 3. Genetic basis of compensation across resistant backgrounds.

Mutations identified through population sequencing in each evolved antibiotic-resistant background. Mutations detected at least in one population map at the position indicated in the circular plot (see also S1 Table). The blue track and labels indicate mutations found in the 12 independently evolved Rif^R populations. The green track and labels indicate mutations found in the 12 independently evolved Str^R populations. The grey track and the black labels indicate mutations found in the 24 independently evolved Rif^R Str^R. Mutations found in both a single- and the double-resistant backgrounds are indicated by yellow brackets (* indicates a back-mutation). Mutations in regulatory regions are indicated with (Reg). Mutations selected for allelic reconstruction and competition experiments are marked with an arrow. The Venn diagram shows the number of mutations specific to each background and in common between backgrounds (shown as blue, green, and grey).

Fig 4. Mutations specifically compensating for epistasis between Rif^R and Str^R.

Fitness effect of specific compensatory mutations across backgrounds. Strains were constructed with each compensatory allele and competed against their corresponding ancestral (i.e., the resistant or the sensitive strain without the compensatory mutation). Mean values shown in columns (assays based on four independently constructed clones for each allelic replacement, each with at least n = 3 competitions, * p < 0.05, **p < 0.01, ***p < 0.001 from a one sample t-test), error bars indicate twice the Standard Error of the Mean (2SEM). (A) nusE is beneficial in all resistance backgrounds and neutral in the sensitive background (s = 0.04 ± 0.03 [2SEM], p = 0.01 for Rif^R Str^R background; s = 0.03 ± 0.02 [2SEM], p = 0.037 for Str^R background; s = 0.03 ± 0.02 [2SEM], p = 0.008 for Rif^R background; s = 0.003 ± 0.002 [2SEM], p = 0.774 for sensitive background, one-sample t-test). (B) RpoC^Q1126K shows a strong benefit in the double-resistance background (0.25 ± 0.04 [2SEM], p < 0.0001), whilst being neutral in Rif^R (s = −0.02 ± 0.04 [2SEM], p = 0.457) and deleterious in both the Str^R and the sensitive backgrounds (s = −0.05 ± 0.05 [2SEM], p = 0.085 or (s = −0.02 ± 0.01 [2SEM], p = 0.015) for the RpsL^K43T or the sensitive, respectively). (C) nusG is beneficial in the double-resistance background (s = 0.06 ± 0.02 [2SEM], p < 0.0001) and deleterious in all other backgrounds (s = −0.07 ± 0.02 [2SEM], p < 0.0001 for Str^R background; s = −0.07 ± 0.03 [2SEM], p < 0.0001 for Rif^R background; and s = −0.132 ± 0.02 [2SEM], p < 0.0001 for sensitive background). (D) The benefit of RpoC^Q1126K mutation is significantly reduced and nusG is deleterious in minimal media supplemented with glucose (M9+Glucose), where the epistasis between Rif^R and Str^R does not take place. However, nusE, which does not seem to specifically compensate for the epistasis, is beneficial regardless of the media. (E) Schematic representation of the coupling between the RNA polymerase (RNAP) and the ribosome mediated by the NusE–NusG complex. A red cross represents the approximate position of RpoC^Q1126K in the domain of RpoB that NusG binds to.

Fig 5. Location and effects of the mutations in the regulatory regions of nusE and nusG.

(A) Structure of the leader transcript of the nusE operon as determined in [50]. The ribosome binding site and the translation start codon are shown in light blue and green, respectively. The mutated nucleotide (red) is located in a region (marked in yellow) essential for the attenuation mediated by L4 [45,47,51]. (B) Sequence of the promoter region of the secE-nusG transcript. The transcription start site is indicated by the blue arrow. The mutation (red) affects the −10 sequence [44]. (C) Log2 fold change of the expression of nusE (light grey) and nusG (dark grey) genes in each mutant with respect to the wild type, measured by quantitative real-time PCR (RT-qPCR) (from left to right, mean ± twice the standard error of the mean [2SEM], p-values of one sample t-test are: 0.65 ± 0.19, p = 0.020; 0.31 ± 0.17, p = 0.068; 1.88 ± 0.89, p = 0.052; 2.33 ± 0.31, p = 0.004; n = 3).