Addressing the growing threat of antimicrobial resistance (AMR) requires the development of large-scale resources that link bacterial genomic data with phenotypic antimicrobial resistance profiles. Such datasets are essential for advancing genotype-based predictions of resistance to uncover novel resistance mechanisms, as well as identifying and tracking global trends. Here, we describe the development of the ‘Comprehensive Assessment of Bacterial-Based AMR prediction from GEnotypes’ (CABBAGE) database, linking bacterial genomes to associated antibiotic susceptibility data and relevant metadata across WHO Bacterial Priority Pathogens, sourced from both publications and existing databases, and curated into a format that is compatible with and extends both NCBI and ENA’s formats. The resulting CABBAGE database, comprising over 170,000 unique sequenced isolates and approximately 1.7 million genome-phenotype pairs linked to extensive metadata, represents the largest of its kind, consolidating existing AMR phenotype-genotype data into a single unified format. The database encompasses a broad range of antimicrobials, facilitating the analysis of global resistance trends as well as benchmarks of genotype-to-phenotype predictive methods, and empowering further research uses. The database is freely accessible at and is currently being integrated with the BioSample database for easy access. ### Competing Interest Statement The authors have declared no competing interest. Medical Research Council, https://ror.org/03x94j517, MR/Z505547/1, MR/X020258/1

Viruses that infect bacteria, known as bacteriophages or phages, are the most prevalent entities on Earth. Their genetic diversity in nature is well documented, and members of divergent lineages can be found sharing the same ecological niche. This viral diversity can be influenced by a number of factors, including productivity, spatial structuring of the environment, and host-range trade-offs. Rapid evolution is also known to promote diversity by buffering ecological systems from extinction. There is, however, little known about the impact of coevolution on the maintenance of viral diversity within a microbial community. To address this, we developed a 4 species experimental system where two bacterial hosts, a generalist and a specialist phage, coevolved in a spatially homogenous environment over time. We observed the persistence of both viruses if the resource availability was sufficiently high. This coexistence occurred in the absence of any detectable host-range trade-offs that are costly for generalists and thus known to promote viral diversity. However, the coexistence was lost if two bacteria were not permitted to evolve alongside the phages or if two phages coevolved with a single bacterial host. Our findings indicate that a host's resistance response in mixed-species communities plays a significant role in maintaining viral diversity in the environment.

Antibiotics, by definition, reduce bacterial growth rates in optimal culture conditions; however, the real-world environments bacteria inhabit see rapid growth punctuated by periods of low nutrient availability. How antibiotics mediate population decline during these periods is poorly understood. Bacteria cannot optimize for all environmental conditions because a growth-longevity tradeoff predicts faster growth results in faster population decline, and since bacteriostatic antibiotics slow growth, they should also mediate longevity. We quantify how antibiotics, their targets, and resistance mechanisms influence longevity using populations of Escherichia coli and, as the tradeoff predicts, populations are maintained for longer if they encounter ribosome-binding antibiotics doxycycline and erythromycin, a finding that is not observed using antibiotics with alternative cellular targets. This tradeoff also predicts resistance mechanisms that increase growth rates during antibiotic treatment could be detrimental during nutrient stresses, and indeed, we find resistance by ribosomal protection removes benefits to longevity provided by doxycycline. We therefore liken ribosomal protection to a "Trojan horse" because it provides protection from an antibiotic but, during nutrient stresses, it promotes the demise of the bacteria. Seeking mechanisms to support these observations, we show doxycycline promotes efficient metabolism and reduces the concentration of reactive oxygen species. Seeking generality, we sought another mechanism that affects longevity and we found the number of doxycycline targets, namely, the ribosomal RNA operons, mediates growth and longevity even without antibiotics. We conclude that slow growth, as observed during antibiotic treatment, can help bacteria overcome later periods of nutrient stress.

Tradeoffs between life history traits impact diverse biological phenomena, from biodiversity maintenance to combating infectious disease. However, while research into trait correlations has focused on tradeoffs, less attention has been paid to other associations, like `tradeups' where traits are positively correlated. But if tradeoffs are common, tradeups should be too: if traits X and Y tradeoff, so too traits Y and Z, then X and Z will tradeup. Relying on tradeoffs as a mechanism for biodiversity maintenance is therefore problematic because tradeups are thought to purge polymorphisms. We therefore studied a model host-parasite system consisting of bacteriophage lambda and Escherichia coli that is thought to harbour a canonical tradeoff. We investigated several trait correlations in this model where the bacterial hosts should trade increased parasite resistance for diminished absolute fitness. Conversely, parasites should trade increased infectivity for reduced host-range. We report that these previously hypothesised tradeoffs appear, in fact, as tradeups and unexpectedly, phage resistance came with fitness benefits for bacteria. Bacteria-phage coevolution experiments coupled with protein and population dynamics modelling reveal the tradeoffs behind this tradeup. Structural mutations in E. coli alter the phage receptor, conferring resistance, but this slows nutrient uptake. As nutrient uptake negatively impacts biomass yield, some phage resistance mutations incur growth rate increases. Rapidly evolving bacterial systems are thought to operate near physical limits where changes in some traits incur concomitant losses in others and a common interpretation of this logic is that tradeoffs should be ubiquitous in polymorphic systems. We show this reasoning can be flawed because tradeoffs have been viewed as one-dimensional links between two traits, overlooking the freedoms we find in higher-dimensional trait interactions.

Antibiotic resistance represents a growing medical concern where raw, clinical datasets are under-exploited as a means to track the scale of the problem. We therefore sought patterns of antibiotic resistance in the Antimicrobial Testing Leadership and Surveillance (ATLAS) database. ATLAS holds 6.5M minimal inhibitory concentrations (MICs) for 3,919 pathogen-antibiotic pairs isolated from 633k patients in 70 countries between 2004 and 2017. We show most pairs form coherent, although not stationary, timeseries whose frequencies of resistance are higher than other databases, although we identified no systematic bias towards including more resistant strains in ATLAS. We sought data anomalies whereby MICs could shift for methodological and not clinical or microbiological reasons and found artefacts in over 100 pathogen-antibiotic pairs. Using an information-optimal clustering methodology to classify pathogens into low and high antibiotic susceptibilities, we used ATLAS to predict changes in resistance. Dynamics of the latter exhibit complex patterns with MIC increases, and some decreases, whereby subpopulations’ MICs can diverge. We also identify pathogens at risk of developing clinical resistance in the near future.

Developing mathematical models to accurately predict microbial growth dynamics remains a key challenge in ecology, evolution, biotechnology, and public health. To reproduce and grow, microbes need to take up essential nutrients from the environment, and mathematical models classically assume that the nutrient uptake rate is a saturating function of the nutrient concentration. In nature, microbes experience different levels of nutrient availability at all environmental scales, yet parameters shaping the nutrient uptake function are commonly estimated for a single initial nutrient concentration. This hampers the models from accurately capturing microbial dynamics when the environmental conditions change. To address this problem, we conduct growth experiments for a range of micro-organisms, including human fungal pathogens, baker's yeast, and common coliform bacteria, and uncover the following patterns. We observed that the maximal nutrient uptake rate and biomass yield were both decreasing functions of initial nutrient concentration. While a functional form for the relationship between biomass yield and initial nutrient concentration has been previously derived from first metabolic principles, here we also derive the form of the relationship between maximal nutrient uptake rate and initial nutrient concentration. Incorporating these two functions into a model of microbial growth allows for variable growth parameters and enables us to substantially improve predictions for microbial dynamics in a range of initial nutrient concentrations, compared to keeping growth parameters fixed.

To determine the dosage at which antibiotic resistance evolution is most rapid, we treated Escherichia coli in vitro, deploying the antibiotic erythromycin at dosages ranging from zero to high. Adaptation was fastest just below erythromycin's minimal inhibitory concentration (MIC) and genotype-phenotype correlations determined from whole genome sequencing revealed the molecular basis: simultaneous selection for copy number variation in three resistance mechanisms which exhibited an "inverted-U" pattern of dose-dependence, as did several insertion sequences and an integron. Many genes did not conform to this pattern, however, reflecting changes in selection as dose increased: putative media adaptation polymorphisms at zero antibiotic dosage gave way to drug target (ribosomal RNA operon) amplification at mid dosages whereas prophage-mediated drug efflux amplifications dominated at the highest dosages. All treatments exhibited E. coli increases in the copy number of efflux operons acrAB and emrE at rates that correlated with increases in population density. For strains where the inverted-U was no longer observed following the genetic manipulation of acrAB, it could be recovered by prolonging the antibiotic treatment at subMIC dosages.

We sought global patterns of antibiotic resistant pathogenic bacteria within the AMR Research Initiative database, Atlas. This consists of 6.5M clinical minimal inhibitory concentrations (MICs) observed in 70 countries in 633k patients between 2004 and 2017. Stratifying MICs according to pathogens (P), antibiotics (A) and countries (C), we found that the frequency of resistance was higher in Atlas than other publicly available databases. We determined global MIC distributions and, after showing they are coherent between years, we predicted MIC changes for 43 pathogens and 827 pathogen-antibiotic (PAs) pairings that exhibit significant resistance dynamics, including MIC increases and even decreases. However, many MIC distributions are multi-modal and some PA pairs exhibit sudden changes in MIC. We therefore analysed Atlas after replacing the clinical classification of pathogens into ‘susceptible’, ‘intermediate’ and ‘resistant’ with an information-optimal, cluster-based classifier to determine subpopulations with differential resistance that we denote S and R. Accordingly, S and R clusters for different PA pairs exhibit signatures of stabilising, directional and disruptive selection because their respective MICs can have different dynamics. Finally, we discuss clinical applications of a (R, dR/dt) ‘phase plane’ whereby the MIC of R is regressed against change in MIC (dR/dt), a methodology we use to detect PA pairs at risk of developing clinical resistance.

Antimicrobial resistance frequently carries a fitness cost to a pathogen, measured as a reduction in growth rate compared to the sensitive wild-type, in the absence of antibiotics. Existing empirical evidence points to the following relationship between cost of resistance and virulence. If a resistant pathogen suffers a fitness cost in terms of reduced growth rate it commonly has lower virulence compared to the sensitive wild-type. If this cost is absent so is the reduction in virulence. Here we show, using experimental evolution of drug resistance in the fungal human pathogenCandida glabrata,that reduced growth rate of resistant strains need not result in reduced virulence. Phenotypically heterogeneous populations were evolved in parallel containing highly resistant sub-population small colony variants (SCVs) alongside sensitive sub-populations. Despite their low growth rate in the absence of an antifungal drug, the SCVs did not suffer a marked alteration in virulence compared with the wild-type ancestral strain, or their co-isolated sensitive strains. This contrasts with classical theory that assumes growth rate to positively correlate with virulence. Our work thus highlights the complexity of the relationship between resistance, basic life-history traits and virulence.

Microbes occupy almost every niche within and on their human hosts. Whether colonizing the gut, mouth or bloodstream, microorganisms face temporal fluctuations in resources and stressors within their niche but we still know little of how environmental fluctuations mediate certain microbial phenotypes, notably antimicrobial-resistant ones. For instance, do rapid or slow fluctuations in nutrient and antimicrobial concentrations select for, or against, resistance? We tackle this question using an ecological approach by studying the dynamics of a synthetic and pathogenic microbial community containing two species, one sensitive and the other resistant to an antibiotic drug where the community is exposed to different rates of environmental fluctuation. We provide mathematical models, supported by experimental data, to demonstrate that simple community outcomes, such as competitive exclusion, can shift to coexistence and ecosystem bistability as fluctuation rates vary. Theory gives mechanistic insight into how these dynamical regimes are related. Importantly, our approach highlights a fundamental difference between resistance in single-species populations, the context in which it is usually assayed, and that in communities. While fast environmental changes are known to select against resistance in single-species populations, here we show that they can promote the resistant species in mixed-species communities. Our theoretical observations are verified empirically using a two-species Candida community.

Eukaryotes and prokaryotes exploit the ability of genetically identical cells to exhibit different phenotypes in order to enhance their survival. However, the mechanisms by which cells transition from one phenotype to another remain unclear. Canonical models of this dynamic posit that molecular fluctuations provide the noise that drives the cell out of one stable state and into another. Stochastic processes generated by canonical models should, therefore, be good descriptors of phenotype dynamics and between-state transitions should become more likely at greater noise amplitude, for instance at higher extracellular temperatures. To test these predictions, we observed temporal expression dynamics of the promoter of a flagellum gene, , in a microfluidic device using serovar Typhimurium and green fluorescent protein (GFP). Our observations show that while cells can exhibit multistable phenotypes, including stable -OFF and -ON states characterised by low and high GFP levels, respectively, between-state transitions can exhibit oscillatory dynamics whose return statistics do not conform to canonical theories. For example, here the -ON state was more frequent following a temperature increase. To better understand our data we developed different dynamical frameworks to predict expression data. We conclude that a stochastic dynamical system tailored to the genetic network of is better suited to our data than prior theories where dynamical features, like oscillations and pulsing, are driven by inevitable delays in the post-translational regulation of . Thus, while transcriptional noise promotes phenotypic heterogeneity, as we show here, regular features like oscillatory heterogeneity can result from delays that fundamental molecular processes impose upon a cell’s gene regulatory architecture.

Fisher suggested advantageous genes would spread through populations as a wave so we sought genetic waves in evolving populations, as follows. By fusing a fluorescent marker to a drug efflux protein (AcrB) whose expression provides Escherichia coli with resistance to some antibiotics, we quantified the evolution and spread of drug-resistant E. coli through spacetime using image analysis and quantitative PCR. As is done in hospitals routinely, we exposed the bacterium to a gradient of antibiotic in a ‘disk diffusion’ drug susceptibility test that we videoed. The videos show complex spatio-genomic patterns redolent of, yet more complex than, Fisher’s predictions whereby a decelerating wave front of advantageous genes colonises towards the antibiotic source, forming bullseye patterns en route and leaving a wave back of bacterial sub-populations expressing AcrB at decreasing levels away from the drug source. qPCR data show that E. coli sited at rapidly-adapting spatial hotspots gain 2 additional copies of acr , the operon that encodes AcrB, within 24h and imaging data show resistant sub-populations thrive most near the antibiotic source due to non-monotone relationships between inhibition due to antibiotic and distance from the source. In the spirit of Fisher, we provide an explicitly spatial nonlinear diffusion equation that exhibits these properties too. Finally, linear diffusion theory quantifies how the spatial extent of bacterial killing scales with increases in antibiotic dosage, predicting that microbes can survive chemotherapies that have been escalated to 250× the clinical dosage if the antibiotic is diffusion-limited.

We treated Escherichia coli with the antibiotic erythromycin from zero to high dosages to determine how the evolutionary dynamics of antibiotic resistant phenotypes and genotypes depend on dose. The most rapid increase in resistance was observed just below erythromycin’s minimal inhibitory concentration (MIC) and genotype-phenotype correlations determined from whole genome sequencing revealed the molecular basis of this: simultaneous selection for copy number variation in 3 resistance mechanisms which shared an ‘inverted-U’ pattern of dose-dependent selection with several insertion sequences and an integron. Many genes did not conform to this pattern, however, because of changes in selection as dose increased: media adaptation at zero-to-low dosages gave way to drug target (ribosomal RNA operon) amplification at mid dosages whereas prophage-mediated drug efflux dominated at higher dosages where population densities were lowest. All dosages saw E. coli amplify the efflux operons acr and emrE at rates that correlated strongly with changes in population density that exhibited an inverted-U geometry too. However, we show by example that inverted-U geometries are not a universal feature of dose-resistance relationships.

In the version of this Article originally published, the following sentence was missing from the Acknowledgements: “R.E.B. is an EPSRC Healthcare Technologies Impact Fellow EP/N033671/1; I.G. is funded by ERC Consolidator grant 647292 MathModExp; A.J.P.B., N.A.R.G. and A.T. were funded by BBSRC grant BB/F00513X/1; K.H., I.G., S.N. and E.C. were funded by BBSRC grant BB/F005210/2.” This text has now been added.

The spread of antibiotic resistance is always a consequence of evolutionary processes. The consideration of evolution is thus key to the development of sustainable therapy. Two main factors were recently proposed to enhance long-term effectiveness of drug combinations: evolved collateral sensitivities between the drugs in a pair and antagonistic drug interactions. We systematically assessed these factors by performing over 1,600 evolution experiments with the opportunistic nosocomial pathogen Pseudomonas aeruginosa in single- and multidrug environments. Based on the growth dynamics during these experiments, we reconstructed antibiotic combination efficacy (ACE) networks as a new tool for characterizing the ability of the tested drug combinations to constrain bacterial survival as well as drug resistance evolution across time. Subsequent statistical analysis of the influence of the factors on ACE network characteristics revealed that (i) synergistic drug interactions increased the likelihood of bacterial population extinction—irrespective of whether combinations were compared at the same level of inhibition or not—while (ii) the potential for evolved collateral sensitivities between 2 drugs accounted for a reduction in bacterial adaptation rates. In sum, our systematic experimental analysis allowed us to pinpoint 2 complementary determinants of combination efficacy and to identify specific drug pairs with high ACE scores. Our findings can guide attempts to further improve the sustainability of antibiotic therapy by simultaneously reducing pathogen load and resistance evolution.

a mixed-species community Robert E. Beardmore1+∗, Emily Cook, Susanna Nilsson, Adam R. Smith, Anna Tillmann, Brooke D. Esquivel Ken Haynes1†, Neil A. R. Gow, Alistair J. P. Brown, Theodore C. White and Ivana Gudelj1∗ 1Biosciences, University of Exeter, Exeter, UK, 3School of Biological Sciences, University of Missouri at Kansas City, Kansas City, MO, USA 4Institute of Medical Sciences, Foresterhill, University of Aberdeen, Aberdeen, UK +Authors contributed equally; ∗To whom correspondence should be addressed; E-mails: r.e.beardmore@exeter.ac.uk; i.gudelj@exeter.ac.uk † Passed away 19th March 2018

The worth of antibiotics to human health is without bound, but antimicrobial resistance poses a serious threat to this treasured resource. 1 Hollis A Ahmed Z Preserving antibiotics, rationally. N Engl J Med. 2013; 369: 2474-2476 Crossref PubMed Scopus (111) Google Scholar Reduction of the selection pressure on pathogens through the rational use of antimicrobials, and discernment of which antibiotic treatment strategies achieve this goal, are now international priorities. Antibiotic stewardship encompasses different evidence-based measures to improve the appropriate use of antibiotics by promoting selection of optimal drug regimens, including dosing, duration of therapy, and route of administration. 2 Society for Healthcare Epidemiology of AmericaInfectious Diseases Society of AmericaPediatric Infectious Diseases SocietyPolicy statement on antimicrobial stewardship by the Society for Healthcare Epidemiology of America (SHEA), the Infectious Diseases Society of America (IDSA), and the Pediatric Infectious Diseases Society (PIDS). Infect Control Hosp Epidemiol. 2012; 33: 322-327 Crossref PubMed Scopus (479) Google Scholar , 3 Baur D Gladstone BP Burkert F et al. Effect of antibiotic stewardship on the incidence of infection and colonisation with antibiotic-resistant bacteria and Clostridium difficile infection: a systematic review and meta-analysis. Lancet Infect Dis. 2017; 17: 990-1001 Summary Full Text Full Text PDF PubMed Scopus (388) Google Scholar There have been important recent advances in our understanding of these approaches, but we are still some way from truly knowing how to best use antibiotics in the setting of patients with sepsis in critical care. Although there is a strong consensus that, at the patient level, prescribers need to ensure the right antibiotic, at the right dose, at the right time is used in cases where antibiotics are truly warranted, 4 The LancetNICE antimicrobial stewardship: right drug, dose, and time?. Lancet. 2015; 386: 717 Summary Full Text Full Text PDF PubMed Scopus (11) Google Scholar treatment strategies at the unit level remain unclear. Among manifold others, two such strategies are cycling and mixing of antibiotics. 5 Beardmore RE Pena-Miller R Gori F Iredell J Antibiotic cycling and antibiotic mixing: which one best mitigates antibiotic resistance?. Mol Biol Evol. 2017; 34: 802-817 PubMed Google Scholar Antibiotic cycling refers to prescription of a specific antibiotic drug preferentially as first-line therapy during a prespecified period, with subsequent rotation to another antibiotic drug with different selective properties later on. Antibiotic mixing refers to the changing of antibiotics each time a new patient in need of this drug arrives. The effects of antibiotic cycling and mixing on antibiotic resistance in intensive care units: a cluster-randomised crossover trialAntibiotic cycling does not reduce the prevalence of carriage of antibiotic-resistant, Gram-negative bacteria in patients admitted to the ICU. Full-Text PDF

Can we exploit our burgeoning understanding of molecular evolution to slow the progress of drug resistance? One role of an infection clinician is exactly that: to foresee trajectories to resistance during antibiotic treatment and to hinder that evolutionary course. But can this be done at a hospital-wide scale? Clinicians and theoreticians tried to when they proposed two conflicting behavioral strategies that are expected to curb resistance evolution in the clinic, these are known as "antibiotic cycling" and "antibiotic mixing." However, the accumulated data from clinical trials, now approaching 4 million patient days of treatment, is too variable for cycling or mixing to be deemed successful. The former implements the restriction and prioritization of different antibiotics at different times in hospitals in a manner said to "cycle" between them. In antibiotic mixing, appropriate antibiotics are allocated to patients but randomly. Mixing results in no correlation, in time or across patients, in the drugs used for treatment which is why theorists saw this as an optimal behavioral strategy. So while cycling and mixing were proposed as ways of controlling evolution, we show there is good reason why clinical datasets cannot choose between them: by re-examining the theoretical literature we show prior support for the theoretical optimality ofmixing was misplaced. Our analysis is consistent with a pattern emerging in data: neither cycling or mixing is a priori better than the other at mitigating selection for antibiotic resistance in the clinic.

When bacteria evolve resistance against a particular antibiotic, they may simultaneously gain increased sensitivity against a second one. Such collateral sensitivity may be exploited to develop novel, sustainable antibiotic treatment strategies aimed at containing the current, dramatic spread of drug resistance. To date, the presence and molecular basis of collateral sensitivity has only been studied in few bacterial species and is unknown for opportunistic human pathogens such as Pseudomonas aeruginosa. In the present study, we assessed patterns of collateral effects by experimentally evolving 160 independent populations of P. aeruginosa to high levels of resistance against eight commonly used antibiotics. The bacteria evolved resistance rapidly and expressed both collateral sensitivity and cross-resistance. The pattern of such collateral effects differed to those previously reported for other bacterial species, suggesting interspecific differences in the underlying evolutionary trade-offs. Intriguingly, we also identified contrasting patterns of collateral sensitivity and cross-resistance among the replicate populations adapted to the same drug. Whole-genome sequencing of 81 independently evolved populations revealed distinct evolutionary paths of resistance to the selective drug, which determined whether bacteria became cross-resistant or collaterally sensitive towards others. Based on genomic and functional genetic analysis, we demonstrate that collateral sensitivity can result from resistance mutations in regulatory genes such as nalC or mexZ, which mediate aminoglycoside sensitivity in beta-lactam-adapted populations, or the two-component regulatory system gene pmrB, which enhances penicillin sensitivity in gentamicin-resistant populations. Our findings highlight substantial variation in the evolved collateral effects among replicates, which in turn determine their potential in antibiotic therapy.