Skip to main content
NCBI home page
Infection and Drug Resistance logoLink to Infection and Drug Resistance
. 2026 May 25;19:596309. doi: 10.2147/IDR.S596309

A Fitness Burden Drives the Reversion of in vitro Acquired Doxycycline Resistance in V. cholerae O1

Dayana H Shoukair 1,2,3,*, Ahmad Z Turk 1,2,3,*, Abdullah Kurdi 4, Ghassan M Matar 1,2,3, Antoine G Abou Fayad 1,2,3,✉
PMCID: PMC13220821  PMID: 42222030

Abstract

Introduction

Antibiotic exposure in both clinical and environmental settings can select for resistant isolates with expanded cross-resistance, including last-resort drugs. In Vibrio cholerae, such patterns emerge under sustained doxycycline exposure; however, the evolutionary dynamics and long-term stability of this adaptation remain unresolved.

Methods

Here, we applied long-term experimental evolution to investigate the induction, fitness-costs, and reversion of doxycycline resistance in V. cholerae O1. Isolates were exposed to sub-inhibitory doxycycline concentrations over 13 weeks, followed by phenotypic, genomic, and phylogenetic analyses. Resistance stability was assessed after a 2-week subculture period in the absence of doxycycline.

Results

Doxycycline-susceptible V. cholerae O1 isolates from the 2022 Lebanon outbreak evolved high-level resistance within 13 weeks of sustained exposure. The evolved resistance extended to colistin, while showing collateral sensitivity to azithromycin. Resistance imposed a marked fitness burden characterized by impaired growth and reduced yield. Notably, drug withdrawal restored both doxycycline susceptibility and fitness. Whole-genome sequencing revealed recurrent high-impact mutations in tetR and LysR-family regulators and efflux-associated pathways. Revertants were characterized by the loss of these variants and the accumulation of compensatory mutations in metabolism and membrane structure. Phylogenetic analysis showed that resistance evolved within the original lineages, with evolved isolates clustering closely with globally distributed resistant clinical strains.

Conclusion

Doxycycline resistance in V. cholerae O1 emerges rapidly under antibiotic pressure but represents a transient, high-cost adaptation that is reversible upon drug withdrawal. These findings highlight the fitness burden associated with resistance and demonstrate reproducible evolutionary trajectories, with important implications for antimicrobial stewardship strategies.

Keywords: Vibrio cholerae, doxycycline resistance, collateral sensitivity, reversion, fitness cost

Introduction

Antimicrobial resistance (AMR) is a growing global public health threat, compromising the efficacy of frontline therapies and driving millions of deaths annually worldwide.1 Resistance may be intrinsic, arising from inherent structural or physiological traits, or acquired via mutations or horizontal gene transfer (HGT).2 It can also emerge in vitro by exposing bacterial populations to selective pressure from antibiotics or non-antibiotic inducers.2 Laboratory evolution studies have provided key insights into resistance mechanisms, yet have rarely addressed the stability of acquired resistance once drug pressure is removed.3 This knowledge gap is critical, since resistance mutations often impose fitness costs, such as reduced growth rate or metabolic burden, which strongly shape resistance persistence, however, these costs remain poorly characterized.2,4 Addressing this uncertainty forms the foundation of our investigation in V. cholerae.

Cholera remains a major global public health concern, particularly in regions with limited access to clean water and proper sanitation. Between January and June 2025, over 305,000 cholera cases and 3,522 deaths were reported across 28 countries.5 It is caused by the bacterium V. cholerae, with serogroup O1 responsible for nearly all recent epidemics.6,7 Treatment primarily involves rehydration therapy, while antibiotics are reserved for severe cases to shorten the duration of diarrhea and bacterial shedding. Doxycycline is widely used as a first-line treatment for cholera due to its efficacy, ease of administration, and long half-life.6 Despite its extensive use during outbreaks, clinical reports of doxycycline-resistant V. cholerae are rare.7 However, resistance to doxycycline has been reported in several outbreaks, indicating that its emergence is possible.8 Moreover, the stability and fitness-costs of doxycycline resistance in V. cholerae remain poorly understood, particularly under non-selective conditions, suggesting that resistance may be disadvantageous once selective pressure is removed.2

In this study, using two V. cholerae O1 strains isolated during the 2022 cholera outbreak in Lebanon, we investigated the evolution, stability, and fitness costs of doxycycline resistance through long-term serial passaging under sustained doxycycline exposure followed by drug-withdrawal.

Methods

Vibrio cholerae Isolates and Growth Conditions

Study Design and Isolate Source

This study employed an experimental evolution design to investigate the acquisition and stability of doxycycline resistance in V. cholerae over a four-month period. Five clinical isolates of V. cholerae O1 serotype Ogawa, biotype El Tor (7, 11, 42, 62, and 71) were obtained from cholera patients during the 2022 outbreaks in Syria, and Lebanon. Isolates 7 and 11 originated from Lebanon and belonged to the predominant multidrug-resistant (MDR) strain, whereas isolates 42, 62, and 71 were collected in Syria and belonged to the extensively drug-resistant (XDR) strain, as previously described in the Lebanon outbreak study.9 MDR and XDR were defined based on established criteria. MDR was defined as non-susceptibility to at least one agent in three or more antimicrobial classes. XDR was defined as non-susceptibility to at least one agent in all but two or fewer antimicrobial classes.

Isolate Storage, Processing and Growth Conditions

All isolates were stored at −80°C in 50% glycerol stock. For experiments, isolates were revived in Mueller–Hinton (MH) broth 2 (Nutriselect 70191) and incubated at 37°C with shaking at 160 rpm for 16–18 hours. Isolates were then cultured on non-selective media, including Mueller–Hinton (MH) broth 2 (Nutriselect 70191), Mueller–Hinton (MH) agar (Bio-Rad #64884), and Luria-Bertani (LB) agar (Neogen NCM0142A), as well as on the selective medium Vibrio  chromagar™ (CHROMagar, Paris, France, VB912), and single colonies were selected for downstream assays to ensure culture purity and consistency. A reference V. cholerae strain ATCC 14035 (strain 8021) was included as a quality control in antimicrobial susceptibility testing to ensure assay accuracy and reproducibility. Culture purity was verified by colony morphology and growth characteristics.

In vitro Induction of Resistance by Serial Passaging

Experimental evolution of V. cholerae isolates was performed by serial passaging under sub-inhibitory doxycycline pressure. For each isolate, two independent evolutionary lineages were propagated in parallel (biological duplicates). Colonies grown on CHROMagar™ Vibrio plates for 16–24 hours at 37°C were inoculated into 3 mL of MH broth 2 containing doxycycline at 1/2 MIC (initial concentration). Cultures were incubated at 37°C with shaking at 160 rpm for 24 hours. Each day, 10–20 μL of culture was transferred into fresh MH broth containing the same doxycycline concentration or updated concentration corresponding to a daily bottleneck of approximately 105–106 CFU. MIC values were reassessed weekly by broth microdilution, and the updated MIC guided the doxycycline concentration used for subsequent passaging. This procedure was repeated for 13 consecutive weeks, corresponding to approximately 800–1,000 bacterial generations. At the end of each week, 500 μL of culture was mixed with 500 μL of 50% glycerol and stored at −80°C. Fresh CHROMagar™ Vibrio plates were prepared weekly to maintain access to uncontaminated stocks. This assay was conducted in duplicate, following established in vitro resistance evolution protocols.10

Antimicrobial Susceptibility Testing

Antimicrobial susceptibility was assessed using broth microdilution according to Clinical and Laboratory Standards Institute (CLSI) guidelines.11 Serial two-fold dilutions of doxycycline and tetracycline (0.125 to 256 µg/mL) were prepared in 96-well plates containing MH broth 2. A standardized bacterial inoculum of 5 × 106 CFU/mL was added to each well. Plates were incubated at 37°C with shaking at 160 rpm for 18–24 hours.

The MIC was defined as the lowest antibiotic concentration preventing visible bacterial growth. Assays were repeated if duplicate MIC values varied by more than one dilution or if control wells failed. Isolates with MICs at or above the resistance breakpoint were classified as resistant to doxycycline, and cross-resistance to tetracycline was assessed in parallel.

Growth Curve Assays

Bacterial fitness was assessed using growth curve analysis and comparative growth under doxycycline exposure. Overnight cultures of V. cholerae grown on MH Agar were adjusted to a 0.5 McFarland standard and diluted into MH broth to obtain 5 × 106 CFU/mL. Ninety microliters of MH broth and 10 μL of standardized bacterial suspension were dispensed into each well of a sterile 96-well plate (final volume = 100 μL). Negative control wells containing only broth were included to account for background absorbance. Plates were incubated in a ThermoScientific Multiskan SkyHigh plate reader (model 1550–801475) at 37°C with continuous shaking for 24 to 48 hours. Optical density at 600 nm (OD600) was automatically recorded hourly. All experiments and measurements were performed in 3 independent biological replicates. Growth curves were generated by plotting OD600 values over time. Fitness parameters, including maximum growth rate (µ), time for duplication (Td), and area under the curve (AUC), were calculated manually using Microsoft Excel. Fitness parameters were compared between isolates based on growth dynamics.

Comparative Growth Under Doxycycline Exposure

To assess relative fitness under doxycycline exposure. Growth of baseline (W0) isolates and doxycycline resistant isolates (W13) was evaluated in MH broth supplemented with doxycycline (16 µg/mL) as shown in Figure 1E and 1F. Each condition was tested in three independent biological replicates. Mean values and standard deviations (SD) were calculated for each time point. MH controls were included to account for background absorbance.

Figure 1.

A composite of 4 line graphs and 2 bar charts showing growth and minimum inhibitory concentration trends. The image A showing a line graph with x-axis label Time (h) ranging 1 to 13 and y-axis label OD600 (a.u.) ranging 0.00 to 0.90. Legend: 7W0, 7W13, 7W13R. 7W0 rises from about 0.00 at 1 to 3 h to about 0.20 at 5 h, about 0.50 at 8 h, about 0.65 at 10 h and about 0.66 at 13 h. 7W13 stays near 0.00 to about 5 h, then rises to about 0.05 at 6 h, about 0.32 at 8 h, about 0.55 at 11 h and about 0.70 at 13 h. 7W13R rises from about 0.00 to about 4 h to about 0.12 at 5 h, about 0.42 at 8 h, about 0.48 at 10 h and about 0.52 at 13 h. The image B showing a line graph with x-axis label Time (h) ranging 1 to 13 and y-axis label OD600 (a.u.) ranging 0.00 to 0.90. Legend: 42W0, 42W13, 42W13R. 42W0 rises from about 0.00 to about 6 h to about 0.35 at 7 h, about 0.45 at 8 h, about 0.60 at 10 h and about 0.78 at 13 h. 42W13 stays near 0.00 to about 6 h, then rises to about 0.10 at 7 h, about 0.40 at 9 h and about 0.45 at 13 h. 42W13R rises from about 0.00 to about 4 h to about 0.25 at 6 h, about 0.40 at 8 h and about 0.46 at 13 h. The image C showing a grouped bar chart titled V. cholerae 7. X-axis categories: μmax (h superscript negative 1), Td (h), AUC (OD600 times h). Y-axis label Value (units as indicated) ranging 0 to 14. Legend: W0, W13, W13R. μmax bars are about 0.5 (W0), about 0.9 (W13), about 0.6 (W13R). Td bars are about 1.4 (W0), about 0.8 (W13), about 1.2 (W13R). AUC bars are about 12.2 (W0), about 11.2 (W13), about 10.6 (W13R). The image D showing a grouped bar chart with x-axis categories μmax (h superscript negative 1), Td (h), AUC (OD600 times h) and y-axis unlabeled with range about 0 to 16. Legend: W0, W13, W13R. μmax bars are about 0.6 (W0), about 1.0 (W13), about 0.3 (W13R). Td bars are about 1.2 (W0), about 0.7 (W13), about 2.2 (W13R). AUC bars are about 15.0 (W0), about 7.5 (W13), about 9.8 (W13R). The image E showing a line graph with x-axis label Time (h) ranging 1 to 37 and y-axis label OD600 (a.u.) ranging 0.000 to 0.070. Legend: 7W0 (16 μg/mL), 7W13 (16 μg/mL). 7W0 stays near 0.004 to 0.008 from 1 to about 30 h, then rises to about 0.010 at 37 h. 7W13 stays near 0.004 to 0.008 from 1 to about 30 h, then rises to about 0.020 at 34 h and about 0.060 at 37 h. The image F showing a line graph with x-axis label Time (h) ranging 1 to 25 and y-axis label OD600 (a.u.) ranging 0.000 to 0.250. Legend: 42W0 (16 μg/mL DOX), 42W13 (16 μg/mL DOX). 42W0 stays near 0.000 to 0.005 from 1 to 25 h, ending near 0.015. 42W13 stays near 0.000 to about 15 h, then rises to about 0.080 at 18 h, about 0.110 at 20 h, peaks near 0.210 at 24 h and ends near 0.190 at 25 h. The image G showing a line graph with x-axis labels W13, plus 1 wk, plus 2 wk and y-axis label MIC (μg/mL) with tick labels 2 superscript negative 1, 2 superscript 0, 2 superscript 1, 2 superscript 2, 2 superscript 3, 2 superscript 4. A dashed horizontal line is labeled Breakpoint (4 μg/mL). Legend Isolates: 7, 11, 42, 62, 71. At W13, isolates 7, 42 and 62 are at 2 superscript 4, isolate 71 is at 2 superscript 4 and isolate 11 is at 2 superscript 4. At plus 1 wk, isolate 7 is at 2 superscript 2, isolate 71 is at 2 superscript 3, isolate 42 is at 2 superscript 4, isolate 62 is at 2 superscript 4 and isolate 11 is at 2 superscript 4. At plus 2 wk, isolate 7 is at 2 superscript negative 1 and isolates 11, 42, 62 and 71 are at 2 superscript 0.

Open in a new tab

Growth dynamics and fitness parameters of V. cholerae O1 isolates upon doxycycline selection and withdrawal. (A and B) Growth curves of isolates 7 and 42 at baseline (W0), after 13 weeks of doxycycline exposure (W13), and after 2 weeks of drug-free passage (W13R) (OD600, mean ± SD over 14 h). (C and D) Fitness parameters derived from growth curves, including maximum specific growth rate (μmax, h−1), doubling time (Td, h), and overall yield (AUC, OD600 × h). Bars = means, dots = replicates. (E and F) Growth curves of 7 and 42 isolates W0 and W13 in the presence of doxycycline (16 μg/mL). (G) MIC values (μg/mL) for isolates 7, 11, 42, 62, and 71 at W13 and after 1 and 2 weeks of doxycycline-free subculture. The dashed line marks the CLSI breakpoint (4 μg/mL).

Whole Genome Sequencing and Variant Calling

Genomic DNA was extracted using the Quick-DNA™ Fungal/Bacterial Miniprep Kit (Zymo Research, Irvine, CA, D6005) and purified with the Genomic DNA Clean & Concentrator™ Kit (Zymo Research, D4003), following the manufacturer’s instructions. DNA libraries were prepared with the Illumina DNA Prep Kit, along with the IDT for Illumina DNA UD indexes (Illumina, San Diego, CA). Library concentrations were measured using the Qubit dsDNA High Sensitivity assay kit (Thermo Fisher Scientific, Waltham, MA) on a Qubit 4 fluorometer (Thermo Fisher Scientific, Waltham, MA). The library was then sequenced on either the Illumina MiSeq (SY-410-1003) or NextSeq 500 platform, depending on instrument availability for a coverage of 100x. Paired-end sequencing was performed with a read length of 2 × 150 bp. Raw sequencing data were processed using standard bioinformatics pipelines for quality control,12 adapter trimming,13 and downstream analysis.

Variant Calling

 Sequencing quality reports were generated for the raw FASTQ files using FastQC to check for low-quality bases and presence of adapters. After that, the VariantDetective14 pipeline was used to call variants. Briefly, the pipeline aligns the raw reads to the reference genome V. cholerae O1 biovar El Tor N16961 strain downloaded from Ensembl using BWA.15 Aligned reads are saved as SAM files. After that, single nucleotide polymorphisms (SNPs) and insertions and deletions (INDELs) are called using FreeBayes,16 GATK HaplotypeCaller17 and Clair318 tools using the SAM files as input. High confidence variants were retained based on a consensus between at least two callers. Additional filtering criteria including a minimum read depth (DP ≥ 10), minimum base quality (Phred score ≥ 30), and removal of variants located in low-complexity or poorly covered regions. Variants with low support or conflicting calls between tools were excluded to reduce false positives. The final output is a VCF file for each isolate that contains the consensus variant calls. Ensembl-VEP19 was run to annotate the called variants using v_cholera_ASM674v1 version 60 from Ensembl as reference annotation and a tabulated form of the variants was generated using bcftools20 (+split-vep plugin).

Phylogenetic Analysis

An SNP-based phylogenetic tree was generated using the MinTyper command-line pipeline.21 The resulting tree in Newick format was used to generate the tree through the ggtree R/Bioconductor package.22

Statistical Analysis

All statistical analyses were conducted using R23 (version 4.2.2). Statistical analyses were applied to antimicrobial susceptibility data. The Mann-Kendall trend test24 was used to assess monotonic increases in doxycycline MIC values over the 13-week exposure period. To compare MIC values between week 0 and week 13, the non-parametric Wilcoxon signed-rank test was used. A significance threshold of p < 0.05 was applied for all tests. Biological replicates for each experiment are specified in the corresponding methods sections.

Results

Rapid Induction of Doxycycline Resistance by Serial Passaging

Serial passaging of all V. cholerae isolates under sub-inhibitory concentrations of doxycycline induced progressive resistance to doxycycline. At baseline (week 0), all isolates (7, 11, 42, 62, and 71) were susceptible to doxycycline (MIC = 0.25 μg/mL; CLSI breakpoint ≤ 4 μg/mL). MIC increased rapidly during the first three weeks, followed by a limited rise over the next six weeks. A sharp increase was then observed reaching 16 μg/mL, a 64-fold increase in MIC value by week 13. Detailed MIC values are provided in Supplementary Table S1. Representative MIC trajectories for isolates 7 and 42 are shown in Figure 2. The Mann–Kendall test confirmed a significant upward trend in MIC values over time for all isolates (p ≤ 0.005), reflecting a consistent increase in MIC of doxycycline over time. However, Wilcoxon signed-rank test indicated no significant differences among independent replicate lineages (p = 0.0625), supporting reproducibility. Together, these results demonstrate that V. cholerae can evolve high-level doxycycline resistance in vitro upon continuous exposure to sub-MIC over 13 weeks.

Figure 2.

A line graph of doxycycline minimum inhibitory concentration over weeks for V. cholerae 7 and 42. A line graph compares V. cholerae 7 and V. cholerae 42's minimum inhibitory concentration (MIC) over 13 weeks. The x-axis spans weeks 0 to 13, while the y-axis, on a log base 2 scale, ranges from 0.25 to 32 micrograms per milliliter. A dashed line at 4 micrograms per milliliter marks the breakpoint. V. cholerae 7 starts at 0.25, rises to 4 by week 3, stays at 4 until week 8, then exceeds the breakpoint, reaching 24 by week 13. V. cholerae 42 also begins at 0.25, climbs to 4 by week 3, remains at 4 through week 8 and surpasses the breakpoint, ending at 24 by week 13. Both series show similar trends, increasing from 0.25 to 4 by week 3, maintaining 4 until week 8 and rising above the breakpoint thereafter.

Open in a new tab

Induction of doxycycline resistance in V. cholerae O1 isolates during 13 weeks of in vitro selection. Mean MIC trajectories (μg/mL, log2 scale) of isolates 7 and 42 are represented across weekly doxycycline passages. Solid lines indicate mean values across the main and replicate lineages; shaded areas represent ±SD. The dashed horizontal line is the clinical resistance breakpoint (4 μg/mL).

Resistance Evolution is Accompanied by Fitness Trade-Offs

To quantify the fitness cost of resistance, we compared growth among baseline (W0), resistant (W13), and revertant (W13R) isolates. Resistant isolates exhibited reduced fitness relative to W0 isolates, represented by extended lag phase, and reduced overall yield (Figure 1A–1D). Corresponding OD600 measurements are provided in Supplementary Tables S2A and S2C. In contrast, resistant isolates showed substantially improved growth in the presence of doxycycline (16 μg/mL) relative to their susceptible progenitors (Figure 1E–1F; Supplementary Tables S3A–S3B). These findings highlight a trade-off between drug tolerance and growth capacity: while resistance evolution confers selective advantage under antibiotic pressure, they impose a fitness cost that reduces competitive growth in the absence of antibiotic pressure.

Reversion of Resistance Parallels Fitness Recovery

Withdrawal of doxycycline led to a rapid and reproducible reversion of resistance. Upon two consecutive weeks of drug-free passaging of evolved isolates (W13), MIC values declined sharply across all isolates (Figure 1G). For example, MIC of isolates 7 and 42 reverted from 16 to 1–2 μg/mL. The same MIC trend was detected among the remaining isolates (11, 62, and 71), falling below the CLSI resistance breakpoint. In parallel, revertant isolates (W13R) restored growth capacity comparable to baseline (W0) isolates (Figure 1A and 1B), with shorter lag phase and greater overall growth yield (AUC) (Figure 1C and 1D; Supplementary Tables S4A–S4E) relative to resistant isolates (W13). Notably, the reversion of resistance aligned with regaining fitness in revertant isolates (W13R), supporting that the fitness burden of resistance drives its instability in the absence of drug pressure. These results indicate that doxycycline resistance in V. cholerae is reversible in the absence of antibiotic selection, underscoring the role of fitness trade-offs in shaping resistance trajectories.

Resistant Isolates Develop Broader Antimicrobial Resistance

Doxycycline resistant isolates (W13) acquired expanded resistance profiles across multiple antibiotics classes (Figure 3). After 13 weeks of serial passaging, resistant isolates developed cross-resistance to different classes of antibiotics such as β-lactams, quinolones, sulfonamide and trimethoprim, phenicols, and polymyxins. MIC values of resistant isolates increased significantly against ampicillin, cefotaxime, chloramphenicol, Bactrim, Ciprofloxacin, and colistin, exceeding CLSI resistant breakpoints in isolates that were susceptible at baseline. However, MIC values remained unchanged for isolates that were already resistant for these antibiotics. Interestingly, collateral sensitivity was noticed for azithromycin in isolates 7, 11, and 71, while for 62 and 42 showed no change. These results demonstrate that resistance evolution in V. cholerae expands the multidrug resistant spectra, highlighting the potential of single-drug exposure to drive the emergence of multidrug resistance. MIC values are listed in Supplementary Table S5.

Figure 3.

V. cholerae O1 MIC heatmap: pre/post doxycycline, antibiotic MIC values, asterisks indicate resistance. The heatmap displays MIC values in micrograms per milliliter for V. cholerae O1 isolates before (W0) and after (W13) doxycycline exposure. Rows represent isolates 7, 62, 42, 11 and 71, with columns for antibiotics: Azithromycin, Ampicillin, Tazocin, Ceftazidime, Cefotaxime, Meropenem, Colistin, Nalidixic acid, Levofloxacin, Ciprofloxacin, Chloramphenicol and Bactrim. Antibiotic classes include beta lactams, Carbapenems, Polymyxins, Quinolones, Phenicols and Sulfonamides. The color gradient indicates MIC levels, with higher values represented by darker shades. Asterisks denote resistance per CLSI breakpoints. Notable changes include isolate 7(13) showing increased MIC for multiple antibiotics and isolate 62(13) with heightened resistance to beta lactams and quinolones. The heatmap highlights resistance evolution post-exposure, with specific isolate-antibiotic pairs shifting from low to high MIC values.

Open in a new tab

MIC heatmap of V. cholerae O1 isolates before (W0) and after doxycycline exposure (W13). MIC values (μg/mL) are shown across antibiotic classes, resistant values (per CLSI breakpoints) are marked with an asterisk (*).

Genetic Variants Underlying Resistance and Reversion

Whole genome sequencing (WGS) followed by variant calling identified distinct genetic variants associated with the evolution of resistance and its subsequent reversion.

Variants Associated with Resistance (W13) Isolates

Recurrent mutations were consistently detected across nearly all lineages (7, 11, 42, 62, and 71). Common variants were identified in the following genes: VC0166 (TetR family regulator), VCA1055 (LysR family regulator), VC0398 (MSHA biogenesis protein), VC1258 (DNA gyrase subunit A), VCA1100 (ABC transporter permease), VC0321 (Elongation factor Tu-A), Table 1. Moreover, these variants were homogeneous across replicate populations within each isolate, reflecting reproducibility. Notably, tetR mutations were of high impact alterations, including frameshift and stop-gained variants, suggesting that disruption of transcriptional regulation contributes to resistance induction. Missense variant mutations observed in LysR, and EF-TuA, inaddition to downstream modifications in transport (ABC transporter) and biofilm-associated genes further suggest altered efflux and motility as contributors. Additional lineage specific mutations were detected, including ribosomal S3 protein (RpsC, isolate 71) and a frameshift mutation in LpxH (isolate 11), which could be linked to colistin resistance. PyMOL modeling and molecular docking of RpsC mutation predicted to have no effect on doxycycline binding. Collectively, these findings implicate regulatory and efflux-related pathways as central to resistance acquisition. Common variants are summarized in Supplementary Tables S6A–S6E.

Table 1.

Common Genetic Variants Detected in V. cholerae W13 Isolates and Their Functional Annotations

Gene ID Variant Type Annotation Gene Function Isolate Count
VC_0166 Frameshift variant,stop gained variant Transcriptional regulator, TetR-family Transcriptional regulator 5/5
VC_0321 Missense variant Elongation factor Tu-A Translation 3/5
VC_0398 Frameshift variant MSHA biogenesis protein Biofilm/motility 5/5
VC_1258 Downstream gene variant DNA gyrase subunit A DNA topology 5/5
VC_A1055 Missense variant Transcriptional regulator, LysR family Transcriptional regulator 5/5
VC_A1100 Downstream gene variant ABC transporter, permease protein Transport (multidrug efflux, nutrient uptake) 4/5
Open in a new tab

Variants Associated with Reversion (W13R) Isolates

Revertant isolates (W13R) were characterized by the loss of tetR and lysR variants while acquiring distinct set of mutations (Table 2). Recurrent mutations were identified in VCA0129 (ribose ABC transporter), VCA0795 (DNA recombination), VC2210 (Iron acquisition), and VC139 (outer membrane protein). The disappearance of resistance related genes, coupled with the appearance of growth compensatory variants, coincided with restoration of fitness and reversion of doxycycline resistance. These observations indicate that while regulatory disruptions underlie the evolution of doxycycline resistance, metabolic and structural adaptations that restore fitness derives resistance instability in the absence of antibiotic-selective pressure. Complete variant list is shown in supplementary tables S7A–S7D.

Table 2.

Common Genetic Variants Detected in V. cholerae Isolates of Week 13 Revertant and Their Functional Annotations

Gene ID Variant Type Annotation Gene Function Isolate Count
VC_A0795 Downstream gene variant Resolvase, putative DNA recombination/repair (putative) 4/4
VC_A0129 Frameshift + missense Ribose ABC transporter, permease protein Transport (sugar uptake) 3/4
VC_2210 Upstream gene variant Vibriobactin utilization protein ViuB Iron acquisition 3/4
VC_1329 Downstream gene variant Opacity protein-related protein Outer membrane protein (putative) 3/4
Open in a new tab

Lineage-Bound Resistance and Baseline Reversion

SNP-based phylogenetic tree revealed doxycycline resistance remained confined within the original lineages. Doxycycline-resistant isolates (W13) clustered with their corresponding baseline progenitors (W0) rather than forming distinct clade (Figure 4B). Notably, revertant isolates (W13R) clustered closely with baseline W0 than resistant W13 isolates, highlighting genomic reversion toward baseline configuration. As an example, isolates 11 and 62 showed their revertant to be proximal to W0, while resistant counterparts are more distant within the same lineage. Significantly, this pattern was observed across the five isolates, demonstrating reproducibility of resistance evolution and reversion. Overall, this indicates that the evolved doxycycline resistance in V. cholerae is a transient, high-cost adaptation, governed by fitness trade-offs.

Figure 4.

Phylogenetic tree of V. cholerae showing doxycycline resistance patterns. The image has two parts: Part A is a circular phylogenetic tree of V. cholerae isolates, showing SNP-based global phylogeny. Surrounding rings indicate year, country, source and tetracycline resistance, with the outermost ring for year, color-coded by a legend. The second ring shows country, the third source and the fourth tetracycline resistance. Part B is a rectangular phylogenetic tree of baseline (W0), doxycycline-resistant (W13) and revertant (W13R) isolates. Resistant isolates (W13) cluster with baseline progenitors (W0), while revertant isolates (W13R) are closer to baseline than resistant ones. Isolates 11 and 62 highlight revertant forms near W0, indicating genomic reversion. This pattern across five isolates suggests transient resistance due to fitness trade-offs.

Open in a new tab

Phylogenetic context of doxycycline resistance in V. cholerae. (A) Global SNP-based phylogenetic tree positioning 2022 outbreak isolates within the seventh pandemic El Tor lineage alongside worldwide clinical isolates. Concentric rings indicate year, country, source, and tetracycline resistance phenotype. (B) SNP-based phylogeny of baseline (W0), doxycycline resistant (W13), and revertant (W13R) isolates from this study. Dashed arrows indicate isolates used in the experimental evolution analysis. Metadata and BioSample accession numbers of isolates used for SNP-based phylogenetic tree are listed in Supplementary Table S8.

Clinical Relevance of Doxycycline Resistance

To assess whether these evolutionary trajectories mirror patterns observed in clinical isolates, we positioned our evolved isolates within a global V. cholerae O1 phylogeny SNP-based including outbreak-associated isolates collected between 1988 and 2022, most of which are tetracycline resistant (Figure 4A). Lebanese isolates used in this study, clustered within the seventh pandemic El Tor lineage alongside diverse clinical isolates from Asia and Africa. Interestingly, one doxycycline resistant clinical isolate clustered closely with our laboratory-evolved isolates, revealing a shared genomic background between experimental and natural resistance. These results demonstrate that doxycycline resistance in V. cholerae tends to follow reproducible genetic path that is clinically relevant and evolutionary reversible, supporting its transient adaptation nature. BioSample accession numbers of isolates included in the global V. cholerae phylogeny are provided in Supplementary Table S8. The accession numbers of the individual samples used in this study are listed in Supplementary Table S9.

Discussion

The rapid induction of doxycycline resistance under sub-inhibitory concentrations (Figure 2) poses a major public health concern, particularly given the widespread presence of low-level antibiotic residues in both clinical and environmental settings. In cholera-endemic regions, V. cholerae is constantly exposed to such selective pressures through incomplete treatment courses,7 prophylactic use25 and contaminated water sources.26,27 Our findings demonstrate that sustained low-level doxycycline exposure can rapidly drive resistance acquisition, supporting the environmental relevance of this phenomenon. These observations are consistent with previous experimental studies showing that sub-inhibitory antibiotic concentrations promote resistance evolution in V. cholerae and other bacterial species through regulatory adaptation and efflux activation,28,29 often leading to multidrug resistance.30 These experimental findings mirror epidemiological observations that have documented persistent tetracycline resistance in environmental V. cholerae O1/O139 isolates,26 and highlighted significant global heterogeneity in tetracycline resistance among outbreak-associated strains.7 Importantly, doxycycline resistance has also been reported in non-O1/O139 V. cholerae isolates, underscoring the broader relevance of these dynamics.31 In light of both our findings and accumulating evidence of multidrug evolution under sustained doxycycline exposure, continued reliance on doxycycline as frontline therapy raises the need for urgent re-evaluation.

A key feature in the evolutionary trajectory we observed was the reversion of susceptibility upon doxycycline withdrawal, which occurred consistently across all lineages (Figure 1G). This reversion coincided with measurable fitness costs in resistant isolates, manifested as prolonged lag phases and reduced overall yield (Figure 1C and 1D). Notably, the growth rate (μ) of the resistant isolates was higher than baseline and revertant isolates, as represented by steeper slopes during the exponential phase (Figure 1A and 1B). This pattern suggests that although resistant bacteria initiate division with a higher rate once the lag phase is overcome, they appear to exhaust resources or cellular capacity more quickly, resulting in lower overall yield compared to baseline isolates. Such altered dynamics show that fitness costs extend beyond simple growth impairments to include phase-specific trade-offs in bacterial physiology.

Fitness costs associated with resistance are well established in the literature.32,33 For example, Wen et al reported that E. coli exposed to sub-inhibitory doxycycline exhibited growth defects associated with resistance acquisition.32 However, the study by Narendrakumar et al, which similarly evolved V. cholerae under doxycycline pressure, did not assess the long-term stability of resistance, leaving unresolved whether resistance is maintained once drug pressure is removed.30 Experimental studies suggest that compensatory mutations can reduce fitness costs without reducing resistance, potentially enabling resistant isolates to remain both highly fit and resistant.34 Yet clinical evidence is mixed: reduced macrolides use in Finnish hospitals resulted in a steady decline in erythromycin-resistant streptococcus pyogenes,31 whereas reduced sulfamethoxazole use in UK did not produce a similar decline.35 Our findings closely mirror the former scenario, showed that reversion coincided with susceptibility and fitness restoration. This aligns with the work done by Dunai et al, who demonstrated that antibiotic-free evolution of resistant E. coli strains partially restored wild type fitness and reduced resistance levels.34 To our knowledge, this is the first study to initiate evolution from sub-inhibitory concentrations, track both resistance induction and stability, and simultaneously monitor fitness outcomes. Previous studies typically examined fitness cost in resistant isolates or stability in pre-existing resistant strains, but not the entire trajectory.30,34 This comprehensive approach, demonstrating the central role of fitness in shaping doxycycline resistance and highlights reversion as a plausible outcome when selective pressure is removed.

Beyond doxycycline resistance, we observed broadened resistance to unrelated antibiotic classes as well as collateral sensitivity (Figure 3). Several lineages displayed collateral sensitivity to azithromycin in consistent with studies in other bacteria where resistance was associated with increased susceptibility to macrolides.36 Notably, both azithromycin and doxycycline target the bacterial ribosome, yet bind to different subunits, doxycycline to 30S subunit and azithromycin to 50S subunit.30 The observed collateral sensitivity suggests that adaptation to doxycycline may expose vulnerabilities in ribosomal function exploitable by macrolides. Conversely, we observed increased resistance against colistin reflecting that exposure to one antibiotic could increase resistance against last resort antibiotic. Importantly, the reversible nature of doxycycline resistance suggests it may not persist without antibiotic pressure, supporting the potential use of antibiotic cycling strategies. Exploiting collateral sensitivity may help limit resistance evolution and should be explored across other drugs and other bacterial species.

Whole genome sequencing provided mechanistic insight into both resistance acquisition and its subsequent reversion. Doxycycline-resistant isolates consistently accumulated high impact mutations in transcriptional regulators, most notably truncations in tetR and missense variants in lysR (Table 1). TetR family proteins are well characterized transcriptional repressors that regulate tetracycline resistance determinants, including efflux systems such as tetA by binding to operator sequences and inhibiting efflux pump expression in the absence of antibiotic pressure.37 Upon exposure to tetracycline, tetR undergoes conformational changes that relieve repression and allow expression of resistance genes.29 Beyond the canonical tetA system, tetR family regulators control broader networks of genes including multidrug efflux, and stress response pathways. This includes RND and MATE transporters such as AcrAB and NorM which are present in V. cholerae and contribute to multidrug resistance. In this context, truncating mutations observed in tetR are consistent with a loss of function mechanism that leads to constitutive expression of efflux pumps such as tetA, reduced intracellular antibiotic accumulations, and enhanced multidrug efflux expression, ultimately contributing to the observed cross-resistance phenotype.30,38,39

Similarly, LysR transcriptional regulators are global regulators that modulate several cellular processes, including efflux pump expression, oxidative stress responses and central metabolism. The missense mutations identified in lysR likely alter its regulatory activity leading to misexpression of efflux and stress-response genes which favors doxycycline resistance and enhances survival under drug pressure.38,39 However, such regulatory dysregulation is associated with a significant metabolic burden. Constitutive expression of efflux systems and stress-response pathways require high energy demands and can impair bacterial growth. This provides a mechanistic explanation for the reduced fitness observed in resistant isolates. Additional variants were also detected in ABC transporters, elongation factor Tu-A, DNA topology and biofilm-associated proteins, suggesting broader adaptive remodeling across multiple functional pathways (Table 1). Remarkably, lineage-specific variants in lpxH provide a potential mechanistic link to the observed collateral resistance to colistin. Together, these findings support the multimodal architecture of tetracycline resistance previously described in Gram-negative bacteria.30,38,39

Interestingly, resistant-associated mutations identified in doxycycline resistant isolates were no longer detected following drug withdrawal in revertant isolates. Instead, compensatory variants emerged in genes associated with DNA recombination, iron acquisition and outer membrane remodeling (Table 2). These findings suggest that resistance reversion is not driven by simple back mutations but rather by a combination of loss of resistance-associated mutations and acquisition of compensatory changes that restore bacterial fitness. The specific loss of tetR and lysR variants in revertants supports their role in mediating doxycycline resistance. However, the appearance of alternative mutations highlights broader metabolic and structural adaptations that counteract the fitness-cost imposed by resistance mutations. These interpretations align partly with Dunia et al (2019), who showed that phenotypic reversion can be mediated by the development of compensatory mutations while maintaining original resistant mutations34 In contrast, our findings indicate that reversion involves both loss of resistance-associated mutations and compensatory adaptation. Further experimental validation using a larger sample size is required to fully confirm the direct role of the identified mutations in resistance acquisition and reversion.10

Phylogenetic reconstruction provided an evolutionary view of how doxycycline resistance emerges and reverts in V. cholerae O1 (Figure 4). Resistant and revertant isolates remained confined within their baseline lineages, suggesting that resistant evolution is shaped by intrinsic genetic background. This pattern, consistent among both outbreak strains, suggests that each lineage follows its own adaptive route and reverts toward baseline, once the antibiotic pressure is removed (Figure 4B). Notably, a clinical doxycycline resistant isolate (S7A; Biosample accession SAMN45125187) originating from South Africa (2013) clustered closely with our in vitro evolved isolates (Figure 4A). Given its distinct geographic and temporal origin, this proximity suggests convergent evolutionary trajectories under doxycycline pressure. This finding supports the reproducibility of these evolutionary trajectories across diverse settings. A limitation of this study is that fitness was assessed using growth curve analysis alone, without competition assays. While growth curves provide insights into growth dynamics, competition assays offer more sensitive and ecologically relevant measure of relative fitness. Incorporating such approaches in future work would refine the quantification of fitness costs associated with resistance.

Although most clinically resistant V. cholerae isolates acquire tetracycline and doxycycline resistance via horizontal gene transfer of plasmid-borne tet genes like tetA and tetB.40,41 However, our findings show that bacterium can de novo develop resistance through chromosomal mutations in regulatory genes after sustained exposure to antibiotic pressure. This resistance was unstable and reverted once drug pressure was removed due to the high fitness costs associated with these mutations.32,34 These results could explain why plasmid encoded determinants – typically more stable and often compensated by secondary mutations – are the dominant mechanism of resistance in clinical and environmental isolates.10,42 Even so, transient chromosomal mutations could still affect therapy outcome especially when developing into multidrug resistance that is resistant to the last-resort antibiotics such as colistin.43 It could also persist for a long period until it captures plasmid-borne tet genes through horizontal gene transfer.44 Through mimicking the selective pressures encountered in cholera-endemic regions,45 our study reveals the remarkable adaptability of V. cholerae and raising concern about the continued reliability of doxycycline as a frontline therapy.

Conclusion

In summary, our findings reveal that doxycycline resistance in V. cholerae-examined here using the two outbreak strains from Lebanon-is a transient, high-cost adaptation that evolves under sustained selective pressure but reverts once the pressure is removed. The reproducible nature of these trajectories across independent lineages, and its clinical relevance, reflects a conserved and reversible path shaped by the balance between survival and fitness. Understanding these dynamics from the evolution of resistance and its persistence or disappearance is highly important which can help guide smarter antibiotic use and preserve the effectiveness of frontline treatments like doxycycline in cholera outbreaks.

Data Sharing Statement

Short reads were uploaded to the NCBI Sequence Read Archive (SRA)46 under Bio Project accession number PRJNA613441. All datasets are publicly accessible and can be retrieved through the SRA portal. Individual sample accession numbers are listed in Supplementary Table S9.

Ethical Statement

This study was based exclusively on previously collected, de-identified V. cholerae O1 isolates obtained through nationwide surveillance during the 2022 outbreak, coordinated by the Lebanese Ministry of Public Health and partner institutions, as described in a prior published study using the same isolates.9 No patient-identifiable data or direct human participation were involved. In accordance with the national regulations governing public health surveillance, ethical approval and informed consent were not required.

Disclosure

The authors report no conflicts of interest in this work.

References

  • 1.Laxminarayan R, Duse A, Wattal C, et al. Antibiotic resistance—the need for global solutions. Lancet Infect Dis. 2013;13:1057–13. doi: 10.1016/S1473-3099(13)70318-9 [DOI] [PubMed] [Google Scholar]
  • 2.Andersson DI, Hughes D. Microbiological effects of sublethal levels of antibiotics. Nat Rev Microbiol. 2014;12:465–478. doi: 10.1038/nrmicro3270 [DOI] [PubMed] [Google Scholar]
  • 3.Maeda T, Furusawa C. Laboratory evolution of antimicrobial resistance in bacteria to develop rational treatment strategies. Antibiotics. 2024;13:94. doi: 10.3390/antibiotics13010094 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Vanacker M, Lenuzza N, Rasigade J-P. The fitness cost of horizontally transferred and mutational antimicrobial resistance in Escherichia coli. Front Microbiol. 2023;14. doi: 10.3389/fmicb.2023.1186920 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.World Health Organization. Multi-Country Outbreak of Cholera. External Situation Report. 2025;28:–24. [Google Scholar]
  • 6.World Health Organization. Cholera treatment guidelines. Available from: https://www.who.int/. Accessed 2025.
  • 7.Ahmadi MH. Global status of tetracycline resistance among clinical isolates of Vibrio cholerae: a systematic review and meta-analysis. Antimicrob Resist Infect Control. 2021;10:115. doi: 10.1186/s13756-021-00985-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Blair JMA, Webber MA, Baylay AJ, Ogbolu DO, Piddock LJV. Molecular mechanisms of antibiotic resistance. Nat Rev Microbiol. 2015;13:42–51. doi: 10.1038/nrmicro3380 [DOI] [PubMed] [Google Scholar]
  • 9.Abou Fayad A, Rafei R, Njamkepo E, et al. An unusual two-strain cholera outbreak in Lebanon, 2022-2023: a genomic epidemiology study. Nat Commun. 2024;15:6963. doi: 10.1038/s41467-024-51428-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Andersson DI, Hughes D. Antibiotic resistance and its cost: is it possible to reverse resistance? Nat Rev Microbiol. 2010;8:260–271. doi: 10.1038/nrmicro2319 [DOI] [PubMed] [Google Scholar]
  • 11.Clinical and Laboratory Standards Institute (CLSI). Methods for Antimicrobial Dilution and Disk Susceptibility Testing of Infrequently Isolated or Fastidious Bacteria; Approved Guideline –. Third Edition (M45). Wayne, PA; 2015. [Google Scholar]
  • 12.Andrews S. FastQC: a quality control tool for high throughput sequence data. 2010.
  • 13.Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30:2114–2120. doi: 10.1093/bioinformatics/btu170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Charron P, Kang M. VariantDetective: an accurate all-in-one pipeline for detecting consensus bacterial SNPs and SVs. Bioinformatics. 2024;40:btae066. doi: 10.1093/bioinformatics/btae066 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25:1754–1760. doi: 10.1093/bioinformatics/btp324 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Garrison E, Marth G. Haplotype-Based Variant Detection From Short-Read Sequencing. Preprint at ed. 2012. [Google Scholar]
  • 17.McKenna A, Hanna M, Banks E, et al. The genome analysis toolkit: a mapreduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20:1297–1303. doi: 10.1101/gr.107524.110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Zheng Z, Li S, Su J, et al. Symphonizing pileup and full-alignment for deep learning-based long-read variant calling. Nat Comput Sci. 2022;2:797–803. doi: 10.1038/s43588-022-00387-x [DOI] [PubMed] [Google Scholar]
  • 19.McLaren W, Gil L, Hunt SE, et al. The Ensembl Variant Effect Predictor. Genome Biol. 2016;17:122. doi: 10.1186/s13059-016-0974-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Danecek P, Bonfield JK, Liddle J, et al. Twelve years of SAMtools and BCFtools. Gigascience. 2021;10:giab008. doi: 10.1093/gigascience/giab008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Hallgren MB, Overballe-Petersen S, Lund O, Hasman H, Clausen PTLC. MINTyper: an outbreak-detection method for accurate and rapid SNP typing of clonal clusters with noisy long reads. Biol Methods Protoc. 2021;6:bpab008. doi: 10.1093/biomethods/bpab008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Yu G. Using ggtree to Visualize Data on Tree-Like Structures. Curr Protoc Bioinformatics. 2020;69:e96. doi: 10.1002/cpbi.96 [DOI] [PubMed] [Google Scholar]
  • 23.R Core Team. R: a language and environment for statistical computing. (2022).
  • 24.McLeod AIK. Kendall rank correlation and Mann–Kendall trend test. 2022.
  • 25.World Health Organization. Global Task Force on Cholera Control (GTFCC). Interim Technical Note: Use of Antibiotics for the Treatment of Cholera. Available from: https://www.gtfcc.org/wp-content/uploads/2025/02/gtfcc-technical-note-on-use-of-antibiotics-for-the-treatment-of-cholera.pdf. Accessed May 21, 2026.
  • 26.Yuan X, Li Y-M, Vaziri AZ, et al. Global status of antimicrobial resistance among environmental isolates of Vibrio cholerae O1/O139: a systematic review and meta-analysis. Antimicrob Resist Infect Control. 2022;11:62. doi: 10.1186/s13756-022-01100-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Laviad-Shitrit S, Sharaby Y, Izhaki I, Peretz A, Halpern M. Antimicrobial susceptibility of environmental non-o1/non-o139 vibrio cholerae isolates. Front Microbiol. 2018;9. doi: 10.3389/fmicb.2018.01726 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Wistrand-Yuen E, Knopp M, Hjort K, et al. Evolution of high-level resistance during low-level antibiotic exposure. Nat Commun. 2018;9:1599. doi: 10.1038/s41467-018-04059-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Gullberg E, Cao S, Berg OG, et al. Selection of resistant bacteria at very low antibiotic concentrations. PLOS Pathogens. 2011;7:e1002158. doi: 10.1371/journal.ppat.1002158 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Narendrakumar L, Chandrika SK, Thomas S. Adaptive laboratory evolution of Vibrio cholerae to doxycycline associated with spontaneous mutation. Int J Antimicrob Agents. 2020;56:106097. doi: 10.1016/j.ijantimicag.2020.106097 [DOI] [PubMed] [Google Scholar]
  • 31.Seppälä H, Klaukka T, Vuopio-Varkila J, et al. The effect of changes in the consumption of macrolide antibiotics on erythromycin resistance in group A streptococci in Finland. Finnish Study Group for Antimicrobial Resistance. N Engl J Med. 1997;337:441–446. doi: 10.1056/NEJM199708143370701 [DOI] [PubMed] [Google Scholar]
  • 32.Wen X, Cao J, Mi J, et al. Metabonomics reveals an alleviation of fitness cost in resistant E. coli competing against susceptible E. coli at sub-MIC doxycycline. J Hazard Mater. 2021;405:124215. doi: 10.1016/j.jhazmat.2020.124215 [DOI] [PubMed] [Google Scholar]
  • 33.Pal A, Andersson DI. Bacteria can compensate the fitness costs of amplified resistance genes via a bypass mechanism. Nat Commun. 2024;15:2333. doi: 10.1038/s41467-024-46571-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Dunai A, Spohn R, Farkas Z, et al. Rapid decline of bacterial drug-resistance in an antibiotic-free environment through phenotypic reversion. eLife. 2019;8:e47088. doi: 10.7554/eLife.47088 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Enne VI, Livermore DM, Stephens P, Hall LM. Persistence of sulphonamide resistance in Escherichia coli in the UK despite national prescribing restriction. Lancet. 2001;357:1325–1328. doi: 10.1016/S0140-6736(00)04519-0 [DOI] [PubMed] [Google Scholar]
  • 36.Ma X, Xi W, Yang D, et al. Collateral sensitivity between tetracyclines and aminoglycosides constrains resistance evolution in carbapenem-resistant Klebsiella pneumoniae. Drug Resist Updates. 2023;68:100961. doi: 10.1016/j.drup.2023.100961 [DOI] [PubMed] [Google Scholar]
  • 37.Du D, Wang-Kan X, Neuberger A, et al. Multidrug efflux pumps: structure, function and regulation. Nat Rev Microbiol. 2018;16:523–539. doi: 10.1038/s41579-018-0048-6 [DOI] [PubMed] [Google Scholar]
  • 38.Hasan CM, Dutta D, Nguyen ANT. Revisiting antibiotic resistance: mechanistic foundations to evolutionary outlook. Antibiotics. 2021;11:40. doi: 10.3390/antibiotics11010040 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Xie P, Liu J, Feng Y, et al. Evolution and characteristics of tetracycline resistance in Mycoplasma pneumoniae. Microbiol Spectr. 2025;13:e0339824. doi: 10.1128/spectrum.03398-24 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Garrine M, Mandomando I, Vubil D, et al. Minimal genetic change in vibrio cholerae in mozambique over time: multilocus variable number tandem repeat analysis and whole genome sequencing. PLoS Negl Trop Dis. 2017;11:e0005671. doi: 10.1371/journal.pntd.0005671 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Weill F-X, Domman D, Njamkepo E, et al. Genomic history of the seventh pandemic of cholera in Africa. Science. 2017;358:785–789. doi: 10.1126/science.aad5901 [DOI] [PubMed] [Google Scholar]
  • 42.Nichol D, Rutter J, Bryant C, et al. Antibiotic collateral sensitivity is contingent on the repeatability of evolution. Nat Commun. 2019;10:334. doi: 10.1038/s41467-018-08098-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Hounmanou YMG, Leekitcharoenphon P, Hendriksen RS, et al. Surveillance and genomics of toxigenic vibrio cholerae o1 from fish, phytoplankton and water in lake victoria, Tanzania. Front Microbiol. 2019;10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Alkatheri AH, Yap PS-X, Abushelaibi A, et al. Microbial genomics: innovative targets and mechanisms. Antibiotics. 2023;12:190. doi: 10.3390/antibiotics12020190 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.World Health Organization. Cholera, 2023: Global Situation Update and Outbreak Response. (2023). Available from: https://www.who.int/publications/i/item/who-cholera-2023-global-situation-update. Accessed May 21, 2026.
  • 46.Leinonen R, Sugawara H, Shumway M; International Nucleotide Sequence Database Collaboration. The sequence read archive. Nucleic Acids Res. 2011;39:D19–21. doi: 10.1093/nar/gkq1019 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Short reads were uploaded to the NCBI Sequence Read Archive (SRA)46 under Bio Project accession number PRJNA613441. All datasets are publicly accessible and can be retrieved through the SRA portal. Individual sample accession numbers are listed in Supplementary Table S9.

Articles from Infection and Drug Resistance are provided here courtesy of Dove Press

RESOURCES