Mapping the Role of AcrAB-TolC Efflux Pumps in the Evolution of Antibiotic Resistance Reveals Near-MIC Treatments Facilitate Resistance Acquisition

Combatting the rise of antibiotic resistance is a significant challenge. Efflux pumps are an important contributor to drug resistance; they exist across many cell types and can export numerous classes of antibiotics.

in pressures that drive resistance (8). With limited novel antibiotics and numerous futile antibiotics, doctors and scientists alike are presented with the challenge of how to best treat infections while keeping the evolution of resistance in check.
Adaptive evolution studies have begun exploring how certain antibiotic pressures influence the evolution of resistance. For instance, studies using a "morbidostat"-a continuous culture device that dynamically adjusts antibiotic concentrations to inhibitory levels-have found numerous targets that can be readily mutated to promote resistance (9)(10)(11), as well as identifying how drug switching can limit the evolution of resistance (12). While these studies have provided pivotal insights for this field, the morbidostat design causes antibiotic concentrations to rise to levels that exceed clinically relevant concentrations due to toxicity for patients (13). In recognition of the drug concentration-dependent nature of evolution, researchers have begun to explore bacterial evolution under treatment conditions with lower antibiotic concentrations as well. Wistrand-Yuen et al. found that bacteria grown at subinhibitory drug concentrations were still able to achieve high levels of resistance (14)(15)(16). Notably, the study identified that the same antibiotic produced unique evolutionary pathways when cells were treated with subinhibitory concentrations as opposed to inhibitory concentrations (14).
One limitation of current studies within the field is that they can be difficult to compare due to variations in experimental parameters, such as species, antibiotics, or other experimental conditions (17). Given the unique evolutionary pathways at different antibiotic concentrations, systematic mapping of these evolutionary landscapes could provide an improved understanding of which conditions pose the highest risk by allowing direct comparisons between different antibiotic concentrations. For instance, Jahn et al. demonstrated that variations in treatment dynamics can significantly alter evolved resistance for some antibiotics, such as tetracycline, but not others, such as amikacin and piperacillin (18). Other evolution experiments that were systematically conducted using a range of concentrations for b-lactams (19) and erythromycin (20) have highlighted the concentration-dependent adaptability of Escherichia coli.
There are many mechanisms by which antibiotic resistance can be achieved, including enzymatic inactivation, alteration of antibiotic binding sites, and increased efflux or reduced influx of antibiotics (21,22). Efflux pumps are omnipresent in prokaryotic and eukaryotic cells alike and are an important contributor to multidrug resistance (23). AcrAB-TolC in E. coli is a canonical example of a multidrug efflux pump, providing broad-spectrum resistance and raising the MIC of at least nine different classes of antibiotics (24). The pump is composed of three types of proteins: the outer membrane channel protein, TolC; the periplasmic linker protein, AcrA; and the inner membrane protein responsible for substrate recognition and export, AcrB (23). Using the proton motive force, AcrB actively exports antibiotics from the cell (23,25). The presence of AcrAB-TolC efflux pumps can increase a strain's MIC from ;2-fold to ;10-fold, depending on the antibiotic (26)(27)(28). Furthermore, genes associated with these multidrug-resistant efflux pumps, including their local and global regulators, are common targets for mutation as strains evolve high levels of drug resistance (15,(29)(30)(31)(32).
Recent studies have indicated that in addition to providing modest increases in the MIC due to drug export, pumps can also impact mutation rate and evolvability of strains, which may ultimately be more important for the acquisition of high levels of drug resistance. For example, Singh et al. found that mutants overexpressing acrAB emerged first, and afterwards these mutants could evolve high levels of quinolone resistance (33). In addition, heterogeneity in efflux pump expression can predispose subsets of bacterial populations with elevated acrAB expression to mutation even prior to antibiotic treatment (34). Deletion of genes associated with efflux pumps, such as tolC, can also reduce evolvability under antibiotic exposure (35). Furthermore, a recent study in Staphylococcus aureus found that higher NorA pump levels increased evolvability and that addition of a pump inhibitor could prevent resistance evolution (36).
These studies provoke the question of how AcrAB-TolC efflux pumps impact the evolution of drug resistance.
Our overall goal in this study was to identify how strains with different AcrAB-TolC genotypes evolve antibiotic resistance over time under a range of chloramphenicol concentrations. Chloramphenicol is both a well-validated substrate of AcrAB-TolC and can serve as a last resort antibiotic in multidrug-resistant infections, as most clinical isolates are still susceptible to this drug (37,38). To identify how AcrAB-TolC impacts the evolution of resistance, we used a turbidostat as an evolutionary platform (39) and measured changes in fitness and resistance. We evolved three strains with different levels of AcrAB-TolC: a wild-type (WT) strain with the native regulatory network controlling AcrAB-TolC expression; a strain that lacks the local regulator AcrR (AcrAB 1 ), which results in a 1.5-to 6-fold increase in expression of the pumps (40)(41)(42); and a strain lacking functional AcrAB-TolC efflux pumps (DacrB). We allowed the cultures to grow and evolve for 72 h in continuous culture while continuously recording growth rates. We periodically sampled the cultures and assessed the population's resistance. We then charted the evolutionary landscapes for each of the three strains under different chloramphenicol concentrations to identify which circumstances gave rise to resistance.

RESULTS
In order to systematically evaluate the evolutionary landscape of efflux pump-mediated antibiotic resistance, we used the eVOLVER, a modular turbidostat capable of growing independent cultures in parallel (39). This platform allowed us to track a culture's fitness by measuring growth rate continuously over multiday experiments. In addition to this, we collected samples at selected intervals and, with these samples, performed antibiotic disc diffusion assays to assess the population's resistance and spot assays to quantify the presence of high-resistance isolates within the population (Fig. 1).
We mapped growth rates over time for cultures subjected to a range of chloramphenicol treatment concentrations ( Fig. 2A; see Fig. S1 in the supplemental material). To compare across strains, we defined MIC 0 parent as the MIC of the parent strain (MIC 0 WT = 2mg/ml, MIC 0 AcrAB1 = 2mg/ml, and MIC 0 DacrB = 0.5mg/ml). We found similar values for MIC 0 WT and MIC 0 AcrAB1 (see Fig. S2 in the supplemental material), which may be due to induction of efflux pump expression in the WT strain in the presence of chloramphenicol. Prior studies have shown that the presence of stress can increase pump expression by 4fold (40,43), which is comparable to the impact of deleting acrR (40-42). We found that treatment with high concentrations of chloramphenicol repressed bacterial growth for multiple days. We observed this growth inhibition at ;10 mg/ml for the WT and AcrAB 1 strains, and at ;2mg/ml for the DacrB mutant. These inhibitory concentrations represent treatments of ;5Â MIC 0 parent for all three strains. We found that cultures grown in lower chloramphenicol concentrations were able to recover growth. For example, when we treated cultures with ;1 to 2Â MIC 0 parent , we observed a significant decrease in the growth rate between 0 and 12 h (see Table S1 in the supplemental material). However, after 12 to 24 h, growth in these populations was partially restored. At lower treatment concentrations (,1Â MIC 0 parent ), all cultures were able to grow, although usually at a deficit compared to the 0 mg/ml chloramphenicol condition. For all three strains, there were qualitatively similar growth recovery patterns, with an initial growth repression phase followed by a partially restored growth phase (Fig. S1). The growth rate results suggested the evolution of drug resistance within the population (9,18). To quantify this, we used an antibiotic disc assay to map the corresponding resistance levels ( Fig. 2B; see Fig. S3 in the supplemental material). We found distinct increases in resistance levels that corresponded to populations that recovered growth. While there were qualitative similarities for the three strains, the timing and level of resistance achieved were dependent on the strain background. We classified populations as resistant when their inhibition zone diameters were smaller than 12 mm, following established standards for antimicrobial susceptibility testing (44). The WT strain gained resistance under a broad range of chloramphenicol treatment concentrations; this resistance emerged within 24 h when cells were treated with ;1 to 2Â MIC 0 WT . The AcrAB 1 strain, where efflux pumps are overexpressed, was able to evolve resistance as well, albeit at a lower rate and at lower levels than the WT. The AcrAB 1 strain achieved resistance within 48 h when treated with 2.5Â MIC 0 AcrAB1 , but the range of chloramphenicol concentrations that resulted in resistance was narrower than for the WT strain. The DacrB cells achieved resistance more slowly, but for the range of ;1 to 2Â MIC 0 DacrB , cells in chloramphenicol cultures were still able to reach resistant levels ( Fig. 2B; Fig. S3).
To compare the ultimate evolved resistance levels, we calculated the final, absolute MIC of the populations at 72 h. When we normalized the treatment concentration by MIC 0 parent , we found that treatments with concentrations of ;1 to 2Â MIC 0 parent evolved the most resistant populations (Fig. 2C). Selective pressures of subinhibitory antibiotic concentrations have often been considered high risk for the evolution of resistance (14,45). Yet, our results indicated that concentrations near or just above MIC 0 parent lead to the highest resistance levels under these conditions. In short, all three strains were able to evolve resistance when treated with ;1 to 2Â MIC 0 parent of chloramphenicol, with the WT strain achieving the highest final, absolute MIC of the three strains. The WT evolved more rapidly than the AcrAB 1 or DacrB strain. Moreover, the relative range of chloramphenicol concentrations that supported the evolution of resistance in the AcrAB 1 strain was narrower than those for the WT or DacrB strains.
We next asked how resistance and growth changed through time. We found that in the absence of antibiotics, the trajectories trended largely toward faster growth, with minimal changes to resistance levels (Fig. 3). With subinhibitory chloramphenicol treatments, we observed that the populations first experienced a slight growth decrease, followed by increased resistance, and then showed restored growth within 48 h. While these populations did gain resistance, they did not tend to reach very high final MIC values in absolute terms, with inhibition zone diameters just at the border of being defined as resistant. In contrast, with inhibitory chloramphenicol treatment, there was a more dramatic reduction in growth within the first 12 h. Although growth was impacted, the populations tended to walk toward high resistance during this period. As depicted in the schematics, the zig-zag patterns trending toward high resistance may be indicative of the cultures acquiring resistant mutations and compensating for the associated fitness costs of these mutations. Finally, at high chloramphenicol concentrations, bacteria first became more susceptible and then stopped growing entirely within 12 h; growth was never restored for these populations. We found that all strains followed similar evolutionary trajectories while balancing the trade-off between growth and resistance. These findings highlight the importance of using antibiotic concentrations that are sufficiently inhibitory.
While these results tell us about the growth rate and resistance of the overall population, it is difficult to determine if subpopulations of cells within the culture have acquired high levels of resistance from disc assays alone. First, because the disc assays do not quantify resistance associated with individual cells in the culture, they cannot reveal the presence of subpopulations of resistant and susceptible cells. Second, beyond a certain resistance level, cells will grow up to the boundary of the disc; thus, it is not possible to quantify resistance increases beyond this. Determining which conditions can give rise to high levels of resistance is important for revealing particularly dangerous treatment regimens. In addition, subpopulations with increased resistance to one antibiotic can promote cross-resistance to other drugs (45).
To quantify the fraction of resistant cells that emerged during our evolution experiment, we conducted a spot assay, in which we measured the fraction of the population capable of surviving on specific chloramphenicol concentrations. For all three strains, we observed subpopulations that were capable of growing on 10mg/ ml chloramphenicol ( Fig. 4A; see Fig. S4 in the supplemental material). Interestingly, these cells primarily emerged from treatment conditions with lower levels of chloramphenicol, and not from conditions where cells were subjected to 10 mg/ml chloramphenicol. For example, at least 0.1% of the population from each of the three WT replicates that were treated with 2 mg/ml chloramphenicol could survive on 10 mg/ml at the end of the experiment. We did find cases where WT cells treated with 10 mg/ml evolved resistance to 10 mg/ml; however, this was less common than at lower treatment concentrations. Thus, cultures were able to evolve resistance to higher levels of chloramphenicol than they were subjected to, a feature that was most pronounced when treatments were just above or at MIC 0 WT . These results closely match trends in FIG 3 Resistance and fitness evolution trajectories. Average diameter of inhibition zone and average growth rate plotted against each other. Lighter purple markers represent trajectories occurring earlier; darker purple markers are later time points. The longer the distance between markers, the greater the change between time points. Colors of boxes indicate the absolute treatment concentration for the depicted trajectories. Schematics summarize patterns for each treatment concentration (ÂMIC 0 parent ). Schematic plots show growth rate in terms of initial growth rate (GR 0 ) and maximum physiological growth rate (GR max ). Resistance is shown in terms of relative diameter of inhibition, where D 0 is the diameter of inhibition at t = 0 h and D min is the diameter of the antibiotic disc. the population's overall resistance (Fig. 2B). We also found isolates capable of growing on 20 mg/ml chloramphenicol, with a reduced frequency relative to 10 mg/ml ( Fig. 4B; Fig. S4).
In contrast, the AcrAB 1 strain was capable of evolving resistance to 10 mg/ml when treated with 5 mg/ml chloramphenicol; yet, surprisingly, the AcrAB 1 strain never produced a subpopulation that was able to grow on 20 mg/ml as the WT did. Meanwhile, despite the higher initial susceptibility of the DacrB strain (MIC 0 DacrB , MIC 0 WT and MIC 0 AcrAB1 ), the DacrB strain consistently produced subpopulations that were able to grow at 20 mg/ml chloramphenicol by 72 h. This subpopulation appeared for chloramphenicol concentrations around 2 mg/ml, similar to the WT strain.
A key question remained: which mutations were responsible for the increases in resistance we observed? To address this, we used whole-genome sequencing to analyze three biological replicates from the 72-h time point for the WT, AcrAB 1 , and DacrB strains (see Table S2 in the supplemental material). For the WT strain, each of the sequenced isolates contained a single point mutation in the DNA binding region of marR, which can upregulate AcrAB-TolC efflux pumps and expression of other stress response genes (46). Two of these point mutations were missense mutations in marR and have been observed in other studies (47)(48)(49)(50)(51). Additionally, one isolate had a missense mutation in the periplasmic encoding region of acrB. The other two isolates had an IS1 or IS5 insertional sequence interrupting acrR, which is known to upregulate acrAB (52). One question these results raise is why the AcrAB 1 strain, where acrR is removed, is outperformed by WT strains with mutations in acrR. A potential explanation for this is that the "marbox" through which acrAB is upregulated sits within acrR (53). The AcrAB 1 strain lacks this marbox (54), while in the sequenced isolates the insertion sequence is located further upstream in acrR and the marbox remains intact, providing global stress response regulation while eliminating the impact of the local repressor. Thus, the exact position of the insertion sequence matters. These sequencing results indicate that strains containing AcrAB-TolC efflux pumps use mutations related to the pumps and their regulation to optimize survival and increase resistance in the presence of chloramphenicol.
When we evolved the AcrAB 1 strain and performed whole-genome sequencing of the most resistant isolates, all isolates had mutations in the noncoding, promoter region of acrAB (Table S2). These mutations indicate that the AcrAB 1 strain might require further tuning of acrAB expression for improved resistance. Furthermore, two of these isolates also had missense mutations in the coding region of acrB as well. Of these, the V139F missense mutation is known to produce high levels of multidrug resistance by accelerating export for a number of AcrAB-TolC substrates (11,18,55,56). We observed acrB (Q569L) evolve from two different parent strains, the WT and AcrAB 1 strains, suggesting it plays a role in chloramphenicol export. Additionally, the evolved AcrAB 1 isolates all had other mutations less directly related to the AcrAB-TolC efflux pump and its regulators, such as genes related to transcription (rpoB and yhjB), fimbria assembly (fimD), or degradation (clpX) ( Table S2).
In contrast, when we evolved the DacrB strain, we found that all three isolates had an insertion sequence located in acrS (Table S2). AcrS is the local regulator of the AcrEF-TolC efflux pump, a homolog to AcrAB-TolC (57). This result agrees with findings from Cudkowicz and Schuldiner, who showed that the DacrB strain gained high resistance by upregulating redundant efflux pumps in E. coli, such as AcrEF-TolC or MdtEF-TolC (11). One of the three isolates also contained a missense mutation in the selenocysteine synthase (selA) and a short insertion sequence in the 16S rRNA of the 30S subunit (rrsG), although whether or how these play a role in chloramphenicol resistance is unclear.

DISCUSSION
In this work, we identified that treatment of strains with antibiotic concentrations close to MIC 0 parent promotes the evolution of resistance; however, the evolvability and ultimate resistance level achieved differed between WT, AcrAB 1 , and DacrB strains. WT populations evolved mutations that conferred high levels of resistance within 24 h after antibiotic exposure. Maximal resistance was evolved at ;1Â MIC 0 WT ; however, 0.25 to 2.5Â MIC 0 WT chloramphenicol treatment concentrations all gave rise to resistance. In contrast, the AcrAB 1 strain evolved resistance, but this was only possible at precise chloramphenicol concentrations at 2.5Â MIC 0 AcrAB1 . The evolved AcrAB 1 populations were less resistant than their WT counterparts, and spot assays determining resistance confirmed this trend. In contrast, the DacrB strain was able to evolve resistance in 1 to 4Â MIC 0 DacrB chloramphenicol treatments and ultimately achieved absolute resistance levels comparable to those observed in the WT strain.
Our results identify that antibiotic treatments near MIC 0 parent are especially prone to evolving resistance. Reding et al. observed this hot spot for adaptability of E. coli in the presence of another antibiotic, erythromycin, just below the MIC of their parent strains (20). While doctors measure resistance of bacterial infections, they sometimes prescribe antibiotic treatment prior to obtaining the results of this assay (58) or use a treatment concentration too low to effectively penetrate the infection site (59). This blind treatment could lead to increased levels of resistance (60,61). These results highlight the presence of regimens that are especially problematic and which should be avoided to limit the evolution of antibiotic resistance.
While we observed that all strains were capable of evolving resistance, sequencing revealed the different pathways that each strain took to achieve this. The WT strain achieved resistance through mutations and insertion sequences in the regulators AcrR and MarR, suggesting that WT cells can fine-tune expression of the AcrAB-TolC pumps to gain resistance to chloramphenicol. Interestingly, these mutations may produce cross-resistance to other antibiotics as well since these regulators control many genes involved in multidrug resistance (62,63). AcrAB 1 cells utilized mutations in acrB and the promoter region controlling its expression to achieve resistance. DacrB populations achieved resistance by targeting homologous efflux pump systems, such as AcrEF-TolC. Although resistance was slow to emerge in this strain compared to the WT or AcrAB 1 strain, this alternative pathway for achieving resistance ultimately resulted in levels comparable to those achieved by the WT strain. By charting evolutionary landscapes across different antibiotic concentrations, we have gained insight into treatments that impact the emergence of antibiotic resistance and the effect of efflux pumps on this process.

MATERIALS AND METHODS
Bacterial strains. We used E. coli strains BW25113 (WT), BW25113 DacrR (AcrAB 1 ), and BW25113 DacrB (DacrB) as the parent strains. The WT strain BW25113 is the base strain for the Keio collection (54). For BW25113 DacrR, we designed primers with homology regions on acrR and amplified the kanamycin resistance marker and FRT (FLP recombination target) sites of pKD13 (54). Primers are listed in Table S3 in the supplemental material. The linear DNA was then treated using a DpnI digest and PCR purification. We electroporated the purified linear DNA into competent BW25113 cells containing the plasmid pSIM6 (64). BW25113 DacrB was derived from Keio collection strain JW0451 (BW25113 DacrB::Kan r ) (26). We removed kanamycin resistance markers from BW25113 DacrR::Kan r and JW0451 following the pCP20 protocol from reference 65.
Determination of MIC. For all experiments, overnight cultures were inoculated from a single colony in 10 ml LB and grown in a 50-ml Erlenmeyer flask at 37°C with 200-rpm orbital shaking. After overnight growth, the optical density at 600 nm (OD 600 ) was measured, and the initial volume was diluted back to OD 600 = 0.1. To determine the MICs of the parent strains (Fig. S2), we added a final concentration of 0, 0.2, 0.5, 1, 2, 4, 8, or 12 mg/ml chloramphenicol to each culture. To determine the MICs of the evolved strains (see Fig. S5 in the supplemental material), we added 0, 0.5, 1, 2, 5, 10, 20, or 50 mg/ml to each culture. Chloramphenicol stocks were prepared with 100% ethanol. The samples were sealed with evaporation-limiting membranes (Thermo Scientific AB-0580) and grown in 24-well plates at 37°C with 200-rpm orbital shaking. OD 600 readings were taken using a BioTek Synergy H1m plate reader before incubation (t = 0 h) and after antibiotic exposure (t = 24 h). All experiments were performed in triplicate using biological replicates.
Experimental conditions in the eVOLVER. In the eVOLVER, cultures were inoculated from a single colony in LB at 37°C. A stir bar mixed the cultures on a medium setting, or approximately 1,000 rpm (39). The LB was supplemented with the detergent Tween 20 (Sigma-Aldrich catalog no. P1379) at 0.2% (vol/ vol) to reduce spurious OD 600 measurements caused by biofilm growth on the flask. As Tween 20 is a detergent and a potential substrate of the AcrAB-TolC efflux pumps, we also conducted the toxicity curve experiments with Tween 20 at our working concentration of 0.2% (vol/vol). We found there was no significant change in resistance for any of the strains in the presence of Tween 20 (see Fig. S6 and Table S4 in the supplemental material).
Cells were inoculated in the eVOLVER overnight (t % 216 to 214 h) prior to the beginning of the experiment (t = 0 h) to establish steady-state exponential growth. We set the eVOLVER using an upper OD 600 bound of 0.2 and a lower bound of 0.1; thus, cultures were grown to a turbidity of 0.2 and then diluted back to 0.1 to maintain the turbidostat at an approximately constant cell density. Samples were collected during the experiment at set time points (t = 0, 1, 3, 6, 12, 24, 48, and 72 h) and used for downstream analysis. All experiments were performed in triplicate using biological replicates.
At t = 0 h, we introduced chloramphenicol at a predetermined final treatment concentration (0, 0.2, 0.5, 1, 2, 5, 10, or 20 mg/ml). This introduction was implemented by switching the source of the medium from one containing 0 mg/ml chloramphenicol to another containing the final treatment concentration; in addition, we spiked the samples directly with the treatment concentration of chloramphenicol at the same time to avoid a delay due to the time required for cycling of the medium in the turbidostat.
Downstream assays and data collection from eVOLVER samples. (i) Growth rate measurements. Growth rate measurements were calculated after each dilution event using the following equation: The growth rate between each dilution was then averaged across sampling time points to compare against disc diffusion assays and spot assays. For example, the growth rate given at t = 0 h is the growth rate from t = 26 h to t = 0 h. To evaluate statistically significant differences in growth rate between two time points, we used the paired t test; to evaluate statistically significant differences in growth rate between two strains, we used the t test (Table S1A).
(ii) Antibiotic disc diffusion assay. We aliquoted samples from the eVOLVER, where the OD 600 from each sample was between 0.1 and 0.2. We used cotton swabs to cover LB agar plates with a layer of the sample (66). An antibiotic disc containing chloramphenicol (30 g) (Thermo Fisher Scientific catalog no. CT0013B) was then placed on the plate. The plate was incubated for 24 h at 37°C. The diameter of the zone of inhibition around each disc was then measured. Diameters of inhibition zones were classified as susceptible, intermediate, or resistant based on reference 44. Additionally, we calculated the MIC using a mapping between the MIC and the diameter of inhibition zone for our samples (Fig. S5) (67). To evaluate statistically significant differences in diameter of inhibition zones or resistance between two time points, Efflux Pumps and Resistance Evolution we used the paired t test; to evaluate statistically significant differences in resistance between two genotypes, we used the t test (Table S1B).
(iii) Spot assay. The samples from the eVOLVER experiment were diluted in phosphate-buffered saline (PBS) in the following dilution series: 1, 10 21 , 10 22 , 10 23 , 10 24 , and 10 25 . We then plated 2.5 ml of each dilution on LB agar plates containing 0, 0.5, 1, 2, 5, 10, and 20 mg/ml chloramphenicol. The plates were then incubated for 24 h at 37°C. To count colonies, we identified the dilution factor with the most countable colonies and recorded the number of CFU and dilution factor (d). The number of CFU/ml for each sample was then calculated as CFU/ml = (CFU Â d)/V, where V is the volume plated. We also calculated the proportion of the population able to grow on different concentrations of chloramphenicol by calculating the CFU/ml from LB agar plates containing 0, 0.5, 1, 2, 5, 10, and 20 mg/ml chloramphenicol.
Whole-genome sequencing. DNA was extracted from single isolates and parent strains using the Qiagen DNeasy PowerBiofilm kit. For each strain, we selected three isolates to sequence; each of these isolates originated from a different biological replicate that was evolved under the same experimental conditions (i.e., each isolate comes from a different eVOLVER culture). Samples were sequenced at the Microbial Genome Sequencing Center (MiGS) in Pittsburg, PA, USA, who conducted library preparation and multiplexing using the Illumina Nextera kit series and then sequenced using a NextSeq 550 platform with 150-bp paired ends and an average coverage of 50 reads. We analyzed reads using version 0.35.1 of breseq (68). Reads were aligned to the BW25113 Keio reference genome (accession no. CP009273) in consensus mode. The treatment concentrations and isolation concentrations used to select each isolate are listed in Table S2.
Data availability. Whole-genome sequencing data for the parent strains and the isolates are available in GenBank under BioProject no. PRJNA666010 and accession no. CP062239 to CP062250. Other data sets generated during this study are available from the corresponding author upon request.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.